Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Review - Intervention

Psychological interventions for needle‐related procedural pain and distress in children and adolescents

Authors' declarations of interest

Version published: 04 October 2018 Version history

https://doi.org/10.1002/14651858.CD005179.pub4
Collapse all Expand all

Background

This is the second update of a Cochrane Review (Issue 4, 2006). Pain and distress from needle‐related procedures are common during childhood and can be reduced through use of psychological interventions (cognitive or behavioral strategies, or both). Our first review update (Issue 10, 2013) showed efficacy of distraction and hypnosis for needle‐related pain and distress in children and adolescents.

Objectives

To assess the efficacy of psychological interventions for needle‐related procedural pain and distress in children and adolescents.

Search methods

We searched six electronic databases for relevant trials: Cochrane Central Register of Controlled Trials (CENTRAL); MEDLINE; PsycINFO; Embase; Web of Science (ISI Web of Knowledge); and Cumulative Index to Nursing and Allied Health Literature (CINAHL). We sent requests for additional studies to pediatric pain and child health electronic listservs. We also searched registries for relevant completed trials: clinicaltrials.gov; and World Health Organization International Clinical Trials Registry Platform (www.who.int.trialsearch). We conducted searches up to September 2017 to identify records published since the last review update in 2013.

Selection criteria

We included peer‐reviewed published randomized controlled trials (RCTs) with at least five participants per study arm, comparing a psychological intervention with a control or comparison group. Trials involved children aged two to 19 years undergoing any needle‐related medical procedure.

Data collection and analysis

Two review authors extracted data and assessed risks of bias using the Cochrane 'Risk of bias' tool. We examined pain and distress assessed by child self‐report, observer global report, and behavioral measurement (primary outcomes). We also examined any reported physiological outcomes and adverse events (secondary outcomes). We used meta‐analysis to assess the efficacy of identified psychological interventions relative to a comparator (i.e. no treatment, other active treatment, treatment as usual, or waitlist) for each outcome separately. We used Review Manager 5 software to compute standardized mean differences (SMDs) with 95% confidence intervals (CIs), and GRADE to assess the quality of the evidence.

Main results

We included 59 trials (20 new for this update) with 5550 participants. Needle procedures primarily included venipuncture, intravenous insertion, and vaccine injections. Studies included children aged two to 19 years, with few trials focused on adolescents. The most common psychological interventions were distraction (n = 32), combined cognitive behavioral therapy (CBT; n = 18), and hypnosis (n = 8). Preparation/information (n = 4), breathing (n = 4), suggestion (n = 3), and memory alteration (n = 1) were also included. Control groups were often 'standard care', which varied across studies. Across all studies, 'Risk of bias' scores indicated several domains at high or unclear risk, most notably allocation concealment, blinding of participants and outcome assessment, and selective reporting. We downgraded the quality of evidence largely due to serious study limitations, inconsistency, and imprecision.

Very low‐ to low‐quality evidence supported the efficacy of distraction for self‐reported pain (n = 30, 2802 participants; SMD −0.56, 95% CI −0.78 to −0.33) and distress (n = 4, 426 participants; SMD −0.82, 95% CI −1.45 to −0.18), observer‐reported pain (n = 11, 1512 participants; SMD −0.62, 95% CI −1.00 to −0.23) and distress (n = 5, 1067 participants; SMD −0.72, 95% CI −1.41 to −0.03), and behavioral distress (n = 7, 500 participants; SMD −0.44, 95% CI −0.84 to −0.04). Distraction was not efficacious for behavioral pain (n = 4, 309 participants; SMD −0.33, 95% CI −0.69 to 0.03). Very low‐quality evidence indicated hypnosis was efficacious for reducing self‐reported pain (n = 5, 176 participants; SMD −1.40, 95% CI −2.32 to −0.48) and distress (n = 5, 176 participants; SMD −2.53, 95% CI −3.93 to −1.12), and behavioral distress (n = 6, 193 participants; SMD −1.15, 95% CI −1.76 to −0.53), but not behavioral pain (n = 2, 69 participants; SMD −0.38, 95% CI −1.57 to 0.81). No studies assessed hypnosis for observer‐reported pain and only one study assessed observer‐reported distress. Very low‐ to low‐quality evidence supported the efficacy of combined CBT for observer‐reported pain (n = 4, 385 participants; SMD −0.52, 95% CI −0.73 to −0.30) and behavioral distress (n = 11, 1105 participants; SMD −0.40, 95% CI −0.67 to −0.14), but not self‐reported pain (n = 14, 1359 participants; SMD −0.27, 95% CI −0.58 to 0.03), self‐reported distress (n = 6, 234 participants; SMD −0.26, 95% CI −0.56 to 0.04), observer‐reported distress (n = 6, 765 participants; SMD 0.08, 95% CI −0.34 to 0.50), or behavioral pain (n = 2, 95 participants; SMD −0.65, 95% CI −2.36 to 1.06). Very low‐quality evidence showed efficacy of breathing interventions for self‐reported pain (n = 4, 298 participants; SMD −1.04, 95% CI −1.86 to −0.22), but there were too few studies for meta‐analysis of other outcomes. Very low‐quality evidence revealed no effect for preparation/information (n = 4, 313 participants) or suggestion (n = 3, 218 participants) for any pain or distress outcome. Given only a single trial, we could draw no conclusions about memory alteration. Adverse events of respiratory difficulties were only reported in one breathing intervention.

Authors' conclusions

We identified evidence supporting the efficacy of distraction, hypnosis, combined CBT, and breathing interventions for reducing children’s needle‐related pain or distress, or both. Support for the efficacy of combined CBT and breathing interventions is new from our last review update due to the availability of new evidence. The quality of trials and overall evidence remains low to very low, underscoring the need for improved methodological rigor and trial reporting. Despite low‐quality evidence, the potential benefits of reduced pain or distress or both support the evidence in favor of using these interventions in clinical practice.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Psychological strategies to reduce pain and distress for children and adolescents getting needles

Bottom line

Psychological strategies help reduce children's pain, distress, and fear of needles. Distraction and hypnosis are helpful, although specific breathing (such as inflating a balloon), and combining multiple psychological strategies can also help.

Background

Psychological strategies affect how children think or what they do before, during, or after a needle. They can be used by children or with support from parents or medical staff, like nurses, psychologists, or child life specialists. The information applies to children aged from two to 19 years who are healthy or ill, undergoing all types of needle procedures at the hospital, in a clinic, or at school.

Key results

For this update, in September 2017, we searched for clinical trials looking at psychological strategies for reducing pain and distress of children and teens getting a needle. We found 59 trials including 5550 children and teens. Twenty of these trials were new for this update. We found six psychological strategies, four of which help reduce children's pain and distress during needles. These include distraction, hypnosis, specific breathing, and combining multiple strategies (‘combined cognitive behavioral’). Ways to distract children and teens during needles include reading, watching a movie, listening to music, playing video games, or virtual reality. Hypnosis involves deep relaxation and imagery, and is usually taught to a child by a trained professional. Examples of strategies that can be combined include distraction, breathing, relaxation, positive thoughts, having the child learn or practice the steps of the needle procedure, and coaching parents about ways to support their child. Other psychological strategies have been tested but do not seem helpful on their own. For example, children do not have less pain or distress when they are only told what is going to happen before or during the needle ('providing information or preparation or both) or when someone merely suggests to the child that something is being done to help them. One other strategy is helping children to remember their previous needles more positively. There is not enough information yet to know if this is helpful.

Quality of evidence

We rated the quality of the evidence from studies using four levels: very low, low, moderate, or high. The quality of the evidence from this review is very low to low, as results may be biased by including only small numbers of children or by children knowing what psychological intervention they received. This means that we are uncertain about the results.

Authors' conclusions

Implications for practice

The findings of this review support the efficacy of distraction, hypnosis, combined CBT, and breathing interventions for reducing children’s needle‐related pain and distress. Despite very low‐ to low‐quality evidence across psychological interventions, the potentially desirable outcomes and value of reduced pain or distress or both provide justification for their use in clinical practice (Andrews 2013a; Andrews 2013b; Higgins 2017).

Clinically, distraction interventions can be used during a variety of routine and more invasive needle procedures across a range of outpatient, inpatient, community, and emergency‐care settings. Findings suggest that a variety of distractors are efficacious; thus, the type of distraction intervention (e.g. books, verbal distraction, virtual reality) does not seem to matter. The evidence supports clinicians using distraction interventions as appropriate for their setting, availability, resources, and expertise. These are particularly useful psychological interventions for even low‐resourced areas and can be feasibly delivered by any healthcare provider. Although parents could potentially deliver distraction interventions, their efficacy is questionable among parents with higher levels of anxiety (Dahlquist 2005). We also assert that these interventions may not be useful for children with significant needle fears who require more intensive, exposure‐based psychological interventions (McMurtry 2016).

Findings also continue to support the use of hypnosis for the management of children’s pain and distress during needle procedures. Hypnosis is particularly applicable in the context of more invasive needle‐related procedures for children with cancer (e.g. bone marrow aspirations, lumbar punctures). Practically speaking, hypnosis requires clinician training for professionals or parents or training in self‐hypnosis for children, and probably greater resources (e.g. time) to implement, which may limit the feasibility of these interventions in some clinical settings.

While we found preliminary support for the use of some basic breathing interventions (e.g. balloon inflation, deep breathing), three adverse events were reported in one of these trials, with children discontinuing intervention due to respiratory difficulties (Pourmovahed 2013). Indeed, the appropriateness of each intervention may differ across settings and clinical populations.

Although we found combined CBT interventions to be efficacious in reducing observer‐reported pain and behavioral distress, findings did not support the use of combined CBT for other outcomes, most notably self‐report. This suggests that increased efforts (time, resources) towards combining elements of CBT in multi‐pronged interventions may not be worthwhile or grounded in evidence. Single‐strategy interventions, i.e. hypnosis, distraction, and breathing, were the only interventions found to be efficacious in reducing child self‐report of pain in this setting. Providing preparatory information alone or merely suggesting something was being done to reduce pain or distress was not effective for any outcomes.

Although the findings here are drawn from studies including children from two to 19 years old, most studies focused on children aged 12 and younger. It is therefore difficult to ascertain the efficacy of interventions for children of particular ages, particularly adolescents. Overall, the implications for practice described here are most directly applicable to children aged 12 and under.

Research knowledge of the efficacy of psychological interventions to reduce needle pain and distress is of little value unless that knowledge is disseminated and taken up into clinical practice. Researchers, clinicians, policy‐makers, parents, and other knowledge users are strongly encouraged to use the evidence reviewed here to inform dissemination and implementation of efficacious psychological strategies to needle procedure and pain management practice. Examples include clinical practice guidelines (Taddio 2015; McMurtry 2016), standards of care (Flowers 2015), institutional policies (ChildKind International; www.childkindinternational.org; Schechter 2008; Schechter 2010a; WHO 2015), and videos and pamphlets targeting healthcare providers and parents (Chambers 2013; www.youtube.com/watch?v=KgBwVSYqfps; Chambers 2016; www.parentscanada.com/health/expert-advice-to-help-make-vaccinations-easier). It is critical to ensure that research evidence reaches those who need it and will use it.

Implications for research

1. General Implications

Examining psychological interventions to reduce needle‐related pain and distress experienced by children and adolescents is a large and very active area of research. For those interventions that are shown to be efficacious, it is time to turn our research attention to ensuring that dissemination of such knowledge is timely, meaningful, and effective. We are referring to the need for empirical evaluation of factors that increase the uptake of intervention strategies into clinical practice, a field known as implementation science (Bauer 2015; Curran 2012; Wittmeier 2015). This is a critical and worthy research endeavor. Several new studies included in this updated review continue to include standard care for needle procedures with children that exclude evidence‐based pain management strategies identified in our original review (Uman 2006). This reflects a lag of more than 11 years to move research evidence into practice, and highlights the need for concerted efforts to reduce this evidence‐to‐practice gap (Morris 2011).

We continue to assert that future studies comparing distraction interventions to standard care are of little value, given continued support for their efficacy. To advance treatment research in this area, studies should include head‐to‐head comparisons of different types of distraction interventions (Birnie 2014a). This has seldom been done, with some exception (Aydin 2017; Bellieni 2006; Miller 2016; Sahiner 2016). We also note a clear lack of evidence and need for future trials of interventions that are frequently used in clinical practice, widely accessible, and potentially deliverable through e‐health or m‐health platforms (e.g. smartphones, tablets). More trials are needed for those interventions covered only by single trials (e.g. memory alteration) and for those where there has been little new evidence in the past several years (e.g. hypnosis). Given the integral role of parents in children’s experience of needle‐related pain and fear, and their presence during these procedures, additional trials involving parents are needed.

2. Design

At this point in time, strong emphasis should be placed on clinical equipoise when designing clinical trials in this area. Robust evidence exists for the efficacy of particular psychological interventions (distraction, hypnosis) in reducing children’s needle‐related distress and pain; there should therefore not be any new trials in which a comparison group receives less than standard care (i.e. some form of evidence‐based pain relief, such as topical anesthetic). The use of placebo or no‐analgesia control groups has been strongly condemned in clinical trials of analgesic treatments for procedural pain in infants, with calls for ethics committees and journals not to approve or publish such trials (Bellieni 2016). Given the evidence in this review, it is reasonable to apply the same expectations that trials stop implementing no‐treatment comparison groups to children and adolescents. Although some trials have moved to including pain management interventions in their standard‐care control groups (distraction, Crevatin 2016; topical anesthetics, Miller 2016), we continue to note inconsistencies in this regard across the evidence base.

Researchers should carefully follow the CONSORT guidelines (Moher 2010; www.consort-statement.org) when reporting randomized trials to ensure that details relevant to randomization, allocation, and blinding are adequately addressed. We continue to strongly encourage researchers to draw from available standards and guidelines for designing pediatric trials (StaR Child Health) informing recruitment, consent (Caldwell 2012), sample size estimation (Van der Tweel 2012), minimizing risk of bias (Hartling 2012), considering developmental stage (Williams 2012), and outcome selection, assessment, and reporting (Sinha 2012; PedIMMPACT McGrath 2008; CONSORT, Altman 2001). All trials should be registered (clinicaltrials.gov or www.who.int.trialsearch).

3. Measurement

Trial reporting should include all outcomes necessary for meta‐analysis (e.g. means, SDs, participant cell sizes). Manuscripts should also clearly report the types and characteristics of all interventions used in sufficient detail to be understood and replicated by another clinician or researcher (Birnie 2014a). Findings from this second review update suggest that treatment efficacy differs across different outcomes; trials should therefore include a variety of primary outcomes beyond self‐report of pain, such as emotional response (e.g. distress, anxiety, fear, unhappiness), satisfaction with treatment, physical recovery, economic factors (e.g. costs), and adverse events and symptoms (PedIMMPACT; McGrath 2008). We suggest that self‐report be included among children who can provide valid and reliable self‐report (e.g. children aged four to five years and upwards; Von Baeyer 2017); however, researchers should also strive to include reports of child pain and distress by others (e.g. parents, behavioral scales). We continue to encourage researchers to take all efforts to blind outcomes wherever possible (e.g. behavioral ratings coded from video).

4. Other

Future research efforts should focus on examining and confirming treatment mechanisms. This includes integration with other types of experimental designs and methods (e.g. neuroimaging, Birnie 2017). This research is particularly important for interventions where the mechanism of effect is unclear or where multiple potential mechanisms are involved. This is critical, as it directly informs the determination and design of most effective treatments, which are now difficult to determine given the significant heterogeneity within intervention categories such as distraction. Further research in this area may also inform more accurate categorization of psychological interventions. For example, breathing interventions in which children are instructed to inflate a balloon may evoke a physiological mechanism, expectancy, or distraction (Gupta 2006; Sahiner 2016).

We encourage researchers to conduct studies informing how to tailor the right intervention to the right child at the right time. Interventions should ideally be matched to child, setting, and context. Individual differences will invariably influence responsiveness to psychological interventions. For example, children with high versus low fear of pain or tendencies to catastrophize about pain may benefit less from distraction (Birnie 2017; Campbell 2017; Verhoeven 2012) and more from hypnosis that involves more intensive training, attentional engagement, and suggestion. The influence of participant preference and choice on treatment efficacy should also be considered (Birnie 2014a). Particularly in the context of vaccination, there is a notable dearth of interventions delivered in the school setting. Given that school‐aged children and adolescents often receive vaccine injections in this setting, research in this unique context would be valuable (Boerner 2014).

