Advertisement

Utilization of NICU Infant Medical Indices to Classify Parental Risk for Stress and Family Burden

Published:September 21, 2019DOI:https://doi.org/10.1016/j.pedhc.2019.07.005

      Abstract

      Introduction

      This study was conducted to examine whether the parental report of objective infant medical indices (e.g., birth weight, length of stay) can be used to identify parents at risk for psychosocial sequelae.

      Methods

      Parents (N = 199) cohabitating with their partner and child who was discharged from a neonatal intensive care unit (NICU) 6 months to 3 years prior to the administration of the study completed an online survey, which included parent-reported infant health, parenting stress, family burden, and family resources.

      Results

      A hierarchical cluster analysis identified the following three clusters of parents at risk for stress and family burden as determined by infant medical severity and access to resources: lowest risk (n = 77), moderate risk (n = 68), and highest risk (n = 8).

      Discussion

      This work highlights how a measure using parent-reported infant health severity, which was developed for this study, can be used to better understand family outcomes following NICU hospitalization.

      Key Words

      About 7% to 15% of newborns delivered each year are admitted to a neonatal intensive care unit (NICU) in the United States (). Parents of hospitalized infants (hereafter referred to as NICU parents) typically describe the NICU experience as an emotional roller coaster because of the fast changes in their child's medical care and the emotions that coincide (
      • Obeidat H.M.
      • Bond E.A.
      • Callister L.C.
      The parental experience of having an infant in the newborn intensive care unit.
      ). For example, parents may learn that their child is gaining weight one day only to find out that they require a feeding tube the next. As a result, parents experience a variety of emotions, including feeling overwhelmed, sad, stressed, tired, relieved, happy, angry, and helpless (
      • Holditch-Davis D.
      • Miles M.S.
      • Weaver M.A.
      • Black B.
      • Beeber L.
      • Thoyre S.
      • Engelke S.
      Patterns of distress in African-American mothers of preterm infants.
      ). Increased parental psychological distress has also been associated with the child's impaired cognitive, social, and behavioral development (
      • Treyvaud K.
      • Anderson V.A.
      • Howard K.
      • Bear M.
      • Hunt R.W.
      • Doyle L.W.
      • Anderson P.J.
      Parenting behavior is associated with the early neurobehavioral development of very preterm children.
      ,
      • Treyvaud K.
      Parent and family outcomes following very preterm or very low birth weight birth: A review.
      ). The health changes of the infant and feelings of stress, guilt, and anxiety about the child's health can lead to adverse mental health outcomes for parents or long-term impairments in parenting styles (
      • Boykova M.
      • Kenner C.
      Transition from hospital to home for parents of preterm infants.
      ). Furthermore, the sights and sounds, as well as psychosocial implications of the NICU, can serve as traumatic stressors for parents, leading to a more difficult adjustment to normal life following the child's hospitalization (
      • Geller P.A.
      • Bonacquistic A.
      • Patterson C.A.
      Maternal expereince of neonatal intensive care unit hospitalization: Trauma exposure and psychosocial responses.
      ).
      The stressors of a NICU hospitalization do not end after discharge. NICU parents must learn to manage their child's ongoing health issues, in addition to typical parenting responsibilities. In fact, in very low-birth-weight populations, parenting stress remains elevated for the first 18 months and does not become similar to stress reported by parents of full-term infants until the child is approximately 2–3 years of age (
      • Treyvaud K.
      Parent and family outcomes following very preterm or very low birth weight birth: A review.
      ). Therefore, the first few years following NICU admission is a critical time to examine parents and families because of the high risk for stress and adverse psychological symptoms. Because of advances in neonatal care over the past 20 years, infants born at lower gestational ages have a higher likelihood of survival (
      • Grunberg V.A.
      • Geller P.A.
      • Bonacquisti A.
      • Patterson C.A.
      NICU infant health severity and family outcomes: A systematic review of assessments and findings in psychosocial research.
      , ). Infants with lower birth weights are at increased risk for complex medical conditions and neurodevelopmental impairments (
      • Gray P.H.
      • Edwards D.M.
      • O'Callaghan M.J.
      • Cuskelly M.
      Parenting stress in mothers of preterm infants during early infancy.
      ,
      • Stephens B.E.
      • Bann C.M.
      • Poole W.K.
      • Vohr B.R.
      Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
      ). Among extremely low-birth-weight infants (<23 weeks), 42% of them experienced moderate to severe neurodevelopmental impairment (e.g., cerebral palsy;
      • Stephens B.E.
      • Bann C.M.
      • Poole W.K.
      • Vohr B.R.
      Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
      ). Infants with increased neurodevelopmental delays may have greater functional limitations. They may require increased medical resources (i.e., rehospitalizations, medical devices, medical appointments) as well as more time and attention from parents and families (
      • Feeley N.
      • Gottlieb L.
      • Zelkowitz P.
      Infant, mother, and contextual predictors of mother-very low birth weight infant interaction at 9 months of age.
      ). Extant literature indicates that parents with infants with greater medical complications and neurodevelopmental impairments report increased stress and greater family burden than parents of healthy infants (
      • Lorenz J.M.
      • Wooliever D.E.
      • Jetton J.R.
      • Paneth N.
      A quantitative review of mortality and developmental disability in extremely premature newborns.
      ,
      • Stephens B.E.
      • Bann C.M.
      • Poole W.K.
      • Vohr B.R.
      Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
      ). Therefore, families with fewer resources report more negative psychosocial sequelae (
      • Doucette J.
      • Pinelli J.
      The effects of family resources, coping, and strains on family adjustment 18 to 24 months after the NICU experience.
      ,
      • Treyvaud K.
      Parent and family outcomes following very preterm or very low birth weight birth: A review.
      ).
      Although research indicates that more severe medical complications impact family outcomes, the specific medical indices of the infant that help identify which families may be at increased risk remain unclear (
      • Grunberg V.A.
      • Geller P.A.
      • Bonacquisti A.
      • Patterson C.A.
      NICU infant health severity and family outcomes: A systematic review of assessments and findings in psychosocial research.
      ). In addition, parental and family outcomes during the first few years following hospitalization remain understudied (
      • Holditch-Davis D.
      • Miles M.S.
      • Weaver M.A.
      • Black B.
      • Beeber L.
      • Thoyre S.
      • Engelke S.
      Patterns of distress in African-American mothers of preterm infants.
      ,
      • Lefkowitz D.S.
      • Baxt C.
      • Evans J.R.
      Prevalence and correlates of posttraumatic stress and postpartum depression in parents of infants in the neonatal intensive care unit (NICU).
      ,
      • Obeidat H.M.
      • Bond E.A.
      • Callister L.C.
      The parental experience of having an infant in the newborn intensive care unit.
      ,
      • Shaw R.J.
      • Bernard R.S.
      • DeBlois T
      • Ikuta L.M.
      • Ginzburg K.
      • Koopman C.
      The relationship between acute stress disorder and posttraumatic stress disorder in the neonatal intensive care unit.
      ). Furthermore, a current systematic review indicates that there is no gold standard measure of infant medical severity within the NICU population (
      • Grunberg V.A.
      • Geller P.A.
      • Bonacquisti A.
      • Patterson C.A.
      NICU infant health severity and family outcomes: A systematic review of assessments and findings in psychosocial research.
      ). Most of the measures that exist require information from the medical team or medical chart, which makes the dissemination of research more difficult. Understanding which indices of infant medical risk help identify parents at risk for adverse psychosocial sequelae is warranted to improve screening and interventions for parents during and after their child's NICU hospitalization. Moreover, an infant health severity measure that only requires parental report would be useful for increased dissemination and accessibility of NICU research.
      The purpose of the present study, which was part of a larger thesis project (
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      ), was to expand on the current literature by determining which specific infant health indices can be used to identify parents at elevated risk for stress and family burden in the aftermath of their child's NICU hospitalization. A secondary aim was to determine whether it is feasible to use the parental report of infant medical risk, given that this can be used to inform the development of a parent measure of infant health issues. This study uniquely examined infant medical risk, parental stress, and family burden using cluster analysis, an analytic technique that can group parents based on shared experiences. Understanding which factors help to differentiate parents at low, moderate, or high risk for negative psychosocial adjustment (e.g., increased stress and greater family burden) may inform screening techniques in the NICU.

