Opioids for procedural pain in neonates

Summary of findings 1. Opioids compared to no treatment/placebo for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to no treatment/placebo for procedural pain in neonates
Patient or population: neonates exposed to procedural pain Setting: neonatal units Intervention: opioids Comparison: no treatment/placebo
Outcomes	Risk with no treatment/placebo	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the PIPP/PIPP‐R pain scale, ranging from 0 to 21 and 0 to 18 in preterm infants < 28 weeks' gestational age and full‐term infants, respectively, during procedure	Mean PIPP/PIPP‐R during procedure ranged 8 to 11.	MD 2.58 lower (3.12 lower to 2.03 lower)	‐	199 (3 RCTs)	⊕⊕⊕⊝ Moderate ¹	Opioids probably reduce pain score assessed with the PIPP/PIPP‐R scale during the procedure compared to placebo.
Pain, assessed using the PIPP/PIPP‐R pain scale up to 30 min after procedure	Mean PIPP/PIPP‐R up to 30 min after procedure ranged 3 to 6.	MD 0.14 higher (0.17 lower to 0.45 higher)	‐	123 (2 RCTs)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale up to 30 min after the procedure compared to placebo.
Pain, assessed using the PIPP/PIPP‐R pain scale 1 to 2 hours after procedure	Mean PIPP/PIPP‐R 1 to 2 hours after procedure ranged 4 to 11.	MD −0.83 (2.42 lower to 0.75 higher)	‐	54 (2 RCTs)	⊕⊝⊝⊝ Very low ³	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale 1 to 2 hours after the procedure compared to placebo.
Pain, assessed using the DAN pain scale, ranging from 0 to 21, 1 to 2 hours after procedure	Mean DAN 1 to 2 hours after procedure was 5.	MD 0.2 lower (2.21 lower to 1.81 higher)	‐	42 (1 RCT)	⊕⊕⊝⊝ Low ⁴	Opioids may result in little to no difference in pain score assessed with the DAN scale 1 to 2 hours after the procedure compared to placebo.
Pain, assessed using the NIPS, ranging from 0 to 7, during procedure	Mean NIPS during procedure ranged 5 to 6.	MD 1.97 lower (2.46 lower to 1.48 lower)	‐	102 (2 RCTs)	⊕⊕⊝⊝ Low ⁵	Opioids may reduce NIPS during the procedure compared to placebo.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of bradycardia	Study population		RR 3.19 (0.14 to 72.69) RD 0.01, (−0.03 to 0.06)	172 (3 RCTs)	⊕⊝⊝⊝ Very low ⁶	The evidence is very uncertain about the effect of opioids on episodes of bradycardia compared to placebo.
Episodes of bradycardia	0 per 1000 (No events in the 3 RCTs)	0 per 1000 (0 to 0)	RR 3.19 (0.14 to 72.69) RD 0.01, (−0.03 to 0.06)	172 (3 RCTs)	⊕⊝⊝⊝ Very low ⁶
Episodes of apnea	Study population		RR 3.15 (1.08 to 9.16) RD 0.07 (0.01 to 0.14) NNTH = 14	199 (3 RCTs)	⊕⊕⊝⊝ Low ⁷	Opioids may result in an increase in episodes of apnea compared to placebo.
Episodes of apnea	30 per 1000	95 per 1000 (33 to 278)	RR 3.15 (1.08 to 9.16) RD 0.07 (0.01 to 0.14) NNTH = 14	199 (3 RCTs)	⊕⊕⊝⊝ Low ⁷
Hypotension	Study population		‐	88 (2 RCTs)	⊕⊝⊝⊝ Very low ⁸	The evidence is very uncertain about the effect of opioids on hypotension compared to placebo.
Hypotension	See comment	See comment	‐	88 (2 RCTs)	⊕⊝⊝⊝ Very low ⁸
Parent satisfaction with care provided in the NICU (as measured by a validated instrument/tool)—not reported	‐	‐	‐	‐	This outcome was not reported.	This outcome was not reported.
*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). CI: confidence interval; DAN: Douleur Aiguë du Nouveau‐né; MD: mean difference; NICU: neonatal intensive care unit; NIPS: Neonatal Infant Pain Scale; NNTH: number needed to treat for an additional harmful outcome; OR: odds ratio; PIPP: Premature Infant Pain Profile; PIPP‐R: PIPP‐Revised; RCT: randomized controlled trial; RD: risk difference; RR: risk ratio
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.
¹Downgraded one level for inconsistency (moderate heterogeneity, I² = 62%); not downgraded for imprecision (narrow CIs). ²Downgraded one level for inconsistency (substantial heterogeneity, I² = 85%) and two levels for imprecision (two small trials with low sample size; CIs overlapping no effect). ³Downgraded one level for study limitations (unclear risk of selection, performance, detection, and reporting bias) and two levels for imprecision (two small trials with low sample size; CIs overlapping no effect). ⁴Downgraded two levels for imprecision (one small trial with low sample size; CIs overlapping no effect). ⁵Downgraded one level for inconsistency (moderate heterogeneity, I² = 60%) and one level for imprecision (two small trials with low sample size; CIs not overlapping no effect). ⁶Downgraded one level for study limitations (unclear risk of selection bias in the informative study) and two levels for imprecision (three small trials with low sample size; CI overlapping no effect). ⁷Downgraded one level for study limitations (unclear risk of selection and reporting bias) and one level for imprecision (wide CIs). ⁸Downgraded one level for study limitations (unclear risk of selection and reporting bias) and two levels for imprecision (two small trials with no events).

Summary of findings 2. Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates
Patient or population: neonates exposed to procedural pain Setting: neonatal units Intervention: opioids Comparison: oral sweet solution or non‐pharmacological intervention
Outcomes	Risk with oral sweet solution or non‐pharmacological intervention	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the CRIES pain scale, ranging from 0 to 10, during the procedure— opioids versus facilitated tucking	Mean CRIES—opioids versus facilitated tucking was 9.	MD 4.62 lower (6.38 lower to 2.86 lower)	‐	100 (1 RCT)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to facilitated tucking.
Pain, assessed using the CRIES pain scale, ranging from 0 to 10, during the procedure— opioids versus sensorial stimulation	Mean CRIES—opioids versus sensorial stimulation was 4.	MD 0.32 higher (1.13 lower to 1.77 higher)	‐	100 (1 RCT)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to sensorial stimulation.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of bradycardia—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of apnea—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Hypotension—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Parent satisfaction with care provided in the NICU—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*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). CI: confidence interval; CRIES: Crying Requires oxygen Increased vital signs Expression Sleep; MD: mean difference; NICU: neonatal intensive care unit; 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.
¹Downgraded two levels for study limitations (high risk of performance bias; unclear risk of bias for the other domains) and one level for imprecision (one small trial with low sample size; CIs overlapping no effect).

See Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.

Summary of findings 3. Opioids compared to other analgesics for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to other analgesics for procedural pain in neonates
Patient or population: procedural pain in neonates Setting: neonatal units Intervention: opioids Comparison: other analgesics
Outcomes	Risk with other analgesics	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the PIPP/PIPP‐R pain scale, ranging from 0 to 21 and 0 to 18 in preterm infants < 28 weeks gestational age and full‐term infants, respectively, during procedure	Mean PIPP/PIPP‐R during procedure ranged 5 to 7.	MD 0.29 lower (1.58 lower to 1.01 higher)	‐	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R during the procedure compared to other analgesics.
Pain, assessed using the PIPP/PIPP‐R pain scale up to 30 min after procedure	Mean PIPP/PIPP‐R up to 30 min after procedure was 5.	MD 1.1 lower (2.82 lower to 0.62 higher)	‐	12 (1 RCT)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R up to 30 min after the procedure compared to other analgesics.
Pain, assessed using the PIPP/PIPP‐R pain scale 1 to 2 hours after procedure	Mean PIPP/PIPP‐R 1 to 2 hours after procedure was 4.	MD 0.17 lower (2.22 lower to 1.88 higher)	‐	12 (1 RCT)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R 1 to 2 hours after the procedure compared to other analgesics.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of apnea—during the procedure	Study population		RR 3.27 (0.85 to 12.58)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on episodes of apnea during the procedure compared to other analgesics.
Episodes of apnea—during the procedure	33 per 1000	109 per 1000 (28 to 419)	RR 3.27 (0.85 to 12.58)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹
Episodes of apnea—after the procedure	Study population		RR 2.71 (0.11 to 64.96)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on episodes of apnea after the procedure compared to other analgesics.
Episodes of apnea—after the procedure	0 per 1000	0 per 1000 (0 to 0)	RR 2.71 (0.11 to 64.96)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹
Episodes of bradycardia—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Hypotension	Study population		RR 1.34 (0.32 to 5.59)	204 (3 RCTs)	⊕⊝⊝⊝ Very low ³	The evidence is very uncertain about the effect of opioids on hypotension compared to other analgesics.
Hypotension	20 per 1000	26 per 1000 (6 to 110)	RR 1.34 (0.32 to 5.59)	204 (3 RCTs)	⊕⊝⊝⊝ Very low ³
Parent satisfaction with care provided in the NICU—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*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). CI: confidence interval; NICU: neonatal intensive care unit; PIPP: Premature Infant Pain Profile; PIPP‐R: PIPP‐Revised; RCT: randomized controlled trial; RR: risk ratio
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.
¹Downgraded two levels for study limitations (high risk of performance bias; unclear risk of selection and other biases) and one level for imprecision (two small trials with low sample size; CI overlapping no effect). ²Downgraded one level for study limitations (unclear risk of selection, performance, and detection bias) and two levels for imprecision (one small trial with low sample size; CIs overlapping no effect). ³Downgraded two levels for study limitations (high risk of performance bias; unclear risk of selection and other biases) and one level for imprecision (three small trials with low sample size; CIs overlapping no effect).

Background

Description of the condition

A painful procedure can be defined as a procedure causing skin or mucosal damage by inserting or removing foreign bodies and disturbing the body integrity through therapeutic or diagnostic methods (Kassab 2019). Most newborn infants undergo painful procedures early in life, including routine therapeutic interventions (such as intramuscular vitamin K injection) and diagnostic testing (such as metabolic screening). Infants that require neonatal intensive care are subjected to numerous invasive and painful procedures, with as many as 14 to 16 painful procedures a day being reported (Courtois 2016; Johnston 2013). The most unwell and the most preterm infants are exposed to a higher number of painful procedures compared to older and healthier infants (Cruz 2016). This is unfortunate as preterm infants are especially vulnerable to the negative effects of pain due to their immature and still‐developing nervous system (Maxwell 2019).

Painful procedures have immediate negative physiological effects as well as long‐term negative effects such as altered pain processing, pain sensitivity, and response to pain (Ranger 2014); reduction in cortical thickness (Ranger 2013); and lower cognitive and motor function (Ranger 2014). Painful procedures can be diagnostic (such as venipuncture and eye screening for retinopathy of prematurity (ROP)), therapeutic (such as tracheal suctioning or bladder catheterization), or surgical (such as circumcision) (Williams 2020). The most common painful procedures that neonates undergo are nasal/tracheal aspiration and heel lance (Carbajal 2008; Cruz 2016), while the procedures that are considered the most painful by nurses and physicians are endotracheal intubation, lumbar puncture, and insertion of a chest tube (Andersen 2010). The likelihood of developing hypersensitivity or persistent pain, or both, later in life increases with the number of painful procedures experienced as a neonate (Williams 2020).

