Design and setting
A retrospective case control study in a tertiary level NICU in Australia.
Ethics approval
This study was approved by the hospital Governance Committee (GEKO7622) and the institutional research screening committee.
Participants : All preterm infants (< 27 weeks gestational age at birth) admitted between January 2010 and December 2015, were enrolled for analysis. Cases were defined as preterm infants who underwent treatment for ROP. Controls defined as preterm infants who had ROP not needing treatment or no ROP. Controls were matched with cases on GA at birth and BW on a 1:4 ratio with equal distribution between no ROP and ROP not needing treatment. Infants who had congenital anomalies or critical cardiac conditions were excluded from the study.
Data collection
The following data was collected from each infant’s medical record and neonatal database using a standardized, pre-piloted data collection form.
Antenatal risk factors: Chorioamnionitis, preeclampsia, diabetes mellitus, oligohydramnios, and premature rupture of membrane, antenatal glucocorticoids
Delivery details: Place of birth, GA, BW, gender, growth status, mode of delivery and Apgar scores.
Post-delivery care: Time to regain BW, duration of mechanical ventilation and noninvasive ventilation, duration of oxygen therapy, treatment for patent ductus arteriosus, necrotizing enterocolitis, postnatal steroids, need for blood transfusion, culture positive sepsis, duration of hospital stay, presence of intraventricular hemorrhage (IVH) and periventricular leukomalacia (PVL) on head scans and death before discharge were collected.
Outcomes
The primary outcome was to determine whether the ‘time taken to regain BW’ can be a useful marker in identifying the infants needing treatment for ROP. The secondary outcomes were to assess the effect of chorioamnionitis, blood transfusion, antenatal steroids, duration of oxygen and culture positive sepsis as risk for ROP needing treatment.
Guidelines for management of ROP in the NICU:
Screening criteria
Neonates born < 31 weeks postmenstrual age, regardless of BW and neonates with BW < 1250 grams, regardless of postmenstrual age at birth were screened for ROP. Neonates born greater than 31 weeks gestation at birth, BW > 1250g, with additional problems, were screened at the discretion of the treating Neonatologist.
The screening examination for ROP followed the guidelines proposed by American Academy
Of Ophthalmology and Pediatrics and the Association for Pediatric Ophthalmology and Strabismus. The first screening examination was performed at 30 to 31week postmenstrual age in infants born at < 27 weeks GA and at 4-5weeks postnatal age in infants born at 27- to 32-week GA. After the first evaluation, if the subject did not have ROP, the neonate was evaluated at 2–4 week intervals until full vascularization. If the patient had active or rapid progressive lesion, the subject was evaluated more frequently, depending on the clinical findings.
Treatment for ROP in the unit was primarily laser surgery and the treatment decisions were based on the Early Treatment ROP guidelines. Laser was performed with advanced ROP particularly stage III with Plus disease, significant pre-Plus or Aggressive Posterior ROP (AP-ROP) after review by the Ophthalmologist and treating Neonatologist. Anti VEGF treatment was used for infants who had Zone 1 disease and in infants who were unstable to undergo laser surgery.
Statistical method
Continuous data were summarized with medians, interquartile ranges (IQR) and ranges (R), and categorical data as frequency distributions. Conditional logistic regression analysis was used to univariately compare characteristics between cases and controls, and multivariable modelling to examine the effect of time to regain BW on ROP needing treatment. An assessment of potential confounders including chorioamnionitis, duration of oxygen, blood transfusion, antenatal steroids and culture positive sepsis was made using a forward selection method to avoid over-parameterization. Univariate and adjusted effects were summarized as odds ratios (OR) and adjusted OR (aOR) respectively, with 95% confidence intervals (CI). The predictive potential of time to regain birthweight on ROP needing treatment was further assessed using area under the curve (AUC) constructed from probabilities generated from the conditional logistic regression models. Time to event data, including duration of continuous positive airway pressure, mechanical ventilation, nasal high flow and oxygen were summarized using Kaplan-Meier survival estimates and compared using the log rank test, with deaths before discharge censored for analysis. IBM SPSS (version 22.0, Armonk, NY) and Cytel Studio (version 8.0, 2007) statistical software were used for data analysis. p-values < 0.05 were considered statistically significant.