Iatrogenic injuries occur commonly in the health care system. In recent years, medication error events received considerable attention because of its substantial mortality, morbidity, and additional health care costs. Many reports indicated that nearly one third of adverse drug events (ADEs) are associated with medication errors and are thus preventable [^1-6]. The frequency of these events was described over several studies. According to Brennan et al [⁷] 3.7% of hospitalized patients experienced an adverse event related to medical therapy in 1984. Of these iatrogenic injuries, 69% were preventable. Later, Lazarou et al [⁸] estimated that over 100.000 hospitalized patients in 1994 had fatal adverse drug reactions. Bates et al [⁹] reported 5 medication errors per 100 medication orders; only 7 in 100 medication errors had significant potential for harm. Other studies largely confirmed that ADEs are common; and have important economic and human consequences [³,¹⁰] as patients suffering from ADEs have about three times the mortality rate compared with matched controls, with an increase of about two days in length of stay, and an increased cost of more than $US2000 of admission. Thus, identification of ADEs and medication errors appears to be a crucial first step in improving patient safety, although the approach seems to be different in research and routine care [¹¹].

Estimates of ADEs rates vary substantially depending on the setting and the data sources used. Intensive care units (ICUs) can be considered as an optimal location for developing voluntary reporting incentives based on the frequency of events. In fact, ICU patients may be at higher risk for ADEs [¹²] because of the higher exposure to medicines and the weaker health conditions compared with other patients. Cullen et al [¹³] reported twice combined incidence of preventable and potential ADEs that rate in non ICU areas. Specific improvements in the medication ordering and processing system may reduce the risk of errors. Studies have demonstrated that some of these interventions can be effective such us physician computer order entry and a computerized clinical decision support program [^14-20].

In medical community, few epidemiological data are available regarding the incidence, type and causes of medication errors in ICUs. Less information is available concerning medication errors and ADEs in developing countries and in Morocco. In this field, a pilot project built by the World Alliance for Patient Safety in collaboration with the World Health Organization program for International Drug Monitoring was initiated in 2007. The Moroccan Pharmacovigilance Centre was assigned as project coordinator. As part of this project, a prospective study in Moroccan medical ICU was initiated showing 7.7% medication errors for 1000 patient-days [¹²].

The aim of the present study was to evaluate incidence, type and consequences of medication errors in a Moroccan medical ICU.

In our study, medication errors were common in the inpatient medical ICU setting. Potential ADEs occurred more frequently than ADEs. The rate of medication errors was high. Errors occurred most commonly at the stage of drug ordering and transcribing. All types were concerned but those essentially increased were drug route and dosing errors. The drug classes associated most frequently with errors were anti infective and anticoagulants.

Comparing these results with those from a study by Bates et al [⁹] using similar methods in an adult patient population, our study had a higher rate of medication errors (10 errors/100 orders in our study vs. 5.3/100 orders). In a clinical review [²⁴], the rate of medical errors among critically ill adults ranges from 1.2 to 947 errors per 1000 patient-days with a median of 106 errors per 1000 patient-days, the incidence of medication errors in our study seems then to be higher. However, a single centre study in a medical ICU [²⁵] using an observation method of medication administration reported a higher rate with 1500 MEs per 1000 patient-days. In fact, a systematic review of ME incidence in different ICU types [²⁶] have found a wide variation in reported rates of MEs. We believe thus, that much of this large variability was due to differences in the definitions used of the same type of event and also in methods used to detect events. Because most MEs do not result in harm, it is logical that MEs are more frequent than ADEs.

The incidence of ADEs and potential ADEs found in our study is consistent with what is expected based upon the literature. Several studies have focused on ME incidence and few data are available concerning ADE and portential ADE incidences [²⁶]. To maximize the yield of events, we used a multifaceted approach to event detection involving chart review supplemented by other methods, based on that undertaken by Cullen et al [¹³]. In our study, the incidence of potential ADEs (222 per 1000 patient-days) was found to be about 20 times higher to that reported by Cullen et al [¹³] (13.5 potential ADEs per1000 patient-days). However, Rothschild et al [²⁷] in a cardiac surgery ICU using also a multifaceted approach reported a rate of 23.8 potential ADEs per1000 patient-days which is lower to ours. Measuring ADE rates is also useful since this identifies actual situations in which patients are harmed and also allows for change for safer policies [²⁶]. Concerning the incidence of ADEs, 15.7 ADEs per 1000 patient-days were noted in our study, Cullen et al [¹³] in a mixed ICU using a multifaceted approach reported a rate of 5.1 ADEs per 1000 patient-days. Rothschild et al [²⁸] using the same method in a mixed ICU reported much higher rate (37.6 ADEs per 1000 patient-days). The exact rate of events is difficult to determine, but it has become widely accepted that the best estimate of the incidence of events requires comprehensive data collection using multiple strategies [²⁶]. The incidence of events is greatly influenced by the definitions used, the method of detection and classification of events, and the study setting [¹⁹]. This makes comparison of reported rates between studies extremely difficult and, thus, any differences in rates should not be interpreted simply as reflecting differing levels of the quality of care between institutions, but more as reflecting differences between study methodologies.

