Gone are the days when physicians were given carte blanche to do what seemed best for their patients. In its place are almost daily challenges to that decision-making autonomy. Pharmacy benefit managers call my office to obtain permission to substitute an alternate prostaglandin. I need to prove to the hospital that I am qualified to perform a new surgical procedure. And now, I need to provide documentation of quality patient-care measures in order to receive my 1.5 percent fee increment (well, actually, it’s a 1.5 percent less decrement). I’m sure you can think of many more examples. These scenarios are mirrored for all of us collectively. Payers demand proof that what we do for our patients makes a difference in their lives, improves outcomes, is done according to the book, and doesn’t cost very much. Especially the latter.
Back in the carte blanche days, we had no proof of these things to offer, had proof been requested. The first shots across our bow came from a 1983 article by David Eddy and associates,1 who asserted that there was no proof that glaucoma treatment made any difference, and then from the bean counters at the actuarial firm Milliman & Robertson who distributed guidelines to U.S. health care organizations calling for tight limits on second eye cataract surgery. Thankfully, our profession got busy. Our NEI-funded trials in glaucoma conclusively showed the benefit of treatment, and other studies showed that getting by on one eye was indeed not enough, if two could be available.
There are two other facets to the problem of proving treatment value, apart from showing improvement in outcome of treated vs. untreated patients. One is to show the cost-utility of treatment. Pioneered in ophthalmology by Drs. Gary and Melissa Brown, the method compares benefits across treatments of all kinds. The common denominator is cost per quality-adjusted life-year gained. With this data, society can rationally choose the most cost-useful treatments in a setting of scarce resources.
The final facet of the problem is a determination of the costs of treatment vs. the social costs of not providing treatment. In the case of ophthalmic conditions, little large-scale hard data has existed heretofore. For example, what are the costs to the Medicare program related to care of the visually disabled? In the February Ophthalmology is a paper that warrants attention. Jonathan Javitt and associates, using the publicly available analytic sample of 5 percent of all fee-for-service Medicare beneficiaries for 1999–2003, tracked costs of all claims including nonophthalmic ones. They identified enrollees with moderate and severe vision loss, and those who were blind, and compared them against beneficiaries with no vision loss. Significantly increased costs were identified in every category of vision loss, especially in those whose vision worsened during the period. Notably, approximately 90 percent of their costs were nonophthalmic in nature. Criticism can be leveled against any study using claims-based data, and against some of the methodology, but it is hard to dismiss such striking results as method bias.
In the meantime, all of us need to be advocates on behalf of our patients for what we do. We need to know where the data for dubious doubters are deposited. Most of all, we need to remind decision makers about the personal and family costs of vision loss and blindness, most often paid in emotional currency and lost opportunity.
1 Surv Ophthalmol 1983;28(3):194–205.