Commentary: Economics of diabetes care must move forwards

BMJ 1995; 311 doi: (Published 16 December 1995) Cite this as: BMJ 1995;311:1599
  1. P D Home, professor of diabetes medicinea
  1. aDepartment of Medicine, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH

    A problem of clinical science, albeit ultimately a happy one, is that it does not stop moving forwards. Even this autumn, the observations by the Gentofte group that percentage survival after onset of diabetic nephropathy over 16 years has quintupled and that median survival time has trebled, in association with lower blood pressure, will lead many to intensify their efforts to control even minor degrees of hypertension (>140/90 mm Hg) in these patients.1 So much is consistent with the current paper, but meanwhile the Gentofte group has also defined a 75% reduction (40% to 10%) over eight years in progression to proteinuria (albumin excretion >300 mg/day) in those patients with type 1 diabetes treated with angiotensin converting enzyme inhibitors who have microalbuminuria but not hypertension (mean blood pressure 127/78 mm Hg).2 Can this be reconciled with the concerns evidently felt by Kiberd and Jindal that microalbuminuria screening may fail the test when it comes to increased quality of life for the sums invested?

    The management of diabetes mellitus is for the most part preventive medicine in that a high proportion of the efforts expended by people with the disease and professionals give no immediate return in health. In the primary prevention of microvascular complications the lead times are particularly great (several decades), and this results in particularly difficult judgments between the present and the future in both health (hypoglycaemia now or blindness later) and economic (intensive insulin therapy now, renal replacement therapy later) terms. Evidently the only appropriate way to answer these questions is to perform modelling along the lines that Kiberd and Jindal describe, but in doing so these authors also illustrate that there are fundamental unmet needs in defining some of the parameters that underlie their model.

    Such is evident from the sensitivity analysis (shown in their table 3) with sixfold variations shown in cost per QALY for microalbuminuria screening methods, and infinite variation for costs for angiotensin converting enzyme inhibitors (lower drug costs actually giving more QALYs for less total health care costs). More troublesome, however, are the assumptions about degree of benefit of this method of treatment, the length of the period of microalbuminuria before hypertension and of hypertension or proteinuria to renal failure, and the survival time and costs of management of renal replacement therapy. Indeed this poses another problem, for we are talking about medicine 20 years hence; will medical inflation continue to exceed underlying inflation (implying that discount rates should be adjusted downwards), or will novel cheaper forms (pig kidneys) of management of end stage renal disease be the rule?

    But even in its present form we can derive some useful messages from the present paper. Firstly, to be cost effective the prevention of complications several decades in the future must use methods (drugs, investigations, consultations) that are similar in cost (fully applied) to those in the cheaper industrialised nations today. Secondly, we need further economic studies, but perhaps preferably those which start from an acceptable cost/QALY, and work back to appropriate costs of management. Thirdly, there are still too few data on which to base reliable estimates of the effectiveness of treatment, as needed for this type of study; the bias against funding such studies of the natural and medical history of disease needs to be reduced.

    Perhaps the newly announced initiative of the International Diabetes Federation will help us to a better understanding of the economics of diabetes care and its needs.3


    View Abstract