To examine the relationship between changes in American Academy of Pediatrics (AAP) guidance and palivizumab use for infants admitted to the NICU. We hypothesized that each change in guidance would be associated with a change in palivizumab usage.
This is a retrospective repeated cross-sectional study of palivizumab usage in defined subgroups of infants discharged between 1999 and 2020 using the Pediatrix Clinical Data Warehouse.
Palivizumab utilization increased in all groups between 1999 and 2003 and remained stable until 2013. Large changes in palivizumab use occurred between 2013 and 2015 followed by slower changes from 2016 to 2020. The largest decrease was in infants born between 29 0/7 and 31 6/7 weeks’ gestational age without chronic lung disease (decreased from 87% to 21%; P < .001). The second largest absolute decrease was infants born at 32 0/7 to 34 6/7 weeks’ gestational age without chronic lung disease and no major anomalies (decreased from 52% to 6%; P < .001). The decrease in term infants with major congenital heart problem was smaller (25 to 17%; P < .001). Even in the most vulnerable infants born between 22 0/7 and 28 6/7 estimated gestational age, palivizumab use declined (88% in 2013 to 74% in 2020; P < .001).
Early AAP guidelines had minor impacts on palivizumab use in infants discharged from the hospital from the NICU. The 2014 guidelines resulted in major changes in palivizumab use and extended into populations for which the AAP guidance remained unchanged.
Outpatient respiratory syncytial virus immunoprophylaxis has decreased, and respiratory syncytial virus hospitalization rates for infants born at <29 weeks’ gestational age appears to be increasing.
Early American Academy of Pediatrics guidelines had minor impacts on palivizumab use in NICU infants. In contrast, the August 2014 guidelines resulted in major changes in palivizumab use in at-risk subgroups of NICU infants, including infants born at <29 weeks’ estimated gestational age.
Palivizumab was licensed in June 1998 by the Food and Drug Administration for children at risk for severe disease because of respiratory syncytial viral infection (RSV) on the basis of the results of a large randomized clinical trial.1 The American Academy of Pediatrics (AAP) published initial guidance for the use of palivizumab in 19982 and subsequently has updated guidance on 3 occasions (Table 1): December 2003, December 2009, and August 2014.3–5 The updates were based on data that provided a better understanding of infants and young children at greatest risk of hospitalization attributable to RSV infection and in general suggested a more restrictive approach to palivizumab prophylaxis.5
The most recent guidance from December 2014 suggested palivizumab prophylaxis be provided to infants born before 29 weeks’ gestation who were younger than 12 months at the start of the RSV season. For infants born at 29 weeks, 0 days’ gestation or later, palivizumab prophylaxis was only recommended if the infant had congenital heart disease (CHD), chronic lung disease (CLD), or another condition that increased their risk of severe RSV infection. Following the most recent AAP guidance, several studies have evaluated changes in RSV hospitalization rates for at-risk infant populations.6,7 Although studies have evaluated RSV hospitalization rates after guidance changes, no study has examined predischarge use of this medication in relevant subgroups of NICU infants, and only a few studies have looked at postdischarge use in extremely preterm infants.7 In August of 2021, Goldstein et al reported that in Medicaid populations, the rate of palivizumab use decreased in infants <29 weeks’ estimated gestational age (EGA) following the 2014 AAP guidelines, which corresponded to a significantly increased risk of RSV hospitalizations in this population.7
Our objective in this study was to examine the relationship between changes in AAP guidance and palivizumab utilization for infants discharged from the NICU. We hypothesized that each change in AAP guidance would be associated with a concomitant change in palivizumab utilization.
Methods
We performed a retrospective repeated cross-sectional study review of the Pediatrix Clinical Data Warehouse (CDW), a deidentified dataset that is generated from BabySteps, a standardized documentation and billing software tool used by participating NICUs in 33 states and Puerto Rico. It is estimated that the CDW captures information on ∼20% of infants admitted to NICUs in the United States. Clinical data on these infants are prospectively recorded during their hospitalization in the NICU. Admission, discharge, and daily progress notes are generated with a proprietary computer-assisted tool, and these data are stored in an electronic database. Clinicians providing care to patients interact with the patient’s data daily to generate progress notes and billing. Each day’s note is stored along with documentation on medications, procedures, and diagnoses. Diagnoses and medications in the CDW are entered by the physicians or nurse practitioners who are caring for the infant. They are chosen from a picklist and verified each day while the infant is in the NICU. The data warehouse does not specify how specific diagnoses were obtained or confirmed. As a deidentified dataset, the authors did not have direct access to the underlying study population. The Pediatrix CDW does not contain linkages to other existing databases. The local data are deidentified, made compliant with the Health Insurance Portability and Accountability Act of 1996 regulations and consolidated within the Pediatrix CDW, which is configured into tables that can be joined and queried for statistical analyses. The details of this data set have been described in detail elsewhere.8 This dataset is approved for queries by the Western Institutional Review Board (Olympia, WA) and the MEDNAX Research Advisory Committee (Sunrise, FL).
