AGENDA ITEM 4.5 3 September 2014 Title of the Health Board Report Professor Stephen Palmer Review of the use of Risk Adjusted Mortality Index (RAMI) data Executive Leads: Medical Director Authors: Mr Kamal Asaad, Medical Director Contact Details for further information: Mr Kamal Asaad kamal.asaad@wales.nhs.uk Purpose of the Health Board Report The purpose of this report is to inform the Board of the outcome of Professor Stephen Palmer s review, commissioned by the Health Minister, of the use of the risk adjusted mortality data (RAMI) in NHS Wales which was published in July 2014. Governance Link to Health Board Strategic Objective(s) Supporting evidence This report aligns fully with the Health Board s Triple Aim; To provide the highest possible service quality and excellent patient experience To improve health outcomes and help reduce inequalities To get high value from all services The information referenced within this report Engagement Who has been involved in this work? The Terms of Reference for Professor Palmer s review were outlined by the Health Minister, Welsh Assembly. In relation to Cwm Taf UHB the review and visits by Professor Palmer were coordinated by the Board Secretary and the Medical Director and members of staff from the Performance and Clinical Coding, Clinical Audit departments, along with the Chairman, and senior clinicians met with Professor Palmer and also provided various related information to inform the review. Professor S Palmer report Page 1 of 5 University Health Board Meeting
Health Board / Committee Resolution (insert ) To; APPROVE ENDORSE DISCUSS NOTE Recommendation The Board is asked to; NOTE and DISCUSS the report. Summarise the Impact of the Health Board Report Equality and diversity Legal implications Population Health Quality, Safety & Patient Experience Resources Risks and Assurance Standards for Health Services Workforce There are no equality and diversity implications of the report. There are no legal implications of this report. There are no specific population health issues. This report aims to provide assurance with regards the quality and safety of care provided to patients. There are no direct resource implications of this report. However, the Board should note the resources being used to ensure robust mortality review processes are in place. This report provides assurance that whilst the inpatient Risk Adjusted Mortality Index shows Cwm Taf UHB as an outlier when compared with other Welsh Health Boards, the mortality case note review process has not found any evidence of a systematic failure in healthcare leading to excess deaths. Crude mortality remains high due to a lack of alternative arrangements for many patients just prior to their death, and due to a population with a high rate of co-morbidities (the presence of one or more disorders in addition to the primary disease or disorder) often presenting late in the course of their illness, often as an emergency, limiting opportunities for effective therapeutic interventions. The details of this report relate mainly to; Standard 1: Governance & Accountability (supports the quality strategy, delivery plan and performance assurance/improvement) Standard 6: Participating in quality improvement activities (Quality Improvement is the essence of implementing the recommendations) Standard 7: Safe and clinically effective care (safe and effective care delivery is core) There are no direct workforce implications. Professor S Palmer report Page 2 of 5 University Health Board Meeting
PALMER STEPHEN REPORT ON MORTALITY 1. SITUATION / PURPOSE OF REPORT The purpose of this report is to inform the Board of the publication of Professor Stephen Palmer s review of the use of risk adjusted mortality data (RAMI) in NHS Wales, a review commissioned by the Health Minister, Welsh Assembly. 2. BACKGROUND / INTRODUCTION The Welsh Government and NHS Wales are committed to transparency of patient outcome and have been publishing an increasing number of quality indicators over the last 18 months. One measure which has caused some public confusion is the use of RAMI, which in Wales is calculated by an external company (CHKS). There continues to be widespread uncertainty among experts and public about the use of a single measure to indicate quality of care which is delivered by a complex range of hospital services. In March 2014, the Minister for Health and Social Services asked Professor Stephen Palmer, a distinguished epidemiologist, to undertake a review to advise the minister about the following questions: To what extent does risk adjusted mortality data provide valid information? How can these measures be interpreted for NHS Wales? Do hospitals with risk adjusted mortality indicators above 100 require further work on data or clinical quality? How are the NHS Wales Health Boards using clinical data for quality improvement? What is the quality of clinical data used by NHS Wales for improvement purposes? Are NHS organisations using clinical data effectively for quality improvement? What do the clinical data tell us about the quality of care in NHS Wales? 