Once again, we argue it is imperative that future trials take into account the child’s age and developmental stage when developing, implementing, and evaluating interventions. There was a notable lack of trials specifically targeting adolescents, and participant samples included a wide range of children spanning several developmental periods. Age can influence treatment efficacy in these interventions (Birnie 2014a). We suggest that studies including a wide age range of children used age‐based analyses and report outcomes by recommended age groups for pediatric clinical trials (Williams 2012). Previous experience with and associated memories of needle procedures also invariably influence the child’s pain experience and treatment responsiveness (Noel 2012). Comparison of intervention efficacy for single (vaccine injection) versus repeated needle procedures (bone marrow aspirations for cancer treatment regimens) is warranted. Relatedly, trials examining intervention implementation and efficacy among youth with developmental disabilities are notably absent from the literature and are needed (Boerner 2014).

Finally, wide dissemination of efficacious interventions is critically important and can be threatened by real‐world concerns about time, resources, and feasibility in busy clinical settings. Given that our findings suggest a variety of distraction interventions are efficacious regardless of cost and level of technology, cost effectiveness should be assessed and considered. This is highly relevant for uptake of evidence‐based pain management in less resourced areas.

Summary of findings

Open in table viewer
Summary of findings 1. Distraction compared to control for needle‐related procedural pain and distress in children and adolescents

Distraction compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 2‐19 years with mixed medical (acute or chronic illness) or generally healthy undergoing venipuncture, immunization, intravenous insertion, lumbar puncture, bone marrow aspiration, routine injection, allergy testing injections, or laceration repair
Setting: hospital (inpatient/outpatient/emergency department), community clinic, or school
Intervention: distraction
Comparison: control (varied across studies)

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with distraction

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 0.65 to 8.32 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with distraction was 0.56 standard deviations lower (0.78 to 0.33 lower).

2802
(30 RCTs)

⊕⊕⊝⊝
LOW a,b

This result equates to a moderate difference in favor of distraction

Self‐reported distress

See comment

The mean level of self‐reported distress with distraction was 0.82 standard deviations lower
(1.45 to 0.18 lower)

426
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a large difference in favor of distraction

Observer‐reported pain

See comment

The mean level of observer‐reported pain with distraction was 0.62 standard deviations lower
(1 to 0.23 lower)

1512
(11 RCTs)

⊕⊝⊝⊝
VERY LOW a,d

This result equates to a moderate to large difference in favor of distraction

Observer‐reported distress

See comment

The mean level of observer‐reported distress with distraction was 0.72 standard deviations lower
(1.41 to 0.03 lower)

1067
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,d

This result equates to a moderate to large difference in favor of distraction

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with distraction was 0.33 standard deviations lower
(0.69 lower to 0.03 higher)

309
(4 RCTs)

⊕⊕⊝⊝
LOW a,c

There is no evidence of an effect of distraction

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with distraction was 0.44 standard deviations lower
(0.84 to 0.04 lower)

500
(7 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a small to moderate difference in favor of distraction

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

d Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.

Open in table viewer
Summary of findings 2. CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents

CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐18 years with mixed medical (acute or chronic illness) or generally healthy undergoing immunization, intravenous insertion, venipuncture, bone marrow aspiration, insulin injection, or dental local anesthetic
Setting: hospital (inpatient/outpatient/emergency department), community clinic, or school
Intervention: CBT‐combined
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with CBT‐combined

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 0.84 to 8.4 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with combined CBT was 0.27 standard deviations lower
(0.58 lower to 0.03 higher)

1359
(14 RCTs)

⊕⊕⊝⊝
LOW a,b

There is no evidence of an effect of combined CBT

Self‐reported distress

See comment

The mean level of self‐reported distress with combined CBT was 0.26 standard deviations lower
(0.56 lower to 0.04 higher)

234
(6 RCTs)

⊕⊕⊝⊝
LOW a,c

There is no evidence of an effect of combined CBT

Observer‐reported pain

See comment

The mean level of observer‐reported pain with combined CBT was 0.52 standard deviations lower
(0.73 to 0.30 lower)

385
(4 RCTs)

⊕⊕⊝⊝
LOW a,c

This result equates to a moderate difference in favor of combined CBT

Observer‐reported distress

See comment

The mean level of observer‐reported distress with combined CBT was 0.08 standard deviations higher
(0.34 lower to 0.50 higher)

765
(6 RCTs)

⊕⊕⊝⊝
LOW a,b

There is no evidence of an effect of combined CBT

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with combined CBT was 0.65 standard deviations lower
(2.36 lower to 1.06 higher)

95
(2 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,d

There is no evidence of an effect of combined CBT

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with combined CBT was 0.40 standard deviations lower
(0.67 to 0.14 lower)

1105
(11 RCTs)

⊕⊕⊝⊝
LOW a,b

This result equates to a small to moderate difference in favor of combined CBT

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.
dDowngraded twice for imprecision: analysis based on < 100 participants per group.

Open in table viewer
Summary of findings 3. Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents

Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐16 years with chronic illness (cancer) or generally healthy undergoing bone marrow aspirations, lumbar punctures, venipuncture, or local dental anesthetic
Setting: hospital (inpatient/outpatient), community clinic
Intervention: hypnosis
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with hypnosis

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 4.17 to 8.6 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with hypnosis was 1.40 standard deviations lower
(2.32 to 0.48 lower)

176
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c,d

This result equates to a large difference in favor of hypnosis

Self‐reported distress

See comment

The mean level of self‐reported distress with hypnosis was 2.53 standard deviations lower
(3.93 to 1.12 lower)

176
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,c,d,e

This result equates to a large difference in favor of hypnosis

Observer‐reported pain

See comment

See comment

See comment

.‐

This outcome was not assessed in any study

Observer‐reported distress

See comment

See comment.

36
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with hypnosis was 0.38 standard deviations lower
(1.57 lower to 0.81 higher)

69
(2 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of hypnosis

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with hypnosis was 1.15 standard deviations lower
(1.76 to 0.53 lower)

193
(6 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c,d

This result equates to a large difference in favor of hypnosis

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded twice for imprecision: analysis based on < 100 participants per group.
d Downgraded once for possibility of publication bias given that almost all trials are from one expert group.
e Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.

Open in table viewer
Summary of findings 4. Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents

Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐12 years with mixed medical (acute or chronic illness) or unclear diagnoses undergoing venipuncture or intravenous insertion
Setting: hospital (outpatient/emergency department) or community clinic
Intervention: preparation/information
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with preparation/information

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 2.6 to 6.12 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with preparation/information was 0.18 standard deviations lower
(0.60 lower to 0.23 higher)

313
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of preparation/information

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

The mean level of observer‐reported pain with preparation/information was 0.40 standard deviations lower
(0.98 lower to 0.18 higher)

259
(3 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of preparation/information

Observer‐reported distress

See comment

See comment

100
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

39
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ distress

See comment

See comment

54
(1 RCT)

‐

This outcome was assessed in one study only

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

Open in table viewer
Summary of findings 5. Breathing compared to control for needle‐related procedural pain and distress in children and adolescents

Breathing compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 6‐15 years who are generally healthy, undergoing surgery, or have leukemia undergoing venipuncture or lumbar puncture
Setting: hospital (outpatient/inpatient)
Intervention: breathing
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with breathing

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 4 to 7.6 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with breathing was 1.04 standard deviations lower
(1.86 to 0.22 lower)

298
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a large difference in favor of breathing

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

See comment

120
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

120
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

‐

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

Open in table viewer
Summary of findings 6. Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents

Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐17 years who are generally healthy or have chronic illness undergoing intramuscular injection, immunization, or venipuncture
Setting: hospital (outpatient) or community clinics
Intervention: suggestion
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with suggestion

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 1.83 to 9.33 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with suggestion was 0.13 standard deviations lower
(0.40 lower to 0.15 higher)

218
(3 RCTs)

⊕⊝⊝⊝
VERY LOW a,b

There is no evidence of an effect of suggestion

Self‐reported distress

See comment

See comment

78
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported pain

See comment

See comment

78
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

20
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded twice for imprecision: analysis based on < 100 participants per group.

Open in table viewer
Summary of findings 7. Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents

Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐18 years with chronic illness (leukemia) undergoing lumbar puncture
Setting: hospital (inpatient)
Intervention: memory alteration
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with memory alteration

Self‐reported pain

See comment

See comment

24
(1 RCT)

‐

This outcome was assessed in one study only

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

See comment

42
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

50
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

50
(1 RCT)

‐

This outcome was assessed in one study only

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

Background

This review is an update of two previous iterations published in the Cochrane Library (2006, Issue 4 and 2013, Issue 10).

Description of the condition

Pain and distress due to medical procedures are common during childhood. Needles are routinely given from the first year of life, particularly for vaccine injections. Current recommendations state that healthy children receive 20 to 30 immunizations before the age of 18 years (CDC 2018; NACI 2018; WHO 2018). Among children with acute or chronic illness, needle procedures are even more frequent for the assessment and management of their conditions (Stevens 2011; Stevens 2012), and are reported as the most distressing part of treatment (Ljungman 1999). Unfortunately, pain and distress associated with medical procedures are often poorly managed in routine care (Berberich 2012; Stevens 2011; Taddio 2009). As well as negatively impacting the child, significant child pain and distress during needle procedures are reported as highly distressing and challenging for parents and healthcare providers (Kennedy 2008).

Failure to adequately manage pain and distress during needle procedures can lead to the development of significant needle fears, which often begin in early to middle childhood and persist into adulthood (McMurtry 2015b). Moreover, fear of needles contributes to vaccine hesitancy (Taddio 2012), and medical non‐adherence (Pate 1996). Thus, needle pain and distress are critical and timely to address, given the growing concern for increasing outbreaks of preventable and infectious diseases and the potential loss of herd immunity (Smith 2014).

In recent years, there has been increasing recognition of the need to adequately manage needle‐related pain and distress. Evidence‐based clinical practice guidelines have been developed for the management of pain and fear during vaccine injections, and include recommended pharmacological, physical, and psychological strategies (McMurtry 2016; Taddio 2015). For psychological interventions, existing guidelines recommend using a variety of cognitive and behavioral interventions that have been deemed efficacious in reducing pain during needle procedures (e.g. blowing bubbles, distraction). Moreover, guidelines recommend not using strategies that have been deemed ineffective in reducing pain (e.g. making reassuring statements like “don’t worry”). Many of these strategies have been recommended by the World Health Organization’s Strategic Advisory Group of Experts for the management of immunizations worldwide (WHO 2015). Additional efforts include hospital certifications (ChildKind International; www.childkindinternational.org; Schechter 2010a) and hospital‐wide policies (Schechter 2008), as well as recommended standards of care for the management of medical procedures, including needles, for youth with cancer (Flowers 2015), and social media efforts targeting parents (#itdoesnthavetohurt; www.itdoesnthavetohurt.ca).

Description of the intervention

Consistent with previous iterations of this review (Uman 2006; Uman 2013), we include only non‐pharmacological psychological interventions for pain that are cognitive‐behavioral in this update. We do not include non‐pharmacological physical interventions such as acupuncture, heat, or cold.

Cognitive interventions include techniques that target negative or unrealistic thoughts to help replace them with more positive beliefs and attitudes. Behavioral interventions include techniques that target negative or maladaptive behaviors to help replace them with more positive and adaptive behaviors. Cognitive behavioral therapy (CBT) uses a combination or variation of strategies targeting cognitions (thoughts) or behaviors, or both (Barlow 1999). CBT for pain management aims to help individuals develop and use coping skills to manage their pain and distress, and often includes a combination of various techniques, such as distraction, relaxation training, deep breathing, hypnosis, preparing for and rehearsing the procedure in advance, using positive reinforcement for adaptive behaviors, making positive coping statements, and receiving coaching to use adaptive strategies (Chen 2000a; Christophersen 2001; Keefe 1992). Engagement and developmental appropriateness are thought to contribute to the efficacy of each intervention (Birnie 2017).

Many of the psychological interventions described in this review do not require highly specialized training and can be administered by non‐psychologists, such as other healthcare providers (e.g. child life specialists, nurses) and parents. However, some trials describe brief training and education in order to maximize the effectiveness of interventions. Furthermore, psychological interventions often have natural appeal, given their tendency to draw on children’s coping tendencies, potential use of minimal or widely available resources, and feasible implementation across clinical care settings.

How the intervention might work

There are several proposed mechanisms through which psychological interventions might work. Cognitive interventions target thoughts whereas behavioral interventions target individual behaviors. CBT uses a combination or variation of both cognitive or behavioral strategies, or both (Barlow 1999). Cognitive, behavioral, and combined cognitive‐behavioral strategies are believed to influence pain and distress through cognitive (e.g. attention, motivation, expectations, suggestibility), learning processes, physiological, or neurobiological or both mechanisms (Accardi 2009; Birnie 2017; Jafari 2017; Noel 2018). As in our previous reviews, all of the cognitive, behavioral, and cognitive‐behavioral strategies described above fall under the overarching category of ‘psychological’ interventions. Psychological interventions for pain management aim to help individuals to develop and use coping skills to manage their pain and distress, and can include various techniques such as distraction, relaxation training, deep breathing, hypnosis, preparing for and rehearsing the procedure in advance, using positive reinforcement for adaptive behaviors, making positive coping statements, and receiving coaching to use adaptive strategies (Chen 2000a; Christophersen 2001; Keefe 1992).

For this second review update, we revised the previous intervention categories that focused on methodological similarities to now reflect theorized or proposed mechanisms of treatment effect versus methodological similarities (e.g. virtual reality interventions are now encompassed within distraction interventions) (Accardi 2009; Birnie 2017; Jafari 2017; Noel 2018). We hoped this would lead to more meaningful conclusions and would also minimize the number of intervention categories that had small numbers of trials.

Why it is important to do this review

Previous narrative, non‐systematic reviews and book chapters on this topic have been published (Alvarez 1997; Blount 2003; Chen 2000a; Christophersen 2001; Kazak 2001; Powers 1999; Young 2005); however, more rigorous systematic reviews examining a variety of psychological interventions for needle pain and distress are essential to draw firmer conclusions about intervention efficacy and guide clinical decision‐making.

Our original Cochrane Review (Uman 2006; Uman 2008) included 28 RCTs, supported the efficacy of several interventions (distraction, hypnosis, and CBT), and led to recommendations about improving the quality of trials in this area (Uman 2010). Our first review update (Uman 2013) expanded this original review, and included 39 RCTs (18 new) and coded each trial for risks of bias. We found evidence for the efficacy of distraction and hypnosis. Overall, we rated the risks of bias of trials as high or unclear, suggesting the need for improvements in methodological rigor and reporting.

In the five years since the publication of our first update, several new trials have been published, examining psychological interventions for needle‐related pain and distress in children and adolescents. Advances and expansion of technology have been made (e.g. humanoid robots, smartphone apps), and peer‐reviewed journals have continued to raise standards for trial reporting and quality. The aim of this second review update was therefore to identify new trials, to synthesize the results of new trials with those previously reviewed, and to extend assessment of trial quality. This enables us to make firmer, more refined conclusions about the existing evidence on the efficacy of these interventions and to strengthen the evidence base for future research and clinical practice.

Objectives

To assess the efficacy of psychological interventions for needle‐related procedural pain and distress in children and adolescents.

Methods

Criteria for considering studies for this review

Types of studies

We included published randomized controlled trials (RCTs) with at least five participants in each study arm. The original version of this review (Uman 2006) included quasi‐randomized trials (e.g. alternating assignment) and unpublished trials (e.g. dissertations). We excluded these from the first review update (Uman 2013) and from this current update, to focus only on the highest‐quality evidence available. We applied no language restrictions during the search, and obtained translations when necessary.