Infants were eligible for inclusion in our study if they were from a participating NICU between January 1, 1999 and December 31, 2020, survived to discharge, were born at a gestational age between 22 and 41 weeks inclusive, and were discharged during typical RSV season fourth quarter (October to December) and first quarter (January to March) of each year. We included all clinical sites that reported at least 20 infants given palivizumab to eliminate sites with low rates of at-risk patients and/or where treatment was provided during follow-up pediatric appointments. We excluded infants who died or were transferred before discharge because palivizumab is given just before discharge to home. We included both inborn and outborn infants and infants with anomalies.
For reporting purposes, consecutive fourth and first quarter were combined to represent 1 RSV season. For example, RSV season 1999 included infants discharged from October 1999 to March 2000.
Based on AAP guidance (Table 1), the infants were grouped by prophylaxis indication into the following categories to describe the changes more discretely in utilization of palivizumab:
Infants born between 22 0/7 and 28 6/7 weeks’ EGA.
Infants born between 29 0/7 and 31 6/7 weeks’ EGA and with no report of CLD.
Infants born between 29 0/7 and 31 6/7 weeks’ EGA who had CLD based on oxygen use at 36 weeks’ postmenstrual age.
Infants born between 32 0/7 and 34 6/7 weeks’ EGA with no report of CLD and no report of major anomalies.
Infants born between 37 0/7 and 41 6/7 weeks’ EGA with report of major CHD (types of congenital heart disease are reported in supplemental table) and no report of CLD.
As a low utilization reference group, we included infants born at 37 0/7 to 41 6/7 weeks’ EGA with no report of major anomalies or CLD.
Data on EGA represented the best estimates based on both obstetrical data and neonatal examination findings. Designation of race was based on the options contained in the database, which were White, Black, Hispanic, Asian, and other.
Chronic lung disease was defined as use of oxygen at 28 days for infants more than 32 weeks’ gestational age and 36 weeks’ postmenstrual age for infants born at or before 32 weeks’ gestational age.
The changes in the proportion of infants treated is shown in Fig 1. For statistical analysis, RSV seasons were combined into multiyear epochs that corresponded to the periods during which the AAP guidance was active. These epochs were 1999 to 2003, 2004 to 2009, 2010 to 2013, and 2014 to 2020. To evaluate time-related changes, we used the linear trend and Cochran-Armitage trend tests to look at changes that occurred between epoch and within each epoch. We used JMP 12 (SAS Institute, Cary, NC) for all statistical analyses to evaluate changes in each of these year groups.
Results
Between January 1, 1997, and December 31, 2020, the total number of infants included in the Pediatrix CDW was 1 612 438 infants. Of these, 1 351 638 were between 22 0/7 and 41 6/7 weeks’ EGA at birth and survived to be discharged from the hospital between January 1, 1999 and December 31, 2020. Infants born before 1999 were excluded because palivizumab was approved in 1998. Additionally, there were 681 915 infants excluded because they were discharged during quarters 2 (April to June) and 3 (July to September) each calendar year when the rate of palivizumab use was significantly lower (3.6% vs 18.9% of the total study eligible patients). There were 281 sites who treated more than 20 infants with palivizumab.
In total, 463 268 infants met our criteria for inclusion in each of our study groups: group 1, 22 0/7 to 28 6/7 weeks’ EGA (n = 39 273; 8.5%); group 2, 29 0/7 to 31 6/7 weeks’ EGA and no CLD (n = 46 783; 10.1%); group 3, 29 0/7 to 31 6/7 weeks’ EGA with CLD (n = 4373; 0.9%); group 4, 32 0/7 to 34 6/7 weeks’ EGA and without CLD or congenital anomalies (n = 141 837; 30.6%); group 5, 37 0/7 to 41 6/7 weeks’ EGA without CLD but with a report of a major congenital heart defect (n = 6468; 1.4%); and group 6, 37 0/7 to 41 6/7 weeks’ EGA without CLD and without any major anomaly (n = 224 534; 48.5%). The demographic characteristics of each group are in Table 2, and the changes in demographics across time are in Table 3.
