  |
|
|
||||||||||||||
|
||||||||||||||||
| Journal of the New Zealand Medical Association,
13-December-2002, Vol 115 No 1167
|
||||||||||||||||
|
Major inequities between district health boards in referred
services expenditure: a critical challenge facing the primary health care
strategy
Laurence Malcolm
Many studies and reviews have demonstrated poor access to
primary health care for disadvantaged New Zealand populations, including
Maori.2–5 They have also demonstrated
wide variation in primary care expenditure between populations at practice,
primary care organisation (PCO) and district
levels.6,7,8 PCOs are a generic term used for
organisations that contract with funders for services delivery, eg, independent
practitioner associations.7 The major component
of this expenditure is referred services, defined in this paper as
pharmaceutical and laboratory services. PCOs are expected to become primary
health organisations (PHOs) within the next year or two and to be equitably
funded for their referred services on the basis of their enrolled
populations.8
The serious health consequences of this inequity and the
potential for reducing health inequalities through better access to
comprehensive primary healthcare have been documented by the National Health
Committee.9 Equity in health services is
defined by Starfield as implying that there are no differences in health
services where health needs are equal, or that enhanced services are provided
where greater heath needs are
present.10
Population-based funding of primary care, as set out in the
Primary Health Care Strategy, is expected to reduce these
inequities.1 The Strategy states (p14) that,
‘Population-based funding will help to reduce inequalities by directing
resources to communities with greatest health
needs.’1 Finding a way to shift resources
to those in greater need will be a key factor in the success of the
Strategy.
However, little work has been done to measure the extent of
the redistribution required including between districts, in part because of the
poor quality of the information available. This study has drawn on recent data
on district expenditure and work on the primary care funding formula designed to
determine budgets for the allocation of primary care expenditure to DHBs on an
equitable basis. The study aimed to compare actual with equitable budget
expenditure on referred services and to analyse population need factors that
might explain variation from equity.
MethodsActual pharmaceutical and
laboratory expenditure for the first five months of the 2001/2002 financial year
was obtained from the Ministry of Health. This was adjusted for seasonality
using Pharmac’s phasing ratios to estimate an actual annual figure. DHB
populations used in calculating DHB budgets by community services card (CSC) and
high use health card (HUHC) were based on data from Work and Income New Zealand
and Health Benefits. DHB data on ethnicity and deprivation (NZDep96) were also
obtained. The GP lists used to identify GP expenditure as compared with total
expenditure (ie, inclusive of specialist prescribing and laboratory use per
DHB), were based on cross-referencing various lists of GPs, including the
Medical Council list as at August, 2001. GP-related referred services
expenditure for each DHB was determined through the location of GPs and their
referred services expenditure linked to their NZMC numbers. This raised the
question of cross boundary flows that is discussed below. The analysis included
only GP referred services expenditure which, based on an analysis of 2000/2001
expenditures, was estimated to be 73% for pharmaceuticals and 75% for laboratory
services, most of the remainder being specialist related. This actual
expenditure was compared with expected equitable expenditure calculated from the
national cost weights for age, gender, CSCs and HUHCs from the Ministry of
Health. These weights are based on the work of Sutton for the Health Funding
Authority in 2000, with data derived from the RNZCGP Research Unit in Dunedin.
The percentage variation of actual from the expected annual pharmaceutical and
laboratory expenditure was calculated for each DHB.
From the population data, the three measures of disadvantage from the funding formula – percentage of Maori, percentage with CSC, and mean NZDep96 score by district – together with population per GP, were calculated. A regression analysis was then undertaken to determine the relationships between the dependent variables percentage variation from equity of pharmaceutical, laboratory and total referred services expenditure and these four measures of disadvantage. A multiple regression analysis was also undertaken to determine the relative importance of each of these measures in predicting variation from equity. ResultsThe percentage variation from equity
for each DHB in pharmaceutical and laboratory services expenditure is shown in
Figure 1. Capital and Coast was found to be 17.0% over budget in
pharmaceuticals, 18.9% over in laboratory expenditure and 17.5% over budget in
total referred services combined. On the other hand, Tairawhiti was found to be
25.2% under budget in pharmaceutical services, 18.8% in laboratory services and
23.9% under budget in total.
Figure 1. Variation from equity between DHBs in
pharmaceutical and laboratory expenditure
 There is a loose but statistically significant correlation
(r= 0.41, p <0.05) between percentage variation in pharmaceutical and
laboratory expenditure. In other words, DHBs over budget in one category also
tend to be over in the other.
Table 1 presents the calculated coefficients between
percentage variation in total referred services expenditure and the measures of
disadvantage used. All are significantly correlated with one another and are
statistically significant at p <0.01. A multiple regression analysis was
carried out to determine the relative and overall predictive power of the four
variables measuring disadvantage and which had the strongest relationship to the
observed variation. The combined variables had a multiple correlation
coefficient of 0.75 and an R2 value of 0.56,
indicating that, combined, they explained 56% of the total variation in actual
expenditure from budget. The variable with the strongest predictive power was
population per GP. Figure 2 shows the relationship between the variation from
equity in different DHBs and population per GP, indicating that the
maldistribution of GPs was a key factor in the associated maldistribution of
referred services expenditure.
Table 1. Correlation coefficients between percentage
variation in total referred services expenditure and measures of disadvantage
(all coefficients p <0.01)
Figure 2. Relationship between percentage variation in
total expenditure and population per GP
 DiscussionThe findings raise questions about a
number of issues needing further discussion, eg, the validity of the funding
formula; the accuracy of the data used to calculate budgets; and the use of data
for only five months of the year to calculate actual expenditure. With regard to
the funding formula used to calculate budgets, this is very dependent upon the
accuracy of CSC and HUHC uptake. For example, the most important factor leading
to Capital and Coast DHB being so far above budget is its low percentage of
CSCs: only 24.7%. Is this the actual figure, or is there a fault in the data
used?