Types of participants

We included RCTs involving children and adolescents aged two to 19 years, undergoing any needle‐related medical procedure. Participants included healthy children and children with chronic or transitory illnesses from both inpatient and outpatient settings. We excluded children under two years old, due to developmental differences either precluding the appropriateness of reviewed psychological interventions in infancy or the qualitatively different application of these interventions in that age group. Furthermore, the efficacy of psychological interventions for procedural pain and distress in infants is thoroughly addressed in another review (Pillai‐Riddell 2011; Pillai‐Riddell 2015). We selected a maximum age of 19 years to be consistent with the World Health Organization (WHO) definition of adolescence that extends to 19 years of age (http://www.searo.who.int/ en/Section13/Section1245_4980.htm). The efficacy of psychological interventions for needle‐related procedural pain in adults is reviewed elsewhere (Boerner 2015). We preferred a broad age range to minimize exclusion of any relevant studies. We excluded studies including any participants beyond two‐ to 19‐year olds unless we could obtain data from authors for the relevant eligible subsample, while also continuing to meet the minimum criterion of five participants per group.

This review focuses on needle‐related procedures performed for medical purposes, which are the most commonly occurring and feared procedures for both healthy and chronically‐ill children (Broome 1990; Ljungman 1999; McMurtry 2015a). A list of common included needle‐related procedures and their definitions can be found in Table 1. We excluded needle procedures performed for non‐medical purposes (e.g. body piercings or tattoos).

Open in table viewer
Table 1. Definitions of medical procedures

Procedure

Definition

Accessing a portacath (also known as a port)

Insertion of a needle into an implanted access device (portacath) which facilitates the drawing of blood and intravenous (or intra‐arterial) injections by not having to locate and insert a cannula into a new vessel. Some ports are connected for intrathecal, intraperitoneal or intracavitary injections.

Arterial blood gas (ABG)

A test which analyses arterial blood for oxygen, carbon dioxide and bicarbonate content in addition to blood pH. Used to test the effectiveness of respiration.

Arterial line (also known as intra‐arterial catheter)

Insertion of a catheter into an artery.

Arterial puncture

A hole, wound, or perforation of an artery made by puncturing.

Bone marrow aspiration (BMA)

The bone marrow is the tissue that manufactures the blood cells and is in the hollow part of most bones. This test is done by suctioning some of the bone marrow for examination.

Bone marrow biopsy (BMB)

The removal and examination of tissue, cells, or fluids from the bone marrow of a living body; usually performed at the same time as a BMA.

Central line (also known as central venous catheter)

Insertion of a catheter into the large vein above the heart, usually the subclavian vein, through which access to the blood stream can be made. This allows drugs and blood products to be given and blood samples withdrawn.

Finger prick/pin

Obtaining blood by puncturing the tip of the finger.

Immunization (also known as immunization)

Protection against a particular disease or treatment of an organism by protecting against certain pathogen attacks; the introduction of microorganisms that have previously been treated to make them harmless.

Injection

The act of forcing a liquid into tissue, the vascular tree, or an organ.

Intramuscular injection

Injection administered by entering a muscle.

IV/catheter insertion

A narrow short, flexible, synthetic (usually plastic) tube known as a catheter, that is inserted approximately one inch into a vein to provide temporary intravenous access for the administration of fluid, medication, or nutrients.

Lumbar punctures (LP) (also known as spinal tap)

The withdrawal of cerebrospinal fluid or the injection of anesthesia by puncturing the subarachnoid space located in the lumbar region of the spinal cord.

Paracentesis

A surgical puncture of a bodily cavity (e.g. abdomen) with a trocar, aspirator, or other instrument usually to draw off an abnormal effusion for diagnostic or therapeutic purposes.

Subcutaneous injection

Injection administered under the skin.

Suture (also known as laceration repair)

A stitch made with a strand or fiber used to sew parts of the living body.

Thoracocentesis (also called thoracentesis)

Aspiration of fluid from the chest.

Venepuncture (also known as venipuncture)

The surgical puncture of a vein typically for withdrawing blood or administering intravenous medication.

We also excluded studies if they specifically included participants with known needle phobias (i.e. diagnosed by a qualified professional such as a psychologist and warranting specific clinical assessment, diagnoses, and targeted intervention). Other reviews address specific evidenced‐based psychological interventions for high levels of needle fears/phobias that are different from those indicated for procedural pain management (e.g. exposure‐based treatment) (McMurtry 2015a; McMurtry 2016).

We excluded children undergoing surgery, given the numerous factors specific to surgery or intensive care units that complicate or interfere with self‐report of pain and distress (e.g. sedation, intubation, more intensive pharmacological interventions, long‐term hospitalization, inability or difficulty attributing pain or distress to a specific medical procedure) (Puntillo 2004). We made an exception for studies evaluating a psychological intervention for a pre‐surgical needle procedure (e.g. intravenous insertion) only when outcomes of interest were completed prior to surgery or sedation.

Types of interventions

Studies had to include at least one trial arm that assessed a primarily psychological intervention. Trials had to include at least one comparator arm (i.e. no treatment, other active treatment, treatment as usual, or waitlist). We placed no restrictions on duration, intensity, or frequency of psychological interventions. Interventions had to occur at some point prior to the needle procedure and the assessment of outcomes of interest. We excluded studies in which psychological intervention(s) were combined with a non‐psychological intervention (e.g. pharmacological, physical), so that the unique effects of the psychological intervention could not be isolated and evaluated.

As specified in the review protocol and original version of this review (Uman 2005; Uman 2006), psychological interventions were broadly defined as those using cognitive, behavioral, or combined cognitive‐behavioral strategies. In brief, cognitive interventions are those primarily targeting thoughts and feelings, whereas behavioral interventions are those primarily targeting overt behaviors (Barlow 1999). Combined cognitive‐behavioral interventions are defined as those including at least one cognitive strategy combined with at least one behavioral strategy. For this update, intervention categories were generally based on key theorized mechanisms of effect or specific distinct strategies, or both (Accardi 2009; Birnie 2017; Jafari 2017; Noel 2018), with similar interventions grouped together based on conceptualizations or intervention descriptions or both, included in published papers. Emphasizing mechanisms of effect in defining intervention categories resulted in some intervention categories from previous iterations of this review being subsumed under other broader categories. We did this to allow more meaningful meta‐analyses, and to avoid intervention categories with very small numbers of studies or single trials only. Specifically, we now include virtual reality interventions as one type of distraction, given their conceptualized application in this context for pain reduction (Kenney 2016). We now include the following interventions from the previous review as combined CBT, as they include at least one cognitive and one behavioral strategy as described in the original review protocol (Uman 2005): parenting coaching plus child distraction, parent positioning plus child distraction, and distraction plus suggestion. As stated in our original review protocol (Uman 2005), the division of psychological interventions into mutually exclusive categories is difficult, given a lack of consistent operational definitions. We feel our emphasis on treatment mechanisms in this second update reflects emerging empirical evidence and contemporary thinking in our understanding of psychological interventions for acute pain and distress.

Studies in this second update fall under one of the following psychological intervention categories:

  • Distraction;

  • Combined CBT;

  • Hypnosis;

  • Preparation/information;

  • Breathing;

  • Suggestion;

  • Memory alteration.

Types of outcome measures

Primary outcomes

Primary outcomes of interest were pain intensity and distress. These are two core outcome domains recommended for clinical trials in pediatric acute pain (PedIMMPACT; McGrath 2008). Distress is broadly defined as any type of negative affect associated with the needle procedure (e.g. anxiety, fear, stress). We extracted pain and distress outcomes separately, as assessed by child self‐report, observer global report (e.g. parents, nurses, researchers, etc. report using single‐item scales), and/or behavioral measurement (e.g. validated rating scales assessing observed pain or distress behaviors or both, displayed by the child).

Secondary outcomes

Secondary outcomes include any physiological measurement that has been associated with pain and distress and that it is practical to quantify in a clinical setting. Examples include heart rate, respiratory rate, blood pressure, oxygen saturation, cortisol levels, transcutaneous oxygen tension (tcPO2), and transcutaneous carbon dioxide tension (tcPCO2) (Jafari 2017; Sweet 1998). We also assessed adverse events.

Timing of outcome assessment

Where possible, we extracted outcomes assessed during the needle procedure. If outcomes during the needle procedure were not evaluated, we selected the next time point occurring closest to the completion of the procedure. If outcomes were assessed both during and following the needle‐related procedure, we included only outcomes assessed during the needle procedure. We did not include outcomes assessed at other times (e.g. pre‐needle outcomes).

Search methods for identification of studies

We identified published studies through electronic database searches, postings to various electronic listservs, and clinical trial registries.

Electronic searches

We developed detailed search strategies through consultation with a reference librarian and assistance from the Cochrane Pain, Palliative and Supportive Care (PaPaS) Group. Definitions for included medical procedures (MedLine 2004) are described in Table 1.

We searched the following six electronic databases for relevant trials:

  • Cochrane Central Register of Controlled Trials (CENTRAL), the Cochrane Library Issue 8 of 12, 2017;

  • MEDLINE and MEDLINE in Process (OVID), March 2013 to 12 September 2017;

  • Embase (OVID), March 2013 to 2016, week 37;

  • PsycINFO (OVID), 2013 to September week 1, 2017;

  • Web of Science (ISI Web of Knowledge), 2013 to 12 September 2017;

  • Cumulative Index to Nursing and Allied Health Literature (CINAHL), March 2013 to September 2017.

Database search terms were consistent with previous versions of this review (Uman 2006; Uman 2013). We conducted updated searches in September 2016 and September 2017 to identify any records published since the last review update in 2013. See appendices for search strategies, keywords, and MeSH terms as appropriate for each database: MEDLINE (Appendix 1), PsycINFO (Appendix 2), CENTRAL (Appendix 3), Embase (Appendix 4), IBI Web of Knowledge (Appendix 5), and CINAHL (Appendix 6).

Searching other resources

We also solicited relevant studies through professional listservs, including:

  1. Pain in Child Health (PICH);

  2. Pediatric Pain;

  3. American Psychological Association’s Society of Pediatric Psychology Division 54;

  4. American Psychological Association’s Health Psychology Division 38.

For this update, we also searched clinical trial registries for any relevant completed trials, including clinicaltrials.gov and the World Health Organization International Clinical Trials Registry Platform (www.who.int.trialsearch). We also included any other relevant studies identified and included in the original review and the previous update.

Data collection and analysis

Selection of studies

Two review authors considered titles and abstracts retrieved from database searches for review inclusion (LU and CC for original review; KB and MN for first and second review updates). Two review authors checked full‐text articles when relevance and eligibility for the current review were unclear from the abstract alone (LU and CC for original review; KB and MN for review updates). We resolved discrepancies through discussion with a third review author.

Included studies had to use true random assignment. We determined this based on the description of participant assignment available in each study’s peer‐reviewed publication. We retrieved and included 28 RCTs in the original review (Uman 2006), although we later excluded seven of these studies from subsequent review updates, including this one, as they were unpublished dissertations or reported quasi‐randomized methods (e.g. alternating assignment). Updated database searches (conducted March 2012 and March 2013) for the last review update (Uman 2013) identified an additional 18 RCTs meeting our inclusion criteria and providing necessary data. Searches for the current (second) update (conducted September 2016 and September 2017) identified an additional 20 RCTs, for a total of 59 RCTs included in this review. We coded all included RCTs in full. References for included studies are provided below in the ’Description of studies’ section.

Data extraction and management

Two review authors extracted data (LU and CC original review; KB and MN for review updates), using a data extraction form designed for the original review. A researcher outside the review team who was fluent in Farsi reviewed one non‐English study in the previous review update, to confirm inclusion eligibility and conduct data extraction. Extracted data included study design, participant demographics, diagnosis (when applicable), type of needle procedure, type of intervention and control conditions, outcomes, as well as other related variables. A third review author was available to resolve coding discrepancies, if needed. If studies reported incomplete data necessary for meta‐analysis, we contacted study authors. We excluded RCTs if the data necessary for data pooling were not available in the published study, could not be identified through contact with the study authors, or could not be calculated based on other data provided. A trained research assistant or another study author, or both, reviewed the extracted data for errors. We analyzed all data suitable for pooling using Review Manager 5 software (RevMan) (RevMan 2014).

Assessment of risk of bias in included studies

Two review authors (KB and MN) independently assessed risks of bias for all included studies, using the criteria outlined in the Cochrane Handbook (Higgins 2017), with any disagreements resolved by discussion. We completed a 'Risk of bias' table for each included study using the 'Risk of bias' tool in RevMan.

We assessed the following for each study.

  • Random sequence generation (checking for possible selection bias). We assessed the method used to generate the allocation sequence as: low risk of bias (any truly random process, e.g. random‐number table; computer random‐number generator); unclear risk of bias (method used to generate sequence not clearly stated). We excluded studies using a non‐random process (e.g. odd or even date of birth; hospital or clinic record number).

  • Allocation concealment (checking for possible selection bias). The method used to conceal allocation to interventions prior to assignment determines whether intervention allocation could have been foreseen in advance of or during recruitment, or changed after assignment. We assessed the methods as: low risk of bias (e.g. telephone or central randomization; consecutively‐numbered sealed opaque envelopes); unclear risk of bias (method not clearly stated). We considered studies that did not conceal allocation (e.g. open list) to have high risk of bias.

  • Blinding of participants and personnel (checking for possible performance bias). We assessed the methods used to blind study participants and personnel from knowledge of which intervention a participant received. We assessed methods as: low risk of bias (study states that it was blinded and describes the method used to achieve blinding); unclear risk of bias (study states that it was blinded but does not provide an adequate description of how this was achieved). We considered studies that were not blinded or when the nature of the psychological intervention precluded participants and personnel from being blinded (e.g. obvious intervention such as watching television or a medical clown in the room) to have high risk of bias.

  • Blinding of outcome assessment (checking for possible detection bias). We assessed the methods used to blind study participants and outcome assessors from knowledge of which intervention a participant received. We assessed the methods as: low risk of bias (study has a clear statement that outcome assessors were unaware of treatment allocation, and ideally describes how this was achieved); unclear risk of bias (study states that outcome assessors were blind to treatment allocation but lacks a clear statement on how this was achieved). We considered studies where outcome assessment was not blinded or when the nature of the psychological intervention precluded outcome assessors from being blinded to have a high risk of bias.

  • Incomplete outcome data (checking for possible attrition bias due to the amount, nature and handling of incomplete outcome data). We assessed the methods used to deal with incomplete data as: low risk (no missing data, reasons for missing data unlikely to be related to true outcome or balanced with similar reasons across groups); unclear risk of bias (insufficient information to permit judgment of risk); high risk of bias (reasons for missing data judged likely to be related to true outcome, with either imbalance in numbers or reasons across groups).

  • Selective reporting (checking for reporting bias). We assessed whether primary and secondary outcome measures were prespecified and whether these were consistent with those reported: low risk of bias (study protocol is available, or is not available but it is clear that the report specified and reported on all expected outcomes); unclear risk of bias (insufficient information to permit judgment of risk); high risk of bias (study did not prespecify and/or report all primary outcomes, one or more outcomes is reported incompletely so that it cannot be entered in meta‐analysis).

  • Other bias (checking for possible biases not covered elsewhere). We assessed other bias in studies as: low risk of bias (study appears to be free of other sources of bias); unclear risk of bias (insufficient information to permit judgment of risk); high risk of bias (at least one important risk of bias likely to impact study findings, such as baseline group differences reported and not accounted for, using unvalidated/unreliable measurement tool, inadequate sample size/study underpowered).

Measures of treatment effect

Given the nature of the outcome measures in this review, all outcome data were continuous (e.g. rating scales). We calculated standardized mean differences (SMDs) with 95% confidence intervals (CIs), which allowed the combination of results across different measurement scales assessing the same outcome (e.g. pain). We applied the following rule of thumb for interpreting SMDs as effect sizes, as suggested by Cochrane (Higgins 2017): 0.2 represents a small effect, 0.5 represents a medium effect, and 0.8 represents a large effect (Cohen 1988). We assessed each category of psychological intervention separately in a meta‐analysis. Within each intervention category, we assessed outcomes and measurement type separately. We pooled all comparators together. We only conducted meta‐analysis when data from more than a single RCT were available. Thus, for each psychological intervention we assessed possible treatment effects for the following seven outcomes:

  • Pain: self‐report;

  • Pain: observer global report;

  • Pain: behavioral measure;

  • Distress: self‐report;

  • Distress: observer global report;

  • Distress: behavioral measure;

  • Physiological measures: each physiological outcome was assessed separately.

