Demographics of patients with retinal re-detachment after primary PPV

The study included 344 consecutive eyes with RRD who underwent primary PPV. There were two cohorts of those consecutive cases: Group 1: the patients without retinal re-detachment after primary PPV (n = 308) and Group 2: the patients with retinal re-detachment after primary PPV (n = 36).

Table 1 summarizes the baseline demographic and clinical characteristics of all the patients including sex, age, diabetes status, with or without 360° ILR, combined operations, endotamponade, surgeons, lens status, macular status PVR grade, RD quadrants, axial length, and duration of follow-ups (shown in Table 1).

Table 1 Demographics and clinical characteristics of patients with retinal redetachment * p < 0.1, ** p < 0.05, *** p < 0.01.

Table 1 displays the demographics and clinical characteristics of patients with retinal re-detachment.  *p < 0.1, **p < 0.05, ***p < 0.01.

In Group 1, the mean age was 62.9 ± 12.9, the proportion of male patients was 202/308 (65.58%), the mean length of axial (mm) was 24.7 ± 1.7, the mean follow-up (months) was 10.9 ± 13.1 (Median: 5.3; Q1, Q3:1.6, 15.8). In Group 2, the mean age was 60.2 ± 11.1, the proportion of male patients was 18/36 (50%), the mean length of axial (mm) was 24.9 ± 1.6, the mean follow-up (months) was 21.2 ± 14.3 (Median: 14.7; Q1, Q3:11.8, 33.6). Median follow-up of total patients was 6.2 months (Q1, Q3:2.0, 17.2). There was no statistically significant difference in those variables between the two groups. There was also no statistically significant difference in lens status, PVR grade, and RD quadrants as well.

Preoperative parameters

In preoperative parameters, 33/308 patients (10.71%) had diabetes in Group 1, while 8/36 patients (25%) in Group 2. Moreover, the macula was detached in 106/308 eyes (34.41%) in Group 1, while in Group 2, the macula was detached in 18/36 eyes (50%). Non-parametric Kaplan-Maier showed the risk factors of retinal re-detachment might involve those two variables [diabetes (p = 0.025) and macular-off (p = 0.0019)]. The results of the hazard ratios Cox regression further confirmed that diabetes and macular off before the primary surgery were associated with higher rate of retinal re-detachment outcome (diabetes: OR = 2.301, 95% CI [1.130–4.687], p < 0.05; macular-off: OR = 2.243, 95% CI [1.212–4.149], p < 0.05).

Intraoperative parameters

In intraoperative parameters, there was no statistically significant difference in combined operations and endotamponade. In Group 1, 360° ILR was performed in 186/308 patients (60.39%), while in Group 2, 360° ILR was performed in 17/36 patients (47.22%). 17/203 eyes (8.4%) in the 360° ILR group had retinal re-detachment after primary PPV, which was significantly lower compared to the focal laser retinopexy group (19/141 eyes, 13.5%) (p < 0.05). Non-parametric Kaplan-Maier showed the risk factors of retinal re-detachment might also involve focal laser without 360° ILR (p = 0.0021). The result of the hazard ratios Cox regression further confirmed that 360° ILR was associated with a significant reduction in the odds of retinal re-detachment (OR = 0.456, 95%-CI [0.245–0.848], p < 0.05).

Survival probabilities of retinal re-detachment

From the previous result, we estimated observed survival probabilities for any combination of the risk factors of retinal re-detachment (shown in Fig. 1). The cumulative incidence of re-detachment at 6,12,24 and 60 months was shown in Table 2.

Fig. 1

Survival probabilities of retinal re-detachment

Table 2 Cumulative incidence of re-detachment at 6,12,24 and 60 months. Kaplan-Meier estimates and their 95% CI.

Table 2: Cumulative incidence of re-detachment at 6,12,24 and 60 months.The differences between these cumulative incidence curves are significant (Log Rank Test p = 0.0021). The multivariate Cox regression was applied to estimate conditional hazard rates. The result of the estimation is summarized in the following table (shown in Table 2).

Table 3 Multivariate Cox regression

Table 3 displays the multivariate Cox regression, which estimated the conditional hazard rates.

From the regression model we calculated the predicted survival probabilities by 12 months shown in Table 4. The riskier the combination, the larger and more red was the font (shown in Table 4).

Table 4 Survival probabilities of retinal re-detachment

Table 3 displays the survival probabilities of retinal re-detachment.

Incidence of epiretinal membrane (ERM) formation

In regard to ERM formation, there were 18/236 cases of ERM formation in the cohort at latest follow-up. 6/118 cases occurred in the 360° ILR group (5.1%) compared to 12/118 (10.2%) in the focal laser group. 360° ILR was associated with a non-significant decrease in the odds of ERM formation (p = 0.219).