Economic evaluations

Nick R Anthonisen

The present issue of the Canadian Respiratory Journal contains two papers dealing with cost evaluations (1,2). Both involve the evaluation of previously published datasets, and, appropriately, both are specific to the Canadian situation. Both use methods of measuring cost that may seem simple, or at least comprehensible, but are, in fact, the kind of thing that requires special expertise, ie, cannot be done by amateurs. Therefore, it is not a coincidence that the lead authors of both papers have their primary degrees in Pharmacy, where they presumably are taught this kind of thing. Fortunately, the Canadian Respiratory Journal also has access to an expert in this area, Dr Andre Lamy, a surgeon (!) from Hamilton, who regularly and graciously reviews this type of study. His contributions are greatly appreciated and it should be noted that he did not review this editorial and cannot be held responsible for it.

The first paper reports Canadian costs for alternative therapies of asthma tested in an international trial sponsored by a member of Big Pharma (1). Although Canadian patients were included in the trial, the whole dataset (more than 2000 patients) was used for the analysis (3), with Canadian costs attached. The evaluation was very thorough. Drug costs were of course measured, as were care costs such as physician and emergency room visits, other health care visits, hospitalizations and losses of wages by both patients and their caregivers. I cannot think of anything that this study might have omitted, and it is hard to argue with the monetary values assigned to each. The study set out to test whether a regimen of a combined budesonide-formoterol inhaler, used as both maintenance and rescue therapy, was as effective and less costly than a fixed dose combined fluticasone-salmeterol inhaler used as maintenance with additional salbutamol for rescue purposes. Combination inhalers are expensive, so that both treatment groups had substantial drug costs. However, they were somewhat less in the budesonide-formoterol group, although this group did not use the relatively cheap salbutamol. Further, there were fewer exacerbations in the former group, and because these are the chief cost-drivers, the difference increased when all costs were considered. My interpretation of these results is that budesonide-formoterol is probably a better rescue agent than salbutamol, presumably because its use increases the intake of inhaled steroids, and that variable dose therapy dictated by the patient is likely to be better than fixed-dose therapy. The latter is the more important point, I think. Intelligent self-management is the best management of asthma or any other chronic disease, and prescriptions and physicians that encourage this behaviour are most likely to be successful. One could regard the ‘fixed dose’ arm of this trial as a bit of a straw man, because it denied the patients this opportunity to some extent. It should also be noted that we all have patients who seem to be incapable of intelligent self-management, and that in such people fixed-dose regimens are eminently sensible.

The second study (2) is much more restricted, involving the costs of diagnosing occupational asthma. It was authored by the groups most active in this area in Canada, which have been world leaders in this diagnosis. Previous techniques for making this diagnosis have largely involved monitoring peak expiratory flow through working and non-working days, as well as specific inhalation challenges with suspected inciting agents. The former technique is relatively cheap, but is nonspecific and very dependent on patient cooperation. The second is difficult and expensive, and, in fact, is carried out as a clinical routine in only a few Canadian centres, including those of the authors. It is, however, the gold standard. Recently, this group has examined the utility of examining induced sputum for eosinophils, and noted that people with occupational asthma had increases of 1% or more in sputum eosinophilia at work (4). In the current study, they look at the costs involved and the results obtained as compared with the gold standard. Sputum examination costs were estimated to be roughly comparable with those of peak expiratory flow monitoring. A large number of other factors were considered but the take home message is clear. Peak expiratory flow monitoring is not very good, making the correct diagnosis in approximately 50% of the cases. Sputum examination was substantially better, leading to a correct diagnosis in 72% of cases. Presumably, this was because it was less dependent on patient cooperation. So the message is that one is better off doing sputum examinations than peak expiratory flow monitoring, in terms of cost-effectiveness, that is, dollars spent for a correct conclusion. There are several caveats. First, sputum eosinophilia is not necessarily an easy thing to do, and comparable cost-effectiveness is dependent on doing it as well as the authors, the world experts. Second, is a 72% success rate adequate for a test of occupational asthma? It would be easy to argue that it is not, and that everyone should have specific inhalation challenge for the diagnosis, but this is nearly impossible in any place but Quebec and Ontario. We are left with a conundrum. I suspect that in many parts of Canada the diagnosis is made on history alone, often giving the benefit of the doubt to the patient. This is, I believe, defensible, but it would probably be better if more centres learned to measure sputum eosinophilia.