Volume of procedures & outcome of treatment: The NHS needs to Understand the relationship more
In his recent editorial, Michael Soljak highlights recent research
suggesting that in the majority of clinical areas studied patients treated
by physicians or hospitals with high volumes have better outcomes than
those with low volumes.(1) This has been particularly the case in surgery,
where the procedure relies primarily on individual physicians. In areas,
where the delivery of care is shared over a group or team, the evidence of
a volume-outcome relationship is less strong.(2)
The editorial also refers to a systematic review undertaken by Halm et al
which describes the current theories given for high volume being related
to better outcomes than low volume are that practice makes perfect, and
the effects of selective referral, where hospitals whose performance is
said to be good are referred to more often than other hospitals.(3) The
review goes on to criticise these theories as lacking substantial
empirical support. That is to say, there is little evidence as to reveal
exactly why bigger volume units may be related to better outcomes. Indeed
there is recognition that other reasons may exist to explain why outcomes
may be better in higher volume hospitals. Specifically it has been
suggested that some of the differences in outcomes between high and low
volume providers could be attributed to the greater uptake of specific
interventions in high than low volume hospitals.(4) There is a natural
extension of this theory that relates to the timing of the introduction of
new therapies or interventions, and how this affects the comparisons of
high and low volume hospitals. It is often the case that clinical trials
of new interventions are conducted in specialist centres or hospitals,
which in turn tend to be hospitals with high volumes. When an intervention
is found to be effective, then those patients randomised to receive that
intervention will have benefited from it, and consequently when these
hospitals are compared with other low volume hospitals it would be
expected that their outcomes appear better. In due course, once the new-
found intervention is introduced it will become generally adopted in all
hospital types. Thus the apparent improvement in the high volume hospitals
will disappear.
The above can be illustrated using an example from research conducted in
neonatal intensive care. In the United Kingdom in 1988-90 high volume
hospitals appeared to have better patient outcomes than low volume
hospitals, but in 1998-99 this effect appeared to have disappeared.(5,2)
In 1988-90 surfactant therapy was slowly being introduced to the UK. Also,
the effects of maternal steroids were being fully realised around this
time too. These two therapies, both of which have substantial proven
positive effects were administered more frequently in the high volume
hospitals than in the low volume hospitals.(6) Consequently, the
different uptakes of these therapies in low and high volume hospitals may
have contributed to the difference in their respective outcomes for the
1988-90 patients. Moving forward 10 years to 1998-99, surfactant and
maternal steroids are now routinely administered and the uptake of these
therapies is similar in high and low volume hospitals.
Furthermore,
despite on-going randomised controlled trials, there have been no new
major interventions in neonatal intensive care during the last 10 years
that have been identified and implemented. Consequently if the major
factor impacting the volume-outcome relationship in neonatal intensive
care is the uptake of effective therapies then there is little surprise
that in 1988-90 there was a difference in outcome and by 1998-99 no such
difference was observed. An alternative explanation could have been that
patients were no longer treated at hospitals whose initial risk-adjusted
mortality was high.(7) This is unlikely, because hospitals in the earlier
study did not receive feedback of their risk-adjusted outcomes until 1993.
In developing ways of improving the performance of the NHS it is important
to be aware of the effect different uptakes of new therapies have on
volume-outcome relationships. There is a need to identify and quantify the
relative importance of all factors that impact on the volume-outcome
relationship. Such knowledge would inform policymakers who also have to
weigh up issues of access to local, usually smaller hospitals and more
distant larger hospitals in a centralised service.
Dr Gareth Parry
Senior Research Fellow
Medical Care Research Unit,
University of Sheffield
Dr Janet Tucker
Senior Research Fellow
Dugald Baird Centre for Research on Women’s Health,
University of Aberdeen
Prof. William Tarnow-Mordi
Westmead and Children’s Hospital at Westmead,
University of Sydney
1. Soljak M. Volume of procedures and outcome of treatment: The NHS
needs to harness the relation more effectively. BMJ 2002;325:787-788
2. UK Neonatal Staffing Study Collaborative Group. A prospective
evaluation of patient volume, staffing and workload in relation to risk-
adjusted outcomes in a random, stratified sample of all UK neonatal
intensive care units. Lancet 2002; 359:99-107
3. Halm EA, Lee C, Chassin MR. How is volume related to quality in health
care? A systematic review of the research literature. Washington, DC:
Institute of Medicine, 2000
4. Thiemann DR, Coresh J, Oetgen WJ, Powe NR. The association between
hospital volume and survival after acute myocardial infarction in elderly
patients. N Eng J Med 1999;340:1640-8
5. International Neonatal Network. The CRIB (clinical risk index for
babies) score: a tool for assessing initial neonatal risk and comparing
performance of neonatal intensive care units. Lancet 1993; 342:193-198
6. Scottish Neonatal Consultants’ Collaborative Study Group, International
Neonatal Network. Trends and variations in use of antenatal
corticosteroids to prevent neonatal respiratory distress syndrome:
recommendations for national and international comparative audit. Br J
Obstet Gynaecol 1996; 103:534-540
7. Hannon EL, Siu AL, Kumar D, Kilburn H, Chassin MR. The decline in
coronary artery bypass graft surgery mortality in New York State: The role
of surgeon volume. JAMA 1995; 273(3): 209-213
Rapid Response:
Volume of procedures & outcome of treatment: The NHS needs to Understand the relationship more
In his recent editorial, Michael Soljak highlights recent research
suggesting that in the majority of clinical areas studied patients treated
by physicians or hospitals with high volumes have better outcomes than
those with low volumes.(1) This has been particularly the case in surgery,
where the procedure relies primarily on individual physicians. In areas,
where the delivery of care is shared over a group or team, the evidence of
a volume-outcome relationship is less strong.(2)
The editorial also refers to a systematic review undertaken by Halm et al
which describes the current theories given for high volume being related
to better outcomes than low volume are that practice makes perfect, and
the effects of selective referral, where hospitals whose performance is
said to be good are referred to more often than other hospitals.(3) The
review goes on to criticise these theories as lacking substantial
empirical support. That is to say, there is little evidence as to reveal
exactly why bigger volume units may be related to better outcomes. Indeed
there is recognition that other reasons may exist to explain why outcomes
may be better in higher volume hospitals. Specifically it has been
suggested that some of the differences in outcomes between high and low
volume providers could be attributed to the greater uptake of specific
interventions in high than low volume hospitals.(4) There is a natural
extension of this theory that relates to the timing of the introduction of
new therapies or interventions, and how this affects the comparisons of
high and low volume hospitals. It is often the case that clinical trials
of new interventions are conducted in specialist centres or hospitals,
which in turn tend to be hospitals with high volumes. When an intervention
is found to be effective, then those patients randomised to receive that
intervention will have benefited from it, and consequently when these
hospitals are compared with other low volume hospitals it would be
expected that their outcomes appear better. In due course, once the new-
found intervention is introduced it will become generally adopted in all
hospital types. Thus the apparent improvement in the high volume hospitals
will disappear.
The above can be illustrated using an example from research conducted in
neonatal intensive care. In the United Kingdom in 1988-90 high volume
hospitals appeared to have better patient outcomes than low volume
hospitals, but in 1998-99 this effect appeared to have disappeared.(5,2)
In 1988-90 surfactant therapy was slowly being introduced to the UK. Also,
the effects of maternal steroids were being fully realised around this
time too. These two therapies, both of which have substantial proven
positive effects were administered more frequently in the high volume
hospitals than in the low volume hospitals.(6) Consequently, the
different uptakes of these therapies in low and high volume hospitals may
have contributed to the difference in their respective outcomes for the
1988-90 patients. Moving forward 10 years to 1998-99, surfactant and
maternal steroids are now routinely administered and the uptake of these
therapies is similar in high and low volume hospitals.
Furthermore,
despite on-going randomised controlled trials, there have been no new
major interventions in neonatal intensive care during the last 10 years
that have been identified and implemented. Consequently if the major
factor impacting the volume-outcome relationship in neonatal intensive
care is the uptake of effective therapies then there is little surprise
that in 1988-90 there was a difference in outcome and by 1998-99 no such
difference was observed. An alternative explanation could have been that
patients were no longer treated at hospitals whose initial risk-adjusted
mortality was high.(7) This is unlikely, because hospitals in the earlier
study did not receive feedback of their risk-adjusted outcomes until 1993.
In developing ways of improving the performance of the NHS it is important
to be aware of the effect different uptakes of new therapies have on
volume-outcome relationships. There is a need to identify and quantify the
relative importance of all factors that impact on the volume-outcome
relationship. Such knowledge would inform policymakers who also have to
weigh up issues of access to local, usually smaller hospitals and more
distant larger hospitals in a centralised service.
Dr Gareth Parry
Senior Research Fellow
Medical Care Research Unit,
University of Sheffield
Dr Janet Tucker
Senior Research Fellow
Dugald Baird Centre for Research on Women’s Health,
University of Aberdeen
Prof. William Tarnow-Mordi
Westmead and Children’s Hospital at Westmead,
University of Sydney
1. Soljak M. Volume of procedures and outcome of treatment: The NHS
needs to harness the relation more effectively. BMJ 2002;325:787-788
2. UK Neonatal Staffing Study Collaborative Group. A prospective
evaluation of patient volume, staffing and workload in relation to risk-
adjusted outcomes in a random, stratified sample of all UK neonatal
intensive care units. Lancet 2002; 359:99-107
3. Halm EA, Lee C, Chassin MR. How is volume related to quality in health
care? A systematic review of the research literature. Washington, DC:
Institute of Medicine, 2000
4. Thiemann DR, Coresh J, Oetgen WJ, Powe NR. The association between
hospital volume and survival after acute myocardial infarction in elderly
patients. N Eng J Med 1999;340:1640-8
5. International Neonatal Network. The CRIB (clinical risk index for
babies) score: a tool for assessing initial neonatal risk and comparing
performance of neonatal intensive care units. Lancet 1993; 342:193-198
6. Scottish Neonatal Consultants’ Collaborative Study Group, International
Neonatal Network. Trends and variations in use of antenatal
corticosteroids to prevent neonatal respiratory distress syndrome:
recommendations for national and international comparative audit. Br J
Obstet Gynaecol 1996; 103:534-540
7. Hannon EL, Siu AL, Kumar D, Kilburn H, Chassin MR. The decline in
coronary artery bypass graft surgery mortality in New York State: The role
of surgeon volume. JAMA 1995; 273(3): 209-213
Competing interests: No competing interests