advertisement
Using state-of-the-art decision-analytic methods, Rein et al. (934) examined two related questions: 1) Assume that we do not reimburse the physician for the glaucoma examination, but conduct such an examination nonetheless, is it cost-effective to treat the patients identified by the examination?; and 2) Assume we do reimburse the physician for the examination, is it cost-effective to test and treat? It is not surprising that the investigators found an answer of 'yes' to the first question. The cost-effectiveness of IOP lowering treatment for people at moderate risk for developing glaucoma was established in 2005 by the OHTS study. This finding is simply an extension of that, indicating that if we are sure that the person has glaucoma, we should treat them. The latter question is far more provocative, as this asking if providing payment for the glaucoma examination (in this context, at U.S. rates) is good use of social resources. The finding of Rein et al. indicates that this
Providing payment for the glaucoma examination is good use of social resourcesis indeed good use of social resources. Thus, for glaucma specialists operating in 'fee for service' payment environments, a compelling case might be made that this reimbursement for this exam is likely to meet accepted standards for cost-effectiveness, although it would likely be necessary to conduct a jurisdiction-specific analysis to confirm this finding, based upon clinical practice and charges for care outside the United States.
The does remain an important unanswered question from Rein's study. He and his colleagues assume that if you do not pay the general ophthalmologist to conduct the glaucoma examination, the ophthalmologist will do it, thus resulting in the estimated yield of patients for treatment. However, there is no empirical evidence that this is the case. Indeed, most economic studies have shown lower usage of tests when they are not reimbursed in a fee-for-service environment. If this is the case, then there would be more patients remaining untreated in the unreimbursed scenario than in the reimbursed one. That being the case, it is likely that the latter scenario is even more likely to meet accepted standards of cost-effectiveness than was found by Rein et al.