(How) do we learn from errors? A prospective study of the link between the ward's learning practices and medication administration errors.

Background: Attention in the ward should shift from preventing medication administration errors to managing them. Nevertheless, little is known in regard with the practices nursing wards apply to learn from medication administration errors as a means of limiting them. Aims: To test the effectiveness of four types of learning practices, namely, non-integrated, integrated, supervisory and patchy learning practices in limiting medication administration errors. Methods: Data were collected from a convenient sample of 4 hospitals in Israel by multiple methods (observations and self-report questionnaires) at two time points. The sample included 76 wards (360 nurses). Medication administration error was defined as any deviation from prescribed medication processes and measured by a validated structured observation sheet. Wards' use of medication administration technologies, location of the medication station, and workload were observed; learning practices and demographics were measured by validated questionnaires. Findings: Results of the mixed linear model analysis indicated that the use of technology and quiet location of the medication cabinet were significantly associated with reduced medication administration errors (estimate = .03, p < .05 and estimate = -.17, p < .01 correspondently), while workload was significantly linked to inflated medication administration errors (estimate = .04, p < .05). Of the learning practices, supervisory learning was the only practice significantly linked to reduced medication administration errors (estimate = -.04, p < .05). Integrated and patchy learning were significantly linked to higher levels of medication administration errors (estimate = -.03, p < .05 and estimate = -.04, p < .01 correspondently). Non-integrated learning was not associated with it (p > .05). Conclusions: How wards manage errors might have implications for medication administration errors beyond the effects of typical individual, organizational and technology risk factors. Head nurse can facilitate learning from errors by ''management by walking around'' and monitoring nurses' medication administration behaviors. [ABSTRACT FROM AUTHOR]