Unit of analysis issues

We included parallel two‐group RCTs as well as cluster‐randomized trials (i.e. groups of individuals randomized together to the same intervention). We included cross‐over trials only when data were available separately for each group following the first treatment arm (i.e. prior to cross‐over). We did this because once psychological interventions have been introduced, it can be difficult to prevent participants from using these strategies themselves at subsequent needle procedures (e.g. distraction). We included studies with multiple treatment groups so long as each treatment group separately met the review inclusion criteria.

Dealing with missing data

We tried to contact study authors in all situations when data necessary for data pooling were not reported in published RCTs (e.g. means, standard deviations (SDs), group sizes). If this was not possible, we used statistical methods for calculating missing data from other reported measures of variation as recommended (e.g. calculating standard deviations from standard errors, confidence intervals, t values, and P values) (Higgins 2017). We excluded studies or outcomes or both from this review when we could not contact the authors or they did not respond, did not have data available, or when we could not calculate data necessary for pooling from available data. We included the number of participants in each group identified in published study results sections. When not otherwise specified by the authors, we assumed there were no study dropouts and used the reported group sizes in the meta‐analyses.

Assessment of heterogeneity

We assessed heterogeneity using both the Chi2 test and the I2 statistic. Given that Chi2 tests often have low statistical power, we used a Type 1 error level of 0.10 for rejecting the null hypothesis of homogeneity. While Chi2 tests are useful for identifying whether heterogeneity is present, it has been argued that there will always be some level of heterogeneity in meta‐analyses, given the clinical and methodological diversity (Higgins 2017). The I2 statistic provides a measure of inconsistency across studies to assess the impact of heterogeneity on the meta‐analysis (Higgins 2017). I2 is expressed as a percentage from 0% to 100%, and we used the following rough interpretation guide: 0% to 40%, might not be important; 30% to 60%, may represent moderate heterogeneity; 50% to 90%, may represent substantial heterogeneity; and 75% to 100% represents considerable heterogeneity (Higgins 2017). Ranges overlap, as the importance of I2 depends on several other factors such as the magnitude and direction of effects, as well as the strength of evidence for the heterogeneity (for example, the P value for the Chi2 test or the confidence interval for the I2 statistic). In cases where we found statistically significant heterogeneity, data were still pooled but should be interpreted with caution. Given significant heterogeneity for several analyses, we analyzed results using a random‐effects model.

Assessment of reporting biases

We used several strategies to overcome publication, language, and outcome reporting bias in this update and previous iterations (Uman 2006; Uman 2013). We imposed no language barriers in database searches, we searched clinical trial registries, we posted requests to listservs in pediatric health and pain regarding any published, unpublished, or in‐progress studies, and tried to obtain any and all missing data from included studies through repeated email requests to study authors or co‐authors. We included any studies in the meta‐analysis that provided completed results (i.e. means, SDs, and cell sizes for both treatment and control groups) for at least one outcome measure of interest. Information related to reporting biases is also captured in the 'Risk of bias' tool. Fourteen of the studies included in this review had authors who responded to requests for missing data (Balan 2009; Bisignano 2006; Caprilli 2007; Cavender 2004; Crevatin 2016; Gupta 2006; Kleiber 2001; Liossi 1999; McCarthy 2010; Meiri 2016; Miguez‐Navarro 2016; Nilsson 2015; Sinha 2006; Sander Wint 2002).

Data synthesis

We calculated SMDs using a random‐effects model separately for all outcomes for each intervention category when necessary data were available. We considered interventions to be efficacious when the SMD and corresponding CIs were negative. The reported P values reflect the strength of the evidence against the null hypothesis.

We combined intervention groups that included variations of the same psychological intervention category (e.g. two types of distraction) to create a single pair‐wise comparison, as recommended by Cochrane (Higgins 2017). When multiple control conditions were available, we selected the condition that could most clearly isolate the active ingredient of the intervention condition. For example, comparing eutectic mixture of local anesthetics (EMLA) + distraction (intervention group) to EMLA only (selected control group) instead of no‐EMLA standard care (not‐selected control group). Another example includes comparing music with headphones (intervention group) to headphones only without music (selected control group) instead of no headphones or music (not‐selected control group).

Many control conditions are defined as standard or routine care groups. Some include cognitive or behavioral techniques, or both. We made an a priori decision to consider these as control groups as conceptualized by the authors themselves, and it is ethical for the conduct of clinical trials not to offer below current standard of care. Less common are studies that report the use of pharmacological strategies, such as topical anesthetics, in standard care. In such cases, we also considered these as a control condition as long as it was offered similarly to the intervention condition, in addition to psychological strategies.

We combined outcomes in cases when multiple observers rated children’s pain or distress or both (e.g. nurses, parents, researchers) or when multiple behavioral measures assessed pain or distress or both (e.g. both the child‐adult medical procedure inventory scale (CAMPIS) and observation scale of behavioral distress (OSBD) for distress). We pooled data using statistical formulae recommended by Cochrane for combining means and SDs: pooled mean = [(mean1 x N1) + (mean2 x N2) / (N1 + N2)] and pooled SD = square root of [SD12 (N1−1)+SD22 (N2−1)]/N1+N2−2.

Quality of the evidence

Two review authors (KB and MN) independently rated the quality of the outcomes. We used the GRADE system as applied to continuous outcomes (Guyatt 2013) to rate the quality of evidence separately for all intervention categories and all outcomes with data from more than one RCT. We used the GRADE profiler Guideline Development Tool software (GRADEpro GDT 2015), GRADE recommendations (Guyatt 2011) and the guidelines provided by Cochrane (Higgins 2017).

The GRADE system uses the following criteria for assigning a quality level to a body of evidence (Higgins 2017):

  • High: randomized trials; or double‐upgraded observational studies

  • Moderate: downgraded randomized trials; or upgraded observational studies

  • Low: double‐downgraded randomized trials; or observational studies

  • Very low: triple‐downgraded randomized trials; or downgraded observational studies; or case series/case reports

Factors that may decrease the quality level of a body of evidence are:

  • Limitations in the design and implementation of available studies suggesting a high likelihood of bias;

  • Indirectness of evidence (indirect population, intervention, control, outcomes);

  • Unexplained heterogeneity or inconsistency of results (including problems with subgroup analyses);

  • Imprecision of results (wide confidence intervals);

  • High probability of publication bias.

Factors that may increase the quality level of a body of evidence are:

  • Large magnitude of effect;

  • All plausible confounding would reduce a demonstrated effect or suggest a spurious effect when results show no effect;

  • Dose‐response gradient.

We decreased the GRADE rating by one (−1) or two (−2) if we identified:

  • Serious (−1) or very serious (−2) limitation to study quality;

  • Important inconsistency: I2 statistic moderate > 45% (−1) or I2 statistic considerable > 90% (−2);

  • Some (−1) or major (−2) uncertainty about directness;

  • Imprecise or sparse data: sample size < 400 (−1) or sample size < 100) (−2);

  • High probability of reporting bias (−1).

For transparency, we documented all reasons for downgrading the GRADE quality of evidence rating. Decreases of 3 or more ratings dropped the GRADE quality levels to 'very low'.

GRADE quality levels are interpreted as follows:

  • High: we are very confident that the true effect lies close to that of the estimate of the effect;

  • Moderate: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of effect, but there is a possibility that it is substantially different;

  • Low: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect; and

  • Very low: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.

'Summary of findings' tables

We include seven 'Summary of findings' tables to present the main findings in a transparent and simple tabular format. In particular, we included key information about the quality of evidence, the magnitude of effect of the interventions examined, and the sum of available data on all primary outcomes:

  • Pain: self‐report;

  • Pain: observer global report;

  • Pain: behavioral measure;

  • Distress: self‐report;

  • Distress: observer global report;

  • Distress: behavioral measure

Subgroup analysis and investigation of heterogeneity

We assessed each category of psychological intervention separately (i.e. distraction, hypnosis, etc.) as consistent with previous versions of this review. For each intervention, we conducted analyses separately by type of outcome (pain and distress) and measurement (self‐report, observer‐report, behavioral, physiological). We assessed different physiological outcomes separately (e.g. heart rate versus blood pressure). As described above, we calculated the Chi2 test and I2 statistic for all outcomes to assess heterogeneity.

Sensitivity analysis

We were unable to conduct all of the sensitivity analyses that we had proposed in our original review, due to insufficient data reported within and across studies, as well as the small number of studies within each intervention category. The main sensitivity analyses conducted in the original review involved comparing the study results when quasi‐randomized trials were added to the analyses. However, in order to strengthen the methodological quality of the findings in subsequent updates, we have limited the included trials to only true RCTs. We have therefore not conducted additional sensitivity analyses.

Results

Description of studies

See Characteristics of included studies; Characteristics of excluded studies.

Results of the search

We conducted six electronic database searches in total: one for the original review (February 2005) (Uman 2006); three for the first review update (December 2010, March 2012, March 2013) (Uman 2013); and two for the current review update (September 2016, September 2017).

Database searches conducted in September 2017 for this review update identified 714 records, and our searches of other resources (i.e. professional listservs and trials registries) identified an additional three studies that appeared to meet the inclusion criteria. After duplicate records had been removed, there remained 572 unique abstracts for review. Of these, 511 records were deemed not eligible. We reviewed 61 manuscripts in full, of which 20 met our inclusion criteria and provided the data necessary for data pooling. We included 39 trials in the previous review update (Uman 2013). This incorporated 21 studies from the original review published prior to 2005 (Uman 2006), plus 18 additional studies published between 2005 and 2013. Although the original review included 28 studies (Uman 2006), we excluded seven of these (reported in eight publications) from subsequent review updates, including this one, due to lack of adequate randomization procedures (Cohen 1997; Cohen 1999; Cohen 2002; French 1994) and being unpublished dissertation theses (Krauss 1996; Posner 1998; Zabin 1982). We found no non‐English studies in the database searches for this review update. Previous review searches identified studies in Portuguese (Santos 2000), German (Hoffman 2011; Kammerbauer 2011), Italian (Bufalini 2009; Lessi 2011), and Farsi (Alavi 2005; Shahabi 2007; Vosoghi 2010) which were either translated or assessed and coded in full by a native language speaker. Thus, 59 studies meet the inclusion criteria for this review update. For a further description of our screening process, see the study PRISMA flow diagram (Figure 1).

Figure 1
Study flow diagram.

Study flow diagram.

Included studies

This review and meta‐analysis includes 59 studies (n = 5550 participants). Of these, 21 studies were identified in the original review (Blount 1992; Cassidy 2002; Cavender 2004; Chen 1999; Eland 1981; Fanurik 2000; Fowler‐Kerry 1987; Gonzalez 1993; Goodenough 1997; Harrison 1991; Katz 1987; Kleiber 2001; Kuttner 1987; Liossi 1999; Liossi 2003; Liossi 2006; Press 2003; Tak 2006; Tyc 1997; Vessey 1994; Sander Wint 2002), 18 studies from the first review update (Balan 2009; Bellieni 2006; Bisignano 2006; Caprilli 2007; Gold 2006; Gupta 2006; Huet 2011; Inal 2012; Jeffs 2007; Kristjansdottir 2010; Liossi 2009; McCarthy 2010; Nguyen 2010; Noguchi 2006; Sinha 2006; Vosoghi 2010; Wang 2008; Windich‐Biermeier 2007), and 20 studies from the most recent searches conducted in September 2017 (Aydin 2017; Beran 2013; Cohen 2015; Crevatin 2016; Ebrahimpour 2015; Kamath 2013; Luthy 2013; Meiri 2016; Miguez‐Navarro 2016; Miller 2016; Minute 2012; Mutlu 2015a; Nilsson 2015; Oliveira 2017; Pourmovahed 2013; Ramírez‐Carrasco 2017; Rimon 2016; Sahiner 2016; Yinger 2016; Zieger 2013). Included trials had two to six study arms. Two studies used cross‐over designs (Nilsson 2015; Oliveira 2017), and there were no cluster‐RCTs.

Of the 59 included studies, nine assessed multiple psychological interventions (Cohen 2015; Fowler‐Kerry 1987; Gupta 2006; Kuttner 1987; Liossi 1999; Miller 2016; Sahiner 2016; Tak 2006; Wang 2008). In one study (Mutlu 2015a), we deemed only one of two interventions eligible for inclusion in the review (i.e. balloon inflation). Four studies assessed multiple types of distraction interventions (Aydin 2017; Bellieni 2006; Miller 2016; Sahiner 2016). Assessed interventions included distraction (n = 32), followed by combined CBT (n = 18), hypnosis (n = 8), preparation and information (n = 4), breathing (n = 4), suggestion (n = 3), and memory alteration (n = 1).

Needle procedures varied, and included venipuncture or blood draw only (n = 20), immunization or injection (n = 11), intravenous insertion (n = 8), lumbar puncture (n = 6), intravenous cannulation or venipuncture (n = 4), bone marrow aspiration (n = 3), local dental anesthetic injection (n = 3), and intramuscular injection, laceration repair, allergy testing involving injection, and insulin injection (n = 1 each). Ages of participating children and adolescents ranged from two to 19 years old. Most studies (n = 34) focused exclusively on children in early childhood (two to five years old) or middle childhood (i.e. six to 12 years old). Only one study focused exclusively on adolescents (i.e. 13 to 15 years old). All remaining studies (n = 24) included children ranging from early childhood to late adolescence (up to 19 years olds). Most studies (n = 33) provided no specific health diagnoses for participating children and adolescents. The remaining studies included children with mixed chronic illness (n = 13), children with cancer (n = 12), or children with diabetes (n = 1). Trials were conducted in a variety of settings, including hospital inpatients, hospital outpatient clinics, emergency departments, community clinics, and schools.

See the ’Characteristics of included studies’ tables for more detail by study and the seven 'Summary of findings' tables for more details by type of intervention.

Excluded studies

Overall, across all three iterations of this review, we excluded 168 studies after reviewing full‐text articles. Of these 168 excluded studies, 51 were excluded from the original review (Uman 2006), 69 from the previous review update (Uman 2013), 41 from the current review update, plus an additional seven included in the original review that we excluded from subsequent review updates, due to studies lacking true randomization or being unpublished dissertation theses. Additionally, we excluded one intervention arm for an otherwise included study (Mutlu 2015b).

Primary reasons for exclusion were:

See ‘Characteristics of excluded studies’ table for reasons of exclusion.

Risk of bias in included studies

See Figure 2 and Figure 3 for a summary of ’Risk of bias’ assessments for all included studies. More detail on the ’Risk of bias’ judgments can be found in the 'Characteristics of included studies' section.

Figure 2
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation

We rated 33 studies (55.9%) as being at unclear risk of bias, as the process for sequence generation was not clearly reported. We rated the other 26 studies (44.1%) as being at low risk of bias, as they reported clear strategies for generating random sequences (e.g. computer‐generated random‐number table). We rated no studies at high risk of bias for this domain.

Allocation concealment

We rated 46 studies (78.0%) as being at unclear risk of bias, as they did not report any detail about allocation concealment strategies. We rated only three studies (5.1%) as being at low risk of bias for clearly reporting use of sequentially‐numbered, opaque, sealed envelopes. We rated the remaining 10 studies (16.9%) at high risk of bias for reporting open allocation strategies.

Blinding

We rated 57 studies (96.6%) at high risk of bias and two (3.4%) at unclear risk of bias for blinding of participants and personnel; we rated no studies at low risk of bias for this domain. This was largely due to the nature of psychological interventions that are often obvious to children and nurses, or they are involved in delivery of the intervention itself. We rated one study (Gonzalez 1993) at unclear risk of bias in circumstances where parents delivered the intervention, making it possible for the child or nurse administering the needle, or both, to be blind to study group, although that was unclearly reported. We rated a second study (Goodenough 1997) at unclear risk of bias, as the intervention was a minor alteration to wording (i.e. suggestion), unlikely to be detected by the child.

We rated 56 studies (94.9%) at high risk of bias, two studies (3.4%) at unclear risk, and one study (1.7%) at low risk for blinding of outcome assessment. Similarly, most studies at high risk of bias were attributable to the obvious nature of psychological interventions. The few instances with unclear or low risk of bias occurred when the psychological intervention was not overtly apparent or when outcome raters were unaware of the group assignment (e.g. blinded observational assessment).