Description of the intervention

Interventions for pain management can be classified as non‐pharmacological or pharmacological. In the policy statement regarding the prevention and management of procedural pain in the neonate, the American Academy of Pediatrics (AAP) recommends the consistent use of non‐pharmacological strategies, coupled with pharmacological strategies when necessary (AAP 2016). Non‐pharmacological interventions like swaddling, positioning, skin‐to‐skin care (SSC), breastfeeding, oral sweet solutions, non‐nutritive sucking, multi‐sensory stimulation, and facilitated tucking have the advantage of not being associated with the short‐ and long‐term side effects caused by analgesic drugs, and have increasingly gained importance in the pain management of this most vulnerable population (Mehler 2013; van den 2016). These interventions to keep infants comfortable could be as effective as analgesics administered for painful procedures (Gomes Neto 2020; Shah 2012). They may be used alone or in adjunct with other interventions to address mild to moderate pain secondary to various procedures in the neonatal intensive care unit (NICU) (Squillaro 2019).

Pharmacological interventions signify the administration of analgesic drugs, of which opioids play a major role. Opioids are well‐known to provide both analgesia and sedation across all age groups, and administration of opioids has traditionally been the first choice for pain management in the NICU, where critically ill infants were often under moderate to severe pain as well as being exposed to numerous painful procedures (McKechnie 2008). However, the optimal regimen for the use of the many opioids available for different procedures is not completely understood. Recent studies have reported conflicting results toward the ongoing use of opioids to control pain: Hartley and colleagues reported that 0.1 mg/kg of body weight of oral morphine, given to non‐ventilated preterm infants one hour prior to an eye examination or heel lance, was not effective and may even be harmful (Hartley 2018), and Gitto and colleagues reported that 1 μg/kg of body weight to 2 μg/kg intravenous fentanyl or sensorial saturation given before every heel lance provided effective analgesia (Gitto 2012). Moreover, past studies have reported adverse effects of opioids in the smallest patients, which raises concerns for safety and emphasizes the need for establishing an effective dosing schedule with minimum side effects (Hartley 2018; Orsini 1996).

Of the opioids, morphine and fentanyl are the most commonly used, and therefore the most studied, in neonates. Other fentanyl derivatives (e.g. remifentanil, alfentanil, sufentanil) and other opioids (e.g. codeine, oxycodone, hydrocodone) are used more sporadically (Hall 2014; Thigpen 2019). Codeine is a prodrug (i.e. a substance that, after administration, is metabolized into a pharmacologically active drug) that is converted to morphine, with one‐tenth the potency of morphine; oxycodone and hydrocodone are structurally similar to codeine but do not need to be metabolized for action. Fentanyl is a purely synthetic opioid that is 50 to 100 times more potent than morphine, and sufentanil is more potent while alfentanil and remifentanil are shorter‐acting. Several administration routes are physiologically possible, but morphine is usually administered by intravenous and oral routes in the NICU, and fentanyl and its derivatives are usually administered intravenously (Thigpen 2019). Though classified into the same drug group, each opioid should be administered in an individualized manner based on the condition of the patient and the pharmacokinetic and pharmacodynamic profiles of the opioid. In addition, neonates ranging from preterm to term have differing liver metabolism and renal clearance, owing to the maturation of enzymes and physiologic processes over time (Tibboel 2005). Due to the historical fact that most drugs used in the NICU started off as 'off‐label' use of those drugs given to adults and the older pediatric population (Balan 2018; Krzyżaniak 2016), pharmacological data in small and preterm infants are still lacking and are continuously being updated (Norman 2019). Morphine has the longest time to onset and the longest half‐life and elimination time compared to fentanyl and remifentanil, while remifentanil exhibits rapid action and elimination with twice the potency of fentanyl (Thigpen 2019). Short‐term adverse effects of opioids include hypotension, bradycardia, respiratory depression, chest‐wall rigidity, gastrointestinal dysmotility, and urinary retention, while studies suggesting negative impact on the developing brain have raised concerns about the possible long‐term adverse effects of opioids (McPherson 2015; Zwicker 2016). Some studies indicate an association between the use of opioids and the development of intraventricular hemorrhage in premature infants (Khanafer‐Larocque 2019; Riskin 2015), while others show no evidence of this (Jiang 2012; McPherson 2015). There might also be a correlation between the use of opioids and necrotizing enterocolitis (Hällström 2003; Riskin 2015; Zvizdic 2019).

Neonatal pain management is challenging because physicians must find the balance between achieving sufficient pain control and minimizing oversedation and negative consequences of opioid use, therefore a multifactorial approach is called for. Non‐opioid alternatives (i.e. non‐pharmacological interventions and non‐opioid pharmacological agents) have emerged over the years as understanding of neonatal pain has improved and practice has evolved. Non‐opioid agents include paracetamol (acetaminophen), non‐steroidal anti‐inflammatory drugs (NSAIDs) (e.g. ketorolac, ibuprofen), and alpha 2‐agonists (e.g. clonidine, dexmedetomidine). Paracetamol has been used for its opioid‐sparing effects in treating mild to moderate pain in neonates, in which the pharmacokinetics and metabolism of this drug have been summarized, but recent reviews found that paracetamol failed to provide effective procedural pain management (Allegaert 2020; Ohlsson 2020). Although a Cochrane Review recently concluded that there is still insufficient evidence to recommend the use of clonidine for pain management—including prevention or treatment of procedural pain (Romantsik 2020)—alpha 2‐agonists such as clonidine are administered in combination with opioids to decrease their required doses for sedation and analgesia in critically ill children (Duffett 2012).

Results from the EUROpean Pain Audit In Neonates (EUROPAIN) prospective cohort study showed wide variations in the clinical practice of neonatal sedation and analgesia across institutions and countries (Carbajal 2015). Not all NICUs implement routine analgesia before daily heel pricks and venipunctures, and pain prevention and treatment among neonates is far from optimal (Bellieni 2018). Considering the ethical and practical challenges of accumulating evidence in the youngest patients, as well as the significant impact of using the evidence for their care, the current situation emphasizes the urgent need to organize the evidence at hand, investigate unanswered questions by well‐designed studies (Dotta 2011), and develop common guidelines, in order to ultimately manage neonatal pain in the safest and most effective evidence‐based way possible.

How the intervention might work

Opioids are used for the treatment of moderate to severe, acute, perioperative, and chronic pain in patients of all ages and provide both analgesia and sedation (Nafziger 2018). Moreover, they also attenuate physiological stress responses and have a wide therapeutic window (Nafziger 2018). The positive and adverse effects of opioids depend on their binding affinity to the different opioid G protein‐coupled receptors (mu, delta and kappa), which are present at virtually all neural loci related to pain in both the peripheral and central nervous systems. When a neonate requiring intensive care is exposed to repeated invasive procedures such as blood sampling and tracheal suctioning, continuous administration of low‐dose opioids or intermittent administration of boluses prior to each procedure may be beneficial (Anand 2005). However, caution has been called for on the neonatal use of opioids, with recommendations to only use them in cases when non‐pharmacological interventions are considered insufficient (Anand 2007).

Several findings have supported the positive impact of opioid use in managing procedural pain, even in the smallest patients. For example, it was shown that endogenous opioids, such as endorphins and mu‐receptor agonists, tend to reduce the stress response and produce stress adaptation by preventing overactivation of the hypothalamic–pituitary–adrenal axis (Bali 2015). A retrospective study found that intranasal midazolam and fentanyl were well‐tolerated in preterm and term infants requiring intensive care (Ku 2019). The effectiveness of opioids in relieving procedural pain has been well‐established, primarily starting from older patients, like any major drug. Sedoanalgesia with oral fentanyl citrate and midazolam has been shown to be highly effective in reducing pain during bone marrow aspiration and biopsy in adult patients with hematological malignancies (Cerchione 2020). Similarly, remifentanil, alfentanil, and midazolam were effective in reducing pain during bone marrow aspiration (Antmen 2005). The effectiveness of these drugs in the youngest patients has been studied vigorously over the years, accumulating evidence related especially to the main opioids, morphine and fentanyl. It has been reported that a single dose of fentanyl to ventilated preterm neonates decreased changes in heart rate and increased growth hormone levels, as well as decreasing behavioral measures of pain and stress (Guinsburg 1998). A number of studies have indicated that continuous morphine infusion to neonates during heel lances and endotracheal tube suctioning decreased pain responses compared to control patients (Anand 1999; Scott 1999). However, some recent studies, including the NEOPAIN (Neurologic Outcomes and Pre‐emptive Analgesia In Neonates) trial, have reported conflicting results regarding the effectiveness of morphine analgesia for acute procedural pain, which raises questions warranting further research (Carbajal 2005; Simons 2003). Results from the Poppi (PrOcedural Pain in Premature Infants) randomized controlled trial (a study investigating the effectiveness of oral morphine to infants before procedures) have provided more reason to be cautious about the potential adverse effects of opioids (Monk 2019).

Why it is important to do this review

Safe and effective management of procedural pain is important for humanitarian and ethical reasons but also to minimize the detrimental effects of repeated pain. Pain should be assessed with validated methods for the specific type of pain and newborn infant undergoing the procedural pain (Giordano 2019; Olsson 2021). A Cochrane Review on the validity and reliability of neonatal pain scales is in preparation (Bruschettini 2022).

Cochrane Reviews on opioid administration in newborn infants for postoperative pain (Kinoshita 2021; Kinoshita 2023), sedation during mechanical ventilation (Bellù 2021), pain or sedation management during therapeutic hypothermia (Bäcke 2022), elective endotracheal intubation (Ayed 2017), and the prevention of pain during endotracheal suctioning (Pirlotte 2019), are available or currently underway. However, no systematic reviews have been conducted on opioids for procedural pain. Of note, painful procedures and inadequate pain management in early life may lead to long‐term negative effects (Walker 2019). It is therefore important to synthesize and appraise the available evidence on opioids for procedural pain in neonates.

Objectives

Methods

Criteria for considering studies for this review

Types of studies

We included prospective randomized controlled trials (RCTs), quasi‐RCTs, cluster‐RCTs, and cross‐over RCTs.

Types of participants

We included preterm and term infants of a postmenstrual age (PMA) up to 46 weeks and 0 days—irrespective of their gestational age at birth—receiving opioids for procedural pain such as during dialysis, extracorporeal membrane oxygenation (ECMO) treatment, before screening for ROP, placement of Broviac catheter, air leak drainage, insertion of a central line, heel lance, lumbar puncture, venipuncture, arterial line placement, and any other painful procedures.

