  |
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
01-December-2006, Vol 119 No 1246
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Preventable kidney failure: the cost of diabetes
neglect?
Zoltán Endre, Don Beaven, Adrian Buttimore
Diabetes mellitus is now the major non-communicable world
health problem,1 with estimates of more than
250 million people likely to be affected in recent global
projections.2
Chronic kidney disease (CKD) remains the second most likely
cause of death and morbidity after cardiovascular disease in diabetes, and CKD
is a major independent risk factor for cardiovascular disease. CKD is
resource-intensive, with sufferers requiring dialysis and or kidney
transplantation for survival.
The outcomes of frequent hospitalisation and organ failure
are seen to be almost inevitable for those already diagnosed with diabetes
unless major new initiatives aimed at prevention, control, and good management
are planned and adequately resourced.3
Despite a well-accepted report commissioned by Diabetes
NZ,4 clear guidelines for
management,5 and multiple regional initiatives,
New Zealand (NZ) has not yet adopted a major strategic control plan for this
national epidemic.3 In contrast, the Canadian
Government has allocated Can$100 million to fund the Canadian Diabetes
Strategy, a comprehensive national plan focussing on prevention of diabetes
through education and health promotion.6
The disproportionate cost of hospital bed days stay from
undiagnosed or inadequately managed diabetes is known to be high: 14% of total
bed days stays in Canterbury Province (New
Zealand),7 13% in
Spain,8 and 16% in the United
States.9
Costs for renal dialysis and kidney transplantation remain
unpublished in NZ, so the particular cost of diabetic nephropathy is uncertain.
However, in Canada, where the 1993 economic cost of diabetes was estimated to be
US$9 billion dollars,10 the annual cost of CKD
as a weighted average of all dialysis and transplantation in year 2000 dollars
(approximately CAN$63,000 per annum) easily outstrips the event cost of all
other complications of type 2 diabetes—including stroke, lower limb
amputation, and myocardial infarction ($35,000, $25,000, and $19,000
respectively).11
In the United Kingdom (UK), a cost of illness model was used
to estimate the number of people with diabetic nephropathy (microalbuminuria,
overt nephropathy, and CKD).12 There is
accurate data in NZ today on the role of diabetes as primary cause of entry into
the dialysis programme from the Australian and New Zealand Dialysis and
Transplant Registry.13 Diabetes has been the
primary cause of 36 to 45% of cases of CKD in NZ patients between 1999 and
2004.13 Based on the UK estimate of 765 million
pounds (US1.4 billion) per annum at 2001 currency
values,12 the cost of diabetic nephropathy
could amount to as much as $146 million per annum in NZ.
We have therefore attempted to analyse the direct health
care costs and financial consequences to NZ of diabetic nephropathy. To do this,
we have extracted and analysed all annual costs associated with renal
replacement therapy, that is, for dialysis programmes and renal transplantation
(initial year only). Both groups of costs have been estimated for a single
district health board and then extrapolated to the whole of NZ and factored for
the number with diabetes entering renal-replacement therapy programmes.
The methods and sources for each result are listed in the
appendices. Several cross correlations with the few available published costs
suggest these are reasonable but conservative estimates.
Results and DiscussionIn Figure 1, the number of patients starting dialysis or
undergoing renal transplantation since 1990 is shown both for New Zealand as a
whole and in the Christchurch subset.13
New patients starting dialysis have increased from 179 in
1990 to 447 in 2004, with a net progressive total increase of over 20 NZ
patients per annum on dialysis (65±5.9 per 3 years, linear regression
r2=0.98, p=0.0016 in Figure 1). At the same
time, the total number of transplants performed yearly has not increased
significantly (slope=4.8 per 3 years±2.2,
r2=0.61, ns). Also shown is the fact that the
number of transplants from deceased donors (both nationally and in Christchurch)
has decreased and the relatively constant total number of renal transplants
performed per annum has been maintained through an increase in live donor
transplantation.
