Gordon D Murray PhD FRSEProfessor of Medical Statistics, University of Edinburgh
Graham M Teasdale FRCSG FRCPLond FRSEPresident, Royal College of Physicians and Surgeons of Glasgow
Correspondence to Professor TeasdaleRoyal College of Physicians and Surgeons 242 St Vincent Street Glasgow GH2 5RJ
email: President@rcpsg.ac.uk
SMJ 2006 51(1): 17-22
Abstract
In planning health service delivery, there are tensions between the desire
to provide as much as possible locally yet take advantage of concentration.
There is, therefore, a need to define the extent of benefits of concentration
and specialisation and to understand the factors involved. Varying views are
expressed, reflecting selective use of the available information. This review
was undertaken to assist in the development of a National Framework for Service
Change in Scotland. From publications between 1997 and 2004, it considered over
500 abstracts and 50 papers in full. The findings are that, compared to earlier
reviews there has been an increase in the quality of research on these topics
and there is now a substantial body of evidence of an association between volume
of activity and outcome of care. The pattern of the relationship is likely to be
influenced by a range of factors, including the frequency and complexity of the
intervention. The choice of index is another factor and an increased use is
expected of more sensitive indicators of the quality of experience as against
simple mortality. Although more
understanding is needed of the causality of the relationships and of the effects
of manipulating service provision to promote concentration and specialisation,
these factors need to be taken into account in planning of future health
services.
Key
words: Health Care, volume, outcome
Introduction
The
history of concern about an interaction between experience, volume and outcome
of surgery dates back in Scotland for almost a century and a half.
In 1869, Sir James Y Simpson surveyed practising ‘Country surgeons’
and small, medium and large sized hospitals in England and Scotland about their
experience of operations for amputation (Simpson
1871, quoted in Neuhauser 2005). Simpson found that, among country surgeons,
greater experience was associated with a lower mortality rate.
In contrast, the bigger the hospital, the worse the mortality – which
Simpson attributed to overcrowding and unsanitary conditions.
These observations neatly illustrate the complexity of the topic and
issues perhaps even more relevant today.
In
1979 Luft and his colleagues refocused attention on the possibility of a
connection between an increased volume of clinical work and improved outcome.
Since then, many further studies have reported this relationship.
Nevertheless, certain issues identified by Luft remain to be clarified so
that controversy, even
conflict continues about the importance of the "complexity" of the
condition and the risks of treatment, the existence of "thresholds” in
the relationship, and the relative roles of individual practitioner versus unit
or hospital volume. The need for
risk stratification and the importance of using an appropriate index of outcome
in investigating these issues has been recognised.
Much
of the relevant information is derived from routine statistics. These allow little opportunity for risk stratification and
provide outcome only as mortality. For
interventions with a low risk of death, large numbers must be studied in order
not to overlook an effect that could be important if the intervention is very
common. For many interventions,
mortality is not an appropriate index and other indicators of effectiveness, of
quality of care and of patient satisfaction are required.
In
1997, the NHS Centre for Reviews and Dissemination at the University of York
published a systematic review of the evidence available on the volume/outcome
relationship in health care, based upon data available up to 1996 (Sowden et al 1997). Although they identified many studies that they
considered suitable, that showed a volume outcome association, in essence, the
conclusion reached was that the bulk of research evidence was methodologically
flawed and of little value in forming decisions about the planning of the
delivery of health services. Since
then, substantial additional literature has been published, including other
comprehensive reviews
This
paper is based on a review was undertaken in response to a request in the summer
of 2004 from the Advisory Group to the National Framework for Service Change in
the NHS in Scotland. The purpose
was to re-review the research evidence, in particular taking account of papers
published since 1997. In view of
the time constraints placed by the timetable of the work of the Service Change
Group, it was accepted that the approach taken would be that of a conventional
narrative review, not a formal, comprehensive systematic review.
Background
The
York Review (Sowden et al, 1997)
reached three main conclusions on volume/outcome relationships:
(i)
Case-mix:
“Most of the existing research,
because it does not sufficiently take account of differences in case-mix,
probably overestimates the impact of volume on the quality of care.”
(Summary Report, Page 10).
(ii)
Causation:
“… because none of the research
indicates that increasing activity over time leads to improvements in clinical
outcome, it is difficult to infer from results of cross-sectional studies which
show better outcomes in higher-volume units that similar differences in outcomes
can be expected by the expansion of an existing unit.” (Summary Report,
Page 10).
