Article Text

Download PDFPDF
PP38 Estimation of the burden of preventable adverse events in england: bayesian evidence synthesis
  1. SI Watson,
  2. YF Chen,
  3. CA Taylor,
  4. RJ Lilford
  1. Warwick Medical School, University of Warwick, Coventry, UK


 Background Deaths and injuries resulting from preventable errors in the course of healthcare provision are a substantial problem potentially leading to the loss of thousands of lives annually. Reducing preventable adverse events (AE) is a major aim of the patient safety movement, however, generalizable estimates of the burden of preventable AEs using multiple sources of evidence do not exist. Our aim was to estimate the annual burden of preventable AEs in England in terms of both direct medical costs and losses to health.

Methods We conducted a Bayesian evidence synthesis in two steps. Firstly, we systematically reviewed the literature on preventable AEs. We identified studies through searches of MEDLINE, Cochrane Library, Embase, Econlit, and Google Scholar for articles published up to November 1st 2014 using combinations of the terms ‘preventable’, ‘adverse event’, ‘costs’, ‘health’, and ‘mortality’. We included studies that (i) reported the incidence/prevalence of AEs in acute care settings, (ii) assessed and reported their preventability, and (iii) associated AEs with costs or health outcomes. Secondly, we specified the marginal (i.e. not dependent on any other events occurring) per-patient distributions of costs and health losses, measured in quality adjusted life years (QALYs), in terms of costs and health losses conditional on different categories of preventable AEs occurring. Appropriate meta-analytic sub models allowing for between-study heterogeneity were then used to combine the data from each study and the final, overall model was estimated using Markov Chain Monte Carlo (MCMC) in one step.

Results Overall, 570 citations were identified, of which 29 were selected for full review, and 19 were included in the quantitative synthesis. From the Bayesian meta-analysis, the median (5th, 95th centiles) preventable AE rate in an acute setting was 5.0% (1.3, 12.7%). The median estimated annual direct medical costs resulting from preventable AEs was £3.014 billion (£0.744 billion, 8.069 billion) in 2012/13 costs, approximately 3.2% of the total NHS England budget. The in-hospital mortality rate from preventable AEs was 0.33% (0.01, 1.93%), approximately 9.1% (0.25, 53.8%) of the overall in-hospital mortality rate. The median estimated annual QALY loss resulting from preventable AEs in England was 160,211 (15,913, 1,078,487).

Conclusion Preventable AEs represent a substantial burden on healthcare systems. Interventions that reduce errors in the course of medical treatment may be considered ‘doubly efficient’ in that they may both directly reduce costs and health losses, however, further research is required on how to model the health and cost implications of such interventions.

  • adverse events
  • burden
  • Bayesian statistics

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.