3. ASSESSMENT / GOVERNANCE AND RISK ISSUES Professor Palmer undertook significant stakeholder engagement and applied his wide knowledge & experience in Public Health and Epidemiology to this task. Professor Palmer visited Cwm Taf UHB twice. On the first occasion meeting with the Chairman, the Medical Director, and the Board Secretary. On the second visit he met with colleagues from information, performance, coding, clinical audit departments. He also met with Dr Jason Shannon who leads the senior clinical mortality review process in the UHB. Professor Palmer reviewed the Board reports on Patient Safety and the CHKS data. He also had a detailed discussion regarding the mortality review process. Subsequent to the two visits, Professor Palmer requested additional information which was sent to him. Professor S Palmer report Page 3 of 5 University Health Board Meeting
CONCLUSION Professor Palmer concludes that RAMI is not a meaningful measure of quality of care, indeed it is misleading. He does, however, recommend the use of meaningful and useful information to measure and describe quality of care in hospitals. With regard to Cwm Taf UHB, Professor Palmer s findings is that the Health Board has a high proportion of deaths occurring in hospital, a high underlying general mortality rate in the population, and a relatively high prevalence of risk factors such as smoking. The reasons for the high proportion of in-hospital deaths are lower provision of hospice care in the community and a trend for end of life patients to be transferred into hospitals from nursing homes. Consequently, in relation to Professor Palmer s understanding of the way that RAMI is calculated, he would expect that both hospitals would have a high RAMI. Professor Palmer considers that Cwm Taf Health Board has an impressive and high visibility clinical case notes mortality review process in place giving considerable reassurance to the Board that high RAMIs are not indicators of poor care. The mortality review process is rigorous with three stages (about 25% of cases referred from stage 1 to stage 2, and a small number requiring stage 3), but it is resource intensive and the timeliness of reviews and the sustainability of the second stage review process are live issues. In recent months due to staff illness a backlog of reviews has built up. The Board has received value from the review process, identifying areas for further work such as coagulation treatment, timeliness of scans, and failure to accurately communicate Do not resuscitate decisions. Importantly, a study has been carried out of the usefulness of the individual risk of death score calculated for each patient by the CHKS system. There was not a good correlation between risk of death and subsequent case note review findings and therefore it was rightly concluded that simply choosing those deaths which had a low probability score to review for avoidable factors would not be a sound approach. The Health Board has acknowledged that it has been relatively weak in clinical coding completeness, timeliness of coding and depth of coding (although there is now an improvement plan in place). All these factors will tend to increase the RAMI relative to hospitals with better coding levels. Professor Palmer confirmed that he had reviewed the Board data on RAMI with the Medical Director. Given his concerns about interpreting RAMI he was reluctant to offer an analysis since that may give unjustified weight to the measure. It is Professor Palmer s opinion that the high RAMIs are a result of several underlying factors that influence the model such as proportion of deaths in hospital rather than sub-optimal treatment. Having reviewed category specific mortality indicators and Quality and Safety Committee reports Professor Palmer could not see any particular issues that require intervention beyond ongoing efforts to improve the quality of coding across the Health Board and finding the resources to sustain the timeliness of clinical case notes review of all hospital deaths. Professor S Palmer report Page 4 of 5 University Health Board Meeting
A copy of Professors Palmer s full report is available via the following link;..\professor Stephen Palmer Reports\Healthcare Quality - Report - Independent Review of Mortality Data - 201.pdf 4. RECOMMENDATIONS The Board is asked to; Note and Discuss the findings outlined in the report. Professor S Palmer report Page 5 of 5 University Health Board Meeting