Appraisal of health care: from patient value to societal benefit

The Load Model: an alternative to QALY

BACKGROUND

QALYs are widely used in health economic evaluation, but remain controversial, largely because they do not reflect how many people behave in practice. This paper presents a new conceptual model (Load Model) and illustrates it in comparison with the QALY model.

METHODS

Load is the average annual weight attributed to morbidity and mortality over a defined period, using weightings based on preference judgements. Morbidity Load is attributed to states of illness, according to their perceived severity. When people are in full health, Load is zero (no morbidity). Death is treated as an event with negative consequences, incurred in the year following death. Deaths may be weighted equally, with a fixed negative weight such as -100, or differ according to the context of death. After death, Load is zero. In a worked example, we use the standard gamble method to obtain a weighting for an illness state, for both Load and QALY models. A judge is indifferent between certainty of 1.5 years' illness followed by death, or a 50/50 chance of 1.5 years' full health or 1-year illness, each followed by death. The weightings calculated are applied to a hypothetical life, 72 years in full health followed by 3 years with illness then death, using both models. Three other hypothetical outcomes are also compared.

RESULTS

For an example life, the relative size of the morbidity component compared with the mortality component is much higher in the Load model than in the QALY model. When comparing alternative outcomes, there are also substantial differences between the two models.

CONCLUSIONS

In the Load model the weight of morbidity, relative to mortality, is very different from that in the QALY model. Given the role of the QALYs in economic evaluation, the implications of an alternative, which generates very different results, warrant further exploration.

[From the design of use study to the assessment of the benefit: with or without pharmaceutical industry?]

The financing of clinical studies by the pharmaceutical industry is a controversial topic both internationally and in here in Germany. The well-known unacceptable shortcomings require no further confirmation. It is, however, indisputable that the pharmaceutical industry and medical science are co-dependent. Neither the marketing of industrial products nor the research and education of clinical scientists could function without this cooperation. Therefore, all partners need suggestions concerning goal orientation and consensus. The aim of this discussion is to formulate just such suggestions. To structure this discussion, we have raised the following questions: Must we always be suspicious of the results of studies financed by the pharmaceutical industry? We have to keep in mind that in Germany all clinical trials leading to approval of a drug were supported by the industry. What, exactly, do we want to achieve with our explicit and often justified criticism of these studies? What should be done to achieve a higher validity of the published data if we avoid answering the decisive question of whether we accept the challenge of continuing to let research and teaching be financed by the pharmaceutical industry or reject this kind of cooperation and support altogether.

[Is "evidence-based medicine" followed by "confidence-based medicine"?]

In an appeal concerning accusations of defamation, the England and Wales Court of Appeal Decisions determined that evidence-based statements are to be judged as opinions and not statements of fact. Since the authors consider it probable that this legal judgment will exert influence on physicians' decisions about the provision of health care services, they have compiled the implications of the judgment and discuss its consequences. The own analyses and considerations lead to the conclusion that confidence-based medicine follows evidence-based medicine. This extension is necessary because evidence-based medicine has not been able to generate the required trust. Therefore, it will be demanded to underpin the existing concept with additional data. These data will be necessary because it is no longer sufficient to convince scientists with data which are obtained under ideal conditions, but to convince critical members of society with additional data which have been obtained under everyday conditions.