Validity of first post-operative day automated refraction following dense cataract extraction

Background: The aim of this study was to assess the refractive stability and validity of automated refraction obtained at day one after surgery for dense cataracts in a developing country setting, where patient follow-up is often erratic. Methods: In a prospective study in rural Kenya, we compared automated refraction and visual acuity (VA) after manual small incision cataract surgery (SICS) and phacoemulsification cataract surgery (Phaco) at day one with autorefraction and subjective refraction obtained at a later visit. Results: Ninety eyes of 83 patients fulfilled the ultimate inclusion criteria (21 SICS eyes and 69 Phaco eyes). The correlation between refraction at day one and at a later follow-up was moderately good for both SICS and Phaco (R1 = 0.61 and R2 = 0.58). Spherical equivalent (SE) values measured at day one were within one dioptre of those obtained later in 68 eyes (72 per cent); (SICS: 67 per cent; Phaco: 74 per cent) and within two dioptres in 83 eyes (92 per cent); (SICS: 86 per cent; Phaco: 94 per cent). At post-operative days 14, 28 and more than 42 for SICS and Phaco eyes, the mean SE was 0.58 and 0.48, 0.55 and 0.19, and 0.9 and 0.31 more hypermetropic than the mean SE of post-operative day one, respectively. First-day accuracy was higher in eyes with better pre-operative VA, softer cataracts and lower post-operative prediction error. The surgical technique (SICS versus Phaco) had no influence on first-day accuracy in our set-up. Conclusion: Refractive data from the early post-operative period can provide useful data for qualitative control and monitoring of post-operative refractive outcomes in a setup, where 'better' data are often not available; however, its accuracy is limited and, ultimately, there is no alternative to good follow up of operated patients.