Introduction

Despite improvements in the medical field seen through advanced technology and computerized systems, we still witness medication errors taking place. Medication errors are defined by the National Coordinating Council for Medication Error Reporting and Prevention (NCCMERP) as “any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer” [1]. Numerous studies have reported that medication errors constitute a major reason for increased hospital stay, delayed medication administration, and an endless increase in medical costs [2–4]. In the United States 7,000 to 9,000 people die each year as a result of medication errors [5]. The global impact of medication errors as reported by the World Health Organization is estimated to be 42 billion USD each year [6]. The cost of a medication error can reach over $100 000 USD per patient [3].

The Majority of medication errors reported in the literature include prescribing errors requiring a pharmacist intervention [4, 7–9]. The reported prevalence of prescribing errors was reported to range from 2% to 94% [10]. These include among others, wrong drug, wrong dose, wrong route, wrong duration, and wrong formulation [7–11]. The results of a systematic review reporting error rates suggest that up to one prescribing error per patient takes place within a hospital setting [4]. Pharmacists are well-positioned to improve medication utilization and improve patient safety [12, 13].

Multiple interventions were tested for efficacy in reducing medication errors and studies have documented that pharmacists play a major role in reducing these medication errors [14–17]. Ward-based clinical pharmacists reduce medication errors by providing real-time advice to physicians rather than recommending changes after prescribing has occurred [18–20]. There is an important role of pharmacists in developing a medication safety strategy within their practice to reduce errors and improve the medication utilization process. One of the key elements of such a strategy is providing education to healthcare practitioners and prescribers since “too often, there is insufficient education of healthcare providers on medication safety topics” [13]. Educational interventions implemented by pharmacists within the pediatric ward have been reported to increase detection and reporting of medication errors while at the same time, reducing the severity of medication errors [21].

Pharmacist-led educational interventions seem to show benefit in reducing medication errors; however, it is not clear how often these educational sessions should run, nor what content should be included to provide optimal benefit in reducing medication errors [22]. To the best of our knowledge, this is the first systematic review and meta-analysis aimed to investigate the effectiveness of pharmacist-led educational interventions provided to healthcare providers to reduce medication errors both quantitatively and qualitatively.

Materials and methods

This systematic review and meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting methodology [23]. The research protocol is available in The International Prospective Register of Systematic Reviews (PROSPERO) under the registration number CRD42019116465 [24].

Database and search strategy

Studies were captured from five databases: Cochrane Library, EBSCO, EMBASE, Medline, and Google Scholar, from inception until June 2020. In addition to databases, references of studies were also manually screened for relevant articles. To identify relevant studies, free text and MeSH terms were used and combined with Boolean operators ‘AND’ to combine terms of different categories and ‘OR’ to combine terms within one category. Table 1 summarizes the terms used for the search strategy. The limits applied include searching within title and abstract and using the English language in reporting the article. The search strategy was applied by two researchers, independently, to ensure consistency. Database hits and references identified from reference screening were then transferred to an excel sheet for removing duplicates. Screening of titles and abstracts was then completed by two independent authors. Any discrepancy was resolved through full-text screening of articles.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Search terms.

https://doi.org/10.1371/journal.pone.0253588.t001

Results

Identification and selection of studies

A total of 9354 articles were identified from the different databases and reference screening. After duplicate studies were removed (n = 6857), the remaining articles were screened by title and abstract by two researchers. A total of 12 articles were included in the systematic review and meta-analysis Fig 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Prisma flow diagram.

https://doi.org/10.1371/journal.pone.0253588.g001

Study characteristics

The characteristics of included studies are presented in Table 2. Studies captured were published from 2007 to 2019. The studies found were conducted in different countries including Egypt (n = 3) [27–29], Australia (n = 2) [30, 31], USA (n = 2) [32, 33] and one each from Pakistan [34], Spain [35], Netherlands [36], Saudi Arabia [37] and Vietnam [38]. The majority of studies included looked at the pediatric population and covered different hospital units including neonatal/pediatric intensive care units (n = 3) [27, 35, 36], children ward [27], and one study was conducted in the pediatric surgery department [28]. Three studies covered multiple hospital departments [30, 32, 38] and three covered emergency/acute care hospital wards [29, 31, 37], while only one study was conducted in an outpatient clinic [33]. Medication error definition varied across the included studies; nonetheless, six studies covered prescribing errors which had a similar definition among the studies and included wrong drug, wrong dose, wrong frequency, wrong concentration, wrong or missed rate of administration, wrong or missed instructions for administration, unclear order or incomplete order [27, 28, 30, 32, 33, 35]. Additionally, medication errors included the use of error-prone abbreviations in prescribing in seven studies [27, 30–33, 35, 37], Other medication errors captured by the studies were administration and preparation errors (n = 4) [29, 34, 36, 38] and medication error occurring at any stage [28].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Study characteristics.