(iii)
Thresholds:
“The most that the research
evidence can support is a conclusion that if there are significant quality gains
from increased volume, these gains appear to be exhausted at relatively low
threshold levels. Volumes of
activity above these thresholds should be achievable without significant
structural changes, but may require a more sharply defined internal division of
labour across consultant staff (which may be consistent with increased
sub-specialisation within disciplines).” (Summary Report, Page 11).
Methods
A
general literature search was undertaken on volume/outcome relationships
together with more focused literature searches on methodological aspects of
volume/outcome relationships, on studies relating to volume/outcome
relationships which evaluated the impact of an intervention to increase
regionalisation, and methodological aspects of assessing clinical ‘learning
curves’. These searches
concentrated on publications from 1997 onwards, although some key references
predate this.
These literature
searches showed that more relevant papers have been published between 1997 and
2004 than were published up to and including 1996, and so it would indeed be a
major undertaking to perform a formal systematic review of the more recent
literature. This review was
therefore selective, based on reading abstracts and obtaining full papers when
appropriate and when they were relatively easily accessed.
As most papers in this area are observational and cross-sectional by
design, and are often very specific to a local area (e.g. a single US state),
they tend to be published in low impact journals which are not held by the local
University libraries. Thus most of
the full papers reviewed are either from high impact journals or from journals
whose contents can be accessed electronically.
Results
Over
500 abstracts were scanned and 50 full papers (Appendix I) reviewed in detail.
The vast majority of the papers relate to surgical procedures with
outcomes assessed in terms of short-term (in hospital or 30 day) mortality.
There is also an increasing number of papers relating to surgical
oncology with outcomes assessed in terms of long-term (5 years plus)
survival.
Many
common procedures have very low surgical mortality rates, so that even if a
volume/outcome relationship does exist for mortality, it would be difficult to
demonstrate and might be of limited clinical relevance.
So studies focussed on morbidity outcomes are also becoming more common.
For example, avoidance of a stoma in bowel surgery has a major impact on
quality of life, making it a very relevant patient-centred outcome measure
(Hodgson et al, 2003) and similar issues apply in prostate cancer (Begg at
al, 2002). Sometimes, there may be a long delay between intervention and
its consequence (eg in obstetrics) and / or the effect may be exerted indirectly
(e.g. screening for cancer) so that ‘intermediary’ indices of patient
outcome are needed.
Methodological
Quality
The
York Review (Sowden et al, 1997) was
rightly critical of the methodological quality of volume/outcome studies
published prior to 1997. There is
clear evidence that case-mix adjustment using clinical data on individual
patients leads to the most reliable results and that case-mix adjustment using
only administrative data tends to overestimate the magnitude of volume/outcome
relationships. Studies which use no
case-mix adjustment at all are likely to overestimate such effects even more so.
This is now widely recognised and methodological standards are higher in
recent studies. For example, in the
Gandjour review (2003), 16 of the ‘best’ 33 hospital volume/outcome studies
published between 1990 and 2000 were published in 1999 or 2000.
Increasingly
(and appropriately!) sophisticated statistical approaches are now being used to
try to disentangle the complex issue of whether it is surgeon volume or hospital
volume which drives the observed volume/outcome relationships (Birkmeyer et
al 2003, Panageas et al 2003).
Findings
Methodologically
flawed studies are still published, but there is now a strong core of
methodologically sound papers which use high quality data and appropriate
statistical methods to explore volume/outcome relationships.
These papers are based on either series of patients with data extracted
from administrative systems (giving very large sample sizes but incomplete
case-mix adjustment) or on series of patients with data extracted from clinical
databases (giving smaller sample sizes but good case-mix adjustment).
Even when one restricts attention to these higher quality studies there
is still very strong evidence of an association between volume and outcome in
the direction that high volume surgeons and high volume hospitals tend to have
superior outcomes compared to low volume surgeons and hospitals.
The magnitude of this effect, and how it depends on the clinical area, is
discussed in Section 4.4.
Two
particularly useful systematic reviews were identified.
Halm et al (2002) reviewed
studies published between January 1980 and December 2000 and Gandjour et
al (2003) reviewed studies published between January 1990 and December 2000.
