To determine the frequency and risk factors for clinic and pharmacy use in preterm infants during the first year after neonatal intensive care unit (NICU) discharge.
We analyzed clinic visits and prescriptions in a cohort of 23 to 32 weeks infants. We constructed multivariable regression models to determine risk factors for high use.
The 892 preterm infants experienced 18 346 pediatric visits (mean 20 visits per infant per year) and filled 2100 prescriptions (mean 5.5 prescriptions per year among infants taking medications). Most visits were non-well child care visits to pediatric primary care providers. Prematurity and bronchopulmonary dysplasia (BPD) are important risk factors: infants at 23 to 26 weeks gestation or infants with BPD had an average 29 visits per year and 9 prescriptions per year among infants taking medication. However, half of the highest using infants were relatively healthy infants at 27 to 32 weeks gestation who escaped BPD, NEC or grade 3 to 4 intraventricular hemorrhage.
Premature infants had frequent pediatric visits and prescription medications. Extreme prematurity and neonatal morbidities are important risk factors; however, half of the highest using infants are moderately preterm without neonatal morbidities.
In 2003, 79 606 infants were born at 32 weeks gestational age (GA) in the United States, including 59 217 very-low-birth-weight infants.1 Advances in neonatology and improved survival means that more preterm infants will be discharged into the care of pediatric primary care providers.2, 3, 4, 5 Many of these infants have experienced significant neonatal morbidities including bronchopulmonary dysplasia (BPD), sepsis, intraventricular hemorrhage (IVH), necrotizing enterocolitis (NEC) and retinopathy of prematurity. These conditions increase the risk for mild-to-severe adverse outcomes after neonatal intensive care unit (NICU) discharge including rehospitalization for respiratory infections or reactive airway disease, poor growth, neurosensory problems, cerebral palsy and behavioral disorders.6, 7, 8, 9, 10, 11 An infant's transition from the NICU to the home is poorly understood, and yet it quite likely represents a critical step in infant growth and development.
The complexity of health-care usage during infancy is poorly understood. The American Academy of Pediatrics recommends a minimum of six well child care (WCC) visits in the first year,12 with additional visits for illnesses. Term newborns without morbidities can expect to have, on average, 12 visits in the first year.13, 14, 15, 16 Descriptive studies often do not capture individual patient level follow-up and have not reported specifically on infants. The 2002 Medical Expenditure Panel Survey study reported four to five health-care visits and three to four prescriptions per year for 0- to 4-year olds.17
Understanding the enormity of medical care for preterm infants after NICU discharge is important. The magnitude and determinants of pediatric visits and pharmacy prescriptions for preterm infants have not been described. Without this information, medical care providers and parents cannot anticipate and optimize the care that preterm infants require during infancy.
The all-inclusive nature of the Kaiser Permanente health program makes each infant's NICU hospitalization history and outpatient health-care records accessible. The goals for this study were (1) to determine the frequency and type of clinic visits and prescription medication use in a population cohort of preterm infants enrolled in a managed care plan during the first year after NICU discharge, and (2) to identify predictive factors for a high number of office visits (>30 visits per year) or a high number of prescriptions (>5 prescriptions per year).
Preterm infants included in this study were identified as the 32 weeks infant subset of the Infant Functional Status Study (IFS) cohort.18 IFS cohort eligibility included all surviving infants born at 32 weeks GA in five Northern California Kaiser Permanente Medical Care Program (KPMCP) hospitals between 1998 and 2001. KPMCP is a group-model managed care organization with integrated information systems whose perinatal outcomes have been described.6, 13, 19, 20, 21, 22 Exclusion criteria included death, major congenital anomalies, need for mechanical ventilation at home, placement of a ventricular peritoneal shunt or loss to follow-up in the Northern California KPMCP within 1 year from discharge. The initial source for our data was the Kaiser Permanente Neonatal Minimum Data Set, which tracks all NICU admissions in the KPMCP.23 We used a common medical record number to link the Neonatal Minimum Data Set records to the KPMCP hospitalization, outpatient visits, pharmacy databases and cost management information system. This project was approved by the institutional review boards of both The Children's Hospital of Philadelphia and KPMCP.
