NDT Advance Access published online on June 2, 2009
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfp257
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Comorbidity data collection by renal registries – a remaining challenge
1 ERA-EDTA Registry, Department of Medical Informatics, Academic Medical Center, University of Amsterdam, The Netherlands 2 CNR-IBIM Clinical Epidemiology and Pathophysiology of Renal Disease and Hypertension, Renal and Transplantation Unit, Ospedali Riuniti, Reggio Calabria, Italy
Correspondence and offprint requests to: Kitty J. Jager; E-mail: k.j.jager{at}amc.uva.nl
Keywords: comorbidity; data quality; registries; survival
 |
Introduction |
|---|
 Top Introduction Comorbidity definitions and data...What do registries need...What more can we...References |
|---|
In the review on case-mix-adjusted international renal registry comparisons, published in this issue of the journal, Karamadoukis and colleagues put their primary focus on the collection of comorbidity data [1]. To this end they ask the readers’ attention for a number of considerations related to the collection of such data.
Ideally, comorbidity data collection should be complete, sufficiently detailed, reliable, as easy as possible for the health care professionals in the renal centres, and last but not least, provide added value for registry analyses in comparisons between renal centres and countries. In this respect, current daily practice may sometimes fall short of expectations.
|
Comorbidity definitions and data completeness |
|---|
|
TopIntroductionComorbidity definitions and data...What do registries need...What more can we...References |
|---|
Karamadoukis et al. report on the comorbidity data collection in four registries and found that the data completeness varied from 55% (UK) to 100% (USRDS). Nevertheless, it is important to note that the latter used a ‘tick if present’ policy in which case it is not possible to distinguish between the absence of comorbidity and missing data. Some years ago, Stel et al. investigated the prevalence of comorbidity in another four registries collaborating within the ERA-EDTA Registry, in addition to that in the UK Renal Registry [2]. In those other four registries, data completeness ranged from
99% (Austria) to 100% (Norway) that also used a ‘tick if present’ policy (Dr Vianda Stel, personal communication). This suggests that in this selection of countries, with exception of the UK, comorbidity data were reasonably complete. In addition, both studies showed that there were some differences between registries in the definitions of the comorbidity items. The problem of substantial comorbidity data incompleteness is that it bears the risk of studying selected populations and that it hampers statistical adjustment for case-mix in any comparisons between centres and registries. Efforts to increase data completeness as much as possible are therefore commendable, including the ones already brought into practice by different registries: returning forms with incomplete comorbidity data to the centres, pre-populating fields with previous years’ returns and teaching of individuals interested in registry reporting in data entry and handling. Nevertheless, such efforts cost time that need to be weighed against the usefulness of having complete data available. Another valuable approach, that of using other databases to update the renal registry database, is only possible in countries where it is allowed to merge them for this specific purpose. In this respect, the USRDS has an advantage over most renal registries in Europe. In relation to ‘good practice’ in comorbidity data collection, we would like to add that it is preferable to make a difference between missing data and the absence of comorbidity. We believe that although used in the majority of studies, for reasons of data quality, ‘tick if present’ policies are less ideal. Standardization of the definitions of comorbidity data remains important especially for international projects, and therefore this task has been included in the work of the ERA-EDTA Registry QUEST Coding and Definitions Working Group [3].
|
What do registries need comorbidity data for? |
|---|
|
TopIntroductionComorbidity definitions and data...What do registries need...What more can we...References |
|---|
As far as international survival comparisons are concerned, the value of having comorbidity data available for adjustment on top of age, gender and primary renal disease (PRD) has not been proven beyond question. Van Manen et al. found that five comorbidities explained only an additional 1.9% in the variance in survival on top of the 14.4% explained by age, gender, PRD, treatment modality and country [4]. Their results also demonstrated that after a first adjustment for age, gender, treatment modality and country, the moderate survival effect of PRD (renal vascular disease versus glomerulonephritis) was only somewhat reduced after further correction for comorbidity, and in other comparisons such as that of countries the effect of adjustment was even less pronounced. From a statistical point of view, this is not surprising as comorbidity is very much associated with age and with PRD, so adjustment for such factors already partially ‘corrects’ for comorbidity. Van Manen et al. also performed analyses to investigate if some comorbidities have more effect than others, as is suggested by Karamadoukis et al., and found that there was no consistency in the importance of comorbidities across countries and age groups and that there is no apparent hierarchy on the basis of which we could decide which comorbidities are crucial to include in registry data collection and which are not. The fact that adjustment for comorbidity affects international survival comparisons only in a very limited manner [4,5] may be due to the fact that, in contrast to what people may expect, the prevalence of comorbidity across countries is remarkably similar. The only exception to that rule is that in countries with a relatively high percentage of diabetic nephropathy, the prevalence of diabetes-related comorbidity like ischaemic heart disease and peripheral vascular disease is higher [2]. Then the question arises why should we collect comorbidity data? Karamadoukis et al. mention ‘emotive’ reasons. Though this reason is not very scientific, and without saying that international survival comparisons are not useful without additional adjustment for comorbidity, we do agree with them as even unjustified distrust in the (lack of) data will create disbelief in any analytic results coming from them. At the national level, there are at least two other reasons for collection of comorbidity data. The first is that similarity in the prevalence of comorbidity may exist across many countries, but this prevalence may be quite different across centres, because referral patterns may be different and patient numbers are much smaller than in country samples. For these reasons, variation in comorbidity between centres is likely, and therefore such data may be more useful in centre comparisons that are the core business of many national and regional registries. Finally, the presence of information on comorbidity will allow researchers to select specific groups of patients. This is important as registries also serve as sampling frames for special studies.