Incomplete outcome data

We rated 48 studies (81.4%) at low risk of bias, seven studies (11.9%) at unclear risk, and four (6.8%) at high risk for incomplete reporting of outcome data. We gave low ratings in circumstances where all outcomes were reported in full and with sufficient detail to be included in meta‐analysis, or where missing data were balanced across groups, or likely not to be related to study outcomes. We gave high risk of bias ratings when reasons for missing data were likely related to study outcomes, or an imbalance in missing data or dropouts between groups. We gave unclear ratings where insufficient information was provided.

Selective reporting

We rated 36 studies (61.0%) at unclear risk of bias, 17 studies (28.8%) at high risk, and six studies (10.2%) at low risk for selective reporting of outcomes. Most studies were given unclear risk of bias ratings when primary and secondary outcomes of interest were not clearly outlined, making it difficult to determine whether study outcomes were fully reported. We gave high risk of bias ratings when one or more outcomes of interest were incompletely reported. We gave low risk of bias ratings in circumstances where primary and secondary outcomes were identified a priori and reported in full.

Other potential sources of bias

We rated 32 studies (54.2%) at high risk of bias, 15 studies (25.4%) at low risk, and 12 studies (20.3%) at unclear risk for other potential sources of bias. Common areas of concern contributing to high risk of bias ratings included studies with small sample sizes or that were underpowered to detect treatment effects, contamination of intervention strategies between groups, use of unreliable or unvalidated outcome measures, or significant group differences that were not controlled for in analyses (e.g. variable number of injections, parental presence).

Effects of interventions

See: Summary of findings 1 Distraction compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 2 CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 3 Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 4 Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 5 Breathing compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 6 Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents; Summary of findings 7 Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents

Distraction

Thirty‐two studies assessed the efficacy of distraction for reducing children’s needle‐related pain and distress, of which four studies evaluated two types of distraction each (Aydin 2017; Bellieni 2006; Miller 2016; Sahiner 2016). Distraction interventions were varied and included watching cartoons or a movie (n = 9) (Bellieni 2006; Cassidy 2002; Cohen 2015; Luthy 2013; Miguez‐Navarro 2016Oliveira 2017; Sahiner 2016; Tak 2006; Wang 2008), listening to music or a spoken story (n = 8) (Aydin 2017; Balan 2009; Caprilli 2007; Fowler‐Kerry 1987; Kristjansdottir 2010; Nguyen 2010; Noguchi 2006; Press 2003), interactive handheld computer or video games (n = 3) (Crevatin 2016; Miller 2016; Minute 2012), distraction cards (n = 3) (Aydin 2017; Inal 2012; Sahiner 2016), virtual reality (n = 2) (Gold 2006; Sander Wint 2002), playing with a toy (n = 2) (Vessey 1994; Vosoghi 2010), parent distraction (n = 2) (Bellieni 2006; Gonzalez 1993), medical clown (n = 1) (Meiri 2016), squeezing a rubber ball (n = 1) (Gupta 2006), or a combination or selection of various distractors such as toys, books, cartoons, games, or music (n = 5) (Aydin 2017; Fanurik 2000; Jeffs 2007; Kuttner 1987; Sinha 2006).

Of these 32 studies assessing distraction, needle procedures included venipuncture or blood draws only (n = 14) (Aydin 2017; Balan 2009; Bellieni 2006; Caprilli 2007; Crevatin 2016; Gupta 2006; Inal 2012; Miguez‐Navarro 2016; Oliveira 2017; Press 2003; Sahiner 2016; Tak 2006; Vessey 1994; Wang 2008), immunization or injection (n = 7) (Cassidy 2002; Cohen 2015; Fowler‐Kerry 1987; Gonzalez 1993; Luthy 2013; Kristjansdottir 2010; Noguchi 2006), intravenous insertion (n = 4) (Fanurik 2000; Gold 2006; Miller 2016; Vosoghi 2010), intravenous cannulation or venipuncture (n = 2) (Meiri 2016; Minute 2012), lumbar puncture (n = 2) (Nguyen 2010; Sander Wint 2002), laceration repair (n = 1) (Sinha 2006), allergy testing involving injection (n = 1) (Jeffs 2007), and bone marrow aspiration (n = 1) (Kuttner 1987). Across the 32 studies, distraction efficacy was assessed in two‐ to 19‐year‐olds. Twenty‐three studies included children 12 years old or younger, of which seven studies exclusively included children aged seven or younger (Cassidy 2002; Cohen 2015; Fowler‐Kerry 1987; Gonzalez 1993; Kuttner 1987; Noguchi 2006; Vosoghi 2010). Seven more studies included samples ranging from early childhood to late adolescence (up to 19 years old) (Caprilli 2007; Crevatin 2016; Fanurik 2000; Jeffs 2007; Press 2003; Sinha 2006; Sander Wint 2002). Only one study exclusively included adolescents (Kristjansdottir 2010).

Thirty studies including 2802 participants (intervention group = 1509) revealed a moderate effect of distraction for self‐reported pain: standardized mean difference (SMD) −0.56, 95% confidence interval (CI) −0.78 to −0.33, Z = 4.83, P < 0.001, I2 = 87% (Analysis 1.1; Figure 4). We found a large effect of distraction relative to control groups in meta‐analysis of four studies including 426 participants (intervention group = 214) for self‐reported distress: SMD −0.82, 95% CI −1.45 to −0.18, Z = 2.52, P = 0.01, I2 = 89% (Analysis 1.2). We also found a moderate effect of distraction in meta‐analysis of 11 studies including 1512 participants (intervention group = 921) for observer‐reported pain: SMD −0.62, 95% CI −1.00 to −0.23, Z = 3.14, P = 0.002, I2 = 91% (Analysis 1.3), and a moderate‐to‐large effect in meta‐analysis of five studies including 1067 participants (intervention group = 635) for observer‐reported distress: SMD −0.72, 95% CI −1.41 to −0.03, Z = 2.05, P = 0.04, I2 = 96% (Analysis 1.4). Analyses reported no evidence of effect for distraction relative to control groups for behavioral pain in four studies including 309 participants (intervention group = 164): SMD −0.33, 95% CI −0.69 to 0.03, Z = 1.80, P = 0.07, I2 = 57% (Analysis 1.5). Analyses revealed a moderate effect of distraction for behavioral distress in seven studies including 500 participants (intervention group = 261): SMD −0.44, 95% CI −0.84 to −0.04, Z = 2.16, P = 0.03, I2 = 75% (Analysis 1.6), and for the physiological measure of heart rate in three studies including 252 participants (intervention group = 126): SMD −0.64, 95% CI −0.89 to −0.38, Z = 4.93, P < 0.001, I2 = 0% (Analysis 1.7). Analysis of two studies including 112 participants (intervention group = 56) revealed a moderate effect of lowered oxygen saturation in control groups relative to distraction: SMD 0.60, 95% CI 0.22 to 0.98, Z = 3.10, P = 0.002, I2 = 0% (Analysis 1.8). Only one study (Nguyen 2010) assessed the efficacy of distraction for physiological outcomes (i.e. respiratory rate, systolic and diastolic blood pressure); we can therefore draw no conclusions about its efficacy for these outcomes. Sample size, means, and SDs for these outcomes are available in Table 2. No adverse events were reported.

Figure 4
Forest plot of comparison: 1 Distraction, outcome: 1.1 Self‐reported pain.

Forest plot of comparison: 1 Distraction, outcome: 1.1 Self‐reported pain.

Open in table viewer
Table 2. Means and standard deviations for outcomes from single trials

Intervention

Outcome

Study

Treatment

Control

N

Mean (SD)

N

Mean (SD

Distraction

Physiological measure ‐ Respiratory rate

Nguyen 2010

20

25.1 (3.6)

20

28.5 (3.86)

Physiological measure ‐ Systolic BP

Nguyen 2010

20

97.1 (8.57)

20

105.6 (15.97)

Physiological measure ‐ Diastolic BP

Nguyen 2010

20

65.2 (6.83)

20

69.8 (11.67)

Hypnosis

Observer‐reported distress

Katz 1987

17

3 (0.9)

19

3.3 (0.6)

Physiological measure ‐ Skin conductance

Ramírez‐Carrasco 2017

20

10.61 (6.42)

20

8.74 (4.63)

Physiological measure ‐ Heart rate

Ramírez‐Carrasco 2017

20

93.57 (12.33)

20

99.3 (18.56)

Preparation/Information

Observer‐reported distress

Harrison 1991

50

2.43 (1.62)

50

3.17 (1.3)

Behavioural measure ‐ Pain

Miller 2016

19

3.21 (3.46)

20

4.90 (3.67)

Behavioural measure ‐ Distress

Tak 2006

26

2.64 (1.1)

28

2.37 (1.12)

Physiological measure ‐ Pulse rate

Harrison 1991

50

84.6 (8.6)

50

88.6 (8.3)

Breathing

Observer‐reported pain

Sahiner 2016

60

2.56 (3.24)

60

3.80 (2.93)

Observer‐reported distress

Sahiner 2016

60

0.59 (1.03)

60

1.25 (1.41)

Suggestion

Self‐reported distress

Goodenough 1997

39

0.7 (1.1)

39

1.1 (1.3)

Observer‐reported pain

Goodenough 1997

39

0.9 (1.6)

39

1.7 (2.3)

Observer‐reported distress

Eland 1981

10

1.8 (0.71)

10

1.8 (0.79)

Memory Alteration

Self‐reported pain
(during procedure change score)

Chen 1999

15

−0.06 (3.9)

9

−0.02 (3.2)

Observer‐reported pain

(during procedure change score)

Chen 1999

20

04. (3.1)

22

−0.1 (1.8)

Observer‐reported distress

(during procedure change score)

Chen 1999

25

−0.2 (2.6)

25

−0.5 (1.9)

Behavioural measure ‐ Distress

(during procedure change score)

Chen 1999

25

−0.62 (3.7)

25

−0.48 (2.0)

Physiological measure ‐ Heart rate

(during procedure change score)

Chen 1999

24

0.1 (26.9)

20

−4.9 (21.8)

Physiological measure ‐ Cortisol

(during procedure change score)

Chen 1999

22

0.01 (0.18)

22

0.01 (0.2)

Physiological measure ‐ Systolic BP

(during procedure change score)

Chen 1999

23

−0.5 (11.8)

19

−5.4 (7.9)

Physiological measure ‐ Diastolic BP

(during procedure change score)

Chen 1999

23

−4.1 (10.9)

19

2.9 (10.2)
Quality of Evidence

The quality of evidence was low for self‐reported pain and behavioral measures of pain; further research is therefore likely to have an important impact on our confidence in the estimate of these effects, and is likely to change the estimates for these outcomes. The quality of evidence was very low for all other outcomes, including self‐reported distress, observer‐reported pain and distress, and behavioral measures of distress. Given this, we are very uncertain of the estimates of effects for these outcomes. See summary of findings Table 1. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias), inconsistency (analyses demonstrated moderate to considerable heterogeneity), and imprecision of results due to small numbers of participants.

Combined CBT

Eighteen studies assessed the efficacy of combined cognitive behavioral interventions for reducing children’s needle‐related pain and distress. The interventions involved different combinations of two or more cognitive and behavioral strategies, including distraction, preparation/information, modeling and rehearsal, breathing, suggestion, relaxation, guided imagery, positive coping statements, cognitive restructuring, positioning, and parent coaching. Interventions were delivered to the child by a music therapist, a medical clown, humanoid robot, electronic device, parent, or healthcare provider. Needle procedures included immunization (n = 6) (Beran 2013; Blount 1992; Cohen 2015; Fowler‐Kerry 1987; Nilsson 2015; Yinger 2016), intravenous insertion (n = 5) (Bisignano 2006; Kleiber 2001; McCarthy 2010; Miller 2016; Tyc 1997), venipuncture (n = 3) (Rimon 2016Wang 2008; Windich‐Biermeier 2007), intravenous insertion or venipuncture (n = 1) (Cavender 2004), bone marrow aspiration (n = 1) (Liossi 1999), insulin injection (n = 1) (Ebrahimpour 2015), or local dental anesthetic injection (n = 1) (Kamath 2013). Across the 18 studies, combined cognitive behavioral strategies were assessed in three‐ to 18‐year‐olds. Fourteen studies included children aged 12 or younger, of which five studies exclusively included children aged seven or younger (Blount 1992; Cohen 2015; Fowler‐Kerry 1987; Kleiber 2001; Yinger 2016). Four more studies included participants ranging from early childhood to late adolescence (up to 18 years old) (Liossi 1999; Rimon 2016; Tyc 1997; Windich‐Biermeier 2007).

Analysis of 14 studies examining combined cognitive‐behavioral strategies with 1359 participants (intervention group = 633) revealed no evidence of an effect for self‐reported pain: SMD−0.27, 95% CI −0.58 to 0.03, Z = 1.74, P = 0.08, I2 = 83% (Analysis 2.1; Figure 5). Six studies examining combined cognitive‐behavioral strategies for self‐reported distress with 234 participants (intervention group = 110) also showed no evidence of an effect: SMD −0.26, 95% CI −0.56 to 0.04, Z = 1.69, P = 0.09, I2 = 24% (Analysis 2.2). A moderate effect of combined cognitive‐behavioral strategies compared to control groups was shown for reduced observer‐reported pain across four studies including 385 participants (intervention group = 191) SMD −0.52, 95% CI −0.73 to −0.30, Z = 4.68, P < 0.001, I2 = 8% (Analysis 2.3), but no effect was shown across six studies including 765 participants (intervention group = 358) examining observer‐reported distress: SMD 0.08, 95% CI −0.34 to 0.50, Z = 0.39, P = 0.70, I2 = 78% (Analysis 2.4). Analysis of two studies including 95 participants (intervention group = 47) showed no evidence of an effect for behavioral pain: SMD −0.65, 95% CI −2.36 to 1.06, Z = 0.74, P = 0.46, I2 = 94% (Analysis 2.5); however, analysis of 11 studies including 1105 participants (intervention group = 529) revealed a moderate effect of combined cognitive‐behavioral strategies for behavioral distress: SMD −0.40, 95% CI −0.67 to −0.14, Z = 3.00, P = 0.003, I2 = 70% (Analysis 2.6). Analysis of three studies including 310 participants examining the physiological outcome of cortisol (intervention group = 150) showed no evidence of an effect: SMD −1.16, 95 CI −3.37 to 1.06, Z = 1.02, P = 0.31, I2 = 98% (Analysis 2.7). No adverse events were reported.

Figure 5
Forest plot of comparison: 2 CBT‐combined, outcome: 2.1 Self‐reported pain.

Forest plot of comparison: 2 CBT‐combined, outcome: 2.1 Self‐reported pain.

Quality of Evidence

The quality of evidence was low for outcomes of self‐reported pain and distress, observer‐reported pain and distress, and behavioral measures of distress. Further research is likely to have an important impact on our confidence in the estimate of these effects and is likely to change the estimates for these four outcomes. The quality of evidence was very low for behavioral measures of pain. We are therefore very uncertain of the estimate of effects for this outcome. See summary of findings Table 2. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias), inconsistency (analyses demonstrated moderate heterogeneity), and imprecision of results due to small numbers of participants.

Hypnosis

Eight studies assessed the efficacy of hypnosis interventions for reducing children’s needle‐related pain and distress. Interventions included both direct (i.e. analgesic) and indirect hypnotic suggestion, led by a therapist or taught to the child (i.e. self‐hypnosis). Needle procedures included bone marrow aspirations (n = 3) (Katz 1987; Kuttner 1987; Liossi 1999), lumbar puncture (n = 2) (Liossi 2003; Liossi 2006), local dental anesthetic injections (n = 2) (Huet 2011; Ramírez‐Carrasco 2017), or venipuncture (n = 1) (Liossi 2009). Across studies, hypnosis was assessed in two‐ to 16 year‐olds.

