Supplementary evidence submitted by Dr
Sally Ruane (ISTC 46A)
I should welcome the opportunity to present
further evidence regarding the introduction of ISTCs into the
UK health care system within the market context. First, I should
like to expand further on the use of research on the US health
care market, conducted against a backdrop of debate in Canada
regarding the introduction of for-profit hospitals. Second I shall
consider the likely impact upon the founding principles of the
NHS of ISTCs within the evolving UK health care market.
NORTH AMERICAN
RESEARCH
The literature surrounding comparisons of different
types of hospitals is, as noted, mixed. However, this does not
mean that all the research conducted is of equally good quality
or should be afforded equal weighting. In the hierarchy of evidence,
the systematic reviews and meta-analyses of Devereaux and colleagues
must surely rank among the best in this field. These studies offer:
systematic review of pre-existing
research and meta-analysis of pooled data;
publication in high quality journals
following a rigorous peer review process;
a high degree of methodological sophistication
and transparency;
findings which suggest applicability
across different health care contexts; and
contribution to an active political
debate surrounding whether a health-care system (Canadian in Devereaux's
case) should open its doors to for-profit hospitals.
These studies have attracted considerable attention
in Canada and have given rise to a debate between authors and
critics some of which is in the public domain. This has shed further
light on methodological judgements.
More specifically, the strengths of the Devereaux
(2002a) research include the following:
The selection process (in Devereaux
et al, 2002a) to determine which pre-existing studies to
include in the systematic review and meta-analysis involved the
specification of eligibility criteria; an extensive literature
search; teams of two independently screening titles and abstracts
for relevance; teams of two independently analysing potentially
relevant studies "blind" to their findings to avoid
bias; implementation of a specified process to resolve any differences
between those analysing studies independently although agreement
on study inclusion was high; exclusion of studies involving hospitals
which had changed profit status; exclusion of studies where further
information and/or data were required from the authors and where
this was not forthcoming or demonstrated that the study was inappropriate
for inclusion.
Adjustment for patients' severity
of illness, for patients' socio-economic status, for hospital
teaching status and other potential confounding variables; non-adjustment
for variables under the control of hospital administrators where
these might be influenced by profit status and affect mortality
such as staffing levels per bed after severity of illness adjusted
for. Most studies used the Health Care Financing Administration
database which includes data on all US hospitals which serve Medicare
patients and generates risk-adjusted mortality rates. Where studies
included unadjusted and adjusted results for disease severity,
the analysis consistently led to effect estimates which were more
favourable to the private not-for-profit hospitals suggesting
they serve a population with greater disease severity. As a result,
Devereaux et al judged that any residual confounding would
make private not-for-profit institutions look worsethat
is, the results of the Devereaux et al study may, if anything,
underestimate the potential increase in mortality associated with
for-profit hospitals.
In most of the studies, patient care
was funded through Medicare; most studies included general acute
care, medical and surgical patients; one study examined maternity
services.
Alongside the very large scale research
(encompassing 38 million patients across 26,000 hospitals) which
compared private for-profit and private not-for-profit hospitals
(Devereaux et al, 2002a), a separate report analysed four
studies including public as well as private hospitals (Devereaux
et al, 2002b). Devereaux concluded that the inclusion of
public hospitals in the first piece of research would have, if
anything, strengthened the findings in relation to the comparatively
poor performance of for-profit institutions.
Comparatively low numbers of highly
trained staff per risk-adjusted bed, the need for investors to
make a rate of return on their investments of 10-15% and the bonus
and financial packages to senior staff are posited as possible
explanations for the observed differences.
One study I am aware of on mortality rates in
a UK market context was conducted by Propper and colleagues who
found a modest but upward effect on mortality rates of competition
in the internal market of the 1990s (Propper et al, 2003).
In relation to payments for care (Devereaux
et al, 2004), complexity of comparison is not confined
to questions of price fix (which in any case may not remain an
unmodified feature of the evolving UK market in health care) but
rather a number of factors need to be taken into consideration.
In weighing up the contribution to the Committee's deliberations
of research based on US studies, the following should be noted:
The data covered and the results
persisted over more than a decade and over significant changes
in the structure of US health care, including changes in payment
systems for Medicare patients.
The five studies showing significantly
higher payments for for-profit care had variations in their sources
of payments (Medicare, insurance and both).
These observations suggest higher
payments for care at private for-profit hospitals are manifest
within a variety of health care contexts.
The inclusion of for-profit health
care institutions in the UK may over time and depending upon policy
development in the UK include the same US hospital chains.
Although the tariff is currently
fixed in the UK, the development of a market dynamic is likely
to lead to pressure for future policy changes or modifications
to which I return below.