We excluded infants:

receiving opioids during mechanical ventilation for respiratory morbidity (assessed in a separate Cochrane Review, Bellù 2021);
receiving opioids pre‐intubation (assessed in a separate Cochrane Review, Ayed 2017);
undergoing endotracheal suctioning (assessed in a separate Cochrane Review, Pirlotte 2019);
receiving opioids for postoperative pain (assessed in separate Cochrane Reviews, Kinoshita 2021; Kinoshita 2023);
treated for neonatal abstinence syndrome (assessed in a separate Cochrane Review, Osborn 2021);
undergoing therapeutic hypothermia (assessed in a separate Cochrane Review, Bäcke 2022);
undergoing invasive procedures during the postoperative period and in other excluded conditions.

Types of interventions

We included studies on any opioids (e.g. morphine, diamorphine, fentanyl, alfentanil, sufentanil, pethidine, meperidine, codeine) for procedural pain. We included any systemic route of administration. We included the following comparisons.

Comparison 1: opioids versus no treatment or placebo.
Comparison 2: opioids versus oral sweet solution or non‐pharmacological intervention (skin‐to‐skin contact, music exposure, non‐nutritive sucking, swaddling, etc.).
Comparison 3: opioids versus other analgesics (e.g. paracetamol) and sedatives (e.g. midazolam and other benzodiazepines).
Comparison 4: head‐to‐head comparison of different opioids.
Comparison 5: different routes for administration of the same opioid.

Types of outcome measures

The following outcome measures did not form part of the eligibility criteria.

Primary outcomes

Pain assessed with the following scales: ABC scale (Bellieni 2005); Bernese Pain Scale for Neonates (Cignacco 2004); Behavioral Indicators of Infant Pain (BIIP) (Holsti 2008); Douleur Aiguë du Nouveau‐né (DAN) (Acute Pain in Newborn infants, APN, English version) (Carbajal 1997); Neonatal Infant Pain Scale (NIPS) (Lawrence 1993); Neonatal Pain, Agitation, and Sedation Scale (N‐PASS) (Hummel 2008); Premature Infant Pain Profile (PIPP)/PIPP‐Revised (PIPP‐R) (Gibbins 2014; Stevens 1996). We planned to report the median and mean values of each pain scale assessed during the procedure; up to 30 minutes after the procedure; and at one to two hours after the procedure.
Any harms.

Secondary outcomes

All‐cause neonatal mortality (death until postnatal day 28).
All‐cause mortality during initial hospitalization.
Use of additional pharmacological intervention for the relief of procedural pain.
Episodes of bradycardia, defined as a fall in heart rate of more than 30% below the baseline or less than 100 beats per minute for 10 seconds or longer.
Episodes of desaturation, defined as a decrease of arterial oxygen saturation (SpO_²) to less than 80%, with no minimum duration specified.
Episodes of apnea (mean rates of apnea).
Hypotension requiring medical therapy (vasopressors or fluid boluses).
Parent satisfaction with care provided in the NICU (as measured by a validated instrument/tool) (Butt 2013).
Intraventricular hemorrhage (IVH; all (grade 1 or 2) or severe (grade 3 or greater) on cranial ultrasound, according to Papile classification (Papile 1978)).
Necrotizing enterocolitis (NEC) (modified Bell stage 2/3; Walsh 1986).
Constipation during the course of treatment, defined as a delay in defecation sufficient to cause significant distress to the infant.
Major neurodevelopmental disability: cerebral palsy, developmental delay (Bayley Scales of Infant Development ‐ Mental Development Index Edition II (BSID‐MDI‐II; Bayley 1993), Bayley Scales of Infant and Toddler Development ‐ Edition III Cognitive Scale (BSITD‐III) (Bayley 2005)), or Griffiths Mental Development Scale ‐ General Cognitive Index (GCI) (Griffiths 1954; Griffiths 1970), assessment greater than two standard deviations (SDs) below the mean, intellectual impairment (intelligence quotient (IQ) greater than two SDs below the mean), blindness (vision less than 6/60 in both eyes), or sensorineural deafness requiring amplification (Jacobs 2013). We planned to separately assess data on children aged 18 to 24 months and those aged three to five years.
Cognitive and educational outcomes in children aged more than five years old.

Search methods for identification of studies

We conducted the searches in December 2021. The Cochrane Sweden and Cochrane Neonatal Information Specialists developed a draft search strategy for Ovid MEDLINE in consultation with the review authors (Appendix 1). This strategy was peer reviewed by an Information Specialist using the Peer Review of Electronic Search Strategies (PRESS) checklist (McGowan 2016a; McGowan 2016b). The MEDLINE strategy was translated, using appropriate syntax, for other databases. Methodological filters based on those developed by Cochrane, Lefebvre 2021; RCT‐Filter EMBASE, and the Canadian Agency for Drugs and Technologies in Health (CADTH), CADTH 2016, were used to limit retrieval to RCTs and quasi‐RCTs, and systematic reviews. We conducted the searches without restriction on language, publication year, publication type, or publication status.

The timeline for this publication was disrupted by the COVID‐19 pandemic and staffing issues at the Cochrane Neonatal editorial base. As a result, publication of this review has been delayed, and the literature search is more than one year old. We will endeavor to undertake an updated search within the next calendar year.

Electronic searches

We searched the following databases.

Cochrane Central Register of Controlled Trials (CENTRAL), via Wiley (2021, Issue 12) (16 December 2021)
MEDLINE via PubMed (1966 to 16 December 2021)
Embase, via Elsevier (1974 to 16 December 2021)
CINAHL Complete, via EBSCOhost (1982 to 16 December 2021)

Searching other resources

We identified trial registration records using CENTRAL and by independent searches of the US National Library of Medicine ClinicalTrials.gov (clinicaltrials.gov/) and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) on 17 December 2021. We identified conference abstracts using CENTRAL, Embase, and the Eastern Society for Pediatric Research (ESPR) (2019; 2018).

We screened the reference lists of included studies for studies not identified by the database searches. We searched for errata or retractions for included studies published on PubMed (ncbi.nlm.nih.gov/pubmed). In addition to searching for related systematic reviews via databases, we searched the Epistemonikos registry of systematic reviews (epistemonikos.org).

We conducted a grey literature search to identify reports of trials conducted by or referenced in research by agencies or organizations. We identified sources by consulting the Technical Supplement of the searching chapter in the Cochrane Handbook for Systematic Reviews of Interventions (Lefebvre 2021).

Data collection and analysis

We collected information regarding the method of randomization, blinding, intervention, stratification, and whether the trial was single‐ or multicenter for each included study. We noted information regarding trial participants, including birthweight, gestational age, number of participants, type of procedural pain, modality of administration, and dose of opioids. We analyzed the clinical outcomes noted above in Types of outcome measures.

Selection of studies

We used Cochrane’s Screen4Me workflow to help assess the search results. Screen4Me comprises three components: known assessments—a service that matches records in the search results to records that have already been screened in Cochrane Crowd and been labeled as an RCT or as Not an RCT; the RCT classifier—a machine learning model that distinguishes RCTs from non‐RCTs; and, if appropriate, Cochrane Crowd—Cochrane’s citizen science platform where the Crowd help to identify and describe health evidence.

For more information about Screen4Me and the evaluations that have been done, please visit the Screen4Me webpage on the Cochrane Information Specialist’s portal. In addition, more detailed information regarding evaluations of the Screen4Me components can be found in the following publications: Marshall 2018; Noel‐Storr 2020; Noel‐Storr 2021; Thomas 2020.

We included all RCTs, quasi‐RCTs, and cluster‐RCTs fulfilling our inclusion criteria. Two review authors (EO, FB) reviewed the results of the search and separately selected studies for inclusion. Any disagreements were resolved by discussion or by involving a third review author when necessary. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and Characteristics of excluded studies table (Moher 2009).

Data extraction and management

Two review authors (EO, FB) independently extracted data using a data extraction form integrated with a modified version of the Cochrane Effective Practice and Organisation of Care Group data collection checklist (Cochrane EPOC Group 2017). We piloted the form within the review team, using a sample of included studies. We extracted the following characteristics from each included study.

Administrative details: study author(s); published or unpublished; year of publication; year in which study was conducted; presence of vested interest; details of other relevant papers cited.
Study: study design; type, duration, and completeness of follow‐up (e.g. greater than 80%); country and location of study; informed consent; ethics approval.
Participants: sex, birthweight, gestational age, number of participants.
Interventions: initiation, dose, and duration of opioids administration.
Outcomes as mentioned above in Types of outcome measures.

Any disagreements were resolved by discussion. We described any ongoing studies identified by our search, detailing the primary author, research question(s), methods, and outcome measures, together with an estimate of the reporting date, in the Characteristics of ongoing studies table.

In the case of queries or where additional data were required, we contacted study investigators/authors for clarification. Two review authors (MB, MK) used Cochrane statistical software Review Manager 5 for data entry (Review Manager 2020). We replaced any standard error of the mean (SEM) by the corresponding SD.

Assessment of risk of bias in included studies

Two review authors (EO, FB) independently assessed the risk of bias (low, high, or unclear) of the included trials using the Cochrane risk of bias tool for the following domains (Higgins 2011).

Random sequence generation (selection bias)
Allocation concealment (selection bias)
Blinding of participants and personnel (performance bias)
Blinding of outcome assessment (detection bias)
Incomplete outcome data (attrition bias)
Selective reporting (reporting bias)
Any other bias

Any disagreements were resolved by discussion or by consulting a third review author (MK). See Appendix 2 for a more detailed description of risk of bias for each domain. We assessed overall risk of bias according to three categories, as follows.

Low risk of bias: we classified the outcome result of a trial as being at low risk of bias overall only if all domains were classified as being at low risk of bias.
Unclear risk of bias: we classified the outcome result of a trial as being at unclear risk of bias overall if one or more domains were classified as being at unclear risk of bias, and no domain was at high risk of bias.
High risk of bias: we classified the outcome result of a trial as being at high risk of bias overall if at least one domain was classified as being at high risk of bias.

Measures of treatment effect

We performed the statistical analyses using Review Manager 5 (Review Manager 2020). We summarized the data in a meta‐analysis if they were sufficiently homogeneous, both clinically and statistically.

Dichotomous data

For dichotomous data, we presented results using risk ratios (RRs) and risk differences (RDs) with 95% confidence intervals (CIs). We calculated the number needed to treat for an additional beneficial outcome (NNTB), or number needed to treat for an additional harmful outcome (NNTH) with 95% CIs if there was a statistically significant reduction (or increase) in RD.

Continuous data

For continuous data, we used the mean difference (MD) when outcomes were measured in the same way between trials. We used the standardized mean difference (SMD) to combine trials that measured the same outcome but used different methods. However, we did not pool in the same analysis pain scores assessed with different scales. Where trials reported continuous data as median and interquartile range (IQR) and data passed the test of skewness, we converted median to mean and estimated the SD as IQR/1.35.

Unit of analysis issues

The unit of analysis was the participating infant in individually randomized trials, and an infant was considered only once in the analysis. The participating neonatal unit or section of a neonatal unit or hospital was to be the unit of analysis in cluster‐randomized trials. We planned to analyze these using an estimate of the intracluster correlation coefficient (ICC) derived from the trial (if possible), or from a similar trial or from a study with a similar population as described in Section 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2020). If we were to use ICCs from a similar trial or from a study with a similar population, we would report this and conduct a sensitivity analysis to investigate the effect of variation in the ICC. In the end, no cluster‐randomized trial was included.