Halm et al was a conventional
systematic review covering 27 procedures and diagnoses.
In the 135 studies that meet their criteria a statistically significant
relationship between higher volume and better outcomes were found for 71% of
studies of hospital volume (Appendix II) and for 69% for clinician volume
(Appendix III). The review of Gandjour et
al covered 34 diagnoses and interventions and included another 26 reports
not analysed by Halm et al. In a
total of 76 studies, higher hospital volume was statistically significantly
better in 51, non-significantly better in 21, non-significantly worse in 3 and
significantly worse in only one. These
authors took the unusual additional approach of identifying the single most
reliable study (based on criteria such as the quality of case-mix adjustment)
for each of a number of procedures. In
20 such ‘best’ studies, high volume was significantly better in 10,
non-significantly better in 6, non-significantly worse in 3 and significantly
worse in one.
Thresholds
Ramsay
et al (2001) undertook a systematic review of methods used to
analyse learning curves in health care, and more recently, Cook et
al (2004) proposed methods for adjusting for learning curve effects in
randomised trials of surgical interventions.
There
was no clear consensus on appropriate ways to analyse learning effects, with one
of the major problems being that as experience is gained in a new technique it
tends to be deployed for higher risk patients, meaning that outcomes can
deteriorate as experience is gained. This
requires careful case-mix adjustment to interpret correctly, but almost by
definition there are insufficient data for such analyses early in the learning
experience.
The
Leapfrog Group, a large US-based consortium of health care purchasers (Birkmeyer
et al 2004) places great importance on volume thresholds.
However, the impression from the literature is that definitions of ‘low
volume’ and ‘high volume’ relate more to potential
volumes than to any objective evidence on the actual
level of activity which is required to achieve and/or maintain competence.
For example, a unit performing 100 carotid endarterectomies per year
could be classified ‘high volume’ whereas a unit performing 400 coronary
artery bypass graft procedures per year could be classified as ‘low volume’
(Gandjour et al 2003, Birkmeyer et
al 2004).
Studies
which present outcome data for a range of activity volumes, as opposed to a
simple low volume/high volume dichotomy, do often report poor outcomes at low
activity levels and then a levelling off with outcomes in moderate volume units
being comparable to outcomes in high volume units.
This is partly the result of a statistical artefact, with greater
variation being observed in the small samples which derive from low volume
units. However when this excess
variation is taken into account there is still evidence of poorer outcomes being
observed in very low activity units (see, for example, the review by Shahian and
Normand, 2003).
Causation
In
spite of there being very strong evidence of an association between increased
volume and better outcomes, there are remarkably few studies which try to assess
if this association is causal. It
was stated in the York review that there was no evidence that increasing the
volume of activity in a given unit would lead to an improvement in outcomes. This reflected a lack of evidence rather than evidence of a
lack of effect. Evidence in this area is still extremely limited, but a number
of studies are beginning to appear which evaluate the impact of interventions
designed to concentrate activity
Trauma
systems are an area with a long history of regionalisation, and where different
approaches adopted by different countries constitute a ‘natural experiment’
on the organisation of trauma care. Nathens
et al (2004) review the history of
trauma management in the US and in France, and demonstrate how outcomes of
trauma victims improved in the US following the introduction of regionalisation.
However, there was a substantial lag period between the introduction of
regionalisation and an observed improvement in outcome.
The
UK Neonatal Staffing Study Group (2002) reviews the evidence for regionalisation
of neonatal intensive care units. The
situation is complex, but evidence of volume/outcome relationships from older
studies is not seen currently, and this is ascribed to lower volume units
adopting developments in treatment which were initially used only in high volume
units. Training and staffing levels
appear to be more important than volume per
se. The study also raises the
caveats that high volume units with a large number of consultant staff had
higher levels of nosocomial bacteraemia, and that units running close to
capacity have worse outcomes then when there is ‘slack’ in the system.
These findings on the importance of staffing levels are consistent with
the analysis of English hospital death rates published by Jarman et
al (1999).
Nobilio
et al (2004) report on the impact of regionalisation of cardiac
surgery in an Italian region. They
looked at patient outcomes, accessibility for patients and the efficiency of
referral systems following the adoption of a “hub & spoke” model.
The study does provide evidence of benefit, and the authors conclude that
their findings suggest “that policies aimed at increasing cooperation rather than competition
among health service providers have a positive impact on quality of care.”