More specifically:
Again, a detailed process was employed
by the research team to identify and select high quality pre-existing
studies which met the inclusion criteria involving eligibility
decisions and data abstraction in duplicate.
Canadian hospitals are publicly funded
and the researchers thus focused upon the policy question: how
much will government pay for care delivered by private for-profit
versus private not-for-profit providers?
Studies were considered methodologically
strong where they adjusted for patient source of payment (as well
as patient age, sex, ethnicity, income education, primary diagnosis
(case mix), co-morbid conditions, severity of illness, the concentration
of hospitals in a region and hospital teaching status).
Various hypotheses to explain variability
in the direction and magnitude of effect across studies were specified
in advance. These included whether patient source of payment was
public or mixed (as well as whether payments for care were per
discharge or per day; whether payments for care were related to
hospital stay or included a period of time after hospital discharge;
whether hospitals evaluated were specialty or general; whether
the patient population was adult or paediatric; and whether the
analysis was adjusted for potential confounders or not).
The eight publications of observational
studies included over 350,000 patients and assessed a median of
324 hospitals per study. Five studies showed significantly higher
levels of payment for care at private for-profit hospitals; 1
showed significantly lower levels of payment for care at for-profit
hospitals. The primary meta-analysis demonstrated that private
for-profit hospitals are associated with higher payments for care;
only one of the pre-specified hypotheses helped explain the large
heterogeneity across study results (general vs specialty). The
studies with the most extensive adjustment for potential confounders
reported statistically higher payments for care at private for-profit
hospitals. Pooled estimates from both the 3 studies evaluating
specialty hospitals and the five studies evaluating general hospitals
showed higher payments of care for the private for-profit hospitals.
Ten publications excluded from the
analysis either because no measure of variance was supplied or
because public and private not-for-profit had been grouped also
all showed higher payments or charges for care at for-profit hospitals,
statistically significant in six out of the 10.
Given the significant variability
in direction and magnitude of effect among studies, the authors
conclude that whilst the inference that for-profit hospitals result
in higher payments for care is secure, the magnitude of the effect
may differ according to circumstances.
The authors believe that the results
may underestimate the association between private for-profit hospitals
and higher payments for care since, because studies in the systematic
review adjusted for case mix, increase in payments resulting from
inappropriate upcoding of patient diagnosis to enhance reimbursement
is not captured; and because issues of fraud are not addressed
(performance of unnecessary surgeries, billing for services not
provided, inappropriate detainment of psychiatric patients for
billing purposes). (The multimillion-dollar fraud lawsuits in
the US have been overwhelmingly against private for-profit hospitals.)
Vaillancourt Rosenau and Linder's (2003) research
is methodologically simpler and arguably less sophisticated than
that of Devereaux and colleagues. It does, however, provide the
first systematic review of data-based, peer reviewed assessments
of the relative performance of for-profit and not-for-profit providers
in the United States. The 149 articles which met the inclusion
criteria reported 179 performance assessments of four common performance
criteria: access, cost/efficiency, quality and amount of charity
care. The full universe of eligible studies (since 1980) was included
in the systematic review and these were given equal weighting;
the synthesis focused upon statements of findings. The authors
assume that technical flaws and consequent errors in the studies
are randomly distributed.
56 of the 149 studies considered
cost, only 13 of which reported for-profit providers as superior.
69 studies sought to compare quality
of care on at least one measure and 41 found not-for-profit providers
as superior with a further 20 finding no difference.
30 compared on access and 20 found
not-for-profit providers superior.
Out of 24 assessments of charity
care performance, 16 found not-for-profits as superior.
ISTCS: THE
MARKET AND
NHS FOUNDING PRINCIPLES
Although ISTC policy must be examined in terms
of its technical details, it cannot be reduced to these and any
overall assessment within the political process must adopt a broader
view, encompassing the political, moral and value basis of the
policy. This means that ISTCs must be clearly understood and evaluated
in the context of the government policy of moving UK health care
away from the NHS as traditionally understood towards a system
of health care based on a competitive, partly commercialised market.
The reason ISTCs pose a threat to the rather
frayed founding principles of the NHS lies in the dynamic they
help create, particularly within the market context, and there
are two key aspects to this on which I wish to comment.