Figure 1. Renal replacement in New Zealand
1990–2004. Incident numbers of patients commencing dialysis (combined
haemodialysis and peritoneal dialysis) and undergoing transplantation (TP) in
New Zealand (Upper Panel) and Christchurch (Lower Panel).
 While the total number of transplants is constant, the
number of deceased donor transplantations (TP-DD) is decreasing while the live
donor (LD) rate is increasing. In contrast, there is a progressive increase in
patients commencing dialysis of approximately 20 per annum in New Zealand (see
Results for further details).
Table 1 shows the costs of chronic dialysis in New Zealand
for each of the main modalities. These are agreed costs, as sourced in Appendix
1. It is possible for each nephrology centre in New Zealand to compare their
audited costs with the estimated cost based on population based funding using
these figures. For example, the audited costs in Christchurch for 2003 are $5.9
million. This compares with $6.9 million which would have been attributed using
the 11.1% of population-based formula.
Table 1. Dialysis in NZ: annual
costs
*Based on 2003 data. See Appendix 2 for details of
costings.
Christchurch costs are thought to be lower than average
because home dialysis and peritoneal dialysis are the only dialysis services
presently offered outside acute dialysis. These cheaper modalities reduce the
average cost for dialysis in our centre. However, it should be noted that these
costs may rise because of the need to provide domiciliary or alternative
dialysis services for the aging dialysis population who are experiencing
difficulties in dialysing in this particular community.
Table 2 summarises the annual cost of renal transplantation
in New Zealand for 2003, calculated according to Appendix 2. We recognise that
costs vary amongst centres principally because of differences in the average
length of hospital stay. We have used the average length of stay in Christchurch
of 5.4 days (say 5.0) and added these to the retrieval costs (for cadaver
kidneys) and transplant surgery costs to obtain the $27,500 average cost per
transplant to derive the total figures shown here. Other Units could
re-interpret these data with their own average length of stay. We suggest the
total figure provided here is a very conservative estimate.
Table 2. Transplantation in NZ: annual costs
*Based on data for 2003. See Appendix 3 for details of cost
derivation.
Table 3 combines the data from Table 1 and 2 to derive
national costs of chronic renal failure in 2003. In addition, the cost of
nephrologist services and of one pharmaceutical item, erythropoietin, have been
included as substantive costs, and an allowance made for unassessed or currently
inaccessible costs to provide the final figure of $90 million per annum as a
conservative figure for New Zealand. The consultant time calculations and
erythropoietin costs are attributed in Appendix 3.
Table 3. Chronic kidney disease in NZ: annual
costs
*Only one drug (erythropoietin) has been included in this
estimate. Note that considerably more is spent on other common medications by
renal patients with diabetes28; †Costs
not assessed include consultation for early renal failure, community
pharmaceuticals, non-hospital dialysis fluids, surgery for vascular access for
dialysis and Tenckoff catheter insertion, live donor expenses, registry support
costs, tissue typing, additional transport, advisory committee and Ministry
costs.
Table 4 shows the ethnic distribution of new patients
commencing dialysis in 2003 in New Zealand. As 40% were diabetic, the maximum
cost attributable to diabetic nephropathy is 40% of the total figure for renal
replacement therapy shown in Table 3 ($90 million per annum) giving an annual
cost of $36 million.
Table 4. New chronic kidney disease patients in
NZ in 2003: ethnicity and diabetes
See Appendix 6.
Given the conservative nature of our preceding calculations,
this figure is also very conservative and probably underestimates the true costs
by a substantial amount. One caveat is that the numbers overestimate the
prevalent population with diabetes in the New Zealand transplant and dialysis
populations because they are based on the 40% of new patients who are diabetic.
While this number has been relatively constant since 1999, prior to this fewer
patients were reported as diabetic, so some dilution of the final dialysis and
transplant populations is likely.
In addition, the higher mortality of diabetic patients with
end-stage renal disease relative to their nondiabetic counterparts will reduce
the prevalence of diabetes in the final surviving total population. We would
like to highlight the difficulties encountered in obtaining the costings
described here, and suggest that mechanisms for regular and more accurate cost
ascertainment will assist short and long term strategy development. There are
several major cost areas that we have not commented on including quality of life
and the economic impact on the community including family, work and leisure.