      METHODS

      Partcipants

      Inclusion criteria

      Biological mothers of living, singleton infants who were admitted to a NICU, and their male partner or spouse, were eligible for participation in this study. Each parent was invited to participate as individual survey respondents, meaning that one or both could be deemed eligible and participate. The infant of potential participants must have been discharged from the NICU 6 months to 3 years prior to the administration of the study. Mothers and partners were required to be in an intimate relationship with each other and cohabitating along with the infant since NICU discharge. The reason parents were surveyed online was to gather a large sample and obtain parents of infants with a variety of health issues. This design helps to capture parents and infants who spent time in level I, II, III, or IV NICUs, which means that the hospitals ranged in severity. Based on the literature, parenting stress remains elevated for the first few years following discharge (
      • Treyvaud K.
      Parent and family outcomes following very preterm or very low birth weight birth: A review.
      ). As the first 6 months are often a time of transitions, we sought to understand how parental stress and family impact were associated with infant medical issues years after NICU admission and discharge. The current study was part of a larger thesis project (
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      ), which was investigating couple functioning as an outcome of interest. The examination of couples also helps increase internal validity given that single parents may report increased stress because of a number of additional responsibilities and potentially less support with parenting. Participants were required to be at least 18 years old at the time of the study, reside in the United States, understand written English, and have Internet access. All participants provided a waiver of consent. This protocol was reviewed and approved by the Drexel Institutional Review Board (approval number: 1507003801).

      Recruitment

      Participants included 199 parents of children who had a NICU hospitalization and who were recruited via the Internet and posted flyers at Drexel University and The Children's Hospital of Philadelphia. The study was advertised on online support communities, and social networking Web sites (e.g., Facebook, Craigslist, YouTube) and on printed flyers, with information relevant to parenting and a link to the 2016 Qualtrics survey (http://www.qualtrics.com). Completion of the entire anonymous survey required at least 20 minutes. Three hundred seventy-two participants began the survey, 136 were ineligible based on inclusion criteria, and 69 were excluded because of missing data. Sociodemographics of participants with incomplete data did not differ from those with complete data. Of the 199 eligible parents who completed the survey, all of them identified as biological mothers or fathers.

      Measures

      Sociodemographics questionnaire

      Information regarding participants’ sex, age, race/ethnicity, religious identity, education level, annual household income, employment status, couple or relationship status, infant's sex, reproductive history, mental health history, and stressful life events that have occurred since the infant's birth were collected using a questionnaire created for this study.