This
latter finding is consistent with data from the Lothian Surgical Audit which was
presented at the recent Consensus Conference of the Association of Surgeons of
Great Britain and Ireland on “Modernising Medical Careers and General
Surgery”. (Robson et al, 2005). In
the Lothian’s experience, restructuring of emergency surgical care, focussed
on sub-specialisation appropriate to upper and lower abdominal conditions, has
led to improved quality of care and outcome.
Examples
of the Magnitude of Volume/Outcome Associations
Halm
et al (2002) summarised volume/outcome effects in terms of absolute
differences in mortality between high and low volume hospitals (See Tables
1 and 2) and Gandjour et
al (2003) presented mortality rates for high volume relative to low volume
hospitals. Absolute differences in
mortality rates of the order of 10% are reported when high volume units are
compared to low volume units in a number of complex high risk surgical
procedures including paediatric cardiac surgery, surgery to repair ruptured
abdominal aortic aneurysms, pancreatic cancer surgery and oesophageal cancer
surgery. Relative differences in
mortality rates of at least 10% are reported in a range of common lower risk
procedures including percutaneous transluminal coronary angioplasty, carotid
endarterectomy, knee replacement and surgery for hip fracture. Associations
between 5 levels of relative volume and outcome, drawn from 2.5 million
procedures for 14 interventions, in the study of Birkmeyer et al 2002, are
illustrated in Figures 1 and
2.
Commentary:
Implications of Findings
The
interpretation of the results of volume/outcome studies is complex. At the time
of the York Review, methodological deficiencies in the evidence base meant that
the studies had little if any relevance to health service planning. Recent improvements in the methodological rigour of at least
a proportion of published volume/outcome studies mean that there is now a large
body of credible evidence. This
shows both substantial effects in a limited range of complex high risk surgical
procedures and modest but clinically relevant effects in a wide range of more
general procedures. Questions
remain about the nature of the effects and their implications for service
planning. Clarification of these
depends upon somewhat different avenues of thinking.
The
effects are likely to be most clear in circumstances where the condition is
complex and its treatment associated with high risk, and where data from
substantial numbers of patients are available, covering a wide spectrum of
levels of volume. This is reflected
in the abundance and consistency of evidence about complex, high risk surgical
procedures. This is already
accepted into service delivery. Indeed,
in specialities such as neurosurgery, cardiac surgery and transplantation, the
debate is not if they should be regionalised but if greater, even national
concentration is appropriate. Furthermore,
the relationship between increased volume and improved outcome in these
circumstances is likely to be continuous, with improvement even at relatively
high levels of experience. One
exception may be if the volume becomes excessive, so that penalties of
"overwork" lead to deterioration in outcome.
For more common, less complex procedures, the improvement in outcome with
increasing volume is likely to diminish beyond a certain threshold.
In practice the majority of services will lie between these two extremes,
which are illustrated graphically in Figure
3.
For
some disorders, even though evidence is less abundant and the extent of the
effect not dramatic (and hence less easy to show and more controversial), the
consequences may still be important. For
example, reduction by a few percent in mortality for myocardial infarction could
save many lives in Scotland.
Another
issue is the fact that the volume/outcome literature looks at average effects.
Although high volume is associated with good outcome in general, there
are low volume hospitals whose outcomes are superior to typical high volume
hospitals and there are high volume surgeons with poor results who work within
high volume hospitals. However this
does not deflect the criticism that, in the NHS, too many operations are still
taking place in hospitals with low volumes and that this may change if patient
choice is effective (Taylor 2004).
Is
it the volume of activity for an individual surgeon or physician which is
important or the volume of the relevant unit or hospital?
More studies have looked at hospital volume than have looked at surgeon
volume. There is evidence that each
can be influential, with perhaps hospital volume stronger but there is no
consensus. A related point is if
the surgeon/hospital needs to be ‘high volume’ for the procedure in
question, or is high volume in general associated with good outcomes for all
procedures? Urbach and Baxter
(2004), for example, argue that volume in general is more important than volume
for the specific procedure.
There
is an increasing focus on indicators of outcome other than mortality.