Previous studies have not reported criteria for defining injury, and other studies have used multiple definitions for the same type of event [²⁶]. The reason for this diversity of definitions is likely related to that fact that no standard definition is accepted by all the major organisations related to medication safety [²⁶]. We used a multidisciplinary approach that examined all aspects of the medication system, from the physician's order through administration of the drug to the patient. Moreover, we encouraged voluntary reporting by emphasizing the role of systems problems in the origin of errors and by nurturing a blame-free environment.

Although 35% of errors in our study occurred in drug orders and 60% in drug transcribing many of these errors were detected and corrected prior to the order reaching the pharmacy. Cogent theories regarding the origin of errors (often categorized as human factor research) have been developed. Most investigators have focused on problems in health care delivery systems that predispose to error, rather than emphasizing the role of individuals [²⁹,³⁰]. Ongoing multidisciplinary analysis of incidents is important for developing further system improvements

There were some limitations of the present study, our study included only 63 admissions but 509 patient-days of admission were noted and 4942 prescriptions episodes were written which can be easily exploitable. The location was a general hospital; there may be some limitations to the generalizability of the results of the study to other types of healthcare facilities. The scope of the study was limited to MEs, ADEs and potential ADEs occurring in hospitalized patients, and did not include outpatients. It is however unlikely that any harmful events would have been missed as such events would have prolonged hospital stay or would have been identified through readmission of the patient. However, our study has the strength point to compare both the results of MEs and ADEs with the literature data.

Our review of the literature highlighted 2 important findings. First, medication error rates vary widely among clinical settings (both ICU and non-ICU settings), patient populations and studies. The reasons for this variation are likely multifactorial, but the reasons may include different patient populations (illness severity, number and type of prescriptions) clinical practice variation, lack of uniformity of definitions, the processes under investigation (e.g., prescription, transcription), methods of reporting and the culture of the different centers reporting their data [³¹]. The lack of standard definitions and reporting techniques make comparisons across organizations, regions or countries difficult. Second, although there are many potential risk factors for medication errors, the strongest evidence that critically ill patients are at increased risk of a medication error are increased severity of illness; failure to document the patient's usual medication list; prescription of cardiovascular, sedative, analgesic, anticoagulant or anti-infective medications; prescription of each additional medication; admission to a medical ICU compared with a surgical ICU [³¹].

Potential strategies to prevent medication errors in the ICU are focused on 7 prevention strategies: eliminating extended physician work schedules, computerizing physician order entry, implementing support systems for clinical decisions, computerizing intravenous devices, having pharmacists participate in the ICU, reconciling medications and standardizing medications [³¹].

The practical approach is to recognize that errors are a reality of medicine and that all health care providers have a responsibility to ensure patient safety and to use caution in promoting interventions. Improved medication safety may be accomplished by optimizing the safety of the medication process, eliminating situational risk factors and adopting strategies to intercept errors and mitigate their consequences.

The medical use process itself is a complex system, providing many opportunities of ME occurring. Our incidence of ME was generally similar to that of others ICU studies which used similar methods of detection events. All types and categories were identified with high prevalence of transcribing error and wrong route, drugs incriminated was antibiotics and anticoagulants. These results can be used to improve quality of health care delivery; some authors recommended focusing on computerized approach like a first step to prevention in ICU in developing and developed countries. We believe that this improvement research is an expensive technology in our developed countries. We recommend standardization of therapeutic protocols and a systematic checking of prescriptions and transcriptions to reduce incidence of ME.

Medication errors in a medical ICU

ADE: adverse drug event; ADR: adverse drug reaction; ICU: intensive care unit; IQR: interquartile range; ME: Medication error; WHO: World Health Organization.

The authors declare that they have no competing interests.

NJ drafted the manuscript, and participated in the acquisition of data. NM drafted the manuscript, participated in the acquisition of data and the study design. RO participated in the acquisition of data. KA helped to draft the manuscript, and participated in the acquisition of data. IK helped to draft the manuscript. JH participated in the coordination of the study. TD participated in the coordination of the study. AAZ participated in the design of the study, and performed the statistical analysis. RA conceived of the study, participated in the design of the study, performed the statistical analysis and interpretation of data, and gave the final approval of the manuscript. All authors read and approved the final manuscript

We would like to thank the medical and paramedical stuff of medical ICU of Ibn Sina University hospital for its nice cooperation in the study.