Early characteristics of infants with pulmonary hypertension in a referral neonatal intensive care unit

Abstract

BackgroundApproximately 8–23% of premature infants develop pulmonary hypertension (PH), and this diagnosis confers a higher possibility of mortality. As a result, professional societies recommend PH screening in premature infants. However, the risk factors for and the outcomes of PH may differ depending on the timing of its diagnosis, and little evidence is available to determine at-risk infants in the referral neonatal population. The objective of this study was to define clinical and echocardiographic characteristics of infants with pulmonary hypertension during the neonatal hospital course and at or near-term.MethodsInfants who had the following billing codes: < 32 weeks, birth weight < 1500 g, neonatal unit, and echocardiograph had records abstracted from a data warehouse at Children’s Healthcare of Atlanta. The outcome was defined as late PH on the final echocardiogram for all patients, and, separately, for patients with multiple studies. Descriptive statistics, univariable, and multivariable models were evaluated, and odds ratios and 95% confidence intervals are expressed below as (OR, CI).Results556 infants were included in the overall study, 59 had PH on their final echocardiogram (11%). In multivariable analyses, atrial septal defect (2.9, 1.4–6.1), and intrauterine growth restriction (2.7, 1.2–6.3) increased the odds of late PH, whereas caffeine therapy decreased PH (0.4, 0.2–0.8). When the analyses were restricted to 32 infants who had multiple echocardiograms during their hospitalization, the association between atrial septal defect (5.9, 2.0–16.5) and growth restriction (3.7, 1.3–10.7) and late PH was strengthened, but the effect of caffeine therapy was no longer significant. In this smaller subgroup, infants with late PH had their final echocardiogram at a median of 116 days of life, and 42–74% of them had right ventricular pathology.ConclusionsEarly clinical variables are associated with PH persistence in a referral neonatal population. Identification of early clinical factors may help guide the ascertainment of infant risk for late PH, and may aid in targeting sub-groups that are most likely to benefit from PH screening.

Identification of gaps in the current knowledge on pulmonary hypertension in extremely preterm infants: A systematic review and meta‐analysis

“…We included two articles by scanning the reference lists. Finally, 25 articles, 16 retrospective and 9 prospective studies, were eligible for inclusion in the study . The flow chart of this excluding process is given in Figure .…”

“…Eight studies that were included provided clearly defined objectives, study populations, and outcome measures . However, 17 of the 25 included studies had a retrospective design and included a selected study population . Four studies reported on a study population of extremely preterm infants with and without BPD, whereas 16 studies reported selectively on infants with BPD and seven studies reported selectively on a study population of infants with moderate/severe BPD .…”

“…Figure shows all reported prevalences of infants with PH at the various ages that PH was diagnosed in the respective studies. In total 25 prevalences were reported in 21 eligible studies, 10 of these 25 prevelances were reported in prospective studies, whereas 15 of 25 prevalences were reported in retrospective studies . Eleven prevalences were derived from studies that reported a well‐defined time of PH‐assessment .…”

“…Eleven prevalences were derived from studies that reported a well‐defined time of PH‐assessment . Fourteen prevalences were derived from studies that did not report a well‐defined time point of assessment of PH …”

“…Six studies provided prevalences of PH in extremely preterm infants with BPD, and meta‐analyses provided an accumulative estimated prevalence of 20% (95% CI 14, 25) (Figure ) . Ten studies reported a prevalence of PH in extremely preterm infants both with and without BPD, giving an accumulative estimated prevalence of 12% (95% CI 9, 15) after meta‐analysis (Figure ) . Ten studies provided prevalences of PH for each BPD classification.…”