Variable and outcome definitions
We defined small for gestational age (SGA) as an infant whose birth weight was less than the 10th percentile for GA,24 BPD as the need for supplemental oxygen support at 36 weeks post-menstrual age,25 NEC as radiological or clinical evidence of Stage IIA NEC or worse from Bell's criteria26 and IVH as the most severe grade on head ultrasound.27 We classified race and ethnicity based on maternal information in the Neonatal Minimal Data Set. We used the median household income of the home zip code of the mother as a proxy measure for family income.28, 29
We used the date, department code and International Classification of Diseases (ICD)-9 codes to classify outpatient visits into six groups: WCC, non-well primary care pediatric visit (NW-P), non-well specialty care visit (NW-S), emergency room or urgent care visit, occupational or physical therapy visit and testing visit. WCC visits were defined as visits to the department of pediatrics or family practice, which included the ICD-9 codes for health supervision of infant (V20.1 healthy infants or child receiving care or V20.2 routine infant or child health check).30 NW-P visits represented either sick visits, follow-up visits, growth visits or visits to a pediatric sub-specialist included within the department of pediatrics, such as gastroenterology, nephrology and cardiology visits. The NW-S visit category primarily represented visits to dermatology, surgery, orthopedics, ophthalmology, allergy, neurology and otolaryngology departments. Testing visits included visits to radiology, echocardiography, electrocardiogram or electroencephalogram.
The Cost Management Information System itemizes each prescription filled by a patient. We collected the name of the medication dispensed, the date the prescription was filled and the amount dispensed. Refill prescriptions were counted as a new prescription each time the medication was dispensed. We excluded rare duplicate prescriptions for the same medication filled on the same or consecutive days. We excluded all prescriptions for topical ointments, over-the-counter medications and vitamin/nutritional medications.
High resource use definitions
We defined the high clinic user group as infants with >30 visits per year because this number approached the 10th percentile and represented 2.5 visits per month, a 2.5-fold increase above the anticipated 12 visits for term infants.13, 14, 15, 16 We defined the high prescription user group as infants with >5 prescriptions because this figure was twice the average number of prescriptions in this cohort and greater than the reported three to four prescriptions per year in 0- to 4-year olds.17
χ2 Test and Student's t-test were used to identify differences between groups for categorical and continuous variables, respectively. For multivariable logistic regression, we started with a base model that contained gender, GA and SGA variables. Additional variables identified as significant in univariable analysis were tested. Variables remained in the final model if they were significant at the 0.05 level.
This IFS cohort included 892 infants at 23 to 32 weeks GA (Table 1). This cohort contained 265 infants born at 23 to 28 weeks GA, of whom 25% developed BPD, 4% developed NEC and 6% developed grade 3 to 4 IVH. This frequency of neonatal morbidities is comparable to that found in national statistics.2, 3, 4, 8 This cohort was 46% Caucasian (white), 21% Asian, 21% Hispanic, 11% African American (Black) and 1% Native American.
Pediatrician and specialist visits
The 892 infants had 18 346 clinic visits. Most of these clinic visits (91%) were to pediatric primary care providers and specialists, including visits for WCC (22%), NW-P (58%) and NW-S (11%). The remaining 9% of visits included emergency room and urgent care visits (2%), occupational and physical therapy visits (6%) or diagnostic test visits (<1%). Infants had a mean of 20 (median 19) clinic visits per year. The top 10th percentile included infants who had more than 33 visits.
Figure 1 shows the mean number of visits per child per year in each visit type by GA and BPD status. Increased GA was associated with decreased visits, but the average total visits per year for 31- to 32-week infants (17 visits per year) remained higher than the anticipated 12 visits per year for infants in the general population.13, 14, 15, 16 The frequency of WCC visits was constant and consistent with guidelines. The extra visits per year for preterm infants were attributed to non-well pediatric and specialty care.
