The past few months have seen the beginnings of a shift in the prevailing NHS wisdom about patient safety and how best to promote and improve it. The emphasis is moving firmly towards greater openness and accountability - a point well made by Frank Burns in his recent HSJ piece (opinion, p16, 12 April). The appetite amongst the public for assurance that hospital care is delivered safely is reflected in the media with the.ever present interest in MRSA stories..
However, the NHS currently lacks measures of patient safety that allow national and local comparisons and that are meaningful to clinicians and patients alike. In the era of choice and, potentially, pay for performance mechanisms, the challenge is for NHS organisations not only to put safety to the top of their agenda but to prove it too. Yet how can they do so without consistent and comparable information?
With our partners at Imperial College London, we have for the first time ever undertaken an analysis of the performance of acute trusts in England against a set of patient safety indicators, based on those originally created by the Agency for Healthcare Research and Quality (AHRQ) and widely used in the US. The value of the AHRQ's indicators is that they are designed to use routine administrative data, similar to hospital episode statistics data, and therefore to allow comparison. They were selected by AHRQ after an extensive research period with the precondition that they should be amenable to prevention by healthcare organisations. They reflect events such as bed sores and foreign bodies left in after procedures.
We have analysed nine out of the AHRQ's full set of 20, collaborating with the Healthcare Commission in translating the codes for each indicator. Individual trusts have been provided with data showing their results against the national picture, and a national overview will be published in our forthcoming Hospital Guidereport. In order to understand the value of these indicators, we have.been asking trusts for their feedback on their data and also on the measures themselves.
This initiative has generated much interest and enthusiasm. The chief question relates to whether and how the results against a given indicator should be interpreted. A trust with a high rate of post-operative hip fracture may have a strong patient safety culture, reflected in a more rigorous approach to recording and coding information about incidents than in other trusts. On the other hand, it may simply have a poor safety record..
If it is the former, then it would seem that trusts may need to be challenging themselves and each other to improve the quality of coding, particularly at secondary level. If it is the latter, then doesn't that suggest the need for light to be shed on how trusts perform on key safety issues and how they compare with each other?
We think these are important questions and we are beginning a programme of work with a number of trusts to explore the issues in more depth. A review of the coding of secondary diagnoses within a number of pilot organizations will give us a better handle on the question of data quality, and the extent to which it needs to improve to make routine reporting and analysis possible. We will also explore the issues around publication and interpretation of the data.
Our objective is to create meaningful measures of patient safety which will not only be useful in prompting improvement but will also enable a more balanced debate around improvement and poor performance. Without the right information and measures, a patient safety culture is unlikely to develop in the NHS.
Helen Rowntree is the policy and research manager at Dr Foster Intelligence. A version of this article appears in Dr Foster Intelligence's Hospital Guide report, which will be published on Monday 23 April 2007.