Methods
Data was obtained by retrospective chart review that included patients recruited from the Children’s Hospital of Philadelphia (CHOP) and Johns Hopkins Children’s Center outpatient BPD clinics, whose initial hospital discharge was between 2008-2023. All patients were born at or less than 32 weeks-gestation and severity of BPD was based on the 2001 NIH consensus definition.(16) Participants with cyanotic congenital heart disease were excluded. Approval was obtained from the Children’s Hospital of Philadelphia Institutional Review Board (IRB # 20-017614_AM18) and Johns Hopkins IRB (Protocol #NA_00051884). All caregivers provided informed consent. Birthweight percentile was corrected by gestational age.(17) Median household income was estimated through residential zip codes using U.S. Census data.
Chi square tests, t-tests, and ANOVA tests compared clinical characteristics and demographic data between participants on outpatient diuretics and those not on outpatient diuretics as well as between participants on one, two or three diuretic agents and between diuretic classes. Cox regressions were performed to identify characteristics associated with slower or faster weaning of diuretic therapy with shared frailty by study site to account for study site-related differences. Results with a P value less than or equal to 0.05 were considered statistically significant. Stata 15.0 was used for data analysis (StataCorp, College Station TX).
Results
Patient Characteristics: Children with reported use of diuretics at their first pulmonary outpatient visit (n= 737) were more likely to have severe BPD (p<0.001), lower birthweight (p=0.001), and born at an earlier gestational age (p=0.012) compared to those not prescribed diuretics (n=487) (Table 1).
Subjects on outpatient diuretics were more likely to be older at the time of initial hospital discharge (p<0.001), however their first pulmonary appointment occurred at a younger age (p<0.001) compared to those not prescribed diuretic therapy at discharge. Further, being on diuretic(s) at the first outpatient pulmonary appointment was associated with an increased likelihood of home supplemental oxygen requirement (p<0.001), tracheostomy (p=0.018), ventilator dependence (p=0.019), gastrostomy tube (p< 0.001), Nissen fundoplication (p< 0.001), pulmonary hypertension (p< 0.001) and being prescribed inhaled corticosteroids (p< 0.001). No association was found between diuretic use at the first pulmonary outpatient visit and median household income, public insurance, race, or sex.
Diuretic use: In children prescribed diuretic therapy, most were either on one (50.1%) or two (47.4%) diuretics (Table 2). For patients receiving one diuretic, the majority were prescribed a thiazide diuretic (92.6%), whereas significantly fewer individuals were prescribed a loop diuretic alone (6.6%). Of those children with BPD receiving dual diuretic therapy, most were prescribed a thiazide and potassium sparing agent, (94.5%)). Site differences were seen in the use of diuretic therapies (data not shown), however from both sites, thiazides were the predominant diuretic used in the outpatient setting (96.3% and 96.0% respectively).
We also examined differences in patient characteristics between children prescribed a potassium sparing diuretic (n=353) versus those not on a potassium sparing diuretic (n=384). We found that children on a potassium sparing diuretic were more likely to be seen in pulmonary clinic at an older age (p=0.010), and more likely to be prescribed an inhaled corticosteroid (p<0.001) and to have a Nissen fundoplication (p=0.014) (Supplemental Table 1). Interestingly, being prescribed a potassium sparing diuretic was not associated with requiring less electrolyte supplementation compared to those prescribed another class of diuretic.
In contrast, children in the outpatient setting on loop diuretics (n=62) were more likely to require electrolyte supplementation (p<0.001) versus those not on loop diuretics (n=675). (Supplemental Table 2). Children on a loop diuretic at their first outpatient pulmonary visit were more likely to have a later age of initial hospital discharge (p<0.001), to have a tracheostomy (p<0.001), to be ventilator dependent (p<0.001), to require home supplemental oxygen (p=0.003), to have a gastrostomy tube (p<0.001) and Nissen fundoplication (p<0.001), and to have pulmonary hypertension (p<0.001); characteristics aligning with more severe respiratory disease.
Only a minority of children were prescribed three diuretics at the time of their first pulmonary visit (n=19), and when compared to those on one or two diuretics they were more likely to be older at the time of their initial hospital discharge (p<0.001), to have a tracheostomy (p<0.001), to be ventilator dependent (p<0.001), to be on home supplemental oxygen (p=0.014), to have pulmonary hypertension (p<0.001), and to have a gastrostomy tube (p<0.001) and a Nissen fundoplication (p<0.001) (Table 3). Additionally, almost all subjects prescribed three diuretics received electrolyte supplementation.
Diuretic weaning: In a multivariate model, the likelihood of weaning off diuretics in a given month was slower in patients with a tracheostomy tube, home supplemental oxygen requirement, pulmonary hypertension, gastrostomy tube, prescribed inhaled corticosteroids, on public insurance, or born at an earlier gestational age (Table 4).
Interestingly, BPD severity did not influence the likelihood of weaning off diuretics in a given month whereas, children who were prescribed electrolyte supplementation had an increased likelihood of weaning off diuretics faster in a given month. Weaning of outpatient diuretics was guided by healthcare providers in over 80% of patients (Table 3). Over 50% of BPD infants prescribed outpatient diuretics were weaned off by 10 months chronological age with over 75% weaned off diuretic therapy by 15 months chronologic age (Figure 1).
Diuretics and supplemental oxygen: For children with BPD on home supplemental oxygen (n=519), the median age of oxygen weaning was 10.1, 10.2, 14.3 and 33.5 months respectively for children on zero, one, two and three diuretics. There were no statistical differences in time to wean off supplemental oxygen between those on one diuretic compared to those on no diuretics (Figure 2). Children with BPD on two diuretics and home supplemental oxygen took longer to liberate from home supplemental oxygen compared to those not prescribed a diuretic (p < 0.001).