Two incorrect assertions about study of hospital standardised mortality ratios in the West MidlandsBMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f1088 (Published 05 March 2013) Cite this as: BMJ 2013;346:f1088
The series of BMJ articles on hospital mortality statistics and Mid Staffordshire NHS Foundation Trust underscores the complex issues involved.1 2 3 Hawkes and Godlee make two key assertions about our work.
Firstly, that there was a conspiracy in the West Midlands to discredit hospital standardised mortality ratios (HSMRs)—this is fiction. The medical director of the West Midlands Strategic Health Authority and regional director of public health (Rashmi Shukla) took the accusation of unnecessary deaths seriously and acted with integrity and the highest ethical and professional standards. The authority approached us because we had a track record in this type of work and an existing NHS levy contract to provide such support. We proposed a comprehensive scientific study based on hypothesis generation and tested using controlled comparisons.4 Our five candidate hypotheses included examining the association between quality of care and HSMRs and the extent to which a high HSMR might signal a “failing” organisation.
We found weak evidence of a link between HSMRs and staff and patient survey variables, suggesting that the link between HSMRs and organisational factors is ambiguous. We found no systematic association between HSMRs and quality of care or evidence of serious methodological bias in the construction of HSMRs. Our subsequent work suggests that about 10 of 11 raised HSMR signals would be false alarms and, more worryingly, 10 of 11 poorly performing hospitals would escape attention.5 We did not set out to discredit HSMRs—our findings did that.
Secondly, the assertion that the Mid Staffordshire scandal invalidates our research findings is fallacious, because the use of HSMRs to screen for “bad” hospitals cannot be upheld on the basis of one true positive finding.
Cite this as: BMJ 2013;346:f1088
Competing interests: None declared.
Log in using your username and password
Log in through your institution
Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
Sign up for a free trial