Five studies including 176 participants (intervention group = 97) revealed a large effect of hypnosis for self‐reported pain: SMD −1.40, 95% CI −2.32 to −0.48, Z = 2.97, P = 0.003, I2 = 85% (Analysis 3.1; Figure 6), and self‐reported distress: SMD −2.53, 95% CI −3.93 to −1.12, Z = 3.53, P < 0.001, I2 = 91% (Analysis 3.2). Two studies including 69 participants (intervention group = 34) were not significant for behavioral pain: SMD −0.38, 95% CI −1.57 to 0.81, Z = 0.62, P = 0.53, I2 = 83% (Analysis 3.3). Six studies including 193 participants (intervention group = 106) revealed a large effect of hypnosis for behavioral distress: SMD −1.15, 95% CI −1.76 to −0.53, Z = 3.66, P < 0.001, I2 = 71% (Analysis 3.4). Single trials only reported the effects of hypnosis for observer‐reported distress (Katz 1987) and physiological outcomes of skin conductance and heart rate variability (Ramírez‐Carrasco 2017). Sample size, means, and SDs for these outcomes are available in Table 2. No adverse events were reported.

Figure 6
Forest plot of comparison: 3 Hypnosis, outcome: 3.1 Self‐reported pain.

Forest plot of comparison: 3 Hypnosis, outcome: 3.1 Self‐reported pain.

Quality of Evidence

The quality of evidence was very low for all meta‐analyzed outcomes, including self‐reported pain, self‐reported distress, and behavioral measures of pain and distress. We are therefore very uncertain of the estimate of effects for these outcomes. See summary of findings Table 3. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias), inconsistency (analyses demonstrated moderate to considerable heterogeneity), imprecision of results due to small numbers of participants, and the possibility of publication bias with most studies from one expert group.

Preparation and information

Four studies assessed the efficacy of preparation/information interventions for reducing children’s needle‐related pain and distress. Preparatory information was provided through picture/photo storybook before venipuncture (Harrison 1991; Tak 2006; Zieger 2013) or through an interactive handheld electronic device before an intravenous cannulation (Miller 2016). All studies included children under 12 years old.

Analysis of four studies examining the effects of preparation/information for self‐reported pain including 313 participants (intervention group = 155) showed no evidence of an effect: SMD −0.18, 95% CI −0.60 to 0.23, Z = 0.86, P = 0.39, I2 = 68% (Analysis 4.1; Figure 7). Similarly, three studies examining observer‐reported pain including 259 participants (intervention group = 129) showed no evidence of an effect of preparation/information compared to control groups: SMD −0.40, 95% CI −0.98 to 0.18, Z = 1.35, P = 0.18, I2 = 80% (Analysis 4.2). Due to the availability of single trials only, we could reach no conclusions about the efficacy of preparation/information for observer‐reported distress (Harrison 1991), behavioral pain (Miller 2016), behavioral distress (Tak 2006), or physiological outcomes (pulse rate; Harrison 1991). No adverse events were reported.

Figure 7
Forest plot of comparison: 4 Preparation/Information, outcome: 4.1 Self‐reported pain.

Forest plot of comparison: 4 Preparation/Information, outcome: 4.1 Self‐reported pain.

Quality of Evidence

The quality of evidence was very low for all meta‐analyzed outcomes, including self‐reported and observer‐reported pain. We are therefore very uncertain about the estimate of effects for these outcomes. See summary of findings Table 4. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias), inconsistency (analyses demonstrated moderate heterogeneity), and imprecision of results due to small numbers of participants.

Breathing

Four studies assessed the efficacy of breathing interventions for reducing children’s needle‐related pain and distress. Interventions included asking children to inflate a balloon during venipunctures (n = 3) (Gupta 2006; Mutlu 2015a; Sahiner 2016) and deep breathing during lumbar puncture (n = 1) (Pourmovahed 2013). These studies included children aged from six to 15 years.

Four studies including 298 participants (intervention group = 149) revealed a large effect of breathing interventions for self‐reported pain: SMD −1.04, 95% CI −1.86 to −0.22, Z = 2.48, P = 0.01, I2 = 90% (Analysis 5.1; Figure 8). Only one study assessed the effects of breathing for observer‐reported pain and distress (Sahiner 2016); we could therefore draw no conclusions about treatment efficacy for these outcomes. Sample size, means, and SDs for these outcomes are available in Table 2. Adverse events were reported in one study (Pourmovahed 2013), where three children discontinued a deep‐breathing intervention due to respiratory difficulties.

Figure 8
Forest plot of comparison: 5 Breathing, outcome: 5.1 Self‐reported pain.

Forest plot of comparison: 5 Breathing, outcome: 5.1 Self‐reported pain.

Quality of Evidence

The quality of evidence was very low for the one meta‐analyzed outcome of self‐reported pain. We are very uncertain of the estimate of effects for this outcome. See summary of findings Table 5. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias), inconsistency (analyses demonstrated considerable heterogeneity), and imprecision of results due to small numbers of participants.

Suggestion

No new studies of suggestion were identified since the previous review update (Uman 2013); there is therefore no change in results for suggestion interventions for reducing children’s needle‐related pain and distress. Three studies assessed the efficacy of suggestion. Interventions included positive suggestions to the child that something was being done to make the needle easier or less painful. Studies were conducted mostly with children aged under seven during intramuscular injection (Eland 1981) or immunization (Fowler‐Kerry 1987). One study included three‐ to 17‐year‐olds undergoing venipunctures (Goodenough 1997).

Analysis of three studies showed no effect of suggestion for self‐reported pain including 218 participants (intervention group = 89): SMD −0.13, 95% CI −0.40 to 0.15, Z = 0.90, P = 0.37, I2 = 0% (Analysis 6.1; Figure 9). Only one study assessed the effects of suggestion for observer‐reported pain and self‐reported distress (Goodenough 1997), and only one study for observer‐reported distress (Eland 1981); we are therefore unable to draw any conclusions about treatment efficacy for these outcomes. Sample size, means, and SDs for these outcomes are available in Table 2. No adverse events were reported.

Figure 9
Forest plot of comparison: 6 Suggestion, outcome: 6.1 Self‐reported pain.

Forest plot of comparison: 6 Suggestion, outcome: 6.1 Self‐reported pain.

Quality of Evidence

The quality of evidence was very low for the one meta‐analyzed outcome of self‐reported pain. We are very uncertain of the estimate of effects for this outcome. See summary of findings Table 6. Primary reasons for downgrading the quality of the evidence were serious study limitations (most trials had unclear or high risk of bias) and imprecision of results due to small numbers of participants.

Memory alteration

No new studies were identified since the previous review update (Uman 2013); there is therefore no change in results for memory alteration interventions. Only one study assessed the effects of memory alteration in children aged three to 18, undergoing lumbar punctures (Chen 1999). This study included outcomes of self‐ and observer‐reported pain, observer‐reported and behavioral distress, and physiological measures (heart rate, cortisol, systolic and diastolic blood pressure). Given only this single study, we can draw no conclusions about treatment efficacy. Sample size, means, and SDs for these outcomes are available in Table 2. No adverse events were reported.

Quality of Evidence

Given this single study, we drew no conclusions about the quality of evidence in this area. See summary of findings Table 7.

Discussion

Summary of main results

This review synthesizes the results of 59 RCTs; 39 identified from previous versions of this review (Uman 2006; Uman 2013) and an additional 20 identified for this second review update. By including only truly randomized controlled trials published in peer‐reviewed journals, we offer a rigorous systematic examination of the efficacy of psychological interventions for reducing needle‐related pain and distress in children and adolescents. Consistent with our first update (Uman 2013), results from this update continue to demonstrate the efficacy of distraction for pain and hypnosis for pain and distress based on very low to low‐quality evidence; however, very low‐quality evidence now also demonstrates the efficacy of distraction for distress outcomes. Other interventions now identified as efficacious for at least one primary outcome include combined CBT for pain and distress, and breathing for pain. No evidence was available to support the efficacy of preparation and information or suggestion interventions for reducing children’s pain and distress. Furthermore, we could draw no conclusions about the efficacy of memory alteration, as only a single trial addressed this intervention. No new trials have assessed the efficacy of suggestion or memory alteration since our first review update (Uman 2013).

Trials support the use of a variety of distraction interventions for reducing self‐reported pain and distress, observer‐reported pain and distress, behavioral measures of distress, and heart rate. This second review update now offers evidence supporting the efficacy of distraction for all distress outcomes. The efficacy of distraction for distress outcomes was not demonstrated in our first update (Uman 2013). This change toward supportive evidence underscores the very low GRADE quality of evidence ratings for these outcomes. Very low ratings reflect very little confidence in these effect estimates, suggesting that subsequent research may contribute to different findings. As we stated in our previous update, we continue to note significant variability in the distraction methods used across the included trials. A growing number of trials included head‐to‐head comparisons of distraction interventions (Aydin 2017; Bellieni 2006Miller 2016; Sahiner 2016), although it remains unclear whether the type of distraction influences its efficacy across child development, or with different needle procedures. Nevertheless, data are available to support the efficacy of distraction interventions from studies including children from two to 19 years old in a wide variety of settings and needle procedures.

Trials continue to support the use of hypnosis for the reduction of self‐reported pain, as well as self‐reported and behavioral measures of distress. We found only one new trial of hypnosis for inclusion since our first update (Ramírez‐Carrasco 2017), enabling additional meta‐analysis of behavioral measures of pain that did not find a significant effect of hypnosis for this outcome. No new trials contributed to any other primary outcomes. Comments we made about hypnosis in our previous update continue to be relevant (Uman 2013). Specifically, although hypnosis continues to demonstrate the largest effect sizes for reducing pain and distress during needle‐related procedures, most of these trials are published by a single research group, predominantly in children with cancer. This raises some concern about generalizability to other contexts, populations, and healthcare providers. Furthermore, higher hypnotizability appears to be related to greater treatment benefit (Liossi 2003), suggesting that hypnosis may not be equally effective for all children. The only two studies assessing hypnosis for behavioral measures of pain, and not demonstrating an effect, were local dental anesthetic injections in otherwise generally healthy children. We restate the need for additional research assessing the efficacy of hypnosis administered by different providers in different pediatric settings. Rigorous assessment of hypnosis interventions delivered remotely by a smartphone applications would be particularly valuable (Sucala 2013).

Low‐quality evidence supports the efficacy of combined CBT interventions for reductions in observer reports of child pain and behavioral measures of child distress. Combined CBT was identified as efficacious in our original review in this area (Uman 2006), but not in our most recent update (Uman 2013). This may have been due to our removal of less rigorously designed studies at that time. It is worth noting that no evidence was available to support the efficacy of combined CBT for children’s self‐reported pain and distress. This is relevant, given the inherently subjective nature of children’s pain experience (IASP 2004), and the many factors contributing bias to the rating of children’s pain and distress by others, such as their parents or nurses (Craig 2010). Given this, children’s self‐reported outcomes have previously been prioritized in the development of clinical practice guidelines for the management of immunization pain and distress (Taddio 2015).

Since our first update (Uman 2013), a number of new trials contribute very low‐quality evidence for the efficacy of breathing interventions for reducing children’s self‐reported pain during needle procedures. We could draw no conclusions about other primary pain or distress outcomes, due to an insufficient number of trials assessing those outcomes. Notably, this was the only type of intervention in this review to report adverse events, in which three of 50 children reported respiratory difficulties when asked to engage in a specialized form of deep breathing (Pourmovahed 2013). It may therefore be prudent to weigh the potential benefits against possible harms of this type of intervention, based on the medical status of each child.

When comparing the findings from this update to those of our first update (Uman 2013), it is important to reiterate changes to the categorization of psychological interventions. As described earlier in this review (How the intervention might work), we grouped the interventions according to key theorized mechanisms of treatment effect, to reflect contemporary thinking in the field (Accardi 2009; Birnie 2017; Jafari 2017; Noel 2018). Several intervention categories that we separately identified in our first update were subsumed under broader intervention categories for this second update (e.g. virtual reality as distraction; parent coaching and child distraction as combined CBT). An additional benefit of this focus on treatment mechanisms was to allow for more meaningful meta‐analyses, while reducing the number of intervention categories from which we could draw no conclusions because of too few trials. Specifically, we could draw no conclusions in our first update about parent positioning plus distraction or about distraction plus suggestion interventions, because of single RCTs (Uman 2013). These are now subsumed under combined CBT interventions, as they include both cognitive and behavioral strategies. Inclusion of these additional studies within the combined CBT category does not account for new findings supporting its efficacy for observer‐reported pain and behavioral measures of distress. Only new studies published since the first update contributed findings to the meta‐analysis for observer‐reported pain. Furthermore, findings for behavioral measures of distress remained significant when studies from these previously separate intervention categories were removed from the meta‐analysis. Findings for distraction did not change, even with the inclusion of virtual‐reality interventions.

Overall completeness and applicability of evidence

The evidence presented in this review is directly relevant to the efficacy of psychological interventions for needle‐related procedural pain and distress in children and adolescents. Included trials reported on a variety of psychological interventions for an array of needle procedures, settings, children's ages, and medical diagnoses. Findings from this review are therefore directly applicable to the clinical context and populations in whom they are studied.

We note that many interventions have much less or limited outcome data beyond self‐report of pain. Distraction and combined CBT are the only interventions for which we could meta‐analyze all six primary pain and distress outcomes; self‐report of pain was the only outcome that we could meta‐analyze for breathing and suggestion. Of the 59 studies in this review, only 35 (59%) include at least one outcome of distress (self‐report, observer report, or behavioral measure). As well as pain, assessment of emotional response, such as negative affect or distress, is a recommended core outcome in pediatric acute pain clinical trials (PedIMMPACT; McGrath 2008). The evidence therefore speaks most directly to experienced or observed pain intensity and, to a lesser extent, to procedure‐related distress. This is notable, as needle‐related distress is associated with a decreased efficacy of psychological interventions typically helpful for reducing pain, as well as increases in fear behaviors that interfere with completion of medical procedures (e.g. flailing, attempts to escape) or avoidance of medical procedures altogether (McMurtry 2015a). As we have described in our previous update, outcomes beyond child pain and distress may be relevant and may influence treatment efficacy. Examples include child preference or choice of intervention (Birnie 2014a), child pain‐related fear (Birnie 2017; Verhoeven 2012), parent distress and parent behaviors (Campbell 2017; Dahlquist 2005).

This review included trials of children aged between two and 19. Most of the evidence is available for children aged 12 or younger, with 34 of 59 trials (58%) focused exclusively on this age group. While several trials (24 of 59; 41%) included samples crossing early childhood to late adolescence, only one trial focused exclusively on adolescents (Kristjansdottir 2010). Furthermore, this trial reported a significant effect of music distraction without headphones, but not with headphones, for adolescents’ self‐report of pain, compared with standard care. The direct applicability of the evidence reviewed here to adolescents is therefore more limited, compared with needle procedures in younger children. Considerations of differences in coping tendencies and preferences across normative development during childhood and adolescence may be relevant (Skinner 2007). It is worth noting that the evidence of psychological interventions for adults undergoing needle procedures is based on a small number of studies with mixed findings. Specifically, meta‐analysis supports the efficacy of breathing interventions for pain in adults, but not for music or visual distraction (Boerner 2015).

Although included trials assessed the efficacy of psychological interventions for needle procedures in a variety of hospital and community settings, there are a few settings for which there is less evidence. Only two studies were conducted within the context of school‐based immunization clinics; one study assessed music distraction in 14‐year‐olds as already noted above (Kristjansdottir 2010), and a second study assessed a combined CBT intervention of relaxation and guided imagery in 11‐ to 12‐year‐olds (Nilsson 2015). This environment is distinct from other settings in which children undergo needle procedures, with additional potential barriers to implementation of psychological interventions. For example, additional factors may include the absence of parents, the presence of teachers or peers or both, possible observational learning of fear (i.e. 'fear contagion'), and limited education about pain management by public health immunizers (Boerner 2014). School‐based immunization programs are a promising strategy promoted by the World Health Organization for offering equitable benefits of immunization to all children, thereby reducing the prevalence of infectious diseases worldwide (WHO 2017). This review identifies gaps in the availability of direct evidence to inform implementation and efficacy of psychological interventions for such programs.

Included RCTs assessed the efficacy of psychological interventions for an array of needle procedures, predominantly venipuncture or blood draws, immunizations, and intravenous insertions. These seem to reflect the most common needle procedures experienced by hospitalized children (Stevens 2011) or generally healthy children (with respect to immunizations only) (CDC 2018; NACI 2018; WHO 2018). This review also includes other needle procedures experienced less frequently in hospital, outpatient, and community settings, such as lumbar punctures, intramuscular injections, and injections for local dental anesthetic and insulin.

