Nabhan, M., Elraiyah, T. B., Brown, D. R., Dilling, J. A., LeBlanc, A., Montori, V. M., Morgenthaler, T. I., et al. (2012). What is preventable harm in healthcare? A systematic review of definitions. BMC Health Services Research, 12(1), 128. Available from: http://www.biomedcentral.com/1472-6963/12/128/abstract
Background: Mitigating or reducing the risk of harm associated with the delivery of healthcare is a policy priority. While the risk of harm can be reduced in some instances (i.e. preventable), what constitutes preventable harm remains unclear. A standardized and clear definition of preventable harm is the first step towards safer and more efficient healthcare delivery system. We aimed to summarize the definitions of preventable harm and its conceptualization in healthcare.
Methods: We conducted a comprehensive electronic search of relevant databases from January 2001 to June 2011 for publications that reported a definition of preventable harm. Only English language publications were included. Definitions were coded for common concepts and themes. We included any study type, either original studies or reviews. Two reviewers screened the references for eligibility and 28 % (127/460) were finally included. Data collected from studies included study type, description of the study population and setting, and data corresponding to the outcome of interest. Three reviewers extracted the data. The level of agreement between the reviewers was calculated.
Results: One hundred and twenty seven studies were eligible. The three most prevalent preventable harms in the included studies were: medication adverse events (33/127 studies, 26 %), central line infections (7/127, 6 %) and venous thromboembolism (5/127, 4 %). Seven themes or definitions for preventable harm were encountered. The top three were: presence of an identifiable modifiable cause (58/132 definitions, 44 %), reasonable adaptation to a process will prevent future recurrence (30/132, 23 %), adherence to guidelines (22/132, 16 %). Data on the validity or operational characteristic (e.g., accuracy, reproducibility) of definitions were limited.
Conclusions: There is limited empirical evidence of the validity and reliability of the available definitions of preventable harm, such that no single one is supported by high quality evidence. The most common definition is “presence of an identifiable, modifiable cause of harm”.