Palivizumab use in group 6 (term infants without chronic lung disease or a major anomaly, the expected low-usage group) was less than 1.3% during 1999 to 2020, and since 2014, the use of palivizumab in this group has been less than 0.1% (Fig 1). There were no clinically meaningful year-to-year trends for term infants without chronic lung disease or a major anomaly in any of the year groups.
During the years between 1999 and 2003, there were clinically important and statistically significant increases in palivizumab use in all premature infant groups and term infants diagnosed with a major congenital heart anomaly (P < .001; Cochran Armitage Trend Test; Fig 1).
During the years between 2004 and 2009, there was a clinically meaningful change in use of palivizumab in only 1 study group. In group 2 (29 0/7 to 31 6/7 weeks’ EGA and no CLD) infants, the usage rate increased from 82% in 2005 to 87% in 2009 (P < .001; Cochran Armitage Trend Test; Fig 1).
During the years between 2010 and 2013, there were no clinically meaningful changes in use of palivizumab in any of the study groups (P > .01; Cochran Armitage Trend Test; Fig 1). In infant groups 1 and 3 (infants born between 22 0/7 and 28 6/7 weeks’ EGA and infants born between 29 0/7 and 31 6/7 weeks’ EGA who had CLD), palivizumab usage slightly increased, but the increase was not statistically significant in either group.
There were large changes in palivizumab utilization between 2013 and 2015 followed by slower changes from 2016 to 2020. The largest decrease was in group 2 (29 0/7–31 6/7 weeks’ EGA who did not have CLD), in which palivizumab use dropped from 87% to 21%. The second largest decrease was in group 4 (32 0/7–34 6/7 weeks’ EGA with no CLD and no major anomalies), in which palivizumab use dropped from 52% to 6%. The third largest decrease was in group 3 (29 0/7–31 6/7 weeks’ EGA with CLD), in which palivizumab use dropped from 84% to 58%. The decrease in group 5 (term infants with a report of a major congenital heart problem) was slower and smaller (25% to 17%; Fig 1). All these changes were clinically meaningful and significant (P < .001; Cochran Armitage Trend Test).
During the years between 2016 and 2020, there was a decrease in palivizumab use for all premature infant groups and term infants diagnosed with a major congenital heart anomaly (P < .001; Cochran Armitage Trend Test; Fig 1), but the decreases were smaller and occurred over a longer period.
Over each year epoch, there were significant changes in palivizumab use in all prophylaxis indication groups (Table 3). There were small but significant increases in the proportion of infants who were outborn, delivered by cesarean delivery, and in whom White race was reported (Table 3). The report of “other” race increased. The proportion of infants on mechanical ventilation on days 0, 1, or 2 and the median maximum oxygen support during days 0, 1, or 2 both decreased. Gender and Hispanic, Black, and Asian race did not change.
Discussion
Following the initial 1998 guidance,2 we expected to see an increase in palivizumab in all study groups reflecting the AAP recommendations, and between 1999 and 2003, there was an increase in palivizumab use for all infant groups recommended to receive palivizumab. In 2003, Feltes et al9 published results from a randomized controlled trial of 1287 children with CHD that showed palivizumab prophylaxis was associated with a reduction in RSV-related hospitalizations and morbidity. In that same year, AAP guidance was updated for the first time.3 There were 2 principal changes in guidance. One was to limit palivizumab use in infants and young children with CHD to those with “hemodynamically significant” compromise. The second change was to limit the use in infants born at 32 to 35 weeks’ gestation to those infants who had “2 or more risk factors” for development of RSV. During the years after the first update, 2004 to 2009, the only significant change was an increase in palivizumab use among infants born between 29 0/7 and 31 6/7 weeks’ EGA who did not have CLD. Palivizumab use in term infants with CHD and infants born between 32 0/7 and 34 6/7 weeks did not change.