Nevertheless, these figures provide a first estimate of the
minimum cost of diabetic renal disease in NZ and allow us to speculate about
potential savings to be generated if only the economic cost of diabetic
nephropathy is assessed. They can be compared with the population-based
comparison with the UK costs referred to earlier, which yielded an estimate of
$146 million for NZ. Even our minimum estimate of $36 million provides a
substantial economic reason for aggressively pursuing a policy of preventing, or
at least retarding, the progression of diabetic nephropathy.
We speculate that probably half of the total expenditure of
$36 million per annum could be saved or shifted to subsequent years by more
strategic management. Savings could be used to fund further application and
research into mechanisms of prevention of diabetic kidney disease and into
community prevention programmes. Given the current epidemic of diabetes these
numbers will increase progressively. This highlights the urgency with which
these issues need to be addressed.
There is no local data on the cost effectiveness of better
detection and treatment. Furthermore, the prevalent population of diabetics
approaching end-stage kidney failure and requiring dialysis in the next few
years, would ensure a delay in seeing any benefit of new prevention or treatment
measures from a dialysis or transplantation perspective. However stabilisation
of diabetic nephropathy incidence at the present new case rate and reducing the
rate of progression to CKD5 and preventing
development of nephropathy will ultimately translate directly into savings at
the same rate as the annual cost of these services less the costs of better
detection and treatment.
Unfortunately the cost of better detection and treatment
cannot be quantified until we have an agreed (national) strategy. One comparator
could be the cost of maintaining a screening or monitoring program such as
Get Checked. The cost of preventative treatment will include the cost
of encouraging and implementing lifestyle intervention plus the costs of therapy
such as renin-angiotensin-system
blockade.29,30.
Since these medications slow the linear annual rate of
decline in renal function by approximately 30%, they will prolong the
dialysis-free time by a factor of 1.4 (ie 1/0.7). The cost of treatment
therefore depends on the timing of detection and intervention. For example,
prolonging the dialysis-free interval by 1 year using candesartan (e.g. at the
relatively high dose of 16 mg daily), would probably require at least 2.4 (1.4 +
1) years of treatment at a cost of 2.4 × 12 months × $34.53
(manufacturer’s monthly price in New Zealand), that is a total of $994.
Apart from direct savings on annual expenses for renal
replacement, additional benefits to the community of maintaining good health
amongst diabetics in the workforce are clearly substantial but beyond the scope
of this paper
In NZ, several regional or community-based strategies
identify diabetes or attempt primary or early secondary intervention,
including:
These programmes are not nationwide, data from
each programme is not readily available or utilised to inform further assessment
and intervention nationally and it is not yet clear whether these will be cost
effective approaches.
For example, the Get Checked programme was launched
by the Ministry of Health in 2000,5 and has the
potential to identify early complications of diabetes. This free yearly review
allows 12 electronically measurable indices to be transferred yearly to regional
registers (70,000 people currently). If applied to the whole community, this
would offer a unique opportunity for analysis, reporting, planning, and
management of this national problem. Already the programme has accumulated 4
years of data.
Diabetes could thus be viewed as our single best performance
indicator in measuring quality of primary care crossing over into the more
costly hospital sector.17,18 Since starting,
the Get Checked record has monitored annually glycated haemoglobin
(HBA1C), lipids, systolic and diastolic blood
pressures, peripheral (foot) sensation, urinary microalbuminuria as well as
retinopathy by retinal photography.
There is excellent evidence showing that early intervention
with angiotensin-converting enzyme inhibitors in Type 1 diabetes or angiotensin
II receptor blockers in Type 2 diabetes can reverse microalbuminuria, reduce
overt proteinuria, and delay progression of renal
failure.19–22 This strategy is thus
recommended in the national guidelines.5
Consequently, the data from the Get Checked programme could provide a
powerful mechanism for delaying or preventing
CKD.23-25 However plasma creatinine and eGFR
had unfortunately been omitted from the programme and will not be included until
later in 2006 (personal communication, Dr S. Dawson, Ministry of Health).
