Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Equity of access to transplantation was not adequately assessed
- Viliami K. Tutone, Mark S. MacGregor, Specialist Registrar, Chris J. Deighan, Consultant, Jonathan G. Fox, Consultant, Robert A. Mactier, Consultant and Scott TW Morris, Consultant (27 January 2004)
-
Equity of access to renal transplant waiting list and renal transplantation in Scotland: Author's reply
- Gabriel C Oniscu, John LR Forsythe, Consultant Transplant Surgeon (4 June 2004)
|
|
|||
|
Viliami K. Tutone, Visiting Specialist Registrar Renal Unit, Walton Building, Glasgow Royal Infirmary, G4 0SF, Mark S. MacGregor, Specialist Registrar, Chris J. Deighan, Consultant, Jonathan G. Fox, Consultant, Robert A. Mactier, Consultant and Scott TW Morris, Consultant
Send response to journal:
|
Oniscu and colleagues (1) report a lack of equity in access to the transplant waiting list for Scottish dialysis patients. This is an important research area, but several aspects of their study cause us concern. Selection for renal replacement therapy is a major bias which is not addressed. For example, a unit with a liberal policy of acceptance for dialysis regardless of co-morbidity, would appear to have a low percentage of patients on the transplant waiting list. A more conservative unit might accept fewer patients for dialysis, and therefore list a higher proportion for transplantation. Acceptance rates for dialysis are known to vary significantly across Scotland (2). Similar issues arise if dialysis is provided disproportionately to any subgroup described in the study, whether this is due to co-morbidity or other factors. When faced with new dialysis patients, co-morbidity is second only to patient choice in considering whether they should be listed for transplantation. Whilst mentioning co-morbidity briefly in their discussion, the authors say they were unable to examine it. We are the largest dialysis provider in Scotland serving 19% of patients. Like several other Scottish units, we have an electronic patient record, which was instituted in 1987 (3). Co-morbidity is easily obtainable from this database. We think the authors should have at least analysed co-morbidity in a subset of their study patients using one of the available co- morbidity scores which have been validated in dialysis patients (4-6). Ignoring co-morbidity makes their study difficult to interpret, and will mask any real inequities in the system. The key co-morbid factor of diabetes mellitus is reported, but in a misleading fashion: “Patients with diabetes have the lowest rate of access to the waiting list for transplantation.” No data are presented to support this statement, which is also repeated in the abstract. Diabetes mellitus is not a variable recorded by the Scottish Renal Registry (7). The data presented are for patients with diabetic nephropathy as the disease causing end-stage renal failure. However, many patients with other primary renal diseases will also have diabetes mellitus. For example, in our current dialysis population of 314 patients, 60 (19%) have diabetes mellitus, but only 36 (11%) have diabetic nephropathy as their primary renal disease. The authors analyse the time till 50% of all dialysis patients are put on the transplant waiting list, and call this analysis “intention to treat”, lending it a spurious legitimacy. There is never any intention to put the majority of dialysis patients on the waiting list for transplantation. A median time to waitlisting of 2.84 years lacks face validity. Adult renal services were externally audited in 2002 by the Clinical Standards Board for Scotland. All units except one met the standard of assessing all patients for transplantation within three months of commencing dialysis (8). We presume the authors treated death of non- waitlisted patients as a censoring event in their survival analysis. Whilst this may be appropriate in conventional survival analysis it is inappropriate in this study. Patients who die within a year would not and should not have been waitlisted. Indeed, current guidelines suggest expected survival of transplant candidates should exceed 5 years (9). Patients who die more than a year after starting dialysis without being waitlisted, are also likely to have significant factors precluding transplantation, and most would probably never have been waitlisted. We think their analysis produces misleading waiting times affected primarily by the rate of death in the non-waitlisted patients. In their more meaningful analysis of patients who were actually waitlisted, gender difference disappears altogether, and the maximum difference between any subgroup is 3 months. Unfortunately, the authors do not include these results in their abstract, nor are they given prominence in the discussion. The predictable result was that the media reported this study in a highly misleading way, emphasizing a gender difference which probably does not exist, and a waiting time in excess of 2 years for the deprived which is a statistical artefact. Similarly, choosing deprivation category 1 as the reference group, when it is the smallest group of patients (<5% of the population) and is an outlier from the trend, seems designed to maximise trivial differences. The authors conclude that research into fairness of access to transplantation should be pursued with dedication. We agree. 1. Oniscu GC, Schalkwijk AAH, Johnson RJ, Brown H, Forsythe JLR. Equity of access to renal transplant waiting list and renal transplantation in Scotland: cohort study. Brit Med J 2003; 327: 1261-3. 2. Metcalfe W, MacLeod AM, Bennett D, Simpson K, Khan IH. Equity of renal replacement therapy utilization: a prospective population-based study. QJM 1999; 92: 637-42. 3. Simpson K, Gordon M. The anatomy of a clinical information system. Brit Med J 1998; 316: 1655-8. 4. Khan IH, Catto GR, Edward N, Fleming LW, Henderson IS, MacLeod AM. Influence of co-existing disease on renal-replacement therapy. Lancet 1993; 341: 415-8. 5. Beddhu S, Bruns FJ, Saul M, Seddon P, Zeidel ML. A simple co- morbidity scale predicts clinical outcomes and costs in dialysis patients. Am J Med 2000; 108: 609-13. 