Over half of the included trials (32 of 59) examined the efficacy of distraction. There remains a clear need for more research examining the efficacy of other types of psychological interventions in RCTs, particularly interventions for which there is limited evidence (e.g. breathing, memory alteration). Despite this, the array of cognitive and behavioral psychological strategies reflected in the included trials is vast. While this review provides valuable knowledge supporting the use of psychological interventions, studies investigating the use of pain management strategies for needle procedures suggest that such interventions are highly underused in clinical practice (Birnie 2014b; Stevens 2011). Furthermore, most studies in this review with ‘standard care’ control groups did not include any psychological pain management strategies, and in many cases did not include any specific pain management strategies whatsoever. Clinical practice guidelines for the management of immunization pain and distress across the lifespan promote a multimodal approach, with recommendations to integrate pharmacological, physiological, procedural, and psychological strategies for optimal pain management (Taddio 2015). While it is recognized that current practice may vary internationally and is, in part, impacted by the cost and availability of interventions, the use of low‐cost and effective psychological interventions for needle procedures in children is supported by this review.

Quality of the evidence

To our knowledge, this review represents the largest, most rigorous and up‐to‐date review of psychological interventions for needle‐related pain and distress in children and adolescents. We include 59 peer‐reviewed published RCTs, covering 5550 children and adolescents aged two to 19 years old. This can be considered the most comprehensive and well‐reported review in the field.

We assessed the quality of the evidence in this review using the GRADE system as applied to continuous outcomes (Guyatt 2013). GRADE considers the quality of evidence pertaining to limitations in the design and implementation, indirectness of evidence, unexplained heterogeneity or inconsistency of results, imprecision of results, and high probability of publication bias. To inform these quality ratings, we also considered ratings from the Cochrane 'Risk of bias' tool for each trial (Higgins 2017). Across all types of psychological interventions and all outcomes, we rated the quality of evidence as very low or low. No interventions or outcomes obtained quality ratings of ‘moderate’ or ‘high’, despite the inclusion of only the highest‐quality evidence (i.e. RCTs). This suggests that overall we have very little to limited confidence in the reported effect estimates in this review, suggesting that the true effects may be substantially different.

In large part, the quality of evidence was consistently downgraded from high to low or very low due to serious limitations in study design or implementation (as evidenced by the lack of blinding of participants or study personnel or of assessment of study outcomes, poor allocation concealment, and selective reporting), serious concerns with unexplained heterogeneity, or inconsistency of results (as evidenced by moderate to considerable heterogeneity in many meta‐analyses), and imprecision (as evidenced by low numbers of participants for several group comparisons). It is important to recognize that it may be difficult, if not impossible, for RCTs of psychological interventions to obtain the highest quality of evidence ratings due to the inherent challenges in blinding of study participants and personnel, and to the assessment of self‐report and observer report of child pain or distress. Many psychological interventions are obvious and, by their very nature, preclude the achievement of true blinding. Despite that, the quality of evidence can be improved through better design, implementation, and reporting of trials. In particular, future trials should strive for use of appropriate sequence generation and allocation concealment, provide clear a priori identification of primary and secondary outcomes, include larger sample sizes, and make available registered clinical trial protocols. Exaggerated treatment effects have previously been shown in trials with inadequate or unclear random sequence generation, allocation concealment, or selective reporting (Savovic 2012); however, exploration of such potential bias in RCTs of distraction for needle procedures in children and adolescents has not been replicated (Birnie 2014a).

Potential biases in the review process

Strengths of this review include comprehensive and updated literature searches, focus on the highest quality of evidence, inclusion of non‐English publications, and contacting authors when relevant data were missing from published reports. An identified concern in previous versions (Uman 2006; Uman 2013) was the number of studies excluded for not providing sufficient data in published reports or through correspondence to allow for data pooling in meta‐analysis (n = 24 trials in the current review). Any associated bias in review findings appears less concerning over time, as we omitted only one study in this second update for not providing necessary data. This improvement may be due to the availability of published guidelines for the reporting of RCTs (Consolidated Standards of Reporting Trials; CONSORT; www.consort-statement.org), which recommend that findings for each primary and secondary outcome be reported separately for each study group. Many higher‐impact peer‐reviewed journals are also requiring published RCTs to follow CONSORT. Nevertheless, to minimize any potential bias, we continue to encourage authors to include all summary statistics (i.e. means, SDs, cell sizes) separately for all groups and for all assessed outcomes, regardless of study results. Specifically, authors should provide all summary statistics even when there are no significant differences between groups. It is possible that some bias is introduced through the exclusion of unpublished trials. This represents a small number of studies overall, and available information suggests their probable exclusion for other reasons. Unpublished dissertation studies were included in our original review (Uman 2006), with no new dissertations excluded in this review update. While we accept that peer review and subsequent publication are not consistent measures of quality, we have standardized the included studies based on the common factor of peer review and publication status. Furthermore, systematic identification of unpublished trials remains difficult at this time, given the general lack of registration of trial protocols. We also encourage journals to publish RCTs with non‐significant findings, to minimize this potential bias.

As we identified in our previous update, we combined assessments of reported pain or distress at varying times during or following the needle procedure available in included trials. Specifically, some of the included studies assessed outcomes during the needle procedure while others assessed outcomes following the needle procedure. This may introduce bias, as the timing of assessment of pain and distress varied across trials. Individual studies and reviews have also included assessment of pre‐procedural anxiety and distress (Birnie 2015), although we excluded those assessments from our review due to the inconsistent timing of psychological intervention delivery. For example, while implementation of some psychological interventions began prior to the needle procedure, this was not consistent across intervention strategies. Not all pre‐procedural measures of anxiety or distress would be expected to be influenced by psychological interventions delivered only during and after the needle procedure itself. We have previously recommended that, at minimum, the standard assessment of post‐needle pain or distress should be conducted immediately following the procedure (i.e. as soon as the needle is removed). We acknowledge that it is possible that we have introduced potential bias by excluding such studies. Potential bias is also introduced by the pooling of studies with various sources of clinical heterogeneity, such as variability in types of distractors, child age, child diagnosis, and healthcare setting. This is a limitation of psychological interventions that are inevitably more heterogeneous than medication interventions. Previous preliminary work suggests factors such as type of distractor and child age may have some influence on treatment efficacy (Birnie 2014a), and should be considered in future work.

It is possible that bias was introduced through the exclusion of trials in which the effects of particular psychological interventions could not be isolated. Practically, this leads to the exclusion of trials using multimodal interventions or with head‐to‐head comparisons of differing psychological interventions without a standard‐care comparison group. We excluded an increasing number of trials from this second update, as we could not isolate the effects of psychological components from a multicomponent intervention (n = 8). Half of these studies included interventions combining vibration with distraction (e.g. Buzzy device or vibrating wand) (Baxter 2011; Benjamin 2016; Moadad 2016; Schreiber 2016). While the intervention was described by several trial authors as distraction, we have chosen to exclude these interventions as the reported mechanisms of treatment effect are varied (likely also physical) and have not been adequately studied. Furthermore, there is clearly a physical component with a physiological mechanism proposed to impact pain processing. To address these concerns, we underscore the need for future research focused on delineating the mechanisms of treatment effect to clearly identify any additional benefit of psychological strategies (e.g. distraction) to a multimodal intervention.

Agreements and disagreements with other studies or reviews

To our knowledge, this is the most comprehensive review of this topic. However, other systematic reviews in related areas are available, which also support the efficacy of various psychological interventions for pediatric pain management. These include, and are not limited to, systematic reviews of psychological interventions for reducing pain and distress during immunizations in children and adults (Birnie 2015; Boerner 2015; Chambers 2009) and medical procedures in children with cancer (Bukola 2017; Flowers 2015), hypnosis for painful procedures in children and adolescents (Accardi 2009; Richardson 2006), memory alteration interventions for pediatric needle procedures (Noel 2018), as well as music (Klassen 2008) and virtual reality interventions (Kenney 2016) for acute and chronic pain in children and adults.

Study flow diagram.

Figures and Tables -
Figure 1

Study flow diagram.

Navigate to figure in ReviewOpen in new tab

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figures and Tables -
Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figures and Tables -
Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 1 Distraction, outcome: 1.1 Self‐reported pain.

Figures and Tables -
Figure 4

Forest plot of comparison: 1 Distraction, outcome: 1.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 2 CBT‐combined, outcome: 2.1 Self‐reported pain.

Figures and Tables -
Figure 5

Forest plot of comparison: 2 CBT‐combined, outcome: 2.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 3 Hypnosis, outcome: 3.1 Self‐reported pain.

Figures and Tables -
Figure 6

Forest plot of comparison: 3 Hypnosis, outcome: 3.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 4 Preparation/Information, outcome: 4.1 Self‐reported pain.

Figures and Tables -
Figure 7

Forest plot of comparison: 4 Preparation/Information, outcome: 4.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 5 Breathing, outcome: 5.1 Self‐reported pain.

Figures and Tables -
Figure 8

Forest plot of comparison: 5 Breathing, outcome: 5.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Forest plot of comparison: 6 Suggestion, outcome: 6.1 Self‐reported pain.

Figures and Tables -
Figure 9

Forest plot of comparison: 6 Suggestion, outcome: 6.1 Self‐reported pain.

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 1.1

Comparison 1: Distraction, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 2: Self‐reported distress

Figures and Tables -
Analysis 1.2

Comparison 1: Distraction, Outcome 2: Self‐reported distress

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 3: Observer‐reported pain

Figures and Tables -
Analysis 1.3

Comparison 1: Distraction, Outcome 3: Observer‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 4: Observer‐reported distress

Figures and Tables -
Analysis 1.4

Comparison 1: Distraction, Outcome 4: Observer‐reported distress

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 5: Behavioral measures‐ Pain

Figures and Tables -
Analysis 1.5

Comparison 1: Distraction, Outcome 5: Behavioral measures‐ Pain

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 6: Behavioral measures‐ Distress

Figures and Tables -
Analysis 1.6

Comparison 1: Distraction, Outcome 6: Behavioral measures‐ Distress

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 7: Physiological measure ‐ heart rate

Figures and Tables -
Analysis 1.7

Comparison 1: Distraction, Outcome 7: Physiological measure ‐ heart rate

Navigate to figure in ReviewOpen in new tab

Comparison 1: Distraction, Outcome 8: Physiological measure ‐ oxygen saturation

Figures and Tables -
Analysis 1.8

Comparison 1: Distraction, Outcome 8: Physiological measure ‐ oxygen saturation

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 2.1

Comparison 2: CBT‐combined, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 2: Self‐reported distress

Figures and Tables -
Analysis 2.2

Comparison 2: CBT‐combined, Outcome 2: Self‐reported distress

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 3: Observer‐reported pain

Figures and Tables -
Analysis 2.3

Comparison 2: CBT‐combined, Outcome 3: Observer‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 4: Observer‐reported distress

Figures and Tables -
Analysis 2.4

Comparison 2: CBT‐combined, Outcome 4: Observer‐reported distress

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 5: Behavioral measures‐ Pain

Figures and Tables -
Analysis 2.5

Comparison 2: CBT‐combined, Outcome 5: Behavioral measures‐ Pain

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 6: Behavioral measures‐ Distress

Figures and Tables -
Analysis 2.6

Comparison 2: CBT‐combined, Outcome 6: Behavioral measures‐ Distress

Navigate to figure in ReviewOpen in new tab

Comparison 2: CBT‐combined, Outcome 7: Physiological measure ‐ Cortisol

Figures and Tables -
Analysis 2.7

Comparison 2: CBT‐combined, Outcome 7: Physiological measure ‐ Cortisol

Navigate to figure in ReviewOpen in new tab

Comparison 3: Hypnosis, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 3.1

Comparison 3: Hypnosis, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 3: Hypnosis, Outcome 2: Self‐reported distress

Figures and Tables -
Analysis 3.2

Comparison 3: Hypnosis, Outcome 2: Self‐reported distress

Navigate to figure in ReviewOpen in new tab

Comparison 3: Hypnosis, Outcome 3: Behavioral measures‐ Pain

Figures and Tables -
Analysis 3.3

Comparison 3: Hypnosis, Outcome 3: Behavioral measures‐ Pain

Navigate to figure in ReviewOpen in new tab

Comparison 3: Hypnosis, Outcome 4: Behavioral measures‐ Distress

Figures and Tables -
Analysis 3.4

Comparison 3: Hypnosis, Outcome 4: Behavioral measures‐ Distress

Navigate to figure in ReviewOpen in new tab

Comparison 4: Preparation/Information, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 4.1

Comparison 4: Preparation/Information, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 4: Preparation/Information, Outcome 2: Observer‐reported pain

Figures and Tables -
Analysis 4.2

Comparison 4: Preparation/Information, Outcome 2: Observer‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 5: Breathing, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 5.1

Comparison 5: Breathing, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab

Comparison 6: Suggestion, Outcome 1: Self‐reported pain

Figures and Tables -
Analysis 6.1

Comparison 6: Suggestion, Outcome 1: Self‐reported pain

Navigate to figure in ReviewOpen in new tab
Summary of findings 1. Distraction compared to control for needle‐related procedural pain and distress in children and adolescents

Distraction compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 2‐19 years with mixed medical (acute or chronic illness) or generally healthy undergoing venipuncture, immunization, intravenous insertion, lumbar puncture, bone marrow aspiration, routine injection, allergy testing injections, or laceration repair
Setting: hospital (inpatient/outpatient/emergency department), community clinic, or school
Intervention: distraction
Comparison: control (varied across studies)

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with distraction

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 0.65 to 8.32 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with distraction was 0.56 standard deviations lower (0.78 to 0.33 lower).

2802
(30 RCTs)

⊕⊕⊝⊝
LOW a,b

This result equates to a moderate difference in favor of distraction

Self‐reported distress

See comment

The mean level of self‐reported distress with distraction was 0.82 standard deviations lower
(1.45 to 0.18 lower)

426
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a large difference in favor of distraction

Observer‐reported pain

See comment

The mean level of observer‐reported pain with distraction was 0.62 standard deviations lower
(1 to 0.23 lower)

1512
(11 RCTs)

⊕⊝⊝⊝
VERY LOW a,d

This result equates to a moderate to large difference in favor of distraction

Observer‐reported distress

See comment

The mean level of observer‐reported distress with distraction was 0.72 standard deviations lower
(1.41 to 0.03 lower)

1067
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,d

This result equates to a moderate to large difference in favor of distraction

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with distraction was 0.33 standard deviations lower
(0.69 lower to 0.03 higher)

309
(4 RCTs)

⊕⊕⊝⊝
LOW a,c

There is no evidence of an effect of distraction

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with distraction was 0.44 standard deviations lower
(0.84 to 0.04 lower)

500
(7 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a small to moderate difference in favor of distraction

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

d Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.

Figures and Tables -
Summary of findings 1. Distraction compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 2. CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents

CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐18 years with mixed medical (acute or chronic illness) or generally healthy undergoing immunization, intravenous insertion, venipuncture, bone marrow aspiration, insulin injection, or dental local anesthetic
Setting: hospital (inpatient/outpatient/emergency department), community clinic, or school
Intervention: CBT‐combined
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with CBT‐combined

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 0.84 to 8.4 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with combined CBT was 0.27 standard deviations lower
(0.58 lower to 0.03 higher)

1359
(14 RCTs)

⊕⊕⊝⊝
LOW a,b

There is no evidence of an effect of combined CBT

Self‐reported distress

See comment

The mean level of self‐reported distress with combined CBT was 0.26 standard deviations lower
(0.56 lower to 0.04 higher)

234
(6 RCTs)

⊕⊕⊝⊝
LOW a,c

There is no evidence of an effect of combined CBT

Observer‐reported pain

See comment

The mean level of observer‐reported pain with combined CBT was 0.52 standard deviations lower
(0.73 to 0.30 lower)

385
(4 RCTs)

⊕⊕⊝⊝
LOW a,c

This result equates to a moderate difference in favor of combined CBT

Observer‐reported distress

See comment

The mean level of observer‐reported distress with combined CBT was 0.08 standard deviations higher
(0.34 lower to 0.50 higher)

765
(6 RCTs)

⊕⊕⊝⊝
LOW a,b

There is no evidence of an effect of combined CBT

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with combined CBT was 0.65 standard deviations lower
(2.36 lower to 1.06 higher)

95
(2 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,d

There is no evidence of an effect of combined CBT

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with combined CBT was 0.40 standard deviations lower
(0.67 to 0.14 lower)

1105
(11 RCTs)

⊕⊕⊝⊝
LOW a,b

This result equates to a small to moderate difference in favor of combined CBT

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.
dDowngraded twice for imprecision: analysis based on < 100 participants per group.

