advertisement
NNT. Treatment for Ocular Hypertension and Early Glaucoma: How Much Benefit and is it Worth the Cost?
Roy Wilson
Two recently published randomized controlled clinical trials have demonstrated the benefit of lowering intraocular pressure to reduce the incidence of developing glaucomatous damage and to reduce the rate of progression of glaucomatous damage. In the Ocular Hypertension Treatment Study (OHTS), the 5-year cumulative probability of developing primary open-angle glaucoma (POAG) was reduced from 9.5% in the ocular hypertension group with no treatment to 4.4% in the ocular hypertensive group that was medically treated. In the Early Manifest Glaucoma Trial (EMGT), progression of newly detected open-angle glaucoma was reduced by immediate treatment from 62% (no treatment or later treatment group) to 45%.
Treatment in the EMGT resulted in a relative risk reduction of 27% (62-45/62), an absolute risk reduction of 17% (62-45), and a number needed to treat, NNT, of 6 (1/.17). Thus, 6 people would need to be treated to obtain one therapeutic success (prevent progression of glaucomatous damage). The patient population in the EMGT had glaucoma that would typically be defined clinically as "early." However, it can certainly be argued that once definite glaucomatous damage is present, the disease process is pathophysiologically at an advanced stage. To treat 6 patients to prevent glaucomatous progression in one is thus certainly desirable.
The situation with ocular hypertension is arguably different. Patients in the OHTS had no demonstrable disease. With the endpoint being development of optic nerve abnormality or visual field defect, treatment resulted in a relative risk reduction of 54% (9.5-4.4/9.5), an absolute risk reduction of 5.1% (9.5-4.4) and a number needed to treat of 20 (1/.051). Whether or not it is desirable to treat as many as 20 patients to prevent just one patient from developing an optic nerve or visual field abnormality is debatable. The optic nerves were assessed by expert readers who had the benefit of comparing high quality stereo photographs. Their expertise in detecting change is probably beyond that of most ophthalmologists in the clinical setting. If a stricter, more clinically applicable definition of demonstrable functional change is used as an endpoint, the number needed to treat soars to 42.
Does the benefit of preventing one person from developing glaucoma justify treating 20 (or 42) persons with ocular hypertension? The risk of topical medical treatment was found to be low and is probably not a factor that needs serious consideration in answering this question. However, there is a monetary cost to topical hypotensive medication and this cost may be substantial. Jeffrey Schultz from Albert Einstein College of Medicine has calculated that treating 42 patients for five years at one dollar per day would yield a medical cost of $76,650. From a societal perspective, it would appear that such a sum is a lot of money to spend to prevent one patient from going from ocular hypertension to early, clinically irrelevant functional loss.
Additional comment by E.L. Greve and R.A. Hitchings
This brings us back to the essential glaucoma management choice: based on RISK or on FACT. The important ethical and financial notion of NNT, Number of patients Needed to Treat in order to prevent a deterioration in ONE patient, is based on RISK evaluation. It is questionable - to say the least - whether an NNT of 42 needed to prevent the earliest unnoticed visual field defects is acceptable to indeed start treatment (including its psychological effect on Quality of Life) in that many patients. Is its cost justifiable? Is there no other option? Of course there is, and it is called management of glaucoma based on FACT. The FACT is a proven RATE of PROGRESSION probably leading to reduced Quality of Vision during the patients lifetime. In that case the NNT is ONE. There is little doubt that in this ONE particular case treatment is needed. The obvious question is: can we afford to delay treatment until we have established the RATE of PROGRESSION. Yes we can if we make a careful distinction between those patients where follow-up is acceptable and those where it is not (after R.A.Hitchings at Gullstrand Meeting, May 23-24, 2003, Uppsala, Sweden). Follow-up assumes the possibility of regular monitoring in accurate patients and easy and sensible interpretation of results. Follow-up may be sensible in elderly patients, when a slow rate of progression is likely, in non-responders, or when an allergy to medication is present. Treatment may be started when it is impossible to detect progression because of poor observer or equipment, when a rapid rate of progression is likely, in young patients, when paracentral loss is present, in one-eyed patients and with high IOP. Why not go for an NNT of ONE whenever possible?