The occurrence of infection or other post-operative complications are
generally applicable indices. Procedure
- specific "clinical" indicators can include if intestinal surgery is
followed by a stoma or if this is avoided by anastomosis, recurrence after
procedures for hernia, varicose vein, and the persistence of reduced mobility
and pain after orthopaedic surgery. Indices
appropriate to ‘medical’ care are well established but rarely if ever
available from routine sources of information. Quality control of diagnostic
work is well established and recent evidence links volume with accuracy of
breast screening. (Théberge et al
2005). In a service increasingly
taking account of patient choice, satisfaction with the whole process of care
will need to be taken into account, with the potential of benefit from high
volume, highly experienced specialised team care to be set against local
familiarity, and convenience.
It
becomes a value judgement to explain the observed volume/outcome relationships. The two widely cited explanations are:- ‘practice makes
perfect’ and ‘selective referral’ (i.e. patients are selectively referred
to clinicians or hospitals that have historically achieved good outcomes).
The former would suggest that volume/outcome associations are causal but
the latter would imply that the observed associations are artefactual rather
than causal. There are also issues such as aspects of process and/or
structure which are associated with high volume (e.g. a large well staffed
intensive care unit) and which might lead indirectly to better outcomes.
A
major current problem in applying these findings is the shortage of evidence
supporting the hypothesis that the volume/outcome association is a causal
association, whereby manipulating volume will have a beneficial impact on
outcome. It should be noted that
the problem is a lack of evidence rather than clear evidence of a lack of a
causal effect. There is some
limited evidence accumulating to support the association as being causal, but a
great deal more research is needed in this area.
The extent to which benefits can be achieved through diffusion of ‘best
practice’ from ‘centres of excellence’ needs to be defined. Rigorous
evidence of the effectiveness of the Clinical Networks established since 1998
would be important. Ultimately, the quality and safety of any service or
procedure can be assured only by focussed audit.
Service
planners may complain of the lack of clear cut, quantitative relationships,
particularly concerning thresholds. In
practice, the responsibility perhaps now lies with planners to specify the
extent of effect that will be crucial in their decision making.
Evidence may need to be stronger if it is the only or main factor for
change in an existing arrangement. Where
reconfiguration is needed in response to other factors, a general presumption of
volume/outcome relationship is a reasonable starting point, and the issue may be
more what level of effect is relevant in the circumstances under consideration.
The more sensitive the indicator of outcome used, and the more common the
condition, the longer the list of interventions to which the volume/outcome
effect will be relevant.
Conclusions
from Review
Returning
to the three conclusions extracted from the York Review in Section 2, it is
clear that the concerns over case-mix adjustment no longer hold. There is now a core of studies of adequate methodological
quality to establish striking volume/outcome associations in certain complex
high risk surgical procedures and more modest but clinically relevant effects in
a wide range of common procedures. The
size of the effect is influenced by the index of outcome and the range of volume
considered.
There
is still only limited evidence to suggest that the observed associations are
causal, and that interventions to manipulate volume can lead to better outcomes.
It is, however, very important to note that the issue here is that
evidence is sparse, rather than there being strong evidence of a lack of a
causal association. The relevance
of the observed volume/outcome relationships to health service planning depends
crucially on how one interprets the underlying mechanisms which generate the
associations.
The
recent literature appears, in general, to provide limited support for the final
conclusion of the York Review. Benefits
arising from manipulation of volume are likely to be most clearly apparent at a
relatively low level of volume. The rate of improvement may then diminish but
further improvement can still occur.
References
Begg
CB, Riedel ER, Bach PB, et al. Variations
in morbidity after radical prostatectomy. N
Eng. J Med 2002; 346: 1138-1144.
Birkmeyer
JD, Siewers AE, Finlayson EV, et al. Hospital
volume and surgical mortality in the United States.
N Eng J Med 2002; 346:
1128-1137.
Birkmeyer
JD and Dimick JB. Potential
benefits of the new Leapfrog standars: effect of process and outcome measures.
Surgery 2004; 135:
569-575.
Birkmeyer
JD, Stukel TA, Siewers AE, et al. Surgeon
volume and operative mortality in the United States.
N Eng J Med 2003; 349:
2117-2127.
Cook
JA, Ramsay CR and Fayers P. Statistical
evaluation of learning curve effects in surgical trials.
Clinical Trials 2004; 1:
421-427.
Gandjour
A, Bannenberg A and Lauterbach KW. Threshold
volumes associated with higher survival in health care. A systematic review. Med
Care 2003; 41: 1129-1141.
