Aim To determine the prevalence and magnitude of medication errors and their association with patients' sociodemographic and clinical characteristics and nurses' work conditions. Design An observational, analytical, cross‐sectional and ambispective study was conducted in critically ill adult patients. Methods Data concerning prescription errors were collected retrospectively from medical records and administration errors were identified through direct observation of nurses during drug administration. Those data were collected between April and July 2015. Results A total of 650 prescription errors were identified for 961 drugs in 90 patients (mean error 7[SD 4.1] per patient) and prevalence of 47.1% (95% CI 44–50). The most frequent error was omission of the prescribed medication. Intensive care unit stay was a risk factor associated with omission error (OR 2.14; 1.46–3.14: p < .01). A total of 294 administration errors were identified for 249 drugs in 52 patients (mean error 6 [SD 6.7] per patient) and prevalence of 73.5% (95% CI 68–79). The most frequent error was interruption during drug administration. Admission to the intensive care unit (OR 0.37; 0.21–0.66: p < .01), nurses’ morning shift (OR 2.15; 1.10–4.18: p = .02) and workload perception (OR 3.64; 2.09–6.35: p < .01) were risk factors associated with interruption. Conclusions Medication errors in prescription and administration were frequent. Timely detection of errors and promotion of a medication safety culture are necessary to reduce them and ensure the quality of care in critically ill patients. Impact Medication errors occur frequently in the intensive care unit but are not always identified. Due to the vulnerability of seriously ill patients and the specialized care they require, an error can result in serious adverse events. The study shows that medication errors in prescription and administration are recurrent but preventable. These findings contribute to promote awareness in the proper use of medications and guarantee the quality of nursing care.

No abstract

Introduction: Reliable patient identification is essential for safe care, and failures may cause patient harm. Identification can be interfered with by system factors, including working conditions, technology, organizational barriers, and inadequate communications protocols. The study aims to explore systems factors contributing to patient identification errors during intrahospital transfers. Methods:We conducted a qualitative study through direct observation and interviews with porters during intrahospital patient transfers. Data were analyzed using the Systems Engineering Initiative for Patient Safety human factors model. The patient transfer process was mapped and compared with the institutional Positive Patient Identification policy. Potential system failures were identified using a Failure Modes and Effects Analysis.Results: A total of 60 patient transfer handovers were observed. In none of the evaluable cases observed, patient identification was conducted correctly according to the hospital policy at every step of the process. The principal system factor responsible was organizational failure, followed by technology and team culture issues. The Failure Modes and Effects Analysis methodology revealed that miscommunication between staff and lack of key patient information put patient safety at risk. Conclusions:Patient identification during intrahospital patient transfer is a high-risk event because several factors and many people interact. In this study, the disconnect between the policy and the reality of the workplace left staff and patients vulnerable to the consequences of misidentification. Where a policy is known to be substantially different from work as done, urgent revision is required to eliminate the serious risks associated with the unguided evolution of working practice.