Had we identified both cluster‐randomized trials and individually randomized trials, we would only combine the results from both if there was little heterogeneity between the study designs, and if the interaction between the effect of the intervention and the choice of randomization unit was considered to be unlikely; however, in the end, no cluster‐randomized trial was included. In the event that we identified cross‐over trials, in which the reporting of continuous outcome data precluded paired analysis, we would not include these data in a meta‐analysis, in order to avoid unit of analysis error. Where carry‐over effects were thought to exist, and where sufficient data existed, we would only include data from the first period in the analysis (Higgins 2021). We planned to acknowledge any possible heterogeneity in the randomization unit and perform a sensitivity analysis to investigate possible effects of the randomization unit. However, no cross‐over trials were included.

Dealing with missing data

Where feasible, we carried out analysis on an intention‐to‐treat basis for all outcomes. Whenever possible, we analyzed all participants in the treatment group to which they had been randomized, regardless of the actual treatment received. If we identified important missing data (in the outcomes), or unclear data, we contacted the original investigators to request the missing data. If a trial contained a mixed population (i.e. postoperative and non‐operative infants were combined in the report), we first assessed whether subgroup results for non‐operative infants were reported. If not, we contacted the trial authors. If results were not available, we included all the trial data if non‐operative infants made up 50% or more of the total trial population. We planned to carry out a sensitivity analysis to assess the impact of including studies with mixed populations, if we were unable to get the subgroup data from trialists; in the end, we were able to get the trial results upon contact with the authors.

We made explicit the assumptions of any methods used to deal with missing data. We planned to perform sensitivity analyses to assess how sensitive results were to reasonable changes in the undertaken assumptions. We also planned to address the potential impact of missing data on the findings of the review in the Discussion section. Ultimately, there were no missing data.

Assessment of heterogeneity

We estimated the treatment effects of individual trials and examined heterogeneity among trials by inspecting the forest plots and quantifying the impact of heterogeneity using the I² statistic (Deeks 2020). We graded the degree of heterogeneity using the following parameters:

0% to 40%: might not represent important heterogeneity;
30% to 60%: may represent moderate heterogeneity;
50% to 90%: may represent substantial heterogeneity;
more than 75%: may represent considerable heterogeneity.

If we noted statistical heterogeneity (indicated by an I² value greater than 50%), we explored the possible causes (e.g. differences in study quality, participants, intervention regimens, or outcome assessments) and considered conducting sensitivity analysis (see Sensitivity analysis).

Assessment of reporting biases

We created and examined a funnel plot to explore possible small‐study biases. When interpreting funnel plots, we examined the different possible reasons for funnel plot asymmetry, as outlined in Section 10 of the Cochrane Handbook for Systematic Reviews of Interventions (Sterne 2017), and related this to the results of the review. We planned that if we were able to pool more than 10 trials, we would undertake formal statistical tests to investigate funnel plot asymmetry (Sterne 2017); however, this was precluded by the number of included studies.

To assess outcome reporting bias, we checked trial protocols against published reports. For studies published after 1 July 2005, we screened the WHO ICTRP for the a priori trial protocol. We evaluated whether selective reporting of outcomes were present.

Data synthesis

We used a fixed‐effect model to combine data where it was reasonable to assume that studies were estimating the same underlying treatment effect. If we judged meta‐analysis to be inappropriate, we analyzed and interpreted individual trials separately. If there was evidence of clinical heterogeneity, we attempted to explain this based on the different study characteristics and subgroup analyses.

Subgroup analysis and investigation of heterogeneity

Tests for subgroup differences in effects need to be interpreted with caution given the potential for confounding with other study characteristics and the observational nature of the comparisons (Deeks 2022). In particular, subgroup analyses with fewer than five studies per category are unlikely to be adequate to ascertain valid difference in effects and would not be highlighted in our results. We conducted stratified meta‐analysis and a formal statistical test for interaction to examine subgroup differences that could account for effect heterogeneity (e.g. Cochran’s Q test, meta‐regression) (Borenstein 2013; Higgins 2020).

Given the potential differences in the intervention effectiveness related to gestational age (extremely preterm infants are more vulnerable), type and route of opioids administration (which might affect the outcomes), presence of co‐interventions (which might interact with opioids), we planned to conduct subgroup comparisons to see if the intervention is more effective for the following groups for subgroup analysis where data were available. We restricted these analyses to the primary outcomes.

Gestational age: term infants (37 weeks' gestation or greater); preterm infants (less than 37 weeks' gestation); extreme preterm (less than 28 weeks' gestation).
Type of administration: with or without loading dose; bolus or continuous infusion.
Route of administration: enteral or intravenous; between other different routes.
With or without other pharmacological sedation/analgesia as co‐interventions.
Within studies that included co‐interventions: studies in which the protocol allowed co‐interventions for sedation/analgesia for one or both of the intervention groups; studies in which the protocol mandated sedation/analgesia with co‐interventions.
Participants with specific clinical conditions, e.g. infants undergoing dialysis or extracorporeal membrane oxygenation.

Sensitivity analysis

Where we identified substantial heterogeneity, we would conduct sensitivity analysis to determine if the findings were affected by inclusion of only those trials considered to have used adequate methodology with a low risk of bias selection and performance bias. We reported results of sensitivity analyses for primary outcomes only.

Summary of findings and assessment of the certainty of the evidence

We used the GRADE approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the certainty of evidence for the following (clinically relevant) outcomes.

Pain assessed with the following scales: ABC scale (Bellieni 2005); Bernese Pain Scale for Neonates (Cignacco 2004); BIIP (Holsti 2008); DAN (Carbajal 1997); NIPS (Lawrence 1993); N‐PASS (Hummel 2008); PIPP/PIPP‐R (Gibbins 2014; Stevens 1996):
- during the procedure;
- up to 30 minutes after the procedure;
- at one to two hours after the procedure.
Any harms.
Episodes of bradycardia, defined as a fall in heart rate of more than 30% below the baseline or less than 100 beats per minute for 10 seconds or longer.
Episodes of apnea (mean rates of apnea).
Hypotension requiring medical therapy (vasopressors or fluid boluses).
Parent satisfaction with care provided in the NICU (as measured by a validated instrument/tool) (Butt 2013).

Two review authors (MK, MB) independently assessed the certainty of the evidence for each of the outcomes above. We planned to include a summary of findings table for each of the five comparisons specified in Types of interventions; however, we could include only three (summary of findings Table 1; summary of findings Table 2; summary of findings Table 3) because no studies were included for the other two comparisons. We considered evidence from RCTs as high certainty, and downgraded the evidence by one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates, and presence of publication bias. We used GRADEpro GDT software to create a summary of findings table to report the certainty of the evidence (GRADEpro GDT).

The GRADE approach resulted in an assessment of the certainty of a body of evidence in one of the following four grades.

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 the 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.
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.

Results

Description of studies

Results of the search

The search identified a total of 6974 references (6919 via databases; 55 via other methods). After removal of 322 duplicates, 6652 records remained. These 6652 records were evaluated using Cochrane’s Screen4Me workflow The results of the Screen4Me assessment process are shown in Figure 1. Screen4Me eliminated 3367 records; the remaining 3285 records were screened by the review authors. We excluded 3250 records based on title/abstract and reviewed 35 full texts. We included 13 studies (Characteristics of included studies); excluded 12 studies (Characteristics of excluded studies); classified 3 studies as awaiting classification (Characteristics of studies awaiting classification); and identified 7 ongoing studies (Characteristics of ongoing studies). Details of study selection are presented in Figure 2.

Figure 1

Screen4Me summary diagram.

Figure 2

Study flow diagram.

Included studies

We included 13 independent studies (enrolling 823 newborn infants), considering the study by Madathil and colleagues as two separate studies (Madathil 2021a; Madathil 2021b), because the doses of fentanyl and ketamine were increased after the first 97 newborn infants were enrolled. For detailed information, see Characteristics of included studies and Table 1.

See: Summary of findings 1 Opioids compared to no treatment/placebo for procedural pain in neonates; Summary of findings 2 Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates; Summary of findings 3 Opioids compared to other analgesics for procedural pain in neonates

Table 1. Overview of included studies, listed by type of comparison

Author	Country	No. of infants in intervention and control group, respectively	GA in intervention and control group, respectively	Procedure	Intervention	Comparison
Comparison 1: Opioids versus no treatment or placebo
Carbajal 2005	France, USA	21/21	27.3 (1.8)/27.2 (1.7)^a	Heel lances	Morphine 100 μg/kg loading dose and 10 to 30 μg/kg/h continuous infusion (intravenously)	Placebo 5% dextrose (intravenously)
Hartley 2018	UK	15/16	28.1 (26.3 to 30.1)/28.6 (27.9 to 29.7)^b	Heel lances and ROP screening examination	Morphine 100 μg/kg single dose (orally)	Placebo (orally)
Manjunatha 2009	UK	6/6	NR	ROP screening examination	Morphine 200 μg/kg single dose (orally)	Placebo (orally)
Fallah 2016	Iran	23/22	37.4 (0.8)/37.6 (0.6)^a	Lumbar puncture	Fentanyl 2 μg/kg single dose (intravenously)	Placebo normal saline (intravenously)
Sindhur 2020	India	56/55	30.7 (1.7)/31.0 (1.7)^a	ROP screening	Fentanyl 2 μg/kg single dose (intranasal)	Placebo normal saline (intranasally)
Lago 2008	Italy	27/27	28 (2)/29 (2)^a	PICC insertion	Remifentanil 0.03 μg/kg/min continuous infusion (intravenously)	Placebo 5% dextrose continuous infusion (intravenously)
Pokela 1994	Finland	42/42	31.6 (25 to 40)/32.9 (24 to 41)^c	Daily routine care procedures^d and tracheal suction	Meperidine 1 mg/kg single dose (intravenously)	0.9% saline single dose (intravenously)
Comparison 2: Opioids versus oral sweet solution or non‐pharmacological intervention (skin‐to‐skin contact, music exposure, non‐nutritive sucking, swaddling, etc.)
Gitto 2012	Italy	50/50/50	NR	Heel lances	Fentanyl 1 to 2 μg/kg bolus injection (intravenously)	Facilitated tucking/sensorial saturation
Sethi 2020	India	29/29	30.3 (2.2)/30.3 (2.4)^a	Laser for ROP	Fentanyl 1 μg/kg/h continuous infusion (intravenously)	24% oral sucrose single dose
Comparison 3: Opioids versus other analgesics (e.g. paracetamol) and sedatives (e.g. midazolam and other benzodiazepines)
Manjunatha 2009	UK	6/6	NR	ROP screening examination	Morphine 200 μg/kg single dose (orally)	Paracetamol 20 mg/kg single dose (orally)
Cordero 1991	USA	15/14	28 (2)/27 (2)^a	Broviac catheter placement	Fentanyl 2 μg/kg single dose (intravenously)	Secobarbital 1 mg/kg single dose (intravenously)
Madathil 2021a	India	51/46	29.7 (1.9)/29.8 (1.5)^a	Laser for ROP	Fentanyl 2 μg/kg followed by a continuous infusion of 1 μg/kg/h increased to a maximum of 3 µg/kg/h (intravenously)	Ketamine 0.5 mg/kg, followed by further intermittent intravenous bolus doses of 0.5 mg/kg to a maximum of 2 mg/kg (intravenously)
Madathil 2021b	India	13/14	30.3 (1.3)/30.5 (2.4)^a	Laser for ROP	Fentanyl 2 μg/kg followed by infusion of 2 μg/kg/h to a maximum of 5 μg/kg/h (intravenously)	Ketamine 1 mg/kg followed by intermittent bolus doses of 0.5 mg/kg to a maximum of 4 mg/kg (intravenously)
Taddio 2006	Canada	38/42	29.6 (4.9)/30 (5.1)^a	PICC placement	Morphine 100 μg/kg single dose (intravenously)	Tetracaine 0.5 g 4% gel applied to the insertion site
We included no studies for the following comparisons: head‐to‐head comparison of different opioids; different routes of administration of the same opioid.