Halm
EA, Lee C and Chassin MR. Is volume
related to outcome in health care? A systematic review and methodological
critique of the literature. Ann Intern Med 2002; 137:
511-520.
Hodgson
DC, Zhang W, Zaslavsky AM, et al. Relation
of hospital volume to colostomy rates and survival for patients with rectal
cancer. J Natl Cancer Inst 2003; 95:
708-716.
Jarman
B, Gault S, Alves B, et al. Explaining
differences in English hospital death rates using routinely collected data.
BMJ 1999; 318:
1515-1520.
Luft
HS, Bunker, JP, Einthoven AC. Should
operations be regionalised? The
relationship between surgical volume and mortality.
N Eng J Med 1979; 301 1364-1369.
Nathens
AB, Brunet FP and Maier RV. Development
of trauma systems and effect on outcomes after injury.
Lancet 2004; 363: 1794-1801.
Neuhauser
D Surgical experience, hospital
size and severity adjusted mortality: James Y Simpson,1869 Qual Saf Health Care 2005: 14: 67 - 68
Nobilia
L, Fortuna D, Vizioli M, et al. Impact
of regionalisation of cardiac surgery in Emilia-Romagna, Italy.
J Epidemiol Community Health
2004; 58: 97-102.
Panageas
KS, Schrag D, Riedel E et al. The
effect of clustering of outcomes on the association of procedure volume and
surgical outcomes. Ann Intern Med 2003; 139:
658-665.
Ramsay
CR, Grant AM, Wallace SA, et al. Statistical
assessment of the learning curves of health technologies.
Health Technology Assessment
2001; 5 (12): 1-98.
Robson
AJ, Richards JMJ, Nixon SJ and Paterson-Brown, S.
Specialisation of emergency surgery in the management of perforated
peptic ulcer disease. Br
J Surg 2005; 92, Suppl 1: 147.
Shahian
DM and ST Normand. The
volume-outcome relationship: from Luft to Leapfrog.
Ann Thorac Surg 2003; 75:
1048-1058.
Simpson
JY Anaesthesia, hospitalism and hermaphroditism. The works of Sir James Y
Simpson Bart. Volumes I and II Edinburgh
Adam and Charles Black
Sowden
AJ, Grilli R and Rice N. The
relationship between hospital volume and quality of health outcomes. CRD report
8, part 1. York: Centre for Reviews
and Dissemination, 1997.
Taylor
R. “Clinical Quality, where
volume makes a difference” Health Service Journal 9 Dec 2004 p23.
Théberge
I, Hébert-Goteau N, Langlois Et al. Volume
of screening mammography and performance in the Quebec population Breast Cancer Screening Programme. CMAJ, 2005, 172:
195-9.
The
UK Neonatal Staffing Study Group. Patient
volume, staffing, and workload in relation to risk-adjusted outcomes in a random
stratified sample of UK neonatal intensive care units: a prospective evaluation.
Lancet 2002; 359:
99-107.
Urbach
DR and Baxter NN. Does it matter
what a hospital is “high volume” for? Specificity
of hospital volume-outcome associations for surgical procedures: analysis of
administrative data. BMJ
2004; 328: 73
Appendix I: List of 50 Articles Reviewed
1.
Bachmann MO, Alderson D, Edwards D, et al.
Cohort study in South and West England of the influence of specialization
on the management and outcome of patients with oesophageal and gastric cancers.
Br J Surg 2002; 89:
914-922.
2.
Begg CB, Riedel ER, Bach PB, et al.
Variations in morbidity after radical prostatectomy.
N Eng. J Med 2002; 346:
1138-1144.
3.
Begg CB, Scardino PT.
Taking stock of volume-outcome studies. J.Clin.Oncol. 2003; 21:
393-4.
4.
Berman MF, Solomon RA,
Mayer SA, et al. Impact of
hospital-related factors on outcome after treatment of cerebral aneurysms. Stroke
2003; 34: 2200-7.
5.
Birkmeyer JD, Dimick JB. Potential
benefits of the new Leapfrog standards: effect of process and outcome measures.
Surgery 2004; 135:
569-575.
6.
Birkmeyer JD, Finlayson
EV, Birkmeyer CM. Volume standards for high-risk surgical procedures: potential
benefits of the Leapfrog initiative. Surgery 2001; 130: 415-422.
7.