Infants experienced weekly visits in the first month after discharge (Figure 2). Despite a gradual decline in visit frequency, infants were still visiting outpatient clinics an average of 1.5 times per month 1 year after NICU discharge.
Pharmacy prescription services
Of the 892 infants, 383 infants (43%) filled 2100 prescriptions during the first year. Refill prescriptions were counted each time a prescription was filled. Infants who filled at least one prescription received an average of 5.5 prescriptions per year. Most prescriptions could be classified as either antibiotics (29%) or respiratory medications (49%), including inhaled medications and nebulizer supplies. The remaining prescriptions included diuretics (4%), gastric acid suppression and pro-kinetic drugs (13%) and central nervous system stimulants (2%).
The extent of prematurity was related to the likelihood of receiving medication and an increased number of prescriptions per infant. Half (52%) of the 23- to 26-week GA infants had a prescription compared with 37% of the 31- to 32-week GA infants. Among infants who filled at least one prescription, 23- to 26-week GA infants had an average of nine prescriptions per year compared with four prescriptions per year for 31- to 32-week GA infants. Infants with BPD who received at least one medication also had nine prescriptions per year.
Risk factors for high number of clinic visits or prescription
We were particularly interested in high user infants, those with >30 clinic visits or >5 prescriptions. High clinic infants had no increase in WCC visits but rather a 2- to 3-fold increase in all other clinic types. These high clinic infants had an average of 28 visits for non-well pediatric or specialist visits, 4 occupational or physical therapy visits and 1.5 emergency room or urgent care visits per year. Using ICD-9 codes, we determined that 25% of the visits were for respiratory symptoms, growth, development, infections or retinopathy follow-up.
We compared the characteristics of infants with heavy use to those without heavy use (Table 1). Prematurity, lower birth weight, BPD and NICU length of stay were each associated with a high number of clinic visits and a high number of prescriptions (P<0.001). A prior history of NEC or grade 3 to 4 IVH was associated only with a high number of clinic visits. Extreme prematurity and neonatal morbidities (BPD, NEC, IVH and grade 3 to 4), although important, only explain 51% of the high clinic infants and 40% of the high prescription infants. Race and Hispanic ethnicity were not associated with high use. Heavy clinic use was associated with heavy prescription use.
We used multivariable logistic regression to determine the risk factors for a high number of clinic visits or prescriptions (Table 2). After controlling for the variables listed in Table 2, the extent of prematurity remained the strongest risk factor for high number of clinic visits and high prescription use. We did not detect a significant interaction between GA and BPD.
Both multivariable models were able to discriminate the high user groups of infants on the basis of the area under receiver operating curve (AU-ROC). When the high clinic visit model included only prematurity groups, it still maintained discriminating ability with an AU-ROC of 0.72. The higher AU-ROC for the clinic model suggests that GA, gender, SGA and morbidities are able to discriminate high clinic users better than high prescription users.
We used the multivariable regression model to predict the likelihood of an infant having >30 clinic visits in the first year after NICU discharge. Half of the 23- to 26-week GA infants with a history of SGA, BPD, IVH grade 3 to 4 or NEC are predicted to have more than 30 health-care visits. Alternatively, 25% of infants in the following three categories are predicted to have >30 visits, namely: (1) 23 to 26 weeks GA without major neonatal morbidities, (2) 27 to 28 weeks GA regardless of neonatal morbidities or (3) 29 to 30 weeks GA with a history of SGA, BPD, NEC or IVH grade 3 to 4.
At the time of NICU discharge, preterm infants must transition from the NICU to the home environment and outpatient pediatric care. The first year in an infant's life is recognized as a critical period of growth and development. This population-based cohort study is the largest multicenter evaluation of clinic use and prescription medication use among preterm infants and the first to describe the enormity of clinic visits and prescription medication during infancy. Preterm infants in this 23- to 32-week GA cohort had an average of 20 visits per year, and prescription medication use was common. This high number of visits is even greater than that reported for a KPMCP NICU infant cohort and much greater than that reported for healthy term infants in a variety of outpatient settings.13, 14, 15, 16 High utilization occurred despite the cohort exclusion of the sickest infants, those who required VP shunts or home mechanical ventilation, and the exclusion of prescriptions for over-the-counter medication, vitamins and nutritional products.