GA: gestational age; NR: not reported; PICC: peripherally inserted central catheter; ROP: retinopathy of prematurity
^aMean (standard deviation).
^bMedian (interquartile range).
^cMean (range).
^dWeighing, washing, temperature measurement, chest roentgenogram.

All studies were performed in a hospital setting, most at a NICU (Carbajal 2005; Fallah 2016; Gitto 2012; Lago 2008; Manjunatha 2009; Pokela 1994; Sindhur 2020; Taddio 2006), two at a neonatal unit (Hartley 2018; Sethi 2020), one in a regional perinatal center (Cordero 1991), and one at a pediatric high‐dependency unit (Madathil 2021a; Madathil 2021b).

We pooled the included studies in three separate comparisons: seven studies in the comparison opioids versus no treatment or placebo (Carbajal 2005; Fallah 2016; Hartley 2018; Lago 2008; Manjunatha 2009; Pokela 1994; Sindhur 2020); two studies in the comparison opioids versus oral sweet solution or non‐pharmacological intervention (Gitto 2012; Sethi 2020); and five studies in the comparison opioids versus other analgesics (e.g. paracetamol) and sedatives (Cordero 1991; Madathil 2021a; Madathil 2021b; Manjunatha 2009; Taddio 2006). Manjunatha 2009 had three groups (morphine, placebo, paracetamol) and is included in both the first comparison (morphine versus placebo) and third comparison (morphine versus paracetamol). Overall, fentanyl and morphine were used in seven studies (Cordero 1991; Fallah 2016; Gitto 2012; Madathil 2021a; Madathil 2021b; Sethi 2020; Sindhur 2020), and four studies (Carbajal 2005; Hartley 2018; Manjunatha 2009; Taddio 2006), respectively; remifentanil, Lago 2008, and meperidine, Pokela 1994, were used in the remaining two studies. Among the two studies in the comparison opioids versus oral sweet solution or non‐pharmacological intervention, the comparator was facilitated tucking/sensorial saturation in Gitto 2012 and 24% oral sucrose in Sethi 2020.

The sample size of the studies ranged from 12 infants, Manjunatha 2009, to 150 infants, Gitto 2012. One study done was stopped after 31 infants were recruited, because predefined stopping boundary was crossed due to occurrence of adverse events in intervention group (Hartley 2018). Most studies enrolled mainly preterm infants; Table 1 reports the values for each study, which were reported as either mean with SD, mean with range, or median with IQR. Four studies were conducted in India (Madathil 2021a; Madathil 2021b; Sethi 2020; Sindhur 2020), two in Italy (Gitto 2012; Lago 2008), two in the UK (Hartley 2018; Manjunatha 2009), and one each in Canada (Taddio 2006), Finland (Pokela 1994), Iran (Fallah 2016), and the USA (Cordero 1991). Carbajal 2005 was an international multicenter study conducted in France and the USA.

Opioids were administered for different indications: laser for ROP or screening for ROP in five studies (Madathil 2021a; Madathil 2021b; Manjunatha 2009; Sethi 2020; Sindhur 2020), central catheter placement in three studies (Cordero 1991; Lago 2008; Taddio 2006), heel stick or heel lance in two studies (Carbajal 2005; Gitto 2012), routine care procedures and tracheal suction in one study (Pokela 1994), and lumbar puncture in one study (Fallah 2016), and in one study, opioids were administered for heel lance and ROP screening examination (Hartley 2018).

No funding was reported for the studies by Cordero 1991, Gitto 2012, Lago 2008, Madathil 2021a, Madathil 2021b, Manjunatha 2009, and Sindhur 2020. Carbajal 2005 received funds from the Foundation CNP and National Institute for Child Health and Human Development grants HD36484 and HD36270. Hartley 2018 received funds from the Wellcome Trust and National Institute for Health Research. Fallah 2016 received a grant from Deputy for Research of Shahid Sadoughi University of Medical Sciences, Yazd, Iran. Pokela 1994 received funding from Foundation for Paediatric Research in Finland, Helsinki and the Alma and K.A. Snellman Foundation, Oulu, Finland. In the study by Sethi 2020, Prof Velpandian, Department of Ocular Pharmacy, AIIMS, New Delhi and Mr Ujjwal provided the 24% oral glucose. Taddio 2006 received funding from the Canadian Society of Hospital Pharmacists, the Canadian Institutes of Health Research New Investigator Award, and the Ontario Student Opportunity Trust Fund ‐ Hospital for Sick Children Foundation Student Scholarship Program.

Ongoing studies

We identified seven ongoing studies (see Characteristics of ongoing studies): five on fentanyl (CTRI/2017/12/011035; CTRI/2020/08/027144; NCT02125201; NCT03718507; NCT03735563), one on morphine (CTRI/2018/04/012926), and one on remifentanil (NCT04073173).

Indications for opioids administration were less invasive surfactant administration (LISA) in four studies, therapy or screening for ROP in two studies, and overall pain prevention in one study.

Studies awaiting classification

We assessed three studies as awaiting classification (ACTRN12612000385842; Gadzinowski 2000; Li 1997); see Characteristics of studies awaiting classification.

Abstract and full text were not available for two of these studies (Gadzinowski 2000; Li 1997). ACTRN12612000385842 is an RCT comparing intravenous infusion of remifentanil with placebo in term and preterm neonates requiring insertion of a central venous catheter. The protocol was registered in 2012; results are not available.

Excluded studies

We excluded 12 studies at full‐text stage; see Characteristics of excluded studies.

Reasons for exclusion were either study design, patient population, or comparator not matching the inclusion criteria of this review, that is:

study design in six studies: Bell 2019 and NCT03897452 are phase II trials; Campbell‐Yeo 2018 and Soffer 2019 are commentaries; Moustogiannis 1996 is not an RCT; and in Axelin 2009 opioids was administered at the very end, without randomization;
patient population in four studies: Chambers 2002 is conducted on surgical pain; NCT00571636 and Valkenburg 2015 on sedation during mechanical ventilation; and Rosen 2000 in infants older than one month;
comparator in two studies that compared high versus low dose of the same opioid (Shin 2013; Shin 2014).

Risk of bias in included studies

The overall risk of bias assessment for each study, including all domain evaluations and justifications for judgement, is displayed in Figure 3 and Figure 4, as well as in the risk of bias section (Characteristics of included studies), and on the right side of all forest plots. The overall quality of studies was good (Figure 3), as nearly two‐thirds of the studies had low risk of bias for each domain in the Cochrane risk of bias tool. The only exception was blinding of participants and personnel, for which about a third of the studies were assessed as high risk of bias.

Figure 3

Risk of bias graph.

Figure 4

Risk of bias summary.

Allocation

The majority of included studies provided details on random sequence generation and allocation concealment. Madathil 2021a; Madathil 2021b; Manjunatha 2009 described that infants were randomized and envelopes were used to conceal allocation, but without further details regarding the specific method of randomization. In contrast, Hartley 2018 stated that a "web‐based facility hosted by the NPEU CTU (National Perinatal Epidemiology Unit Clinical Trials Unit)" was used to randomize infants, but failed to explain how the treatment allocation was masked. Cordero 1991 and Gitto 2012 did not provide sufficient details about study allocation.

Blinding

Blinding of caregivers and assessors to the intervention was established in more than half (7 out of 13) of the included studies; blinding of outcome assessors was described in all but two studies (Gitto 2012; Manjunatha 2009). An exception was Cordero 1991, where blinding was not described but was judged as at low risk of bias for blinding of outcome assessors, given that all study outcomes were objective measurements. The seven double‐blinded trials usually had a pharmacist or a nurse not involved in clinical care prepare the drugs, thereby ensuring the blinding of the clinicians and nurses assisting the infants.

While two studies did not sufficiently explain how the study participants were blinded and were thus judged as unclear risk of bias (Cordero 1991; Manjunatha 2009), four studies could not blind participants due to the different types of intervention (e.g. continuous infusion versus intermittent bolus) and were thus judged as high risk of bias (Gitto 2012; Madathil 2021a; Madathil 2021b; Sethi 2020).

Incomplete outcome data

Follow‐up and outcome data were complete for all studies except Gitto 2012. In Gitto 2012, 150 infants were randomized to receive either intravenous fentanyl, facilitated tucking, or sensorial stimulation, and it was unclear whether secondary outcomes (i.e. levels of cytokines during painful procedures) were reported for all 150 infants.

Selective reporting

Two‐thirds of the included studies reported the trial registration number, and there were no relevant differences between outcomes in the study protocol and those reported in the published article. For five studies a protocol was not available; these studies were judged to be at unclear risk of bias (Carbajal 2005; Cordero 1991; Gitto 2012; Lago 2008; Pokela 1994).

Other potential sources of bias

In Madathil 2021a and Madathil 2021b, the drug regimens were revised upon recommendation from the study steering committee with regard to inadequate analgesia, thus the study was divided into two separate phases, each with smaller numbers of infants than initially planned.

In Pokela 1994, the pain score was modified in the study to be used with intubated neonates, and needs to be further improved and validated.

Effects of interventions

Comparison 1: Opioids versus no treatment or placebo

We included seven studies in this comparison (Carbajal 2005; Fallah 2016; Hartley 2018; Lago 2008; Manjunatha 2009; Pokela 1994; Sindhur 2020). Certainty of the evidence is reported for the seven outcomes specified for the summary of findings table (summary of findings Table 1).

Primary outcomes

Pain assessed with the following scales

PIPP/PIPP‐R during procedure

Three trials reported this outcome (Hartley 2018; Lago 2008; Sindhur 2020). Opioids probably reduce pain score assessed with the PIPP/PIPP‐R scale during the procedure compared to placebo (mean difference (MD) −2.58, 95% confidence interval (CI) −3.12 to −2.03; 199 participants, 3 studies; I² = 62%; moderate‐certainty evidence; Analysis 1.1). As Hartley 2018 reported pain scores following two procedures (ROP screening and heel stick), we halved the number of randomized infants reported to avoid double counting in the meta‐analysis.

