Results
A total of 131 VLBW premature infants consisting of 71 (54%) female and 60 (46%) male were included in the study. Of these 131 infants, 70 of the infants (53.4%) haven’t any sign of ROP, while 61 of the infants (46.6%) had developed ROP and 19 of them (14.5% of overall infants) had severe (type-1) ROP. The mean gestational age (GA) was 28.8±2.1 weeks (min.-max. 23-32 weeks), the mean birth weight (BW) was 1208±208 g (min.-max.730-1520 g) and the median weight gain at 6th week of life was 620 g (25-75 percentile 430-840 g) in all premature infants.
The mean GA was significantly lower in patients with any stage of ROP and Type-1 ROP compared to No-ROP group (p < 0.01).Furthermore, the mean BW and the mean ROPScore values were significantly different among the groups (p < 0.01 and p < 0.01, respectively). The comparison of the mean GA, BW and ROP score among groups were demonstrated in Table 1.
In ROC analysis, the mean AUC for ROPScore had more predictive value than BW and GA in terms of detecting type-1 ROP (p < 0.001) . Figure 2 presents the analysis of AUCs for the ROPscore, GA and BW for any stage or ROP and type 1 ROP.
In the algorithm of CUT-OFF values, the sensitivity and specificity of the scoring system for predicting any stage of ROP was 88.5% (95% CI, 79-94) and 37.7% (95% CI, 26.6-50.2), respectively. However, the scoring system had 100% (95% CI, 82.3-100) sensitivity and 83% (95% CI, 75-88.8) specificity in terms of predicting severe type -1 ROP. Table 2 highlights the cutoff points, the positive and negative predictive values (PPVs and NPVs) of the model for patients with any stage and type-1 ROP. Our customized cut-off values (12.3 for any stage ROP, 14.9 for type-1 ROP) allowed a drop in numbers of false positives from 54 to 23 for any stage ROP and from 33 to 19 for type-1 ROP. ROPScore model in conjunction with 2018 guidelines detected 62 of 70 infants with any stage ROP, more importantly, correctly identified all of infants with type-1 ROP. ROPscore missed only 8 out of 131 (%6) babies in which none of them needed the treatment for ROP.