In addition, despite yearly tests of microalbuminuria, renal
indicator results are not generally reported back to local diabetes teams for
inclusion in the yearly Get Checked review to each District Health
Board.
Even in Canterbury, Ministry of Health clinical indicators
for detection rates by local diabetes services have not met their target rates
among Maori (e.g. the detection rate in 2004 for Maori was 485 cases or 40.7%
instead of the target 953 or 80%7). Until early
intervention is uniformly available, perhaps when detection rates achieve 100%,
an expansion of diabetic complications including nephropathy is inevitable.
These figures highlight the limited awareness at the professional,
administrative or executive level of either the very high
mortality26 or the potential high cost savings
offered by early intervention and effective treatment of microalbuminuria or
proteinuria in diabetic nephropathy.27
We would suggest that monitoring and analysis of the data
available in the Get Checked programme should be used to better inform
our planning for the likely growth in diabetic nephropathy. Alternative
strategies may also be appropriate, but a national initiative is clearly
required.
Conflict of interest statement: The
authors have no conflicts of interest.
Author information: Zoltán Endre,
Professor;1,2 Don Beaven, Emeritus
Professor;1 Adrian Buttimore, Manager Dialysis
Services2
1Department of Medicine,
Christchurch School of Medicine and Health Sciences, University of Otago,
Christchurch1
and
2Department of Nephrology,
Christchurch Hospital, Christchurch
Acknowledgements: We thank Kelvin Lynn,
David McGregor, Richard Robson, Martin Searle (nephrology consultants,
Canterbury DHB); Wei Yoon, Gary Aiken, Theam Chew, Warren Bennett (Canterbury
DHB Financial Services); David Vial (Finance Manager, Adult Health Services,
Auckland City Hospital); Divisional Support Services, Auckland DHB; Julie Harris
(Surgical Services, Counties Manukau DHB); Pauline Hansra (Funding Services,
Counties Manukau DHB), Justin Roake (Professor, Canterbury DHB Surgical
Services); Wayne McNee and other staff at PHARMAC; and Peter Moore and Helen
Lunt (CDHB Diabetes Services).
Correspondence: Professor Zoltán H
Endre, Department of Medicine, Christchurch School of Medicine and Health
Sciences, PO Box 4345, Christchurch. Fax: (03) 364 0935; email: rowena.fisher@chmeds.ac.nz
References:
AppendicesAppendix 1 Figures used to calculate
Figure 1 were obtained from the Australian and New Zealand Dialysis and
Transplantation Registry(2004) 13 and the
Christchurch Nephrology data base (PROTON, Clinical Computing Plc, London,
United Kingdom).
Appendix 2 Figures in Table 1 were
derived as follows:
The number of subjects with renal failure in New
Zealand were obtained from the ANZDATA
Registry13) which has comprehensive data on all
patients who are transplanted or who remain on dialysis for more than 3 months.
Subjects with renal failure are first referred from primary or secondary care
services to one of eight nephrology services. These advise on surveillance or
appropriate management including type of dialysis (haemodialysis or peritoneal),
creation of vascular or peritoneal access and institution of dialysis, and
location (in centre, satellite or home dialysis). Each nephrology service
maintains their own costings and recoveries.
The total cost of each modality is derived by
multiplying the agreed cost for each modality by the number of patients. The
agreed costs are based on 2005 costs for three major District Health Boards
following a meeting of the National Advisory Committee and financial controllers
and purchasing officers. This follows meetings of the Cost Boundary Flow
Consensus Meeting (Interdistrict Health Boards flow indicators from price and
volume). Previous recoveries used the Victorian case-based mix formula of 2001
of $24,500 per patient which clearly failed to meet real annual costs. Satellite
dialysis refers to centres such as Tauranga where dialysis is instituted at a
major centre (eg Hamilton or Auckland) and then continued at a centre without a
major nephrology service.