6. Miskulin DC, Meyer KB, Martin AA, Fink NE, Coresh J, Powe NR, Klag MJ, Levey AS. Choices for healthy outcomes in caring for end-stage renal disease (CHOICE) study. Am J Kidney Dis 2003; 41: 149-61. 7. Scottish Renal Registry. Core dataset. http://www.show.scot.nhs.uk/srr/About/Main.htm Accessed 18th December 2003. 8. NHS Quality Improvement Scotland. Adult Renal Services: Local Reports. http://www.nhshealthquality.org/nhsqis/nhsqis_sub_publications.jsp Accessed 25th January 2003 9. UK Transplant. Transplant list criteria for potential renal transplant recipients. http://www.uktransplant.org.uk/about_transplants/organ_allocation/kidney_(renal)/national_protocols_and_guidelines/protocols_and_guidelines/transplant_list_criteria.htm Accessed 18th December 2003. Competing interests: All the authors assess patients with renal failure for their suitability for renal transplantation. |
|||
|
|
|||
|
Gabriel C Oniscu, Specialist Registrar Tranasplant Unit, Royal Infirmary of Edinburgh, John LR Forsythe, Consultant Transplant Surgeon
Send response to journal:
|
We are encouraged to see that the results of our paper (1) have stimulated an interesting debate, indicating that equity of access is a matter of concern for the transplant community as well as the potential transplant recipient. Tutone and colleagues (2) note that renal patients face a further inequity when receiving care since acceptance rates for dialysis vary from unit to unit. It is highly speculative to suggest that ‘a conservative unit’ might try to recompense for a low acceptance rate to dialysis by a high acceptance rate to transplantation. Patients wish a reply to two simple questions; Will I take longer than any other patient to get on to the waiting list? And; Do I have the same chance as anybody else in finally achieving acceptance on to that list? It is these questions that our paper attempted to address. Unfortunately Tutone and colleagues are wrong in saying that comorbidity is second to patient choice in listing for transplantation. Comorbidity may be so severe that listing would be inappropriate whatever the patient’s choice. Yet there is no doubt that comorbidity is one of the most major factors in this area of research, a fact acknowledged within our paper. Unfortunately validated data for comorbidity is very difficult to obtain despite the undoubted excellence of the electronic records in one centre. There is no agreed comorbidity set for all 4523 patients in our analysis and trawling through patient records to obtain a longitudinal comorbidity profile up to the moment of dialysis or any subsequent time point would have been a daunting, if not impossible task. Subset analysis in one unit could have been carried out but centre differences were noted in our study and therefore there may not be ‘read across’ to all other centres. Tutone and colleagues also state that the key comorbid factor diabetes mellitus, is reported in a misleading way. In fact, as is clearly shown in table 1 and in the Method section, the data presented in our paper refer to patients with diabetes as the cause of renal failure. It is these patients who have the lowest rate of access to the waiting list for transplantation. However we agree that diabetes as a comorbid condition (not leading to renal failure) is a very important topic and this particular group of patients merit separate analysis. The issue of comorbidity highlights yet again the lack of a meaningful core data set in registry databases. We would agree that there should be efforts to create a more comprehensive database such as that in the new Scientific Registry of Renal Transplant Recipients in the USA, which can be linked with other data sources providing socio-demographic and comorbidity details for patients on the registry. Tutone and colleagues also comment on statistical analyses in the paper. In the Scottish Renal Registry there is no variable which can reliably identify patients who will never be a suitable candidate for transplantation. Therefore, our professional statistical co-authors advised that the correct way to treat the analysis was by ‘intention to treat’. They would have felt uncomfortable about any other form of analysis. However, we anticipated the reaction of some clinicians in this regard and therefore carried out a separate analysis taking into account only listed patients. It is noteworthy that gender differences are absent in this analysis but that all other factors, including age, primary renal disease, deprivation category and transplant centre remain as statistically significant. As in any comparative analysis, one has to choose a reference group. The deprivation profile of the study population is identical with that of the general population in Scotland and therefore it seemed logical to use either category 1 or category 7 as a reference group (category 7 also represents less than 5% of the population). If Tutone and colleagues feel that this was done ‘to maximise trivial differences’ then that charge can also be directed at many other epidemiological and population bases analyses which use an identical deprivation distribution. It is natural that apparent inequity of access may sometimes produce a defensive reaction which states that all inequity can be explained away. Similar differences have been reported by the renal registry in England and Wales and elsewhere in the world. Some areas of inequity are relatively easy to explain, some require careful examination of factors (such as comorbidity) and some are almost impossible to explain. We owe it to our patients to define one from the others and to be transparent in the results. 1. Oniscu GC, Schalkwijk AAH, Johnson RJ, Brown H, Forsythe JLR. Equity of access to renal transplant waiting list and renal transplantation in Scotland: cohort study. Brit Med J 2003; 327: 1261-3. 2. Tutone VK, MacGregor MS, Deighan CJ, Fox JG, Mactier RA, Morris STW. Equity of access to transplantation was not adequately assessed. bmj.com 27 Jan 2004. Competing interests: None declared |
|||