Retinopathy of Prematurity: Identifying Geographical Regions of Poor Follow-Up For a Tertiary Care Center

Abstract

Purpose : Poor outcomes in infants diagnosed with retinopathy of prematurity (ROP) have been partly attributed to poor follow-up rates in lower middle income countries like Pakistan. We performed a retrospective long-term analysis at a tertiary care center in Lahore, Pakistan to learn about the association between ROP follow-up and distance from the facility, as well as identify regions of poor follow-up and regions where mothers of premature patients present to LGH.

Methods : Infants born at Lahore General Hospital (LGH) between March 2021 and September 2022 who met screening criteria (gestational age ≤ 35 weeks and birth weight < 2kg) were included. Primary outcomes were complete follow-up (CF) or incomplete follow-up (IF). Addresses were collected from each family and mapped to OpenStreetMaps/GoogleMaps. Python.juptr was used to calculate theoretical linear distance on a map (as the crow flies) and kilometers traveled.

Results : Out of 851 patients who met inclusion criteria, 317 had mappable addresses (37.3%). Patients were presenting to LGH from specific neighborhoods. Areas of good or poor follow-up were identified (Figure 1). Low follow-up rates were identified in specific regions (Garhi Shahu, Old City Lahore, and other areas to the West/Northwest of LGH). The median kilometers traveled was 8.7 km (IQR: 4.0-23.9) and 7.8 km (IQR: 3.5-31.0) for complete follow-up and incomplete follow-up, respectively. Similar values were measured for linear map distance, as shown in Table 1. However, there was no statistical significance between follow-up rates as a function of theoretical linear distance (p=0.791) on a map or kilometers traveled (p = 0.858).

Conclusions : Neighborhoods of poor follow-up were identified and strategies can be devised to increase follow-up. Surprisingly, follow-up rate was not significantly affected by distance traveled. Additionally, neighborhoods from which premature patients present to LGH were identified and can be helpful to devise strategies to decrease prematurity. It would be helpful for other NICUs to identify their geographical at-risk premature populations. Future research should explore alternative reasons for poor follow-up, such as travel modalities and travel time.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.