Estimating the health burden and costs of patient safety events using English hospital data

There is a growing literature that estimates excess costs of hospital inpatient care due to patient safety events. However, most existing studies are focused on hospitals in the US, and are somewhat limited by small sample size and narrow range of medical errors analyzed. This paper aims to estimate the costs of six patient safety events using English hospital administrative admission information in the Hospital Episode Statistics (HES). Safety events can cause significant harm to patients in the form of excess morbidity and even mortality, and in fact the impact to patients may far exceed the costs to healthcare providers. The innovative contribution of our work is that we estimate the burden to both healthcare providers and patients by supplementing HES data with secondary data sources on health-related quality of life estimates for six patient safety events. This allows us to generate a more encompassing estimate of the societal burden of patient safety events, to ultimately give guidance to policy makers on the value of preventing such events with suitable interventions.

We identify patient safety events with the Patient Safety Indicator (PSI) algorithms of the Agency for Healthcare Research and Quality (AHRQ), using English hospital data for the period from 2005 to 2010. This paper focuses on 6 PSIs for which health-related quality of life weights are available: death in low mortality HRGs, decubitus ulcer, selected infections due to medical care, post-operative hip fracture, post-operative pulmonary embolism or deep vein thrombosis, and post-operative sepsis.

We calculate the additional hospital inpatient costs by comparing patients with a patient safety event recorded over similar patients from a control group that is constructed according to AHRQ definitions. Hospital costs are obtained by linking the data from the National Schedules of Reference Costs to inpatient hospitalization information in HES. Inpatient hospitalization costs are modeled with an instrumental variable approach to address potential endogeneity of patient safety events, where depending on the type of patient safety event, weekday and month of the admission and/or workforce characteristics are used as instruments.

The marginal health burden due to the PSI is also calculated by comparing the Quality Adjusted Life years between the two groups of patients. Linking the HES data to the Office of National Statistics’ mortality data, the time durations from admission to final discharge or to death in hospital or within 30 days are examined using parametric survival time models. Quality of life weights are drawn from published literature. 

Our study provides urgently needed guidance to UK policy makers. Improving patient safety has been put at the top of the policy agenda over the past decade in many countries. While cost-effectiveness analyses could provide important evidence when setting safety improvement initiatives, evidence is very limited. This is mainly due to the difficulty of measuring the costs of patient safety events in a comprehensive fashion, which is essential to value the economic benefits of interventions and to inform health policy and patient safety improvement priorities.