Calculating the global burden of disease: time for a strategic reappraisal?

Large organizations responsible for health care face formidable problems in gathering and deploying data relevant to their principal tasks, which are to monitor the health of the communities they serve, and to decide where resources could most effectively be used to improve things. In the short run, health improvements come by choosing and supporting the right health care activities. In the long run, they also come by choosing and supporting the right research activities. The ambitions of the Global Burden of Disease Study [1] are to contribute to all three of these important objectives. Broadly speaking, the argument is that if we knew the impact of each disease or injury upon people’s life expectancy and upon the (health-related) quality of their lives, and if we knew the incidence and prevalence of each disease or injury, we could use this information to monitor population health, to establish priorities between interventions, and to guide research priorities. In what follows, I am going to consider these general claims from three different viewpoints. In the first phase I will abstract from practical difficulties and ask whether, in principle, calculating the global burden of disease is the best way to approach each of these problems. I shall conclude that it is not. This will lead into the second stage, in which I shall ask whether there is anything useful that could be extracted from the global burden of disease calculations as they are actually performed, that would help with the three major problems that have to solved. I shall conclude that there is a little, but that it is not worth the cost, and that resources could, and should, be better targeted on discovering those things that we really need to know. This will lead into my third and final section, where I aim to be rather more constructive, by offering an alternative strategy that would achieve the above objectives more straightforwardly. Although I shall direct my remarks specifically to the Global Burden of Disease Study (henceforth GBD) as conducted under the impressive leadership of Murray and Lopez, much of what I have to say applies in principle to any burden of disease (or cost of illness) approach, applied at any level (global, national, regional or local). But in the second section I will be considering some rather idiosyncratic elements in the Murray– Lopez approach that other calculators of the burden of disease have omitted (for reasons good or bad). And because I regard one particular element in the GBD enterprise, the disability adjusted life year (DALY), as a particular variant of the more general concept of the quality adjusted life year (QALY), my concern here is not with the use of that general concept (which I support wholeheartedly) but with the specific features of the DALY that make it less useful than other types of QALY that might have been used [2].