PIPP/PIPP‐R up to 30 minutes after procedure

Two trials reported this outcome (Manjunatha 2009; Sindhur 2020). The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale up to 30 minutes after the procedure compared to placebo (MD 0.14, 95% CI −0.17 to 0.45; 123 participants, 2 studies; I² = 85%; very low‐certainty evidence; Analysis 1.2).

PIPP/PIPP‐R one to two hours after procedure

Two trials reported this outcome (Carbajal 2005; Manjunatha 2009). The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale one to two hours after the procedure compared to placebo (MD −0.83, 95% CI −2.42 to 0.75; 54 participants, 2 studies; I² = 47%; very low‐certainty evidence; Analysis 1.3).

DAN one to two hours after procedure

One trial reported this outcome (Carbajal 2005). Opioids may result in little to no difference in pain score assessed with the DAN scale one to two hours after the procedure compared to placebo (MD −0.20, 95% CI −2.21 to 1.81; 42 participants, 1 study; I² not applicable; low‐certainty evidence; Analysis 1.4).

NIPS during procedure

Two trials reported this outcome (Fallah 2016; Lago 2008). Opioids may reduce NIPS during the procedure compared to placebo (MD −1.97, 95% CI −2.46 to −1.48; 102 participants, 2 studies; I² = 60%; low‐certainty evidence; Analysis 1.5).

Any harms

No trials reported this outcome.

Secondary outcomes

Mortality

Two trials reported this outcome, with no events (Fallah 2016; Lago 2008). Opioids may result in little to no difference in mortality compared to placebo (risk ratio (RR) not estimable, risk difference (RD) 0.00, 95% CI −0.05 to 0.05; 102 participants, 2 studies; I² for RR: not applicable, I² for RD = 0%; low‐certainty evidence; Analysis 1.6).

Episodes of bradycardia

Three trials reported this outcome (Hartley 2018; Lago 2008; Pokela 1994). The evidence is very uncertain about the effect of opioids on episodes of bradycardia compared to placebo (RR 3.19, 95% CI 0.14 to 72.69; RD 0.01, 95% CI −0.03 to 0.06; 172 participants, 3 studies; I² for RR: not applicable, I² for RD = 0%; very low‐certainty evidence; Analysis 1.7).

Episodes of desaturation

Three trials reported this outcome (Hartley 2018; Lago 2008; Sindhur 2020). The evidence is very uncertain about the effect of opioids on episodes of desaturation compared to placebo (RR 1.82, 95% CI 0.72 to 4.58; RD 0.05, 95% CI −0.02 to 0.12; 199 participants, 3 studies; I² for RR = 0%, I² for RD = 63%; very low‐certainty evidence; Analysis 1.8).

Episodes of apnea

Three trials reported this outcome (Hartley 2018; Lago 2008; Sindhur 2020). Opioids may result in an increase in episodes of apnea compared to placebo (RR 3.15, 95% CI 1.08 to 9.16; RD 0.07, 95% CI 0.01 to 0.14; number needed to treat for an additional harmful outcome (NNTH) = 14; 199 participants, 3 studies; I² for RR = 0%, I² for RD = 78%; low‐certainty evidence; Analysis 1.9). We obtained outcome data from Sindhur 2020 via personal communication with study authors.

Hypotension

Two trials reported this outcome, with no events (Hartley 2018; Lago 2008). The evidence is very uncertain about the effect of opioids on episodes of hypotension compared to placebo (RR not estimable, RD 0.00, 95% CI −0.06 to 0.06; 88 participants, 2 studies; I² for RR: not applicable, I² for RD = 0%; very low‐certainty evidence; Analysis 1.10).

Comparison 2: Opioids versus oral sweet solution or non‐pharmacological intervention

We included two studies in this comparison (Gitto 2012; Sethi 2020); Gitto 2012 and Sethi 2020 reported one and none of the outcomes specified in this review, respectively. Certainty of the evidence is reported for the seven outcomes specified for the summary of findings table (summary of findings Table 2).

Primary outcomes

Pain assessed with the following scales

CRIES during procedure

One trial reported this outcome (Gitto 2012). The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to facilitated tucking (MD −4.62, 95% CI −6.38 to −2.86; 100 participants, 1 study; I² not applicable; Analysis 2.1 (first subgroup)) or sensorial stimulation (MD 0.32, 95% CI −1.13 to 1.77; 100 participants, 1 study; I² not applicable; Analysis 2.1 (second subgroup)).

Any harms

No trials reported this outcome.

Secondary outcomes

Comparison 3: Opioids versus other analgesics and sedatives

We included five studies in this comparison (Cordero 1991; Madathil 2021a; Madathil 2021b; Manjunatha 2009; Taddio 2006). Certainty of the evidence is reported for the seven outcomes specified for the summary of findings table (summary of findings Table 3).

Primary outcomes

Pain assessed with the following scales

PIPP/PIPP‐R during procedure

Two trials reported this outcome (Madathil 2021a; Madathil 2021b). The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R during the procedure compared to other analgesics (MD −0.29, 95% CI −1.58 to 1.01; 124 participants, 2 studies; I² = 0%; very low‐certainty evidence; Analysis 3.1).

PIPP/PIPP‐R up to 30 minutes after procedure

One trial reported this outcome (Manjunatha 2009). The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R up to 30 minutes after the procedure compared to other analgesics (MD −1.10, 95% CI −2.82 to 0.62; 12 participants, 1 study; I² not applicable; very low‐certainty evidence; Analysis 3.2).

PIPP/PIPP‐R one to two hours after procedure

One trial reported this outcome (Manjunatha 2009). The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R one to two hours after the procedure compared to other analgesics (MD −0.17, 95% CI −2.22 to 1.88; 12 participants, 1 study; I² not applicable; very low‐certainty evidence; Analysis 3.3).

Any harms

No trials reported this outcome.

Secondary outcomes

Episodes of apnea

Two trials reported this outcome (Madathil 2021a; Madathil 2021b). The evidence is very uncertain about the effect of opioids on episodes of apnea during the procedure compared to other analgesics (RR 3.27, 95% CI 0.85 to 12.58; RD 0.09, 95% CI 0.00 to 0.19; 124 participants, 2 studies; I² for RR = 6%, I² for RD = 76%; very low‐certainty evidence; Analysis 3.4 (first subgroup)) or after the procedure (RR 2.71, 95% CI 0.11 to 64.96; RD 0.02, 95% CI −0.04 to 0.07; 124 participants, 2 studies; I² for RR and RD = 0%; very low‐certainty evidence; Analysis 3.4 (second subgroup)).

Hypotension

Three trials reported this outcome (Madathil 2021a; Madathil 2021b; Taddio 2006). The evidence is very uncertain about the effect of opioids on hypotension compared to other analgesics (RR 1.34, 95% CI 0.32 to 5.59; RD 0.01, 95% CI −0.04 to 0.06; 204 participants, 3 studies; I² for RR = 67%, I² for RD = 54%; very low‐certainty evidence; Analysis 3.5).

Comparison 4: Head‐to‐head comparison of different opioids

No studies were included in this comparison.

Comparison 5: Different routes for administration of the same opioid

No studies were included in this comparison.

Discussion

Summary of main results

We evaluated the benefits and harms of opioids for procedural pain in newborn infants. Seven studies compared opioids to no treatment or placebo (Carbajal 2005; Fallah 2016; Hartley 2018; Lago 2008; Manjunatha 2009; Pokela 1994; Sindhur 2020), two studies to oral sweet solution or non‐pharmacological intervention (Gitto 2012; Sethi 2020), and five studies to other analgesics and sedatives (Cordero 1991; Madathil 2021a; Madathil 2021b; Manjunatha 2009; Taddio 2006). Manjunatha 2009 has three groups (morphine, placebo, paracetamol) and is included in both the first comparison (morphine versus placebo) and third comparison (morphine versus paracetamol).

Compared to placebo, opioids probably reduce pain score assessed with the PIPP/PIPP‐R scale during the procedure; however, the evidence is very uncertain when pain score is assessed with the PIPP/PIPP‐R scale up to 30 minutes after the procedure or one to two hours after the procedure. Opioids may reduce NIPS during the procedure and may result in little to no difference in DAN scale one to two hours after the procedure. No studies reported harms. The evidence is very uncertain about the effect of opioids on episodes of bradycardia or hypotension. Opioids may result in an increase in episodes of apnea. No studies reported parent satisfaction with care provided in the NICU.

The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to facilitated tucking or sensorial stimulation. No studies in this comparison reported other pain scales; any harms, episodes of bradycardia or apnea, or hypotension; or parent satisfaction with care provided in the NICU.

The evidence is very uncertain about the effect of opioids on pain scores assessed with the PIPP/PIPP‐R scale during or after the procedure compared to other analgesics. No studies in this comparison reported any harms. The evidence is very uncertain about the effect of opioids on episodes of apnea or hypotension. No studies reported bradycardia or parent satisfaction with care provided in the NICU.

We identified no studies comparing different opioids (e.g. morphine versus fentanyl) or different routes for administration of the same opioid (e.g. morphine enterally versus morphine intravenously).

Overall completeness and applicability of evidence

To date, 13 trials comparing opioids versus placebo or other interventions for procedural pain have enrolled 823 infants. Study authors report limited data about the potential adverse effects of opioids and did not report relevant, long‐term outcomes such as major neurodevelopmental disability and cognitive and educational outcomes. We identified seven ongoing trials. High versus low dose of the same opioid was assessed in a study done during mechanical ventilation, which is outside the scope of this review (Shin 2013; Shin 2014). Phase II trials investigate the influence of opioid use on neonates brain activity (Bell 2019; NCT03897452). We could not perform an appropriate a priori subgroup analysis to detect differential effects because of the paucity of included trials.

Quality of the evidence

According to the GRADE approach, the overall certainty of evidence for critical outcomes for opioids administration for procedural pain ranged from moderate to very low (see summary of findings Table 1; summary of findings Table 2; summary of findings Table 3). We downgraded all outcomes except one for imprecision, in most cases by two levels, because of the paucity and small sample size of the included studies. We downgraded most outcomes for limitations in study design (one level in most cases) (i.e. unclear or high risk of bias in different domains, mainly selection, detection, and reporting bias). We downgraded a couple of outcomes for inconsistency because of moderate or substantial heterogeneity. We explored publication bias by means of a funnel plot (Figure 5); however, its value is limited due to the few studies reporting the outcome.

Figure 5

Funnel plot for PIPP/PIPP‐R during procedure, opioids versus placebo.