Figures and Tables -
Summary of findings 2. CBT‐combined compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 3. Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents

Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐16 years with chronic illness (cancer) or generally healthy undergoing bone marrow aspirations, lumbar punctures, venipuncture, or local dental anesthetic
Setting: hospital (inpatient/outpatient), community clinic
Intervention: hypnosis
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with hypnosis

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 4.17 to 8.6 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with hypnosis was 1.40 standard deviations lower
(2.32 to 0.48 lower)

176
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c,d

This result equates to a large difference in favor of hypnosis

Self‐reported distress

See comment

The mean level of self‐reported distress with hypnosis was 2.53 standard deviations lower
(3.93 to 1.12 lower)

176
(5 RCTs)

⊕⊝⊝⊝
VERY LOW a,c,d,e

This result equates to a large difference in favor of hypnosis

Observer‐reported pain

See comment

See comment

See comment

.‐

This outcome was not assessed in any study

Observer‐reported distress

See comment

See comment.

36
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

The mean level of behavioral pain with hypnosis was 0.38 standard deviations lower
(1.57 lower to 0.81 higher)

69
(2 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of hypnosis

Behavioral measures‐ distress

See comment

The mean level of behavioral distress with hypnosis was 1.15 standard deviations lower
(1.76 to 0.53 lower)

193
(6 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c,d

This result equates to a large difference in favor of hypnosis

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded twice for imprecision: analysis based on < 100 participants per group.
d Downgraded once for possibility of publication bias given that almost all trials are from one expert group.
e Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.

Figures and Tables -
Summary of findings 3. Hypnosis compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 4. Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents

Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐12 years with mixed medical (acute or chronic illness) or unclear diagnoses undergoing venipuncture or intravenous insertion
Setting: hospital (outpatient/emergency department) or community clinic
Intervention: preparation/information
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with preparation/information

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 2.6 to 6.12 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with preparation/information was 0.18 standard deviations lower
(0.60 lower to 0.23 higher)

313
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of preparation/information

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

The mean level of observer‐reported pain with preparation/information was 0.40 standard deviations lower
(0.98 lower to 0.18 higher)

259
(3 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

There is no evidence of an effect of preparation/information

Observer‐reported distress

See comment

See comment

100
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

39
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ distress

See comment

See comment

54
(1 RCT)

‐

This outcome was assessed in one study only

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded once for inconsistency due to moderate heterogeneity (I2) > 45%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

Figures and Tables -
Summary of findings 4. Preparation/information compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 5. Breathing compared to control for needle‐related procedural pain and distress in children and adolescents

Breathing compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 6‐15 years who are generally healthy, undergoing surgery, or have leukemia undergoing venipuncture or lumbar puncture
Setting: hospital (outpatient/inpatient)
Intervention: breathing
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with breathing

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 4 to 7.6 (adjusted to a 0 to 10 scale)

The mean level of self‐reported pain with breathing was 1.04 standard deviations lower
(1.86 to 0.22 lower)

298
(4 RCTs)

⊕⊝⊝⊝
VERY LOW a,b,c

This result equates to a large difference in favor of breathing

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

See comment

120
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

120
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

‐

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded twice for inconsistency due to considerable heterogeneity (I2) > 90%.
c Downgraded once for imprecision: analysis based on < 400 participants per group.

Figures and Tables -
Summary of findings 5. Breathing compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 6. Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents

Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐17 years who are generally healthy or have chronic illness undergoing intramuscular injection, immunization, or venipuncture
Setting: hospital (outpatient) or community clinics
Intervention: suggestion
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with suggestion

Self‐reported pain

The mean level of self‐reported pain in the control group ranged from 1.83 to 9.33 (adjusted to a 0 to 10 scale).

The mean level of self‐reported pain with suggestion was 0.13 standard deviations lower
(0.40 lower to 0.15 higher)

218
(3 RCTs)

⊕⊝⊝⊝
VERY LOW a,b

There is no evidence of an effect of suggestion

Self‐reported distress

See comment

See comment

78
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported pain

See comment

See comment

78
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

20
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded once for serious study limitations: most trials had unclear/high risk of bias in blinding, allocation concealment and/or selective reporting of outcomes.
b Downgraded twice for imprecision: analysis based on < 100 participants per group.

Figures and Tables -
Summary of findings 6. Suggestion compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Summary of findings 7. Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents

Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents

Patient or population: children aged 3‐18 years with chronic illness (leukemia) undergoing lumbar puncture
Setting: hospital (inpatient)
Intervention: memory alteration
Comparison: control

Outcomes

Anticipated absolute effects* (95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments**

Risk with control

Risk with memory alteration

Self‐reported pain

See comment

See comment

24
(1 RCT)

‐

This outcome was assessed in one study only

Self‐reported distress

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Observer‐reported pain

See comment

See comment

42
(1 RCT)

‐

This outcome was assessed in one study only

Observer‐reported distress

See comment

See comment

50
(1 RCT)

‐

This outcome was assessed in one study only

Behavioral measures‐ pain

See comment

See comment

See comment

‐

This outcome was not assessed in any study

Behavioral measures‐ distress

See comment

See comment

50
(1 RCT)

‐

This outcome was assessed in one study only

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).

**One 'rule of thumb' for interpreting the relative effect is that 0.2 represents a small difference, 0.5 a moderate difference and 0.8 a large difference.
CI: Confidence interval; SMD: standardized mean difference; RCT: randomized controlled trial

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

Figures and Tables -
Summary of findings 7. Memory alteration compared to control for needle‐related procedural pain and distress in children and adolescents
Navigate to table in Review
Table 1. Definitions of medical procedures

Procedure

Definition

Accessing a portacath (also known as a port)

Insertion of a needle into an implanted access device (portacath) which facilitates the drawing of blood and intravenous (or intra‐arterial) injections by not having to locate and insert a cannula into a new vessel. Some ports are connected for intrathecal, intraperitoneal or intracavitary injections.

Arterial blood gas (ABG)

A test which analyses arterial blood for oxygen, carbon dioxide and bicarbonate content in addition to blood pH. Used to test the effectiveness of respiration.

Arterial line (also known as intra‐arterial catheter)

Insertion of a catheter into an artery.

Arterial puncture

A hole, wound, or perforation of an artery made by puncturing.

Bone marrow aspiration (BMA)

The bone marrow is the tissue that manufactures the blood cells and is in the hollow part of most bones. This test is done by suctioning some of the bone marrow for examination.

Bone marrow biopsy (BMB)

The removal and examination of tissue, cells, or fluids from the bone marrow of a living body; usually performed at the same time as a BMA.

Central line (also known as central venous catheter)

Insertion of a catheter into the large vein above the heart, usually the subclavian vein, through which access to the blood stream can be made. This allows drugs and blood products to be given and blood samples withdrawn.

Finger prick/pin

Obtaining blood by puncturing the tip of the finger.

Immunization (also known as immunization)

Protection against a particular disease or treatment of an organism by protecting against certain pathogen attacks; the introduction of microorganisms that have previously been treated to make them harmless.

Injection

The act of forcing a liquid into tissue, the vascular tree, or an organ.

Intramuscular injection

Injection administered by entering a muscle.

IV/catheter insertion

A narrow short, flexible, synthetic (usually plastic) tube known as a catheter, that is inserted approximately one inch into a vein to provide temporary intravenous access for the administration of fluid, medication, or nutrients.

Lumbar punctures (LP) (also known as spinal tap)

The withdrawal of cerebrospinal fluid or the injection of anesthesia by puncturing the subarachnoid space located in the lumbar region of the spinal cord.

Paracentesis

A surgical puncture of a bodily cavity (e.g. abdomen) with a trocar, aspirator, or other instrument usually to draw off an abnormal effusion for diagnostic or therapeutic purposes.

Subcutaneous injection

Injection administered under the skin.

Suture (also known as laceration repair)

A stitch made with a strand or fiber used to sew parts of the living body.

Thoracocentesis (also called thoracentesis)

Aspiration of fluid from the chest.

Venepuncture (also known as venipuncture)

The surgical puncture of a vein typically for withdrawing blood or administering intravenous medication.

Figures and Tables -
Table 1. Definitions of medical procedures
Navigate to table in Review
Table 2. Means and standard deviations for outcomes from single trials

Intervention

Outcome

Study

Treatment

Control

N

Mean (SD)

N

Mean (SD

Distraction

Physiological measure ‐ Respiratory rate

Nguyen 2010

20

25.1 (3.6)

20

28.5 (3.86)

Physiological measure ‐ Systolic BP

Nguyen 2010

20

97.1 (8.57)

20

105.6 (15.97)

Physiological measure ‐ Diastolic BP

Nguyen 2010

20

65.2 (6.83)

20

69.8 (11.67)

Hypnosis

Observer‐reported distress

Katz 1987

17

3 (0.9)

19

3.3 (0.6)

Physiological measure ‐ Skin conductance

Ramírez‐Carrasco 2017

20

10.61 (6.42)

20

8.74 (4.63)

Physiological measure ‐ Heart rate

Ramírez‐Carrasco 2017

20

93.57 (12.33)

20

99.3 (18.56)

Preparation/Information

Observer‐reported distress

Harrison 1991

50

2.43 (1.62)

50

3.17 (1.3)

Behavioural measure ‐ Pain

Miller 2016

19

3.21 (3.46)

20

4.90 (3.67)

Behavioural measure ‐ Distress

Tak 2006

26

2.64 (1.1)

28

2.37 (1.12)

Physiological measure ‐ Pulse rate

Harrison 1991

50

84.6 (8.6)

50

88.6 (8.3)

Breathing

Observer‐reported pain

Sahiner 2016

60

2.56 (3.24)

60

3.80 (2.93)

Observer‐reported distress

Sahiner 2016

60

0.59 (1.03)

60

1.25 (1.41)

Suggestion

Self‐reported distress

Goodenough 1997

39

0.7 (1.1)

39

1.1 (1.3)

Observer‐reported pain

Goodenough 1997

39

0.9 (1.6)

39

1.7 (2.3)

Observer‐reported distress

Eland 1981

10

1.8 (0.71)

10

1.8 (0.79)

Memory Alteration

Self‐reported pain
(during procedure change score)

Chen 1999

15

−0.06 (3.9)

9

−0.02 (3.2)

Observer‐reported pain

(during procedure change score)

Chen 1999

20

04. (3.1)

22

−0.1 (1.8)

Observer‐reported distress

(during procedure change score)

Chen 1999

25

−0.2 (2.6)

25

−0.5 (1.9)

Behavioural measure ‐ Distress

(during procedure change score)

Chen 1999

25

−0.62 (3.7)

25

−0.48 (2.0)

Physiological measure ‐ Heart rate

(during procedure change score)

Chen 1999

24

0.1 (26.9)

20

−4.9 (21.8)

Physiological measure ‐ Cortisol

(during procedure change score)

Chen 1999

22

0.01 (0.18)

22

0.01 (0.2)

Physiological measure ‐ Systolic BP

(during procedure change score)

Chen 1999

23

−0.5 (11.8)

19

−5.4 (7.9)

Physiological measure ‐ Diastolic BP

(during procedure change score)

Chen 1999

23

−4.1 (10.9)

19

2.9 (10.2)
Figures and Tables -
Table 2. Means and standard deviations for outcomes from single trials
Navigate to table in Review
Comparison 1. Distraction

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1.1 Self‐reported pain Show forest plot

30

2802

Std. Mean Difference (IV, Random, 95% CI)

‐0.56 [‐0.78, ‐0.33]

1.2 Self‐reported distress Show forest plot

4

426

Std. Mean Difference (IV, Random, 95% CI)

‐0.82 [‐1.45, ‐0.18]

1.3 Observer‐reported pain Show forest plot

11

1512

Std. Mean Difference (IV, Random, 95% CI)

‐0.62 [‐1.00, ‐0.23]

1.4 Observer‐reported distress Show forest plot

5

1067

Std. Mean Difference (IV, Random, 95% CI)

‐0.72 [‐1.41, ‐0.03]

1.5 Behavioral measures‐ Pain Show forest plot

4

309

Std. Mean Difference (IV, Random, 95% CI)

‐0.33 [‐0.69, 0.03]

1.6 Behavioral measures‐ Distress Show forest plot

7

500

Std. Mean Difference (IV, Random, 95% CI)

‐0.44 [‐0.84, ‐0.04]

1.7 Physiological measure ‐ heart rate Show forest plot

3

252

Std. Mean Difference (IV, Random, 95% CI)

‐0.64 [‐0.89, ‐0.38]

1.8 Physiological measure ‐ oxygen saturation Show forest plot

2

112

Std. Mean Difference (IV, Random, 95% CI)

0.60 [0.22, 0.98]
Figures and Tables -
Comparison 1. Distraction
Navigate to table in Review
Comparison 2. CBT‐combined

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

2.1 Self‐reported pain Show forest plot

14

1359

Std. Mean Difference (IV, Random, 95% CI)

‐0.27 [‐0.58, 0.03]

2.2 Self‐reported distress Show forest plot

6

234

Std. Mean Difference (IV, Random, 95% CI)

‐0.26 [‐0.56, 0.04]

2.3 Observer‐reported pain Show forest plot

4

385

Std. Mean Difference (IV, Random, 95% CI)

‐0.52 [‐0.73, ‐0.30]

2.4 Observer‐reported distress Show forest plot

6

765

Std. Mean Difference (IV, Random, 95% CI)

0.08 [‐0.34, 0.50]

2.5 Behavioral measures‐ Pain Show forest plot

2

95

Std. Mean Difference (IV, Random, 95% CI)

‐0.65 [‐2.36, 1.06]

2.6 Behavioral measures‐ Distress Show forest plot

11

1105

Std. Mean Difference (IV, Random, 95% CI)

‐0.40 [‐0.67, ‐0.14]

2.7 Physiological measure ‐ Cortisol Show forest plot

3

310

Std. Mean Difference (IV, Random, 95% CI)

‐1.16 [‐3.37, 1.06]
Figures and Tables -
Comparison 2. CBT‐combined
Navigate to table in Review
Comparison 3. Hypnosis

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

3.1 Self‐reported pain Show forest plot

5

176

Std. Mean Difference (IV, Random, 95% CI)

‐1.40 [‐2.32, ‐0.48]

3.2 Self‐reported distress Show forest plot

5

176

Std. Mean Difference (IV, Random, 95% CI)

‐2.53 [‐3.93, ‐1.12]

3.3 Behavioral measures‐ Pain Show forest plot

2

69

Std. Mean Difference (IV, Random, 95% CI)

‐0.38 [‐1.57, 0.81]

3.4 Behavioral measures‐ Distress Show forest plot

6

193

Std. Mean Difference (IV, Random, 95% CI)

‐1.15 [‐1.76, ‐0.53]
Figures and Tables -
Comparison 3. Hypnosis
Navigate to table in Review
Comparison 4. Preparation/Information

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

4.1 Self‐reported pain Show forest plot

4

313

Std. Mean Difference (IV, Random, 95% CI)

‐0.18 [‐0.60, 0.23]

4.2 Observer‐reported pain Show forest plot

3

259

Std. Mean Difference (IV, Random, 95% CI)

‐0.40 [‐0.98, 0.18]
Figures and Tables -
Comparison 4. Preparation/Information
Navigate to table in Review
Comparison 5. Breathing

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

5.1 Self‐reported pain Show forest plot

4

298

Std. Mean Difference (IV, Random, 95% CI)

‐1.04 [‐1.86, ‐0.22]
Figures and Tables -
Comparison 5. Breathing
Navigate to table in Review
Comparison 6. Suggestion

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

6.1 Self‐reported pain Show forest plot

3

218

Std. Mean Difference (IV, Random, 95% CI)

‐0.13 [‐0.40, 0.15]
Figures and Tables -
Comparison 6. Suggestion
Navigate to table in Review