November 15, 2017

expert reaction to paper on health and social care spending and excess deaths in England

A new paper, published in BMJ Open, reports that health and social care spending cuts linked to 120,000 excess deaths in England.

Dr Richard Fordham, Senior Lecturer in Health Economics, University of East Anglia, said:

“This paper tackles a complexity of issues related to historic  health and social care funding levels, macroeconomic conditions and NHS staffing patterns. But necessarily longitudinal studies are fraught with difficulties in proving causality and despite the best statistical efforts of the authors, one should treat their conclusions with some caution. Different time periods are rarely identical (try looking back at your old photographs and see if you recognize yourself and your friends even if only ten years has elapsed), so too practices change and unexpected events occur.

“The main conclusion of the paper is that it is older people, in particular those resident in nursing homes or in need of acute medical care, who have weathered most of the funding crisis. The paper also found nurse numbers declined substantially between the two periods and together with a crunch in hospital spending  (owing to funding not keeping up with rising demand)  may have impacted on overall quality of care and therefore been reflected in rising mortality rates.  This is indeed a plausible hypothesis. But other explanations are available. Patient cohorts may have changed (for example more end-stage, longer-term illnesses); patients may have succumbed to  different or new diseases (eg. MRSA, cirrhosis etc.); or had greater multiple morbidities (asthma plus diabetes plus cancer etc. ) than similar cohorts of the same age before them?

“The nursing home argument is equally alluring but falls over when you consider that privatisation of nursing homes took place largely in this period (granted some of which failed spectacularly). This could have resulted in freeing up money that could have been spent elsewhere in the social care system (possibly on other services for the elderly)? Nurses don’t make up the majority of care home staff in any case and their vanishing numbers would have been insulated by low-paid, usually unskilled care-assistants.  However, an argument that poor care in nursing homes, after delayed hospital discharged or a spell back in a struggling hospital could have resulted (indirectly) in excess deaths, could be made? The study is a worthy attempt to unravel a complex interaction of micro- and macro-economics and highlights that further economic insight is still needed.”

Prof. Martin Roland, Emeritus Professor of Health Services Research, University of Cambridge, said:

“Mortality has been declining in most developed nations for several decades. Commentators have previously noted that deaths have stopped declining in the UK while they continue to decline in other countries. This study suggests that a change happened to cause deaths to stop declining around 2014. This is likely to be a correct finding. However, the link to health and social care spending is speculative as observational studies of this type can never prove case and effect. The projections on ‘what it would cost to save lives’ are particularly speculative, not least because the causal mechanism is unclear – i.e. what would you spend the money on in order to save lives.

“The study finds that the excess deaths are largely in nursing homes for those over 60, with substantially fewer deaths in hospitals, especially for the under 60s. This could in part, as the authors suggest, be due to patients being moved from hospital to care home when near the end of their life.

“Reductions in social care spending are associated with increased deaths in care homes (more than changes in health spending), and reductions in health spending had no effect on hospital mortality. The authors attempt to explain the effects in terms of nurse staffing levels, but without seeing the online appendix the reader can’t tell if these nurses are mainly employed in hospitals or care homes. If it’s the former, then the postulated relationship between nurse staffing levels and deaths would be unconvincing. Furthermore, nursing levels weren’t actually falling during the period of rising mortality (2010 to 2014), just rising less rapidly than before.

“This study presents the classic problems of interpreting observational data. The data are highly complex, with a number of possible explanations (known and unknown) for the change in mortality. Certainly the recent change in previous trends of reducing mortality is of concern. However, the authors overstate the certainty of this link to funding and are highly speculative about the money needed to ‘save lives’ in future.”

* ‘Effects of health and social care spending constraints on mortality in England: a time trend analysis’ by Watkins et al. published in BMJ Open at on Wednesday 15 November. 

Declared interests

Dr Richard Fordham: No conflicts of interest

Prof. Martin Roland: No conflicts of interest