      Infant medical indices

      A questionnaire created for this study measured infant health during NICU admission and at the time of survey completion through parental self-report. Items included infant diagnosis/diagnoses, birth weight, gestational age at birth, length of NICU hospitalization, medical procedures and devices used during the infant's NICU stay (i.e., oxygen, ventilator, feeding tube, tracheostomy, phototherapy, or Replogle Tube), whether the infant was on extracorporeal membrane oxygenation (ECMO) during NICU hospitalization, and medical devices used following discharge (i.e., oxygen, cardiorespiratory monitor, feeding tube, tracheostomy, ventricular shunt, central line).
      The infant's current health was assessed by asking parents to identify additional diagnosis/diagnoses since discharge, developmental disabilities, medical devices currently used (i.e., oxygen, cardiorespiratory monitor, feeding tube, tracheostomy, ventricular shunt, central line, nebulizer), current medications, number of rehospitalizations in the first year postdischarge (or since discharge for infants discharged <1 year ago), and number of health care specialists seen in the first year postdischarge or since discharge (e.g., neurologists, occupational therapist, physical therapist, speech therapist, neurosurgery).
      The number of rehospitalizations and specialists was limited to the first year because assessing these items within a specific time frame helps to standardize these indices for infant health severity. The health indicators were determined based on published literature, consultations with NICU developmental psychologists and a neonatologist, and a review of accumulated knowledge regarding neurodevelopmental outcomes following preterm births (
      • Nosarti C.
      • Murray R.M.
      • Hack M.
      Neurodevelopmental outcomes of preterm birth: From childhood to adult life.
      ). Objective indicators of infant health were selected to minimize the subjective bias of parental reports. The only items that parents were asked to recall included gestational age and birth. All other infant medical questions (e.g., diagnoses, medical devices) included options that parents could select from to minimize recall bias.
      Research suggests that retrospective reports of childhood health are of reasonable reliability to warrant their judicious use in population research (
      • Haas S.A.
      The long-term effects of poor childhood health: An assessment and application of retrospective reports.
      ). Furthermore, those with more education and individuals who identify as non-Hispanic Whites, compared with other racial and ethnic minorities, were more consistent in their reports. Given the much shorter time since hospitalization and the emotional experience of this birth, parents have a higher chance of accurately recalling information about their child's health. Finally, given the same characteristics, they have a greater likelihood of accurately recalling this information (Table 1).
      Table 1Participant sociodemographics variables
      VariablesTotal (N = 199)Females (n = 182)Males (n = 17)
      Ethnic-racial backgroundn (%)n (%)n (%)
      White163 (81.9)148 (81.3)15 (88.2)
      African American/Black13 (6.5)13 (7.1)0 (0)
      Latina/Latino/Hispanic8 (4.0)8 (4.4)0 (0)
      Bi/multiracial/ethnic7 (3.5)7 (3.8)0 (0)
      Other
      Other incudes participants who identified as Asian/Pacific Islander, Native American/Alaskan Native, or Indian.
      7 (3.5)5 (2.7)2 (11.8)
      Not reported1 (0.5)1 (0.5)0 (0)
      Relationship status
       Married169 (84.9)154 (77.3)28 (14.0)
       Unmarried, but living with partner30 (15.1)15 (7.5)2 (11.7)
      Income
       <$25,00021 (10.6)19 (10.4)2 (11.8)
       $25,000–$49,99946 (23.1)44 (24.2)2 (11.8)
       $50,000–$74,99945 (22.6)39 (21.4)6 (35.3)
       $75,000–$99,99939 (19.6)36 (19.8)3 (17.6)
       $100,000–$124,99917 (8.5)16 (8.8)1 (5.9)
       $125,000–$150,00014 (7.0)13 (7.1)1 (5.9)
       >$150,00016 (8.0)14 (7.7)2 (11.8)
       Not reported1 (0.5)1 (0.5)0
      Education
       Some high school2 (1.0)2 (1.1)0 (0)
       High school graduate13 (6.5)13 (7.1)0 (0)
       Some college or associate's degree72 (36.2)63 (34.6)9 (52.9)
       4-year college or bachelor's degree67 (33.7)60 (33.0)7 (41.2)
       Advanced degree
      Advanced degrees includes participants with Master of Science, Master of Arts, Doctor of Medicine, Doctor of Philosophy, or Doctor of Jurisprudence.
      43 (21.6)42 (23.1)1 (5.9)
       Other
      Other included participant who received vocational education.
      1 (0.5)1 (0.5)0 (0)
       Not reported1 (0.5)1 (0.5)0 (0)
      Employment
       Employed full-time77 (38.7)63 (3.5)14 (8.2)
       Employed part-time31 (15.6)31 (17.0)0 (0)
       Home maker74 (37.2)72 (39.6)2 (11.8)
       Full time student5 (2.5)5 (2.7)0 (0)
       Unemployed7 (3.5)7 (3.5)0 (0)
       Other
      Other included individuals who work from home, self-employed, and combination of student and employment.
      5 (2.5)4 (2.2)1 (5.9)
      a Other incudes participants who identified as Asian/Pacific Islander, Native American/Alaskan Native, or Indian.
      b Advanced degrees includes participants with Master of Science, Master of Arts, Doctor of Medicine, Doctor of Philosophy, or Doctor of Jurisprudence.
      c Other included participant who received vocational education.
      d Other included individuals who work from home, self-employed, and combination of student and employment.

