Article Text

Download PDFPDF
Application of AHRQ patient safety indicators to English hospital data
  1. A Bottle,
  2. P Aylin
  1. Dr Foster Unit at Imperial, Department of Primary Care & Social Medicine, Imperial College Faculty of Medicine, London, UK
  1. Dr A Bottle, Dr Foster Unit at Imperial, Department of Primary Care & Social Medicine, Imperial College Faculty of Medicine, St Dunstans Road, London SW7 2AZ, UK; robert.bottle{at}


Background: Patient safety is recognised worldwide as a major healthcare issue. The US Agency for Healthcare Research and Quality developed a series of evidence-based Patient Safety Indicators for use with hospital administrative data, but to date these have not been translated for use in the UK. They measure harm due to treatment and include infections, obstetric tears and foreign bodies left in the patient following surgery. We aimed to apply 10 of the AHRQ indicators for use in English routine hospital admissions data as the first step in validation, and describe their rates in relation to established measures of negative outcome such as mortality.

Methods: Translation of US coding systems into England systems using look-up files and clinical coding advice. Descriptive analysis of rates, length of stay, mortality and emergency readmission. Qualitative feedback from hospitals on their rates and level of interest in the indicators.

Results: The translation presented a number of challenges, particularly regarding the lack of direct correspondence between the two procedure coding systems. There were a total of 35 918 potential adverse events among the nine successfully translated indicators in England in the financial year 2005/6, with wide variation between hospital trusts. Potential adverse events were usually associated with higher mortality and unplanned readmission rates and longer length of stay. Postoperative sepsis, for example, had a rate of 4.2 per 1000 admissions and was associated with a median length of stay of 19 days compared with 6 days for admissions for operations without sepsis recorded, and was associated with a mortality of 14.7% compared with 0.5%.

Conclusions: These indicators have potential for use in tracking progress in harm-reducing programmes, but prospective evaluation of data quality and coding practices is required to fully assess their utility.

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.


  • Funding: AB is 100% and PA is 50% funded by Dr Foster Intelligence via a research grant for the Unit. The Dr Foster Unit at Imperial is affiliated with the Centre for Patient Safety and Service Quality at Imperial College Healthcare NHS Trust which is funded by the National Institute of Health Research.

  • Competing interests: The Unit is funded by a grant from Dr Foster Intelligence (an independent health service research organisation).

  • Ethics approval: Ethics approval was provided by the Patient Information Advisory Group (PIAG) and St Mary’s Local Research Ethics Committee.

  • Patient consent: Obtained.

  • AB and PA conceived the study. AB performed all analyses. AB and PA wrote the manuscript. AB is the guarantor for the study.