The phacoemulsification learning curve: per-operative complications in the first 3000 cases of an experienced surgeon.

PURPOSE: To assess the per-operative complications occurring during the first 3000 phacoemulsification cases performed by an experienced consultant surgeon.
METHODS: A prospective analysis of 3000 consecutive cases performed without supervision between November 1992 and November 1998 was carried out. Data recorded for each case included details of per-operative complications, pre-operative best corrected visual acuity, nuclear density, history of previous pars plana vitrectomy, and whether phacoemulsification was performed as part of a phacotrabeculectomy procedure.
RESULTS: The overall rate of vitreous loss was 1.3%. Nuclear fragments were lost to the vitreous in 6 cases (0.2%). The initial rate of vitreous loss was 4.0% in the first 300 cases falling to 0.7% in the last 300 cases. Capsulorhexis failure was the commonest per-operative complication observed, but the risk of subsequent posterior capsule rupture fell significantly from 9 of 45 (20.0%) in the first 100 cases to 1 of 49 (2.0%) in the next 2000 cases (p = 0.0061, Fisher's exact test). There was a significant increase in risk with denser cataracts, especially for capsulorhexis failure, rising to over 35% in the densest cases. The increases in posterior capsule rupture and vitreous loss were less dramatic but nonetheless very significant. There was no significant increase in the risk of per-operative complications with phacotrabeculectomy, and no increased risk in patients who had previously undergone pars plana vitrectomy. Posterior capsule rupture occurred in 22 of 612 (3.6%) local anaesthesia cases compared with 31 of 2269 (1.4%) topical anaesthesia cases. Per-operative best corrected visual acuity of 6/9 or better was recorded in 2.0% of the first 1000 cases compared with 13.9% of the last 1000 cases.
CONCLUSIONS: (1) Per-operative surgical risks could be reduced to low levels during the learning curve, but complications continued to occur at a low frequency. (2) The risk of per-operative complications was not significantly elevated in previously vitrectomised eyes. (3) Nuclear density correlated significantly with per-operative complication risk. (4) The visual threshold for cataract surgery fell dramatically with increasing experience of phacoemulsification. (5) Topical anaesthesia was not associated with an increased risk of per-operative complications.