      Parental stress

      A 36-item self-report measure of stress associated with parenting, the Parenting Stress-Index, Short Form (PSI-SF) includes three subscales: parental distress, parent–child dysfunction interaction, and difficult child (
      • Abidin R.R.
      Parenting Stress Index: Professional manual.
      ). Higher scores indicate higher levels of parental stress, and a score of 90 or above on the PSI-SF indicates clinically significant parental stress. The PSI-SF has been used previously with mothers of preterm infants and demonstrated a strong reliability and validity (
      • Abidin R.R.
      Parenting Stress Index: Professional manual.
      ), including strong internal consistency for the parental stress total score (Cronbach α = .92;
      • Abidin R.R.
      Parenting Stress Index: Professional manual.
      ). The Cronbach α coefficient for parental stress total score was .94 in the current study. In previous studies, the PSI-SF demonstrated strong criterion, discriminant, and predictive validity (
      • Abidin R.R.
      Parenting Stress Index: Professional manual.
      ).

      Family impact

      The revised version of the Impact on Family (IOF-R) Scale was used to assess the perceived burden of an ill child on the family (
      • Stein R.E.
      • Jessop D.J.
      The impact on family scale revisited: Further psychometric data.
      ). This self-report measure reflects parental perceptions of changes in family life and attribution of those changes to the child's illness. Higher scores indicate more perceived family burden because of the child's illness. The IOF-R total score is highly correlated with the original total score (r = .97), and internal consistency for the IOF-R has been adequate (Cronbach α = .85;
      • Stein R.E.
      • Jessop D.J.
      The impact on family scale revisited: Further psychometric data.
      ). Cronbach alpha coefficient for parental stress total score was .97 in the current study. Higher IOF-R scores have been significantly associated with maternal psychiatric symptoms (r = .47), poor health for the child (r = −.39), and poorer psychological adjustment for the child (r = −.37;
      • Stein R.E.
      • Jessop D.J.
      The impact on family scale revisited: Further psychometric data.
      ).

      Family resources

      Perception of adequate resources needed for the family was assessed using the revised version of the Family Resource Scale (FRS-R;
      • Van Horn M.L.
      • Bellis J.M.
      • Snyder S.W.
      Family Resource Scale-Revised: Psychometrics and validation of a measure of family resources in a sample of low- income families.
      ). Higher scores indicate more perceived family resources (i.e., basic needs, money, time for self, time for family). Factor analyses revealed four distinct subscales: basic needs, money, time for self, and time for family, and the internal consistency for overall FRS score is acceptable (Cronbach α = .72–.84;
      • Van Horn M.L.
      • Bellis J.M.
      • Snyder S.W.
      Family Resource Scale-Revised: Psychometrics and validation of a measure of family resources in a sample of low- income families.
      ). Cronbach alpha coefficient for family resources total score was .94 in the current study. The FRS-R also demonstrated good predictive and external validity in previous investigations (
      • Van Horn M.L.
      • Bellis J.M.
      • Snyder S.W.
      Family Resource Scale-Revised: Psychometrics and validation of a measure of family resources in a sample of low- income families.
      ).

      Data Analysis

      A hierarchical cluster analysis was conducted to determine clusters or groups of parents who shared characteristics with regard to their infants’ health severity and psychosocial functioning (
      • Blashfield R.K.
      Propositions regarding the use of cluster analysis in clinical research.
      ). The goal was to identify groups of parents whose characteristics are similar within groups but different among groups. The variables included in the analysis were infant medical indices during NICU hospitalization and at the time of survey completion, as well as parental functioning variables.
      Infant medical indices variables included gestational age at birth (weeks), birth weight (grams), length of time of NICU hospitalization (weeks), number of medical devices used during NICU stay, number of medical devices being use at time of discharge, number of medical devices currently in use, number of rehospitalizations, number of health care specialists an infant saw during the first-year postdischarge, and the number of medications currently prescribed.
      Parental variables included parental stress, family burden, and family resources (i.e., basic needs, money, time for self, and time for family). Because these variables were measured on different scales (e.g., birth is measured in grams and gestational age measured in weeks), they were scaled (standardized) for analyses. It is important to standardize these variables for analyses because otherwise, the largest raw value (i.e., grams) would carry the most weight when differentiating groups.

