Predictive formula for calculating the probability of LASIK enhancement

Daniel J. Hu, Robert S Feder*, Surendra Basti, Bing B. Fung, Alfred W Rademaker, Pamela Stewart, Michael A Rosenberg

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

31 Scopus citations


Purpose: To develop a formula to predict a patient's need for laser in situ keratomileusis (LASIK) enhancement. Setting: Northwestern Laser Vision Center, Department of Ophthalmology, Northwestern University, Feinberg School of Medicine, Chicago, Illinois, USA. Methods: In this retrospective study, charts of patients who received LASIK with the Visx Star excimer laser for myopia and myopic astigmatism were reviewed. Laser in situ keratomileusis enhancement was performed in 130 of 720 eyes. Variables such as age, keratometry, spherical power, power and axis of astigmatism, and surgeon factor were compared in patients who required retreatment and those who did not. Multivariate logistic regression analysis was used to determine a formula for the probability of enhancement surgery. Results: Age (P<.0001), preoperative cycloplegic sphere (P<.0001), and surgeon (P<.0001) were the statistically significant factors for predicting retreatment. The predictive formula derived from these factors had a sensitivity of 79%, a specificity of 61%, and positive and negative predictive values of 31% and 93%, respectively. Conclusions: Older age, higher preoperative cycloplegic sphere, and surgeon significantly influenced a patient's likelihood for LASIK retreatment. A formula based on these predisposing factors helps to more accurately predict the need for retreatment.

Original languageEnglish (US)
Pages (from-to)363-368
Number of pages6
JournalJournal of cataract and refractive surgery
Issue number2
StatePublished - Feb 1 2004

ASJC Scopus subject areas

  • Surgery
  • Ophthalmology
  • Sensory Systems


Dive into the research topics of 'Predictive formula for calculating the probability of LASIK enhancement'. Together they form a unique fingerprint.

Cite this