The increasing survival of preterm infants in the last few decades has been associated with cumulative numbers of medically fragile infants being released from NICUs [15]. The increased health utilization of these infants post initial hospitalization has long been recognized in many developed countries, with rehospitalization featuring prominently in the smallest of these infants, particularly in the first year or two of life [2–15]. In a recent study of extremely preterm infants 22 to 26 weeks GA born in 2013–2016 in 19 academic centers who were part of the National Institute of Child Health and Human Development Neonatal Research Network (NICHD NRN) in the United States, 49.9% were rehospitalized in the 2 years following discharge [15]. In an earlier NICHD NRN cohort of ELBW infants born in 2002 to 2005, 45% were rehospitalized by 18 to 22 months of age [3]. Strikingly similar rehospitalization rates have been described from other developed countries. In Northern Tyrol 40% of infants born at < 32 weeks were readmitted in the 1st year [4]. In Finland, 49.3% of preterm infants born at < 29 weeks were rehospitalized in the 1st year [5]. In Canada, 57% of infants born at 23–25 weeks were rehospitalized and 49% of infants born at 26–28 weeks were readmitted by 18 months of age [6]. In the EPIPAGE cohort in 9 regions in France, the rehospitalization rate was 47.3% in infants born at < 29 weeks gestational age [9]. However, most reports on rehospitalizations do not provide information on whether these were planned or unplanned admissions. Schell et al [14] showed that up to 30% of admissions are elective, most often for elective surgery such as hernia repair, which means that 70% of readmissions are unplanned.
Many studies have shown that co-morbidities associated with prematurity are common risk factors for readmission [5, 11, 12, 13, 14, 20].In a policy statement in 2008, the American Academy of Pediatrics Committee on Fetus and Newborn recommended that “ To ensure continuity of care after discharge, infants with unresolved medical issues that persist after their hospital stay, such as bronchopulmonary dysplasia or feeding dysfunction, should be co-managed by a neonatologist or other medical subspecialist from the hospital at which most of the care was provided.” [16] Brockli et al, in a survey of academic and private hospital settings, found that many hospitals with level III NICUs have established follow up clinics with a mission of providing a continuum of specialized medical management for graduates of the NICU and to identify deviations of growth, behavior and neurodevelopmental status [17]. Early identification of medical issues and preventative care could reduce high readmission rates. However, this has not been systematically studied.
A retrospective study of cohort of 898 preterm infants discharged from 5 NICUs in California and followed up at 32 outpatient facilities showed that medical and sociodemographic factors explained the largest amount of variation in risk-adjusted readmission rates up to 12 months following discharge, but the outpatient facility where patients received outpatient care explained a significant proportion of the variation in unplanned admissions [21]. It is unclear if these outpatient facilities were primary pediatrician offices or comprehensive neonatal follow up clinics. Outpatient pulmonary follow up has been shown to decrease the rate of readmissions in infants with bronchopulmonary dysplasia [22]. A Transition Home program in Massachusetts has been shown to reduce rehospitalization in preterm infants in the first 90 days after discharge by the third year of establishing the program [ 23]. To our knowledge, there is only one randomized controlled trial of comprehensive follow up care versus routine care in VLBW infants born in a Texas County Hospital between 1988 and 1996, which showed that infants who received comprehensive care had fewer life-threatening illnesses, fewer intensive care admissions and fewer intensive care days compared to infants who received routine care [19]. In our observational cohort study, the overall readmission rate (of unplanned admissions) for our VLBW population in the first year after discharge was 22% (25/126). However, the readmission rate for infants enrolled in the NFC was less than half that of infants who did not attend the clinic (17% versus 41%).
Most reports on readmissions do not include ER and Urgent care visits, which add to the cost of health care utilization. Patients who attended the NFC in our study had a significantly lower rate of ER/Urgent care visits (21% versus 82%). This supports our hypothesis and demonstrates that comprehensive care involving a neonatologist could decrease health care utilization. The average number of clinic visits per patient was 4 ± 2. We were unable to obtain information about the number of visits to the primary care pediatrician in the 1st year of life which may have had some influence on readmission / ER and urgent care visits.
Racial disparities surrounding preterm birth, infant mortality, and birth outcomes have been clearly established [24]. In our study, there was no difference in racial composition of patients who attended or did not attend the clinic (about 60% of patients were born to African American mothers) and there was no difference in the rehospitalization rates by race. Race was not a predominant variable in the risk of rehospitalization in a cohort of ELBW infants born between 1998–2000 and discharged from NICHD NRN nurseries, but medical morbidities and low family income played predominant roles [13].
In a French regional network with a high level of health care, multivariable analysis showed that living in the most deprived neighborhoods was associated with overall rehospitalization and multiple rehospitalizations in infants born at < 32 weeks GA [25]. Medicaid insurance may be a proxy for low income and we did find that Medicaid insurance was a significant risk factor for readmission; this has been described by others as well [10, 23, 26]. Some of our patients were on Medicaid through managed care organizations which were participating providers and could be seen in the NFC. We found that Medicaid patients seen in the clinic had lower readmission rates than Medicaid patients not followed up in the clinic.
Many other social determinants play a role in readmission rates. In our study, we found that smokers in the home and those infants who were exclusively formula fed after discharge had more readmissions. Consistent with the known literature, the VLBW cohort in both clinic and non-clinic groups had readmissions and ER visits most commonly due to respiratory symptoms during the winter.
This single center study does have several limitations. Some degree of recall bias may have confounded the results. The small sample size was due to the three-year study period chosen to mitigate recall bias. Very few parents filled out the survey sent by email which then resulted in multiple call attempts in order to conduct the phone survey. Despite these limitations, there were only ten patients who were lost to follow up because they could not be contacted, either because of inaccurate contact information or screened phone calls. Maternal literacy rates or educational levels which are also important factors in healthcare reutilization, was not studied.