It is well known that CSC uptake is substantially less than
entitlement.11 Some 55% of the population are
entitled to a CSC but the most recent uptake used in the primary care funding
formula is only 38.6%. It is also well known that disadvantaged populations have
a lower uptake of CSCs than more advantaged
populations.11 A more complete uptake of CSCs
in disadvantaged districts would increase the calculated inequities. It can be
calculated that a 1% increase in CSCs would increase the budget for recent
referred services by more than 1%.
There is also a wide variation in, and inadequate uptake of,
HUHCs and this plays an important part in calculating budgets for the DHBs and
PCOs/PHOs. In the data supplied, HUHC percentages in DHBs varied from 0.2 to
2.8%. Studies of PCOs with capitated funding show that, even with the incentive
of substantially higher general medical services (GMS) funding under capitation,
the percentage of patients on HUHCs can still range widely, eg, from 0.1 to
9.7%.12 It can also be calculated that a 1%
increase in HUHCs increases a DHB’s budget by more than 3%. However,
perhaps not surprisingly, given the variability in uptake, no significant
relationship was found in the analysis between percentage HUHCs and variation
from equity.
There are some uncertainties about the GP availability
figure, as it is based on total GPs only, not full-time equivalents. The actual
expenditure data on which this analysis is based covers only a five-month
period. Despite these reservations, the overall variation is so great that more
complete data is unlikely to result in a different pattern. Another uncertainty
relates to cross boundary flows of patients. The actual DHB expenditure is
calculated on the location of practices, not patients who may have crossed
district boundaries to receive care, especially in Auckland. However, a study of
patient cross boundary flows between Auckland districts in 1997 showed only a
net gain of 3.4% to central Auckland.6 Although
this pattern may not have changed much, it needs further
investigation.
The findings confirm that the measures of disadvantage used,
both individually and especially combined, explain more than 50% of the observed
variation between districts in referred services expenditure. It is clear that
serious inequities exist between districts and that they are closely related to
the level of disadvantage in their populations.
The critical importance of population per GP as the main
factor in explaining variation between districts is understandable. Given the
historically low level of GMS funding, GPs have naturally chosen to live and
work in situations in which they can expect to be adequately remunerated by
better-off populations. Such populations can afford to pay not only consultation
but also pharmaceutical co-payments. They tend to have higher expectations and
lower thresholds for seeking attention. Central Auckland, with a relatively
well-off population has one GP to a population of 880. Consultation rates and
consequently rates of use of referred services have been consistently higher
here than for less well-off
populations.6
On the other hand another district seriously below budget,
the West Coast, has one GP to 1613 population, nearly twice the Auckland
population. It is not surprising, therefore, that the West Coast and similar
disadvantaged populations, are experiencing major problems in GP recruitment as
well as lower expenditure on referred services. The figures suggest that the
total number of GPs available in New Zealand may well be adequate, especially
with moves to capitation and hence greater use of practice nurses. The key
problem is, therefore, one of maldistribution, not overall
availability.
The findings also raise important questions as to other
explanations for this variation. Why, for example, is Nelson Marlborough DHB so
low on laboratory expenditure as well as pharmaceuticals. Why, on the other
hand, is Waikato DHB so high on laboratory expenditure? These and other
questions need answers and point to the need for further research to give a
better understanding of what can be done to reduce inequities.
A substantial increase in the level of funding to improve
access to primary care services and paid by capitation could provide the needed
incentive for GPs to move to more disadvantaged districts. Over time, this may
redress the current maldistribution of GPs and the associated inequitable
distribution of referred services expenditure. However, the proposed annual
increase in GMS funding of some $195 million by 2004/2005 will only increase
government payments per consultation to some $25. This is well short of the
estimated $40–45 needed to run an adequate general practice-based primary
healthcare service. Substantial patient co-payments will therefore remain, with
the new funding doing little to redress the serious inequities in referred
services expenditure.
The solution to addressing inequities in referred services
expenditure can only come from redistribution at all levels. DHBs are facing a
major redistribution of resources from well-off to disadvantaged districts. For
DHBs above budget to successfully reduce their expenditure to equity will
require the full support and commitment of their PCOs. A major investment is
therefore required in helping such PCOs to shift to equitable budgets.
The first step in this difficult process would be for PCOs
to examine and address the inequities between practices within their remit.
There is clear evidence of wide variation in referred services expenditure
between practices, far greater than that between
districts.7,8, 12 Recent work has shown that
this variation appears to be related much less to disadvantage than it is to GP
behaviour.12 It has also produced at least
preliminary evidence of lower expenditure being associated with better
quality.12 If this is so, PCOs would have a
significant incentive to face this issue as a quality as well a cost issue.
Savings could be found for redistribution, both within the PCO and to address
wider equity issues.
The lack of ready availability of even the most basic data
needed for this study, eg, the number of FTE GPs and GP expenditure on referred
services by DHB, highlights another fundamental issue facing the Primary Health
Care Strategy. This is the almost total lack of an information system and
associated research and development strategy needed to support implementation,
monitoring and evaluation. Information is needed at national, DHB and PCO level.
Associated with this issue is the need for organisational, managerial,
contracting and research skills at all levels. Investment in these systems and
skills does not yet seem to be a government priority, but is critical not only
to the success of the Primary Health Care Strategy but to achieving better
health outcomes for many disadvantaged New Zealanders.
Author information:
Laurence Malcolm, Professor Emeritus and Consultant, Aotearoa
Health
Acknowledgements:
The author is grateful to the Ministry of Health, in particular Jon Foley, for
supplying the data on which this study is based.
References:
|
|
||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|