Interventions to reduce medication errors in adult intensive care: a systematic review

Abstract

Critically ill patients need life saving treatments and are often exposed to medications requiring careful titration. The aim of this paper was to review systematically the research literature on the efficacy of interventions in reducing medication errors in intensive care. A search was conducted of PubMed, CINAHL EMBASE, Journals@Ovid, International Pharmaceutical Abstract Series via Ovid, ScienceDirect, Scopus, Web of Science, PsycInfo and The Cochrane Collaboration from inception to October 2011. Research studies involving delivery of an intervention in intensive care for adult patients with the aim of reducing medication errors were examined. Eight types of interventions were identified: computerized physician order entry (CPOE), changes in work schedules (CWS), intravenous systems (IS), modes of education (ME), medication reconciliation (MR), pharmacist involvement (PI), protocols and guidelines (PG) and support systems for clinical decision making (SSCD). Sixteen out of the 24 studies showed reduced medication error rates. Four intervention types demonstrated reduced medication errors post-intervention: CWS, ME, MR and PG. It is not possible to promote any interventions as positive models for reducing medication errors. Insufficient research was undertaken with any particular type of intervention, and there were concerns regarding the level of evidence and quality of research. Most studies involved single arm, before and after designs without a comparative control group. Future researchers should address gaps identified in single faceted interventions and gather data on multi-faceted interventions using high quality research designs. The findings demonstrate implications for policy makers and clinicians in adopting resource intensive processes and technologies, which offer little evidence to support their efficacy.

Figure 1

Flowchart of systematic review

Figure 2

Summary statistics and forest plot of studies (n= 22). Note: CPOE refers to computerized physician order entry, CWS refers to changes in work schedules, IS refers to intravenous systems, ME refers to modes of education, MR refers to medication reconciliation, PG refers to protocols and guidelines, PI refers to pharmacist involvement and SSCD refers to support systems for clinical decision-making. The studies by Fraenkel et al. [26] and Langebrake & Hilgarth [34] were not included in the risk ratio analysis as no denominator values were available and no data were available at pre-intervention respectively. Risk ratio M-H refers to Risk ratio Mantel-Haenszel. M-H, Mantel-Haenszel