A new dose-response curve for bilateral medial rectus recessions for infantile esotropia

Purpose: In 1982, Mims et al(1) generated the first rigorous dose-response curve for bilateral medial rectus recessions for infantile esotropia (ET). Curve fitting calculations were hampered by substantial variability of surgical effect for the larger angles of ET, and an exponential curve was chosen for angles below 30 ET and a straight line for angles above 30 ET. In a continuing effort to improve the surgical success rate, a new series of results of 113 bilateral medial rectus recessions performed from 1990 to 1998 for infantile ET has been analyzed to produce a new dose-response curve and to study other potentially useful variables. Methods: The amount of medial rectus recession performed in this series was the original dose-response curve from 1982, with conjunctival recessions performed for angles above 35 ET. Results: The best curve that could be fitted to the effect versus millimeters of bilateral medial rectus recession was a single exponential curve. Other parameters explored with multivariate analysis, including head circumference, developmental delay, and intraoperative angles under anesthesia were not helpful in predicting effect of surgery at 6 weeks postoperative. A total of 87% were aligned at 6 months after surgery. Surgical failure at 6 months was more prevalent with larger preoperative angles (P = .0007) and with developmental delay (age when child first sat alone; P = .0078). Conclusion: Between 1982 and 1990, decreased variability of results of bilateral medial rectus recessions for larger angles of infantile ET enabled the generation of a single exponential dose-response curve for the entire range of angles of infantile ET.