Creative use of existing clinical and health outcomes data to assess NHS performance in England: Part 1—performance indicators closely linked to clinical care
BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7505.1426 (Published 16 June 2005) Cite this as: BMJ 2005;330:1426All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Lakhani and colleagues have confirmed what those of us working in the
data analysis field in hospitals have known for a long time – the data
collected in the hospital Patient Administation Systems (which becomes the
‘HES’ data) are a very valuable source.
The good news is that the quality of these data can only improve as a
result of the need to produce data to secure funding under Payment by
Results, and as a result of the possibility of the data being used to
inform revalidation.
The transfer of the responsibility for doing something useful with the HES
data has passed from the Department of Health to the new Health and Social
Care Information Centre, and it is hoped that this will lead to better
feedback of data analyses to Trusts and commissioners.
All we then need is to ensure that people know how to use the analyses
appropriately. The important point is that people should use them to ask
better questions than would otherwise be asked: if they are used to make
snap judgments then the opportunity to use a tool for quality improvement
in healthcare will have been lost.
Competing interests:
None declared
Competing interests: No competing interests
Table 2 Headers 'Admissions for acute conditions treatable in primary
care' and 'Admissions for chronic conditions treatable in primary care'
are misleading,
1) at face value they are contradictory,
2) if these conditions are not treatable in hospital it may well be that
they cannot be treated in primary care either,
3) admissions from accident and emergency may reflect the management of
self limiting conditions in these departments (which itself may be a
factor influencing attendance even after correct advice and management
in primary care).
The conclusion that this table reflects factors in primary care may be
true but unproven, it could just as easily represent differing cultures
and attitudes of the accident and emergency departments. Therefore it is
misleading to use the above labels.
Indeed the authors allude to this possibility 'there may be other factors
outside the control of primary care.....such as accessability of accident
and emergency departments', there may well be other more pertinent service
factors, as well as those external to the NHS,( eg television viewing
habits), all of which would be valid areas of invesitigation. I suggest
the table is relabelled with the footnotes rather than the bald titles,
the reader is then free to draw their own conclusions, both concerning the
tables and areas of further research.
Nigel Konzon
Competing interests:
None declared
Competing interests: No competing interests
HSMRs as health outcomes data to assess NHS performance in England
We are pleased that the authors acknowledge the role of a summary
indicator of hospital mortality using Hospital Episode Statistics (HES) in
monitoring case fatality, although just such an indicator was suggested in
1999, (reference: Jarman B, Gault S, Alves B, Hider A, Dolan S, Cook A,
Hurwitz B, Iezzoni L. Explaining differences in English hospital death
rates using routinely collected data. BMJ 1999, 318; 1515-1520 5 June
1999) and has been published annually since January 2001. There are some
acknowledged differences in the detail between Lakhani et al’s suggested
case fatality indicator and the HSMR published by Dr Foster. Lakhani et al
suggest that the inclusion of the linked death registration data from the
Office for National Statistics is an improvement. We agree and can include
these in future analyses now that they have been made available to us.
However, in-hospital mortality is the important outcome that we wish to
measure and, also, the linked data are not available for analysis of more
recent records provided directly from the Nation Wide Clearing Service
(NWCS), and as such can be up to 20 months out of date following release
of the annual HES data. Moreover, we have compared Scottish data (where
out of hospital mortality has been available for some time) and the
inclusion of linked mortality makes little difference to the HSMR (the
correlation coefficient between HSMRs with the inclusion of all deaths
within 30 days and only in-hospital deaths when calculating HSMRs for
Scottish hospital data 2000/1 to 2002/3 is 0.924, p<_0.001. the="the" authors="authors" agree="agree" with="with" our="our" suggested="suggested" method="method" of="of" adjustment="adjustment" by="by" diagnosis="diagnosis" but="but" choose="choose" to="to" use="use" all="all" diagnoses="diagnoses" instead="instead" that="that" lead="lead" _80="_80" deaths="deaths" which="which" is="is" measure="measure" we="we" in="in" order="order" compare="compare" similar="similar" case-mixes="case-mixes" hospitals.="hospitals." they="they" do="do" not="not" provide="provide" any="any" evidence="evidence" what="what" difference="difference" this="this" makes="makes" overall="overall" measure.="measure." same="same" goes="goes" for="for" exclusion="exclusion" cancer="cancer" data.="data." p="p"/>Certainly when we have worked with individual trusts, removal of
malignancies from their HSMR calculation made little difference to their
overall HSMR value. We are always open to proven improvements in the
methodology and will continue to actively explore and sometimes revisit
the incorporation of other factors.
Brian Jarman
Paul Aylin
Imperial College Faculty of Medicine.
Competing interests:
We calculate Hospital Standardised Mortality Ratios at the Dr Foster unit of Imperial College Faculty of Medicine
Competing interests: No competing interests