by Juliette Malley and Lisa Trigg
The Care Quality Commission (CQC) in England recently announced that it will consider monitoring deaths in care homes as a way of uncovering poor quality care. On the face of it, this seems a good idea: inspections can only do so much and death rates have been found to highlight poor practice in hospitals such as Mid Staffs – albeit retrospectively. But is measuring death rates a solution to the problem of monitoring care home quality? Here are five reasons why it might not be.
1. People move to care homes at the end of their lives
The main role of hospitals is to improve the health of their patients and in that context death can be a sign that something has gone wrong. In contrast, most (but not all) people move to care homes permanently to live out the rest of their lives. The best outcome can often be that they are able to die with peace and dignity in their own room at the care home.
2. Using death rates to measure quality will miss an important part of the picture
The care that people in care homes receive should aim to keep them safe, but also to maximise their quality of life in the face of declining health. Poor care may or may not have an impact on survival, by hastening death. However it will always have an immediate and more significant effect on a person’s quality of life and can cause a great deal of pain and distress.
3. Care homes have too few residents to generate reliable statistics
Care homes have on average 25 places, and many have fewer than ten. Small resident numbers undermine the reliability of statistics because high (or low) death rates in any given year can be simply down to chance. This means that the reputation of care homes could be unfairly affected.
4. Identifying care homes with ‘too many’ deaths is complex
The scandal at Mid Staffs was uncovered because hospital death rates were higher than they should have been. However, using hospital-wide death rates in this way is controversial and subject to criticism. This stems from the difficulty in identifying how many of the deaths, from a myriad of causes, were actually preventable and therefore ‘too many’. In care homes, it is not preventable deaths that we are trying to identify but deaths that happened more quickly than expected. Differences in death rates will depend in part on the characteristics of residents, known as ‘case mix’ in the jargon. Large differences in death rates between, for example, nursing homes and residential care homes or between homes catering for people with learning disabilities and older people, would not be surprising. Simple comparisons that do not distinguish between care homes types or case mix are likely to be flawed. Complicated statistical manipulation can provide an insight into this problem but it is not a failsafe way of identifying bad care homes.
5. Focusing on deaths may do more harm than good
CQC may plan to use high deaths solely as a trigger to investigate a care home, but this does not make it any less important that death rates are an accurate indicator of poor quality. Investigations are time-consuming and potentially reputation-damaging and to avoid them some care homes may choose not to act in the interests of the resident. For example, care home providers may choose not to engage with programmes designed to improve end-of-life care because of concerns about increasing the number of deaths on-site, preferring instead to move residents to hospital in the final weeks of life. Another consequence may be that care homes refuse to accept new residents who are already extremely frail and unwell (which is more common, resulting from efforts to keep older people in their own homes for as long as possible), as they would be expected to die more quickly.
If not deaths, then what?
We welcome the CQC’s appetite to use data to inform their regulatory activity. There is very little routinely collected data for care homes, particularly when compared to NHS Trusts. This makes it extremely challenging for CQC to continuously monitor the quality and safety of care. While we recognise that there may be some instances where death rates are able to highlight poor practice, it is our opinion that (for the reasons set out above) most of the time this will not be the case.
Fortunately, there is a wealth of experience of measuring and monitoring quality in care homes, from the Adult Social Care Outcomes Toolkit (ASCOT) developed by PSSRU to the clinically-oriented measures used in the US (and many other countries). The US experience is noteworthy since standardised assessments and records were introduced through legislation following a series of scandals. There, quality indicators, including unexpected weight loss and pressure ulcers, are generated from a standardised assessment tool (the Resident Assessment Instrument) on a quarterly basis. Small numbers and case mix are still a challenge for these measures, but they are better indicators of quality and can alert CQC to poor practice earlier than death rates.
To solve the problem of continuously monitoring care home quality, there is a case for requiring providers to collect relevant data on residents’ health and quality of life. In this respect, the Transparency and Quality Compact, which has been voluntarily adopted by larger UK providers and contains measures similar to those used in the US, is a good starting point. Future development of the Transparency and Quality Compact, and quality measures in general, will continue to benefit from the experiences of testing and implementing quality measures here and internationally.
About the authors
Juliette Malley is Research Fellow within the Personal Social Services Research Unit at both the London School of Economics and Political Science and the University of Kent. She is also a member of the Quality and Outcomes of person centred care (QORU) policy research unit.
Lisa Trigg is Research Fellow within LSE Health at the London School of Economics and Political Science and is currently working on the FP7 European Union Cross Border Care Collaboration Project, conducting research into long-term care for older people with five country partners.