ISTCs contribute to a differentiation of health
care experience. In itself, this differentiation does not necessarily
imply inequality. However, within a competitive market context,
differentiation is likely to be characterised by inequality. This
is because of the instability and disequilibrium created by a
market which results in inequalities among providing units and
potential destabilisation of some NHS hospitals (this relates
to the question about destabilisation posed by the Committee Chairman
at the second hearing). Where NHS hospitals lose their incomeeither
because patients choose to go elsewhere or because their income
has, as a matter of policy, been diverted to the establishment
and maintenance of the for-profit providersthe scope and
quality of the health care they offer is compromised. This is
particularly so where they are left with responsibility for more
disproportionately complex and costly cases.
One of the less often quoted founding principles
of the NHS is that patients should be able to access equally good
care wherever they obtain it. This is one aspect of the principle
of equity which underpins the concept of the NHS (Whitehead, 1994).
Whilst difficult to achieve in practice, this principle poses
a test and reference point by which to evaluate policy initiatives.
The more health care becomes differentiated, the greater the danger
that patient experiences will become systematically unequal. The
current focus upon inequalities in health care around ISTC policy
centres, on the one hand, upon whether procedures carried out
in ISTCs are of equally good quality (compared with those in the
NHS) given existing recruitment practices and modes of service
delivery (eg Wallace, 2006) and, on the other, upon whether the
quality of care across NHS hospitals is becoming unequal as a
result of the destabilising dynamic of the market with the resulting
loss of income, training opportunities, surgical skills etc. in
hospitals affected by ISTCs on their patch. Evidence to date is
limited but concerns expressed so far include the viability of
eye and endoscopy services, with implications for holistic and
comprehensive care; compromises to the quality of care where the
management of follow-up and complications is thwarted by fragmentation
and poor communication and where MRI scans have been poorly and
remotely conducted (eg SCR, 2006; RCO, 2006). The ability of these
affected NHS units in the future to offer good quality health
care will be influenced by the outcomes of the currently evolving
policy on purchasing training from ISTCs and the impact of this
and other loss of investment in NHS services and capacity because
resources have instead been spent on temporary private contracts.
At present, we do not know whether patients accessing different
providing units differ in terms of class, sex, ethnicity and age.
This is not to say that planned systems automatically
guarantee equally good care wherever it is accessed but rather
that markets automatically tend towards difference and inequality;
moreover, costly regulation and incentives may be insufficient
to redress this.
There is another dynamic, however, which poses
a threat to the founding principles of the NHS, and this is a
political one. The introduction of for-profit companies into the
routine workings of the NHS allows the representatives of those
companies a seat at the policy making table. We are already seeing
calls for the ISTCs to be more integrated and for the private
companies to become involved in workforce planning as an example.
For-profit companies will have a role in shaping health policy
and it is reasonable to suppose that they will do this in a way
which advances their interests. In practical terms, this will
involve seeking to secure higher rates of profit through an increased
share of the market and/or through increased prices. The tendency
of for-profit providers to accrue ever greater shares of public
expenditure, to foster dependency upon their services and to resist
effective regulation has been well documented in relation (for
example) to PFI and long-term care for older people (Pollock,
2004).
At the same time, the dynamic of the market
itself could exert an upward pressure on costs, as Nigel Edwards
(2005) has pointed out. This arises from such factors as the administration
of a competitive market (contracting, billing, strengthened governance
arrangements for increasingly independent providers; trying to
hold back supplier-induced demand, via structures of demand management,
treatment pre-authorisation and service use reviews, as providers
try to secure a higher share of health care expenditure); provider
behaviour such as differentiating market "products"
through advertising and information given to patients; and wastage
arising from duplication and redundant facilities (as Kevan Jones
MP reported to the House of Commons in October relation to MRI
scanners, for instance). The steps taken to try to control the
undesirable consequences of the workings of a market (including
fragmentation) are costly and may not succeed. The implementation
of financial incentives is problematic, particularly in the profoundly
moral context of health care (eg Marshall and Harrison, 2005)
and the current political and economic climate is moving away
from regulation, not towards it as the prolonged political battle
to remove health care from the EU draft Directive on Services
in the Internal Market illustrated.
This combination of rising costscosts
which, it should be noted, are not directly spent on health care
and arguably contribute little to improved health outcomesand
effective political lobbying and policy influence exercised by
the for-profit companies will, it is not unreasonable to anticipate,
lead to political pressure to introduce additional charges, such
as through top-up fees for above basic service or co-payments.
This should not be regarded as fanciful or scaremongering. For
instance, the prospect has already been discussed within the pages
of the BMJ (Donaldson and Ruta, 2005); proposals surrounding more
privatised modes of health care funding have been developed by
right of centre think-tanks (eg Booth, 2002); and a voucher scheme
formed part of Conservative Party health policy at the 2005 General
Election.