AAP guidance was updated for a second time in 2009, and guidelines regarding CHD were further clarified, and the recommendation based on gestational age was narrowed.4 In the 2003 guidance, 32 to 35 weeks’ gestation was “defined as 32 weeks 1 day through 35 weeks 0 days”. In 2009, the new definition became 32 weeks 0 days through 34 weeks 6 days (Table 1). From 2010 to 2014, palivizumab use did not change significantly for any subgroup of infants.
In August of 2014, the AAP discontinued their recommendation for use of palivizumab in infants born between 32 0/7 and 34 6/7 weeks’ EGA with any RSV transmission risk factors.5 The 2014 update also further restricted the recommendation in infants born between 29 0/7 and 31 6/7 weeks to only include infants with CLD. The 2014 AAP statement was published in August 2014, before the fourth quarter of 2014 and the first quarter of 2015. The effect of 2014 AAP change in guidance was rapid and dramatic, with palivizumab use in the 32 0/7-to-34 6/7 EGA group dropping from 52% in 2013 to 6% in 2015. Palivizumab use in infants of 29 0/7 to 31 6/7 weeks’ EGA without CLD also dropped precipitously from 87% in 2013 to 21% in 2015. Surprisingly, this downward trend in palivizumab use extended into populations for which the AAP guidance remained unchanged. In infants of 29 0/7 to 31 6/7 weeks’ EGA with CLD, palivizumab use declined from 84% in 2013 to 58% in 2015. Even in the most vulnerable infants born between 22 0/7 and 28 6/7 weeks’ EGA, palivizumab use also declined from 88% in 2013 to 74% in 2020. The reason for the declined palivizumab use in these vulnerable populations is unclear. The AAP reaffirmed its 2014 policy in 2019.
Our study is unique, as it shows the impact of the AAP guidelines in relevant subgroups of infants cared for in the NICU and represents a large sample of at-risk infants in a contemporary data set. Our data present more precise timing of changes in palivizumab use and support findings of other studies showing significant declines in palivizumab use following the AAP 2014 recommendations.7,10 Decreases in palivizumab use have been associated with increased rates of RSV hospitalization, disease severity, and hospital costs in infants at risk for RSV infections.7,10
The primary limitation is that our data are from an administrative dataset reliant on user-entered data. We do not have data on hemodynamic severity for infants with congenital heart disease. Our findings may be subject to underreporting errors or data entry errors, although we had internal data consistency checks, and trends in medications use have been validated against external data sets.11 The AAP did not clearly define CLD until 2014, when the authors wrote, “CLD of prematurity was defined as gestational age <32 weeks, 0 days and a requirement for >21% oxygen for at least the first 28 days after birth”. We have chosen to use a definition that is more restrictive and only included infants still on oxygen at 36 weeks’ postmenstrual age. Although these differences might influence absolute rates, we do not expect that it should influence trends in usage. Our data set does not include postdischarge data, precluding analysis of subsequent palivizumab use or rehospitalization rates. Although other reports7,10 show that decreases in palivizumab usage is associated with an increase in RSV hospitalization rates and hospital costs, those reports and our data cannot be used to calculate the number needed to treat to prevent the observed increase and/or to provide a pharmacoeconomic analysis of the changes in practice. Our data set does not contain information about insurance, limiting our ability to examine how insurance status may act as a cofactor.
RSV infections began to increase in the spring of 2021, and RSV remains an important cause of mortality and morbidity in infants.12,13 The recent change in interseasonal increase is different from the typical RSV epidemiology. It is hypothesized that the relaxation of nonpharmacologic interventions that were previously implemented to prevent the spread of SARS-CoV-2 may have caused the increase in RSV emergency department visits and hospitalizations of infants and children. It is also possible that the decreased pharmacological prophylaxis may have been associated with the increased in RSV. However, we cannot eliminate the possibility that SARS-CoV-2 infections may increase the vulnerability of infants to RSV infections, especially among high-risk infants. Decreases in immunoprophylaxis may only make this problem worse. Research is urgently needed to determine if RSV prophylaxis in preterm infants with a single dose of nirsevimab could be a more effective way to protect high-risk infants.14 Our data provide a detailed picture of palivizumab use in a contemporary data set. We believe our data will help inform ongoing efforts to contain RSV.
Dr Ahmad, Mr Clark, and Dr Tolia conceptualized and designed the study, drafted the initial manuscript, reviewed and revised the manuscript, and supervised data review and data analysis; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no conflicts of interest relevant to this article to disclose.