Potential biases in the review process

We used the standard methods of Cochrane Neonatal in conducting this systematic review. It is unlikely that the literature search missed relevant trials. We are confident that this systematic review summarizes all the currently available evidence from randomized trials on opioids for procedural pain in neonates. We applied no language restrictions. We succeeded in obtaining additional information from some study authors. Following full‐text screening, studies were excluded because of study design (e.g. phase II trial), characteristics of patient population (e.g. surgical pain or sedation during mechanical ventilation), or type of comparator (e.g. comparing high versus low dose of the same opioid). As prespecified in the protocols (Types of participants), multiple indications for opioids administration to neonates were excluded (i.e. during mechanical ventilation for respiratory morbidity, pre‐intubation or endotracheal suctioning, for postoperative pain, neonatal abstinence syndrome, and therapeutic hypothermia). One study done was stopped prematurely due to safety concerns in the opioid group, thus potentially affecting the point estimate (Hartley 2018). We reported pain scores at three time points: during the procedure; up to 30 minutes after the procedure; and at one to two hours after the procedure. The last might be too late for accurate detection of pain; however, the sickest infants can have a delayed reaction to pain and be affected by the pain long after the procedure. The authors of this Cochrane Review were not involved in any of the included trials.

Agreements and disagreements with other studies or reviews

This is the first systematic review on opioids for procedural pain in neonates. A systematic review with considerably broader inclusion criteria, including both opioids and alpha‐2‐agonists, and assessing any type of analgesia and sedation in newborn infants is ongoing (Kinoshita 2020). One Cochrane Review on propofol for procedural sedation/anesthesia in neonates included one study with 63 neonates (Prakeshkumar 2011). The authors' conclusions did not seek to directly affect current practice, but time to complete procedure and for recovery to previous clinical status was shorter in the propofol group compared to morphine‐atropine‐suxamethonium; no difference was detected in clinically significant side effects; however, the number of events was small.

Figure 1

Screen4Me summary diagram.

Figure 2

Study flow diagram.

Figure 3

Risk of bias graph.

Figure 4

Risk of bias summary.

Figure 5

Funnel plot for PIPP/PIPP‐R during procedure, opioids versus placebo.

Analysis 1.1

Comparison 1: Opioids versus no treatment/placebo, Outcome 1: PIPP/PIPP‐R during procedure

Analysis 1.2

Comparison 1: Opioids versus no treatment/placebo, Outcome 2: PIPP/PIPP‐R up to 30 min after procedure

Analysis 1.3

Comparison 1: Opioids versus no treatment/placebo, Outcome 3: PIPP/PIPP‐R 1 to 2 hours after procedure

Analysis 1.4

Comparison 1: Opioids versus no treatment/placebo, Outcome 4: DAN 1 to 2 hours after procedure

Analysis 1.5

Comparison 1: Opioids versus no treatment/placebo, Outcome 5: NIPS during procedure

Analysis 1.6

Comparison 1: Opioids versus no treatment/placebo, Outcome 6: Mortality

Analysis 1.7

Comparison 1: Opioids versus no treatment/placebo, Outcome 7: Episodes of bradycardia

Analysis 1.8

Comparison 1: Opioids versus no treatment/placebo, Outcome 8: Episodes of desaturation

Analysis 1.9

Comparison 1: Opioids versus no treatment/placebo, Outcome 9: Episodes of apnea

Analysis 1.10

Comparison 1: Opioids versus no treatment/placebo, Outcome 10: Hypotension

Analysis 2.1

Comparison 2: Opioids versus oral sweet solution or non‐pharmacological intervention, Outcome 1: CRIES

Analysis 3.1

Comparison 3: Opioids versus other analgesics, Outcome 1: PIPP/PIPP‐R during procedure

Analysis 3.2

Comparison 3: Opioids versus other analgesics, Outcome 2: PIPP/PIPP‐R up to 30 min after procedure

Analysis 3.3

Comparison 3: Opioids versus other analgesics, Outcome 3: PIPP/PIPP‐R 1 to 2 hours after procedure

Analysis 3.4

Comparison 3: Opioids versus other analgesics, Outcome 4: Episodes of apnea

Analysis 3.5

Comparison 3: Opioids versus other analgesics, Outcome 5: Hypotension

Summary of findings 1. Opioids compared to no treatment/placebo for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to no treatment/placebo for procedural pain in neonates
Patient or population: neonates exposed to procedural pain Setting: neonatal units Intervention: opioids Comparison: no treatment/placebo
Outcomes	Risk with no treatment/placebo	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the PIPP/PIPP‐R pain scale, ranging from 0 to 21 and 0 to 18 in preterm infants < 28 weeks' gestational age and full‐term infants, respectively, during procedure	Mean PIPP/PIPP‐R during procedure ranged 8 to 11.	MD 2.58 lower (3.12 lower to 2.03 lower)	‐	199 (3 RCTs)	⊕⊕⊕⊝ Moderate ¹	Opioids probably reduce pain score assessed with the PIPP/PIPP‐R scale during the procedure compared to placebo.
Pain, assessed using the PIPP/PIPP‐R pain scale up to 30 min after procedure	Mean PIPP/PIPP‐R up to 30 min after procedure ranged 3 to 6.	MD 0.14 higher (0.17 lower to 0.45 higher)	‐	123 (2 RCTs)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale up to 30 min after the procedure compared to placebo.
Pain, assessed using the PIPP/PIPP‐R pain scale 1 to 2 hours after procedure	Mean PIPP/PIPP‐R 1 to 2 hours after procedure ranged 4 to 11.	MD −0.83 (2.42 lower to 0.75 higher)	‐	54 (2 RCTs)	⊕⊝⊝⊝ Very low ³	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R scale 1 to 2 hours after the procedure compared to placebo.
Pain, assessed using the DAN pain scale, ranging from 0 to 21, 1 to 2 hours after procedure	Mean DAN 1 to 2 hours after procedure was 5.	MD 0.2 lower (2.21 lower to 1.81 higher)	‐	42 (1 RCT)	⊕⊕⊝⊝ Low ⁴	Opioids may result in little to no difference in pain score assessed with the DAN scale 1 to 2 hours after the procedure compared to placebo.
Pain, assessed using the NIPS, ranging from 0 to 7, during procedure	Mean NIPS during procedure ranged 5 to 6.	MD 1.97 lower (2.46 lower to 1.48 lower)	‐	102 (2 RCTs)	⊕⊕⊝⊝ Low ⁵	Opioids may reduce NIPS during the procedure compared to placebo.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of bradycardia	Study population		RR 3.19 (0.14 to 72.69) RD 0.01, (−0.03 to 0.06)	172 (3 RCTs)	⊕⊝⊝⊝ Very low ⁶	The evidence is very uncertain about the effect of opioids on episodes of bradycardia compared to placebo.
Episodes of bradycardia	0 per 1000 (No events in the 3 RCTs)	0 per 1000 (0 to 0)	RR 3.19 (0.14 to 72.69) RD 0.01, (−0.03 to 0.06)	172 (3 RCTs)	⊕⊝⊝⊝ Very low ⁶
Episodes of apnea	Study population		RR 3.15 (1.08 to 9.16) RD 0.07 (0.01 to 0.14) NNTH = 14	199 (3 RCTs)	⊕⊕⊝⊝ Low ⁷	Opioids may result in an increase in episodes of apnea compared to placebo.
Episodes of apnea	30 per 1000	95 per 1000 (33 to 278)	RR 3.15 (1.08 to 9.16) RD 0.07 (0.01 to 0.14) NNTH = 14	199 (3 RCTs)	⊕⊕⊝⊝ Low ⁷
Hypotension	Study population		‐	88 (2 RCTs)	⊕⊝⊝⊝ Very low ⁸	The evidence is very uncertain about the effect of opioids on hypotension compared to placebo.
Hypotension	See comment	See comment	‐	88 (2 RCTs)	⊕⊝⊝⊝ Very low ⁸
Parent satisfaction with care provided in the NICU (as measured by a validated instrument/tool)—not reported	‐	‐	‐	‐	This outcome was not reported.	This outcome was not reported.
*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). CI: confidence interval; DAN: Douleur Aiguë du Nouveau‐né; MD: mean difference; NICU: neonatal intensive care unit; NIPS: Neonatal Infant Pain Scale; NNTH: number needed to treat for an additional harmful outcome; OR: odds ratio; PIPP: Premature Infant Pain Profile; PIPP‐R: PIPP‐Revised; RCT: randomized controlled trial; RD: risk difference; RR: risk ratio
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.
¹Downgraded one level for inconsistency (moderate heterogeneity, I² = 62%); not downgraded for imprecision (narrow CIs). ²Downgraded one level for inconsistency (substantial heterogeneity, I² = 85%) and two levels for imprecision (two small trials with low sample size; CIs overlapping no effect). ³Downgraded one level for study limitations (unclear risk of selection, performance, detection, and reporting bias) and two levels for imprecision (two small trials with low sample size; CIs overlapping no effect). ⁴Downgraded two levels for imprecision (one small trial with low sample size; CIs overlapping no effect). ⁵Downgraded one level for inconsistency (moderate heterogeneity, I² = 60%) and one level for imprecision (two small trials with low sample size; CIs not overlapping no effect). ⁶Downgraded one level for study limitations (unclear risk of selection bias in the informative study) and two levels for imprecision (three small trials with low sample size; CI overlapping no effect). ⁷Downgraded one level for study limitations (unclear risk of selection and reporting bias) and one level for imprecision (wide CIs). ⁸Downgraded one level for study limitations (unclear risk of selection and reporting bias) and two levels for imprecision (two small trials with no events).

Summary of findings 1. Opioids compared to no treatment/placebo for procedural pain in neonates

Summary of findings 2. Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates
Patient or population: neonates exposed to procedural pain Setting: neonatal units Intervention: opioids Comparison: oral sweet solution or non‐pharmacological intervention
Outcomes	Risk with oral sweet solution or non‐pharmacological intervention	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the CRIES pain scale, ranging from 0 to 10, during the procedure— opioids versus facilitated tucking	Mean CRIES—opioids versus facilitated tucking was 9.	MD 4.62 lower (6.38 lower to 2.86 lower)	‐	100 (1 RCT)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to facilitated tucking.
Pain, assessed using the CRIES pain scale, ranging from 0 to 10, during the procedure— opioids versus sensorial stimulation	Mean CRIES—opioids versus sensorial stimulation was 4.	MD 0.32 higher (1.13 lower to 1.77 higher)	‐	100 (1 RCT)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the CRIES scale during the procedure compared to sensorial stimulation.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of bradycardia—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of apnea—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Hypotension—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Parent satisfaction with care provided in the NICU—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*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). CI: confidence interval; CRIES: Crying Requires oxygen Increased vital signs Expression Sleep; MD: mean difference; NICU: neonatal intensive care unit; 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.
¹Downgraded two levels for study limitations (high risk of performance bias; unclear risk of bias for the other domains) and one level for imprecision (one small trial with low sample size; CIs overlapping no effect).

