Treating all patients with ocular hypertension may not be cost effective

With today's emphasis on evidence-based medicine, outcomes measurement, and cost effectiveness of treatment, the number of studies looking at these factors is growing.

This study from the University of South Carolina applies a well-known model of medical cost-effectiveness, the Markov model, to findings from the Ocular Hypertensive Treatment Study (OHTS). Cost estimates were derived from Blue Cross/Blue Shield data in the South Carolina region over a five-year period.

The Markov model assesses cost effectiveness in association with the quality of life gained, and the unit of benefit used for this analysis was the prevention of disease progression to glaucoma. An important and controversial assumption in this model is that an incremental cost effectiveness ratio (ICER) measuring less than $50,000 indicates that the treatment in question is cost effective.

In this study, the ICER generated by treating all patients with ocular hypertension to prevent one case from progressing to glaucoma was $89,072. By using adjustments for risk factors for progression in OHTS to determine the levels of cost effective treatment the following ICER levels were determined:

• Treating patients 20 years older than the average of 56 years, $45,155.
• Treating patients with 4 mmHg above the average of 25 mmHg, $46,748. 
• Treating patients with 40 microns less than the average central corneal thickness of 573 microns, $36,683. 
• Treating patients with a vertical cup-to-disc ratio of 0.2 wider than the average of 0.4, $35,633. 

This study is an interesting and important exercise for several reasons. It provides a unique way of examining the variation in the separate risk factors in the OHTS study in terms of their impact on the increased risk of developing glaucoma, as well as how they would interact in an economic model. It's acknowledged that one of the important conclusions of the OHTS is that treating all patients with ocular hypertension was not a reasonable approach. However, recommendations on which patients warrant treatment vary. This analysis provides another way to look at the data to at least make an attempt to determine the characteristics of ocular hypertension patients who warrant treatment.

The authors acknowledge that clinical decisions based on this type of model are less important than medical judgment. Other limitations of this study included the size of the OHTS trail. The Markov model is typically applied to much larger clinical trials in internal medicine and meta-analysis studies. Also, the five-year data horizon is relatively short time period, and this study did not take into account the indirect costs of treatment.