Birkmeyer JD, Siewers AE, Finlayson EV, et
al. Hospital volume and
surgical mortality in the United States. N Eng J Med 2002; 346:
1128-1137.
8.
Birkmeyer JD, Stukel TA, Siewers AE, et
al. Surgeon volume and
operative mortality in the United States. N Eng J Med 2003; 349:
2117-2127.
9.
Birkmeyer JD, Warshaw AL, Finlayson SR, et
al. Relationship between
hospital volume and late survival after pancreaticoduodenectomy. Surgery
1999; 126: 178-183.
10.
Brown DL. Analysis of the
institutional volume-outcome relations for balloon angioplasty and stenting in
the stent era in California. Am.Heart J. 2003; 146: 1071-1076.
11.
Chang R-KR, Klitzner TS.
Can regionalization decrease the number of deaths for children who
undergo cardiac surgery? A theoretical analysis.
Paediatrics 2002; 109:
173-181.
12.
Chen EW, Canto JG,
Parsons LS, et al. Relation between hospital intra-aortic balloon
counterpulsation volume and mortality in acute myocardial infarction complicated
by cardiogenic shock. Circulation 2003; 108: 951-957.
13.
Cook JA, Ramsay CR and Fayers P. Statistical
evaluation of learning curve effects in surgical trials.
Clinical Trials 2004; 1:
421-427.
14.
Cowan JA Jr, Dimick JB, Leveque JC, et
al. The impact of provider
volume on mortality after intracranial tumor surgery. Neurosurgery
2003; 52: 48-53.
15.
Daley J. Invited
commentary: quality of care and the volume-outcome relationship--what's next for
surgery? Surgery 2002;131:16-8.
16.
Dimick JB, Cowan JA, Jr.,
Ailawadi G, et al. National variation
in operative mortality rates for esophageal resection and the need for quality
improvement. Arch.Surg. 2003; 138: 1305-1309.
17.
Dimick JB, Cowan JA, Jr.,
Upchurch GR, Jr., Colletti LM. Hospital volume and surgical outcomes for elderly
patients with colorectal cancer in the United States. J.Surg.Res. 2003; 114:
50-66.
18.
Dimick JB, Stanley JC,
Axelrod DA, et al.
Variation in death rate after abdominal aortic aneurysmectomy in the
United States: impact of hospital volume, gender, and age.
Ann Surg 2002; 235:
579-585.
19.
Feasby TE, Quan H, Ghali
WA. Hospital and surgeon
determinants of carotid endarterectomy outcomes.
Arch of Neurology 2002; 59:
1877-1881.
20.
Finlayson EV,.Birkmeyer
JD. Effects of hospital volume on life expectancy after selected cancer
operations in older adults: a decision analysis. J.Am.Coll.Surg. 2003; 196:
410-417.
21.
Gandjour A, Bannenberg A, Lauterbach KW.
Threshold volumes associated with higher survival in health care.
A systematic review. Med
Care 2003; 41: 1129-1141.
22.
Glance LG, Dick AW,
Mukamel DB, Osler TM. Is the hospital volume-mortality relationship in coronary
artery bypass surgery the same for low-risk versus high-risk patients? Ann.Thorac.Surg.
2003; 76: 1155-1162.
23.
Goodney PP, O'Connor GT,
Wennberg DE, Birkmeyer JD. Do hospitals with low mortality rates in coronary
artery bypass also perform well in valve replacement? Ann.Thorac.Surg.
2003; 76: 1131-1136.
24.
Gutow AP. Surgeon volume
and operative mortality. N.Engl.J.Med. 2004; 350:1256-8.
25.
Halm EA, Lee C, Chassin MR. Is
volume related to outcome in health care? A systematic review and methodological
critique of the literature. Ann
Intern Med 2002; 137: 511-520.
26.
Hannan EL, Wu C, Ryan TJ,
et al. Do hospitals and surgeons with higher coronary artery bypass graft
surgery volumes still have lower risk-adjusted mortality rates? Circulation
2003; 108: 795-801.
27.
Ho V.
Certificate of need, volume, and percutaneous transluminal coronary
angiography outcomes. Am
Heart J 2004; 147:
442-448.
28.
Hodgson DC, Fuchs CS,
Ayanian JZ. Impact of patient and
provider characteristics on the treatment and outcomes of colorectal cancer. J.Natl.Cancer
Inst. 2001; 93: 501-515.