Frequent non-well visits to the pediatrician or specialist for non-well care explained the extra number of visits and most prescriptions were for antibiotics and respiratory medication. Clearly, prematurity confers medical consequences that extend through the first year of life. We can now anticipate the extra health-care use of the most extremely preterm infants or infants with a history of SGA, BPD, NEC and IVH grade 3 to 4. The frequency of respiratory medication use in this cohort in infants with and without BPD suggests that pediatricians need to be screening for respiratory symptoms in all preterm infants.
The enormity of office visits and prescriptions among the moderately preterm infants was striking and concurs with emerging data on the high degree of morbidity in babies born at 30 to 36 weeks GA.31, 32, 33 Extra medical attention for moderately preterm infants is warranted. These infants are at risk for gastroesophageal reflux disease, poor weight gain, infections, chronic respiratory difficulties or neurological delays. Frequent health-care visits can offer more active surveillance and preventive strategies to minimize medical consequences that linger through infancy.
High health-care utilization has been associated with parental anxiety reflected in the vulnerable child syndrome. Parents of preterm infants have been noted to perceive their infants as medically vulnerable, even in the absence of indicators of medical vulnerabilities.34, 35, 36 High parental perception of child vulnerability is associated with high health-care utilization along with an increased risk of behavior problems and altered parent–child interaction.37, 38, 39, 40 One recent study of preterm infants suggested that higher parental perception of child vulnerability is correlated with worse developmental outcomes at 1 year adjusted age.34 Pediatricians need to be alert to parental anxiety and provide support and reassurance to these families. Understanding more about how to optimize care and lessen anxiety through the transition from the NICU to the home may diminish the risk of the vulnerable child syndrome.
Families of infants who have more than 30 visits per year to a medical center would benefit from a coordinated schedule of visits and a clear mechanism of communication between and among physicians and the family. For some infants, home nursing visits and follow-up phone communication may serve to support and educate parents in the care of their infants while also reducing frequency of visits and parental anxiety.41, 42 Research is needed to determine if providing frequent, coordinated care for preterm infants improves infant growth and development and/or decreases hospitalization.
The all-inclusive nature of the Kaiser Permanente health program makes each infant's NICU hospitalization history and outpatient health-care records accessible. However, studies using administrative data sets have limitations. The most important limitation is that we did not have information regarding the purpose or appropriateness for clinic visits or prescriptions. We have no information on prescriptions not filled in the pharmacy. We did not have individual socioeconomic status information for families. Finally, a small number of infants were excluded from this cohort if they died or were lost to follow-up in the KPMCP system. These infants were not demographically or medically different from the study cohort and, as such, the impact is felt to be minimal.
Our results indicate that preterm 23- to 32-week GA infants in the Kaiser Permanente Medical Care Program had a very high number of office visits and pharmacy prescriptions. Extreme prematurity and neonatal morbidities were notable risk factors for increased clinic and pharmacy resources. However, even moderately preterm infants who escaped common neonatal morbidities remained at high risk. If frequent pediatric visits are an important component of optimal outcomes for preterm infants, then considerable effort will be needed to reach indigent or rural families. Pediatricians should be alert to parental anxiety and the risk of the vulnerable child syndrome among preterm infants as a potential factor contributing to frequent health-care utilization.
We acknowledge the help of Marla Gardner and John D Greene for help with the creation of the data set, Joan Verdi for her assistance with developing the protocol, Orit Even-Shoshan for her oversight of the project and Karin McGowan, PhD, for her insightful comments on draft versions of this manuscript. This work was supported by a grant by the Maternal and Child Health Branch of the Health and Human Resources Services Administration Grant 1R40MC00236 (JHS).
About this article
The authors state no conflict of interest.