Summary of findings 2. Opioids compared to oral sweet solution or non‐pharmacological intervention for procedural pain in neonates

Summary of findings 3. Opioids compared to other analgesics for procedural pain in neonates

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Opioids compared to other analgesics for procedural pain in neonates
Patient or population: procedural pain in neonates Setting: neonatal units Intervention: opioids Comparison: other analgesics
Outcomes	Risk with other analgesics	Risk with opioids	Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
Pain, assessed using the PIPP/PIPP‐R pain scale, ranging from 0 to 21 and 0 to 18 in preterm infants < 28 weeks gestational age and full‐term infants, respectively, during procedure	Mean PIPP/PIPP‐R during procedure ranged 5 to 7.	MD 0.29 lower (1.58 lower to 1.01 higher)	‐	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R during the procedure compared to other analgesics.
Pain, assessed using the PIPP/PIPP‐R pain scale up to 30 min after procedure	Mean PIPP/PIPP‐R up to 30 min after procedure was 5.	MD 1.1 lower (2.82 lower to 0.62 higher)	‐	12 (1 RCT)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R up to 30 min after the procedure compared to other analgesics.
Pain, assessed using the PIPP/PIPP‐R pain scale 1 to 2 hours after procedure	Mean PIPP/PIPP‐R 1 to 2 hours after procedure was 4.	MD 0.17 lower (2.22 lower to 1.88 higher)	‐	12 (1 RCT)	⊕⊝⊝⊝ Very low ²	The evidence is very uncertain about the effect of opioids on pain score assessed with the PIPP/PIPP‐R 1 to 2 hours after the procedure compared to other analgesics.
Any harms	‐	‐	‐	‐	‐	This outcome was not reported.
Episodes of apnea—during the procedure	Study population		RR 3.27 (0.85 to 12.58)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on episodes of apnea during the procedure compared to other analgesics.
Episodes of apnea—during the procedure	33 per 1000	109 per 1000 (28 to 419)	RR 3.27 (0.85 to 12.58)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹
Episodes of apnea—after the procedure	Study population		RR 2.71 (0.11 to 64.96)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹	The evidence is very uncertain about the effect of opioids on episodes of apnea after the procedure compared to other analgesics.
Episodes of apnea—after the procedure	0 per 1000	0 per 1000 (0 to 0)	RR 2.71 (0.11 to 64.96)	124 (2 RCTs)	⊕⊝⊝⊝ Very low ¹
Episodes of bradycardia—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
Hypotension	Study population		RR 1.34 (0.32 to 5.59)	204 (3 RCTs)	⊕⊝⊝⊝ Very low ³	The evidence is very uncertain about the effect of opioids on hypotension compared to other analgesics.
Hypotension	20 per 1000	26 per 1000 (6 to 110)	RR 1.34 (0.32 to 5.59)	204 (3 RCTs)	⊕⊝⊝⊝ Very low ³
Parent satisfaction with care provided in the NICU—not reported	‐	‐	‐	‐	‐	This outcome was not reported.
*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). CI: confidence interval; NICU: neonatal intensive care unit; PIPP: Premature Infant Pain Profile; PIPP‐R: PIPP‐Revised; RCT: randomized controlled trial; RR: risk ratio
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.
¹Downgraded two levels for study limitations (high risk of performance bias; unclear risk of selection and other biases) and one level for imprecision (two small trials with low sample size; CI overlapping no effect). ²Downgraded one level for study limitations (unclear risk of selection, performance, and detection bias) and two levels for imprecision (one small trial with low sample size; CIs overlapping no effect). ³Downgraded two levels for study limitations (high risk of performance bias; unclear risk of selection and other biases) and one level for imprecision (three small trials with low sample size; CIs overlapping no effect).

Summary of findings 3. Opioids compared to other analgesics for procedural pain in neonates

Table 1. Overview of included studies, listed by type of comparison

Author	Country	No. of infants in intervention and control group, respectively	GA in intervention and control group, respectively	Procedure	Intervention	Comparison
Comparison 1: Opioids versus no treatment or placebo
Carbajal 2005	France, USA	21/21	27.3 (1.8)/27.2 (1.7)^a	Heel lances	Morphine 100 μg/kg loading dose and 10 to 30 μg/kg/h continuous infusion (intravenously)	Placebo 5% dextrose (intravenously)
Hartley 2018	UK	15/16	28.1 (26.3 to 30.1)/28.6 (27.9 to 29.7)^b	Heel lances and ROP screening examination	Morphine 100 μg/kg single dose (orally)	Placebo (orally)
Manjunatha 2009	UK	6/6	NR	ROP screening examination	Morphine 200 μg/kg single dose (orally)	Placebo (orally)
Fallah 2016	Iran	23/22	37.4 (0.8)/37.6 (0.6)^a	Lumbar puncture	Fentanyl 2 μg/kg single dose (intravenously)	Placebo normal saline (intravenously)
Sindhur 2020	India	56/55	30.7 (1.7)/31.0 (1.7)^a	ROP screening	Fentanyl 2 μg/kg single dose (intranasal)	Placebo normal saline (intranasally)
Lago 2008	Italy	27/27	28 (2)/29 (2)^a	PICC insertion	Remifentanil 0.03 μg/kg/min continuous infusion (intravenously)	Placebo 5% dextrose continuous infusion (intravenously)
Pokela 1994	Finland	42/42	31.6 (25 to 40)/32.9 (24 to 41)^c	Daily routine care procedures^d and tracheal suction	Meperidine 1 mg/kg single dose (intravenously)	0.9% saline single dose (intravenously)
Comparison 2: Opioids versus oral sweet solution or non‐pharmacological intervention (skin‐to‐skin contact, music exposure, non‐nutritive sucking, swaddling, etc.)
Gitto 2012	Italy	50/50/50	NR	Heel lances	Fentanyl 1 to 2 μg/kg bolus injection (intravenously)	Facilitated tucking/sensorial saturation
Sethi 2020	India	29/29	30.3 (2.2)/30.3 (2.4)^a	Laser for ROP	Fentanyl 1 μg/kg/h continuous infusion (intravenously)	24% oral sucrose single dose
Comparison 3: Opioids versus other analgesics (e.g. paracetamol) and sedatives (e.g. midazolam and other benzodiazepines)
Manjunatha 2009	UK	6/6	NR	ROP screening examination	Morphine 200 μg/kg single dose (orally)	Paracetamol 20 mg/kg single dose (orally)
Cordero 1991	USA	15/14	28 (2)/27 (2)^a	Broviac catheter placement	Fentanyl 2 μg/kg single dose (intravenously)	Secobarbital 1 mg/kg single dose (intravenously)
Madathil 2021a	India	51/46	29.7 (1.9)/29.8 (1.5)^a	Laser for ROP	Fentanyl 2 μg/kg followed by a continuous infusion of 1 μg/kg/h increased to a maximum of 3 µg/kg/h (intravenously)	Ketamine 0.5 mg/kg, followed by further intermittent intravenous bolus doses of 0.5 mg/kg to a maximum of 2 mg/kg (intravenously)
Madathil 2021b	India	13/14	30.3 (1.3)/30.5 (2.4)^a	Laser for ROP	Fentanyl 2 μg/kg followed by infusion of 2 μg/kg/h to a maximum of 5 μg/kg/h (intravenously)	Ketamine 1 mg/kg followed by intermittent bolus doses of 0.5 mg/kg to a maximum of 4 mg/kg (intravenously)
Taddio 2006	Canada	38/42	29.6 (4.9)/30 (5.1)^a	PICC placement	Morphine 100 μg/kg single dose (intravenously)	Tetracaine 0.5 g 4% gel applied to the insertion site
We included no studies for the following comparisons: head‐to‐head comparison of different opioids; different routes of administration of the same opioid.
GA: gestational age; NR: not reported; PICC: peripherally inserted central catheter; ROP: retinopathy of prematurity ^aMean (standard deviation). ^bMedian (interquartile range). ^cMean (range). ^dWeighing, washing, temperature measurement, chest roentgenogram.

Table 1. Overview of included studies, listed by type of comparison

Comparison 1. Opioids versus no treatment/placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 PIPP/PIPP‐R during procedure Show forest plot	3	199	Mean Difference (IV, Fixed, 95% CI)	‐2.58 [‐3.12, ‐2.03]

1.2 PIPP/PIPP‐R up to 30 min after procedure Show forest plot	2	123	Mean Difference (IV, Fixed, 95% CI)	0.14 [‐0.17, 0.45]

1.3 PIPP/PIPP‐R 1 to 2 hours after procedure Show forest plot	2	54	Mean Difference (IV, Fixed, 95% CI)	‐0.83 [‐2.42, 0.75]

1.4 DAN 1 to 2 hours after procedure Show forest plot	1	42	Mean Difference (IV, Fixed, 95% CI)	‐0.20 [‐2.21, 1.81]

1.5 NIPS during procedure Show forest plot	2	102	Mean Difference (IV, Fixed, 95% CI)	‐1.97 [‐2.46, ‐1.48]

1.6 Mortality Show forest plot	2	0	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

1.7 Episodes of bradycardia Show forest plot	3	172	Risk Ratio (M‐H, Fixed, 95% CI)	3.19 [0.14, 72.69]

1.8 Episodes of desaturation Show forest plot	3	199	Risk Ratio (M‐H, Fixed, 95% CI)	1.82 [0.72, 4.58]

1.9 Episodes of apnea Show forest plot	3	199	Risk Ratio (M‐H, Fixed, 95% CI)	3.15 [1.08, 9.16]

1.10 Hypotension Show forest plot	2	0	Risk Ratio (M‐H, Fixed, 95% CI)	Not estimable

Comparison 1. Opioids versus no treatment/placebo

Comparison 2. Opioids versus oral sweet solution or non‐pharmacological intervention

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 CRIES Show forest plot	1		Mean Difference (IV, Fixed, 95% CI)	Subtotals only

2.1.1 Opioids vs facilitated tucking	1	100	Mean Difference (IV, Fixed, 95% CI)	‐4.62 [‐6.38, ‐2.86]
2.1.2 Opioids vs sensorial stimulation	1	100	Mean Difference (IV, Fixed, 95% CI)	0.32 [‐1.13, 1.77]

Comparison 2. Opioids versus oral sweet solution or non‐pharmacological intervention

Comparison 3. Opioids versus other analgesics

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 PIPP/PIPP‐R during procedure Show forest plot	2	124	Mean Difference (IV, Fixed, 95% CI)	‐0.29 [‐1.58, 1.01]

3.2 PIPP/PIPP‐R up to 30 min after procedure Show forest plot	1	12	Mean Difference (IV, Fixed, 95% CI)	‐1.10 [‐2.82, 0.62]

3.3 PIPP/PIPP‐R 1 to 2 hours after procedure Show forest plot	1	12	Mean Difference (IV, Fixed, 95% CI)	‐0.17 [‐2.22, 1.88]

3.4 Episodes of apnea Show forest plot	2		Risk Ratio (M‐H, Fixed, 95% CI)	Subtotals only

3.4.1 During the procedure	2	124	Risk Ratio (M‐H, Fixed, 95% CI)	3.27 [0.85, 12.58]
3.4.2 After the procedure	2	124	Risk Ratio (M‐H, Fixed, 95% CI)	2.71 [0.11, 64.96]
3.5 Hypotension Show forest plot	3	204	Risk Ratio (M‐H, Fixed, 95% CI)	1.34 [0.32, 5.59]

Comparison 3. Opioids versus other analgesics