29.
Hodgson DC, Zhang W, Zaslavsky AM, et
al. Relation of hospital volume
to colostomy rates and survival for patients with rectal cancer. J Natl Cancer Inst
2003; 95: 708-716.
30.
Jarman B, Gault S, Alves B, et al.
Explaining differences in English hospital death rates using routinely
collected data. BMJ
1999; 318: 1515-1520.
31.
Kizer KW.
The volume-outcome conundrum. N.Engl.J.Med. 2003;349:2159-61.
32.
Liu JH, Etzioni DA,
O’Connell JB, et al.
Using volume criteria: Do Californian hospitals measure up?
J Surg Research
2002; 113: 96-101.
33.
Matz PG. Editorial
comment--Spontaneous subarachnoid hemorrhage: volume, experience, and outcome. Stroke
2003; 34:2206-7.
34.
Meyerhardt JA, Catalano
PJ, Schrag D, et al.
Association of hospital procedure volume and outcomes in patients with
colon cancer at high risk of recurrence. Ann
Intern Med 2003; 139: 649-657.
35.
Nathens AB, Brunet FP and Maier RV.
Development of trauma systems and effect on outcomes after injury.
Lancet 2004; 363:
1794-1801.
36.
Nobilio L, Fortuna D, Vizioli M, et
al. Impact of regionalisation
of cardiac surgery in Emilia-Romagna, Italy.
J Epidemiol Community Health
2004; 58: 97-102.
37.
Nobilio L,.Ugolini C.
Different regional organisational models and the quality of health care:
the case of coronary artery bypass graft surgery. J.Health Serv.Res.Policy
2003; 8: 25-32.
38.
Panageas KS, Schrag D, Riedel E et
al. The effect of clustering of
outcomes on the association of procedure volume and surgical outcomes.
Ann Intern Med
2003; 139: 658-665.
39.
Parks RW, Bettschart V, Frame S, et
al. Benefits of specialisation
in the management of pancreatic cancer: results of a Scottish population-based
study. Br
J Cancer 2004; 91: 459-465.
40.
Peterson ED, Coombs LP,
DeLong ER, et al.
Procedural volume as a marker of quality for CABG surgery. JAMA
2004; 291: 195-201.
41.
Ramsay CR, Grant AM, Wallace SA, et
al. Statistical assessment of
the learning curves of health technologies.
Health Technology Assessment
2001; 5 (12): 1-98.
42.
Shahian DM and ST Normand. The
volume-outcome relationship: from Luft to Leapfrog.
Ann Thorac Surg 2003; 75:
1048-1058.
43.
Smink DS, Finkelstein JA, Kleinman K, Fishman SJ.
The effect of hospital volume of pediatric appendectomies on the
misdiagnosis of appendicitis in children. Pediatrics 2004; 113:
18-23.
44.
Soljak M.
Volume of procedures and outcome of treatment. BMJ 2002; 325:
787-788.
45.
The UK Neonatal Staffing Study Group.
Patient volume, staffing, and workload in relation to risk-adjusted
outcomes in a random stratified sample of UK neonatal intensive care units: a
prospective evaluation. Lancet
2002; 359: 99-107.
46.
Urbach DR, Baxter NN.
Does it matter what a hospital is “high volume” for?
Specificity of hospital volume-outcome associations for surgical
procedures: analysis of administrative data.
BMJ 2004; 328: 737-740.
47.
Vaughan Sarrazin
MS,.Rosenthal GE. Hospital volume and outcome after coronary angioplasty: Is
there a role for certificate of need regulation? Am.Heart J. 2004; 147:
383-385.
48.
Wibe A, Eriksen MT, Syse
A, et al.
Effect of hospital caseload on long-term outcome after standardization of
rectal cancer surgery at a national level.
Br J Surg 2004; 92:
217-224.
49.
Wigmore SJ, Madhavan K,
Currie EJ, et al.
Does the subspeciality of the surgeon performing primary colonic
resection influence the outcome of patients with hepatic metastases referred for
resection? Ann Surg 1999; 230:
759-765.
50.
Yeast JD, Poskin M,
Stockbauer JW, et al.
Changing patterns in regionalization of perinatal care and impact on
neonatal mortality. Am
J Obstet Gynecol 1998; 178:
131-135.