The audited $5.9 million for Christchurch, with which
these figures are compared, are based on 109 dialysis patients in the year
2003-4 and includes other attendances and organ co-ordination but not inpatient
bed days. The total figure of $63 million may be an underestimate because of
differences amongst accounting procedures and differing assumptions within each
DHB. However, the comparisons between our CDHB sample and audit suggests that
these assumptions are reasonable.
Depreciation, security, heating, lighting, overheads of
accommodation for dialysis including training facilities has been built into the
nationally assessed costs for Board recoveries at 35% of building and equipping
new facilities. These are included in the agreed costs in Table 1. Similarly,
dialysis machines which are purchased under contract and yearly costs have been
assessed using the agreed figure rather than total asset costs, new machine
costs and depreciation.
Appendix 3 Total transplant numbers
for New Zealand are obtained from the ANZDATA registry. After surgery, the cost
of immunosuppression is the dominant cost for functioning kidney transplants.
The figure of $8 million is based on best practice immunosuppression. The cost
of immunosuppression for patients within each nephrology service is difficult to
estimate accurately because the PHARMAC funding policy separates “in
community” pharmaceuticals from “in hospital” prefunding. We
are nevertheless grateful for the assistance of Wayne McNee and the staff of
PHARMAC who have been able to break down the costs of 61 different drugs
classified as immunosuppressives to those used in kidney transplantation
management. The costs compare with best practice use of both cyclosporin and
mycophenylate but not for other drugs used for transplantation. Auditing of the
Christchurch figures produced a figure within 10% of these estimates.
The other costs of transplantation in the three
transplanting centres in New Zealand (Auckland, Wellington and Christchurch) are
estimated to average $27,000 per patient treated. Clearly costs will vary per
centre, and this cost is based on an average theatre cost of $20,200, a
retrieval cost for each cadaver kidney of $2,116 and an average length of stay
based on Christchurch data of 5 days. Different centres have different lengths
of stay as alluded to earlier, and these are confidential to each centre. In
addition, the total cost is based on cadaveric transplantation. The cost of live
donor transplantation will be greater because of the additional length of stay
and theatre costs. We therefore regard our estimate as very conservative,
underestimating the true cost by at least 100% .
Appendix 4 The data in table 3 are
based on a consultant time estimation of 70% spent in the management of patient
evaluation, outpatient visits, planning, inpatient management, assistance with
transplantation and supervision of all in and outpatient dialysis. A survey of
each of the 3.5 FTE consultants in Christchurch (total expenditure $686,000)
provided a conservative figure of 70% of FTE giving an annual cost of 515 hours
at $100 spent in the management of 109 patients. Assuming that this reflects
only 11% of the national CKD work load, a conservative total figure for New
Zealand would be $5 million.
The data for erythropoietin were obtained through
courtesy of PHARMAC. We note that the use of erythropoietin requires appropriate
authorisation by a nephrologist and is utilised in both patient approaching CKD
as well as in patients on dialysis. The nonassessed costs include consultations
for early renal failure, community pharmaceuticals, nonhospital provider
solutions including dialysate and intravenous fluids, costs of arteriovenous
fistula formation, costs of live donor assessment and costs over and above
Canterbury donor transplantation, costs of the Registry, costs of Advisory
Committees and central ministry costs and other unallocated or unmeasureable
costs such as patient transport, tissue typing etc. We have allowed 7% of total
renal failure costs for these and believe this to be very conservative. This is
7% of $84 million, giving a total of $6 million.
Appendix 6 ANZDATA Registry figures
were used to assess ethnicity and attribution of diabetes as primary cause of
chronic renal failure13. The 40% figure was
obtained for the year 2003. This number is unlikely to decrease as the
prevalence of diabetes in New Zealand is increasing. There are currently no
checked data audited to assess the incidence of increasing albuminuria.
Appendix 7 Work related costs, loss of
quality of life for patients or their relatives and the economic impact of this
has not been attempted in this survey.
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|