Decision Analysis With Markov Processes Supports Early Surgery for Large-Angle Infantile Esotropia

Abstract

To assess tradeoffs in time spent with aligned visual axes or stereopsis and risked reoperations between three strategies for the treatment of large-angle infantile esotropia: first surgery at 6, 24, or 48 months of age. Decision analysis. We simulated three hypothetic cohorts of children with this condition until their eighth year. Potential achievement of stereopsis, the diagnosis of dissociated vertical divergence or inferior oblique overaction, and various complications were modeled. Probability estimates were obtained from a comprehensive literature review. Analyses aimed to determine the strategy that was associated with longer time spent with aligned eyes and/or stereopsis and fewer reoperations. Eight-year-old children are expected to spend 72.6 vs 58.0 vs 37.8 months with aligned eyes and 26.8 vs 9.0 vs 1.5 months with stereopsis and to receive on average 1.76 vs 1.67 vs 1.46 surgeries, if they first underwent operation at 6 vs 24 vs 48 months, respectively. Children at 8 years are expected to have stereopsis at the rate of 36.1% vs 17.2% vs 5.1%, respectively. Operation at 6 months instead of 48 months is expected to yield an additional 9% of total follow-up time in health states with eye alignment and 11% in health states with stereopsis, per 5% increase in risk for extra operations (5% and 7%, respectively, per 5% increase in risk for operation at 6 vs 24 months). Wide-range sensitivity analyses and long-term projections point to the same direction. Given the benefit/risk tradeoffs, it is probably warranted to attempt to correct large-angle infantile esotropia as early as possible.