Table 4.
Result of cox proportional hazard analysis for assessing risk factors of surgical failure (total of 70 eyes).
| Variables | Univariate | Multivariate | ||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| Preoperative PRP (versus not done) | 0.38 (0.14–1.02) | 0.056 | 0.40 (0.15–1.09) | 0.072 |
| Presence of NVA (versus absence) | 0.91 (0.20–4.20) | 0.907 | ||
| PAS more than 50% (versus less than 50%) | 2.33 (0.96–5.62) | 0.061 | 2.25 (0.92–5.47) | 0.075 |
| IVB group (versus control) | 0.47 (0.20–1.11) | 0.085 | ||
| Number of IVB injection (per 1-time injection) | 0.60 (0.35–1.03) | 0.062 | ||
| Baseline IOP (per 1 mmHg) | 1.03 (0.99–1.08) | 0.144 | ||
| Baseline BCVA (per 1 LogMAR) | 1.23 (0.78–1.93) | 0.368 | ||
HR = hazard ratio; CI = confidence interval; PRP = panretinal photocoagulation; NVA = neovascularization of angle; PAS = peripheral anterior synechiae; IVB = intravitreal bevacizumab; IOP = intraocular pressure; BCVA = best-corrected visual acuity; MAR = minimal angle of resolution. Variables with P < 0.10 in univariate analysis (PRP, PAS extent, IVB, and number of IVB injection) were entered into multivariate analysis. Multivariate model using a backward elimination approach based on likelihood ratio; variables were entered in the model if P < 0.05 and removed if P > 0.10 in the saturated multivariate model.