https://doi.org/10.1371/journal.pone.0253588.t002

Quality of included studies

Two raters with previous experience with using the CCAT [26] appraised each of the 12 studies, which resulted in 24 independent CCAT evaluations. The average score was reported (Table 3). The total score ranged from 56.25% to 92.5% and the overall assessment mean for all studies was 74.68%. Within the CCAT sections, the highest scores were for preliminary and introduction (4.42/5), while the lowest were for sampling (3.17/5) and ethics (2.67/5) (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Cumulative quality assessment results.

https://doi.org/10.1371/journal.pone.0253588.t003

Effect of educational interventions

All studies included in the systematic review were eligible for meta-analysis (Fig 2). Ten of twelve studies indicated a clear and significant decrease in the incidence of medication errors after implementing a pharmacist-led educational intervention [27, 29–31, 33–38]. The pooled OR (n = 115 058 participants) across all studies was 0.38 (95% CI 0.22 to 0.65) (P = 0.0004). However, the results of these studies were substantially heterogeneous (Fig 2). Only two studies [28, 32] conducted in the USA and Egypt showed no significant difference in the occurrence of medication errors after applying the intervention. The study conducted in the USA had a very low rate of medication errors relative to other included studies and was divided into two phases, in which a significant reduction was observed during the intervention, but was not maintained one month post-intervention [32]. The educational intervention for the study conducted in Egypt consisted of one educational session, without follow-up sessions or distributing a printed copy of the lecture, and it showed a nonsignificant reduction of medication errors after the study intervention [28].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Random effect model of pharmacist-led education and its association with medication errors.

https://doi.org/10.1371/journal.pone.0253588.g002

Discussion

Majority of pharmacist-led interventions—targeting medication errors—reported in the literature focus on medication reconciliation or education provided to patients rather than healthcare providers [39–41]. This paper has shown an overall better outcome in favor of pharmacist-led educational interventions provided to healthcare providers on the rate of medication error occurrence.

Although most included studies showed improvement, two papers did not show a significant difference after implementing the intervention. The educational intervention in the first study conducted by Fawaz et al consisted of one lecture delivered to physicians without providing any printout or posters to summarize the content of the lecture which can serve as a constant reminder to recipients [28]. In the second study which was conducted by Peeters et al., pharmacists delivered frequent lectures to residents throughout six months, yet the benefit was only noted during the intervention period but was not maintained one month post-intervention [32]. This could be due to the initial very low rate of medication errors relative to other included studies in this review. Additionally, similar to Fawaz et al. study, the study by Peeters et al. did not provide any form of constant reminder (e.g. printouts or posters) as part of the educational intervention. Multiple studies showed that various educational strategies, such as brochures or training activities, have beneficial outcomes including improving different aspects of knowledge and skills of healthcare workers [42–44]. This benefit is also highlighted within the four studies that used printed handouts which resulted in significant reduction in medication error rates, including the trial conducted by Alagha et al., which used only one educational session [27, 30, 31, 34, 37]. In addition to the printed materials, Alagha et al. and Winder et al. provided individualized reports (written and through Electronic Medical Record (EMR) respectively) to the resident physicians about their own errors. This approach might have contributed to the significant reduction in medication errors in these studies because each resident would avoid the factors that lead to their prescribing errors; hence minimize them [27, 33].

The most intensive intervention was the one conducted by Winder et al. [33]. Within this study, researchers implemented multi-faceted interventions to reduce prescribing errors. This included four phases: a pre-intervention error assessment; followed by an optional educational intervention delivered to resident physicians; and error notification intervention through an audit and direct feedback to residents electronically and verbally depending on the error severity. This phase also included sending weekly newsletters of common errors over a period of three months and post-intervention error assessment conducted after one year. Within this study, the pediatric dosing errors were 36% lower among residents who attended the educational sessions compared to those who did not. This approach highlights the importance of continued audit and direct feedback to improve patient safety to be conducted along with healthcare practitioners’ education. The importance of continuous electronic audit and feedback was, similarly, emphasized upon by a study conducted in the UK which covered over 40 general practices [45].

Additionally, the frequency and duration of educational intervention can play a role in maintaining good outcomes. It was noted that studies that used more frequent educational sessions such as the one reported by Campino and Gursanscky providing 15 and 12 sessions within four months and one month respectively, had better outcomes compared to other studies [30, 35]. Repeated encounters–as reported by Bruner–improves information retention [46]. It must be highlighted though that the mere repetition of the sessions might not be sufficient in improving practice. Brabeck et al. highlighted that deliberate practice significantly improves cognitive gain and brings about the motivation for more learning, hence, better performance [47]. Therefore, the frequent educational sessions should always have a component of the application to have a better outcome.

The results of this meta-analysis highlight that many factors can lead to a significant reduction in medication errors post educational sessions. Medication errors can significantly reduce if pharmacists invest in providing education to healthcare providers that are supplemented with printed handouts summarizing the lecture content. Additionally, the provision of individualized medication error reports and the use of ward-based posters addressing preparation and administration errors can also help reduce prescribing and medication administration errors. For ease of access, pharmacists can also work on providing flashcards with instructions for good prescribing including high-risk abbreviations when addressing prescribing errors. Finally, having frequent application-based educational sessions can be of great benefit to healthcare providers.