The encouragement to relate to the NHS and other
public services as consumers rather than citizens is likely to
undermine the sense among the public that health care should be
considered in terms of equity and need rather than want and personal
satisfaction. Further, whilst the necessary funds for this more
expensive market system could be raised through curbing tax avoidance
schemes, only limited action so far has been taken (Tax Justice
Network, 2006) and raising taxes on the very well off does not
appear to be on the political agenda at all. The conditions would
then be created in which the principle of health care free at
the point of deliverythe NHS founding principle which is
often cited by ministersis seriously undermined. Whatever
the undertakings given by the current government, there will be
pressure on future political leaders to respond to this, with
potentially very unequal outcomes for health care users. Thus,
this other pillar of the NHS principle of equity is also ruinously
weakened.
ISTCs have been located geographically not merely
or even principally, so far as we can tell, on the basis of additional
capacity needs but rather to kick-start the new market in health
care. Markets are typically not appropriate for securing policy
goals such as equity. The growing role and influence of commercial
interests in the NHS has moved the institution in a particular
direction over the past decade or two; the "direction of
travel" intended by government is increasingly clearly articulated.
The Committee's assessment of ISTCs must have regard to this growing
contradiction between a market embracing commercial providers
and the principle of equity institutionalised by the NHS for,
as Bevan put it, the serenity of our society.
REFERENCES
I. P Booth (2002) Getting your Health back:
Rebate Financing for Medical Care. London: Adam Smith Institute.
II. PJ Devereaux, PTL Choi, C Lacchetti,
B Weaver, HJ Schunemann, T Haines, JN Lavis, BJB Grant, DRS Hasla
million, M Bhandari, T Sullivan, DJ Cook, SD Walter, M Meade,
H Khan, N Bhatnagar, GH Guyatt (2002a) "A systematic review
and meta-analysis of studies comparing mortality rates of private
for-profit and private not-for-profit hospitals", in Canadian
Medical Association Journal, 28 May, 166(11): 1399-406.
III. PJ Devereaux, HJ Schunemann, N Ravindran,
M Bhandari, AX Garg, PTL Choi et al (2002b) "Comparison
of mortality between private for-profit and private not-for-profit
hemodialysis centers: a systematic review and meta-analysis",
in Journal of the American Medical Association, 288(19): 2449-57.
IV. PJ Devereaux, D Heels-Ansdell, C Lacchetti,
T Haines, KEA Burns, DJ Cook, N Ravindran, SD Walter, H McDonald,
SB Stone, R Patel, M Bhandari, HJ Schunemann, PTL Choi, AM Bayoumi,
JN Lavis, T Sullivan, G Stoddart, GH Guyatt (2004) "Payments
for care at private for-profit hospitals and private not-for-profit
hospitals: a systematic review and meta-analysis", in Canadian
Medical Association Journal, 8 June, 170(12): 1817-24.
V. C Donaldson and D Ruta (2005) "Should
the NHS follow the American way?", in British Medical Journal,
331: 1328-30.
VI. N Edwards (2005) "Using markets
to reform health care", in British Medical Journal, 17 December,
331:1464-1466.
VII. M Marshall and S Harrison (2005) "It's
about more than money: financial incentives and internal motivation",
in Quality and Safety in Health Care, 14: 4-5.
VIII. A Pollock (2004) NHS plc, London:
Verso.
IX. C Propper, S Burgess, K Green (2004)
"Does competition between hospitals improve the quality of
care? Hospital death rates and the NHS internal market",
in Journal of Public Economics, 88: 1247-72.
X. Royal College of Ophthalmologists (2006)
"Written evidence" Health Committee: Independent Sector
Treatment Centres, Fourth Report 2005-06 Vol II, London: The Stationary
Office.
XI. Society and College of Radiographers(2006)
"Written evidence" Health Committee: Independent Sector
Treatment Centres, Fourth Report 2005-06 Vol II, London: The Stationary
Office.
XII. RCO (2006) Evidence to the Health Committee.
XIII. Tax Justice Network (2006) Brown misses
"last chance" to deliver on tax avoidance promises,
Tax Justice Network: budget 2006 Press Release, March 22.
XIV. P Vaillancourt Rosenau and SH Linder
(2003) "Two decades of research comparing for-profit and
nonprofit health provider performance in the United States",
in Social Science Quarterly, 84(2): 219-241.
XV. A Wallace (2006) Independent sector
treatment centres: how the NHS is left to pick up the pieces,
in British Medical Journal, 332: 614.
XVI. M Whitehead (1994) "Who cares
about equity in the NHS?", in British Medical Journal, 308:
1284-7.
Dr Sally Ruane
Health Policy Research Unit
De Montfort University, Leicester
March 2006
|