The case against annual profiles
following AMI is different from the heart failure at week 47 post-AMI. Again, from an epidemiological point of view it is desirable to have a separate weight for AMI that can be applied for the estimated duration of six weeks and another weight for heart failure, preferably with distinc- tions for mild, moderate and severe. These weights for heart failure can then also be applied to heart failure arising from other diseases such as cardiomyopathy or valvular heart disease.
Another issue is related to chronic diseases of an episodic nature, which include a lot of mental disorders (major depression, bipolar disorder, schizophrenia, a lot of the anxiety disorders), asthma and epilepsy. These conditions differ considerably in the length of episodes: minutes in the case of epilepsy, hours to days in asthma, months to years in the mental disor- ders. Furthermore, the length of an episode may vary between countries and over time. For example, the proportion of time spent in a psychotic episode during five years of follow-up in the International Pilot Study of Schizophrenia of WHO (Leff et al. 1992) varied considerably between countries. Thus, one disability weight for schizophrenia assuming an av- erage split between time spent in psychosis and in between psychotic epi- sodes would be time and context specific and thus unable to reflect differences between countries and over time. This problem is circumvented by having a weight for the symptomatic episode and another weight for the time in between episodes if there is considered to be “rest-disability”. Existing differentials between countries and over time can then be reflected in the estimates of incidence and duration. This becomes more problem- atic the shorter the average duration of an episode. In the case of asthma there is some information on the average time spent with symptoms of wheeze and shortness of breath (e.g. Bauman et al. 1992) that would al- low valuation of time spent while asthmatic with symptoms (with an appropriate disability weight) and time without symptoms (at an appro- priate much lower weight). This would allow capturing changes over time. For instance, if through a concerted effort the coverage of preventive therapy for asthma with steroid inhalers dramatically increases one would expect the average proportion of time asthmatics are symptomatic to decrease and you would want your burden of disease methodology to be able to capture that change.
An alternative method for the disease with short episodic intervals is to develop weights for different levels of severity. An example is the fol- lowing staging for epilepsy: i) mild, asymptomatic for most of the time; ii) moderate, moderate symptoms most of the time; iii) severe, uncon- trolled. A similar staging for asthma is also possible: i) mild, asymptom- atic for most of the time; ii) moderate, moderate symptoms at least half of the time, e.g. coughing at night, wheezing at any time and feeling of breathlessness; iii) severe symptoms most of the time, necessitating re- peated hospitalizations or visits to emergency departments. I would pre- fer the first option with weights for symptomatic asthma and asymptomatic periods as that would give more flexibility to combine with