Additionally, majority of the included studies used errors reported before the educational sessions to inform the content. Such an approach can contribute to the benefit observed in reduced medication errors. Based on the results of this systematic review, the authors recommend using frequent educational sessions supplemented with individualized prescriber reports to significantly minimize medication errors. The session content should be site-specific based on the types of medication errors encountered and summary material should be provided to attendees. Many articles emphasized the importance of continuous monitoring and reporting of errors to maintain a safe practice and reduce patient harm [48–50]. As such, these continuous reports should inform the content of the educational sessions.

The quality of reporting of the included studies was of an acceptable level. Nonetheless, the authors believe it can always be improved through the utilization of reporting tools. Within the CCAT, the item which scored the least was ethics which focuses on reporting of consent, ethics approval, and anonymity of data. This was followed by “sampling”, which includes three components: sampling method, sample size, and sampling protocol. Sampling is a very important component that can affect the study power to detect a difference and can also affect the generalizability of the results [51]. Research on medication errors should maximize its quality through the use of reporting tools and checklists available in the literature and the Enhancing the QUAlity and Transparency of health Research (EQUATOR) Network [52].

Our results are consistent with previous systematic reviews and meta-analyses that studied the effectiveness of different clinical pharmacist interventions on medication error rates in adults and pediatric patients at diverse hospital departments [26, 41, 53–57]. The studies included multiple interventions such as reviewing orders, implementing a ward-based pharmacist, or designing educational programs. However, none of these papers focused on educational interventions or had a subgroup analysis to investigate their effectiveness in decreasing medication errors. Nonetheless, Manias et al. published a recent systematic review of all interventions to reduce medication errors in adult medical and surgical units which described similar outcomes to this systematic review. Pharmacist-led interventions were of great benefit and educational interventions to healthcare providers proved to be effective in reducing medication errors [58]. This emphasizes the importance of having drug experts take lead in delivering education to their colleagues.

Limitations within this study need to be acknowledged. Significant heterogeneity in the studies included in the meta-analysis was reported. This is due to major differences in the delivery of the educational sessions, variation in the settings, and differences in the type and definition of the targeted medication error. Additionally, almost all studies reported positive results with the pharmacist-led educational intervention and we cannot exclude the risk of publication biases. Moreover, the researchers within the included studies might have chosen the areas that are likely to show positive results since they included acute care, pediatric, and emergency settings. The off-label, unlicensed use of medicines, weight-related dose adjustment, and other dosing calculations are widespread in these settings particularly in pediatrics, which may increase the risk of avoidable medication-associated harm. Therefore, they are more likely to report positive results. In addition, the setting for this review was mainly inpatient with only one study reporting findings in the outpatient setting. As a result, our findings are unlikely to be generalizable to all healthcare systems. Finally, the authors acknowledge that excluding non-interventional studies might have precluded many potentially good and relevant articles. However, the authors focused on interventional studies which are generally recognized as stronger evidence generating studies with less confounding [59].

Future investigations and randomized controlled studies are required to evaluate the impact of pharmacist-led educational intervention on other types of medication errors not covered in the studies included in this review, such as transcribing, dispensing, and monitoring errors. This will provide a broader picture of the capability of the pharmacists in providing effective education to healthcare providers that minimize all types of medication errors. The effect of pharmacist-led education to healthcare providers on medication errors in outpatient settings should gain more attention in further studies. This might include outpatient clinics, home care, and community pharmacies. The elderly population may also encounter special issues related to medication errors. For example, people living in care homes are often frail with multiple health conditions and take multiple medications. The administration of medication in this environment often differs to patients’ own homes as it is provided by nursing staff or other personnel, thus raising particular issues around dispensing, administration, and monitoring problems, as well as staff training. Additionally, many pharmacist-led interventions target medication errors apart from providing education to healthcare providers, including providing medication reconciliation; including pharmacists inwards; and involving them in team discussions [39–41]. A study that compares the benefit of these interventions and determining priority intervention will be of significant benefit especially in areas where resources are limited. Medication errors are known to be linked to morbidity and mortality as highlighted by the well-cited report “To Err is human” [60], therefore, future studies should consider the outcome of morbidity and mortality as an extended outcome to medication error rates. This will provide a better understanding of the consequences of interventions done to tackle medication errors. Finally, the economic impact of this intervention and associated cost saving studies related to education can support resource allocation and improve patient safety.

Conclusion

This systematic review and meta-analysis demonstrated that pharmacist-led educational interventions to healthcare providers are effective in reducing medication errors. The educational intervention was observed to be most effective when supplemented with printed handouts that summarize the session content, providing posters or pocket-sized flashcards to prescribers, and when healthcare providers receive an individualized written or electronic report about their medication errors. Additionally, frequent educational sessions were observed to be more effective than one-time sessions. These combined approaches serve as constant reminders to the healthcare practitioners to minimize medication errors at different stages of the medication-use process. Further research should consider the effect on morbidity, mortality related to medication errors and the economic impact of pharmacist-led education provided to healthcare providers.

Supporting information