|
What more can we do to increase data completeness? |
|---|
|
TopIntroductionComorbidity definitions and data...What do registries need...What more can we...References |
|---|
Given the time-consuming character of data collection on paper or even by web-based data entry, the need for standardized information technology (IT) systems that allow extraction of the data needed for the registry is high. This is more easily said than done, as the basis for data extraction is formed by electronic patient management systems that, in order to include all data needed, preferably need to have more or less fully replaced paper patient records. In this respect, it is notable that the highest degree of incompleteness in comorbidity data is present in the UK where there is an automated data extraction system. An advanced stage of renal IT development is being approached in some countries, like the UK and some regions in Italy, but is far off in many countries, let alone the fact that in most countries where electronic patient records are being used, there is a great variety in the systems implemented. For registries to collect data from this ‘mishmash’ of systems, standardization of data extraction for the registry is crucial. A contribution to the solution of this problem is made by the NephroQUEST project [6].
As Karamadoukis and colleagues suggest, the introduction of performance-linked payments could indeed act as an incentive to improve reporting of comorbidity. We, however, think that as long as the explained variance in survival of RRT patients by the data collected in registries or studies is so low, usually in the order of 16–35% [4,7] with an occasional outlier as high as 55% when biomarkers are included [8], basing payments on outcomes like adjusted mortality would be unfair to nephrologists and centres, may introduce gaming instead of quality improvement [9] and should therefore be avoided. Furthermore, research shows that outcomes frequently do not have a one-to-one relationship with quality of care [9]. On the other hand, we do fully support payments or other advantages, including certification of dialysis centres, linked to the provision of high-quality data to a registry or linked to good practice based on clinical process data. In addition, we support disincentives based on proven malpractice, as such initiatives may also contribute to improvement of quality of care.
In conclusion, we feel that collection of comorbidity data by renal registries is valuable for a number of reasons and that the way forward is to stimulate centres to collect such data, for example by making it easier and rewarding for them to do so.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopIntroductionComorbidity definitions and data...What do registries need...What more can we...References |
|---|
- Karamadoukis L, Ansell D, Foley R, et al. Towards case-mix adjusted international registry comparisons: how can we improve data collection practice? Nephrol Dial Transplant (2009) doi 10.1093/ndt/gfp096 [epub ahead of print].
- Stel VS, van Dijk PC, van Manen JG, et al. Prevalence of co-morbidity in different European RRT populations and its effect on access to renal transplantation. Nephrol Dial Transplant (2005) 20:2803–2811.
[Abstract/Free Full Text] - Jager KJ, Zoccali C. QUality European STudies (QUEST)—a step forward in the quality of RRT care. Nephrol Dial Transplant (2005) 20:2005–2006.
[Free Full Text] - van Manen JG, van Dijk PCW, Stel VS, et al. Confounding effect of comorbidity in survival studies in patients on renal replacement therapy. Nephrol Dial Transplant (2007) 22:187–195.
[Abstract/Free Full Text] - Goodkin DA, Bragg-Gresham JL, Koenig KG, et al. Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States: the Dialysis Outcomes and Practice Patterns Study (DOPPS). J Am Soc Nephrol (2003) 14:3270–3277.
[Abstract/Free Full Text] - Jager KJ, Zoccali C. Quality of care in end-stage renal disease: the importance of comparing ‘apples with apples’. Nephrol Dial Transplant (2008) 23:1116.
[Free Full Text] - Jager KJ, Merkus MP, Dekker FW, et al, for the NECOSAD Study Group. Mortality and technique failure in patients starting chronic peritoneal dialysis: results of the Netherlands Cooperative Study on the Adequacy of Dialysis. Kidney Int (1999) 55:1476–1485.[CrossRef][Web of Science][Medline]
- Mallamaci F, Tripepi G, Cutrupi S, et al. Prognostic value of combined use of biomarkers of inflammation, endothelial dysfunction, and myocardiopathy in patients with ESRD. Kidney Int (2005) 67:2330–2337.[CrossRef][Web of Science][Medline]
- Lilford R, Mohammed MA, Spiegelhalter D, et al. Use and misuse of process and outcome data in managing performance of acute medical care: avoiding institutional stigma. Lancet (2004) 363:1147–1154.[CrossRef][Web of Science][Medline]
Accepted in revised form: 5. 5.09
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Hommel, S. Rasmussen, M. Madsen, and A.-L. Kamper The Danish Registry on Regular Dialysis and Transplantation:completeness and validity of incident patient registration Nephrol. Dial. Transplant., October 26, 2009; (2009) gfp571v1. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||