Prediction error in iris suture fixated intraocular lenses and long-term stability

Importance: Little evidence exists for prediction error in iris-sutured intraocular lenses. Background: To determine the magnitude of prediction error in iris-sutured intraocular lenses, associated factors and their long-term stability. Design: Retrospective, nonrandomized, noncomparative case series conducted at the Wilmer Eye Institute (Baltimore, Maryland, United States). Participants: Adult patients with subluxated intraocular lenses that underwent iris-suture fixation between January 2000 and December 2014 by a single surgeon. Pregnant women, children (below the age of 18) and cases with follow-up under 1 month were excluded. Methods: Prediction error was calculated in 60 eyes and survival analysis was performed on 99 eyes. Main Outcome Measures: Prediction error (the difference between the postoperative manifest refraction in spherical equivalent and the spherical equivalent predicted by the Barrett Universal II, Holladay 2 and SRK/T formulas), preoperative and postoperative distance-corrected visual acuity, manifest refraction, frequency of postoperative complications and time until re-subluxation. Results: Mean prediction error using the Barrett formula was −0.35 ± 1.0 D. Higher axial length (≥25.5 mm) was associated with greater prediction error (−0.72 ± 1.11 D vs −0.18 ± 0.91 D, P =.048). Twelve re-subluxations occurred over a mean follow-up period of 30.28 ± 41.86 months. The predicted 50% survival of iris-sutured lenses was 114.25 months. Conclusions and Relevance: Iris-suture fixation may require moderate lens power adjustment to compensate for prediction error, especially in eyes with higher axial length. Longer follow-up demonstrates that iris-suture fixation remains a viable technique, yet re-subluxations require routine monitoring of such eyes.