      RESULTS

      Participant Sociodemographics

      Descriptive statistics for all sociodemographics of the sample were examined (Table 1). Of the 199 eligible participants who completed the survey, most were female (n = 182, 91.5%) and identified as non-Hispanic White (n = 163, 81.9%). There was a range regarding estimated annual household income and educational attainment across participants, with the greatest percentage earning $25,000–$74,999 (n = 91, 45.70%) and having achieved up to, but not past, a bachelor's (4-year college) degree (n = 67, 33.70%). Participants’ mental health and reproductive history are provided in Table 2. More than three fourths (n = 88, 44.4%) of the parents reported having a history of mental illness, 25.9% (n = 51) engaged in mental health treatment previously, and 35.4% (n = 70) endorsed a history of pregnancy loss.
      TABLE 2Participant mental health and reproductive history (N = 199)
      History of mental illnessn (%)
       Yes88 (44.4)
       No110 (55.6)
      History of mental health treatment
       Yes51 (25.9)
       No146 (74.1)
      Number of children
       195 (47.7)
       259 (29.6)
       331 (15.6)
       47 (3.5)
       56 (3.0)
       61 (0.5)
      History of pregnancy loss history
      Men reported on female partner's reproductive history.
       Yes70 (35.4)
       No128 (64.6)
      History of fertility treatments
      Men reported on female partner's reproductive history.
       Yes22 (11.1)
       No176 (88.9)
      a Men reported on female partner's reproductive history.

      Infant Health Characteristics

      Mean gestational age at birth was 31.54 weeks (SD = 4.64) and ranged from 23 to 42 weeks of gestation. Mean birth weight was 1,791.54 g (SD = 1,001) and ranged from 454–5,687 g. On average, infants were in the NICU for 7.63 weeks (SD = 6.30) with a range of 0.29–30 weeks. The average length of time since the infant was discharged from the NICU was 66.68 weeks (1.28 years, SD = 44.65) and ranged from 26 to 156 weeks (6 months to 3 years).
      During the NICU stay, 67 infants were on oxygen, ventilator, and a feeding tube (33.6%), and almost all infants were on some combination of these three devices. Most infants were not on ECMO (n = 185, 97%). Most infants were discharged without medical devices (n = 133, 66.8%). For infants discharged on a medical device, a cardiorespiratory monitor was the most common (n = 20, 10.0%). Moreover, most infants were not prescribed medical devices at the time of study completion (n = 176, 88.4%). For infants using medical devices postdischarge, a feeding tube was the most common (n = 9, 4.5%).
      Common diagnoses for infants during NICU admission included reflux (n = 65, 32.7%), apnea (n = 62, 31.2%), respiratory distress syndrome (n = 55, 27.9%), and chronic lung disease (n = 39, 19.6%). A total of 54 infants were diagnosed with an additional medical condition after discharge (27.4%), and 47 were diagnosed with a developmental disability (23.6%). Of the 54 infants with additional diagnoses, 21 were diagnosed with a developmental disability (39.6%). Most infants were not rehospitalized following discharge (n = 146, 73.4%) and were not currently prescribed medication at the time of the survey (n = 138, 69.3%). Most infants saw at least one additional health care specialist during the first year following discharge (n = 161, 80.9%).

      Descriptive Statistics

      Mean parental stress, as measured by PSI-SF, was 73.72 (SD = 22.14) and ranged from 36 to 156. Thirty-seven participants (20.4%) had a score of >90, indicating severe parental stress in the current sample. Mean family burden, as measured by IOF-R, was 16.31 (SD = 17.25) and ranged from 0 to 59. Mean family resources, as measured by FRS-R, was 80.10, with an SD of 17.10 and a range of 3–105.

      Main Findings

      Data were analyzed using RStudio (version 1.0.136). For statistical evaluation, complete data sets were defined as those in which the participant answered every question for all measures. Of the 199 participants, 153 had complete data, and only participants with complete data were included in analyses. Sociodemographics of participants with incomplete data did not differ from those with complete data.
      Five clusters were initially selected based on the Dendrogram, a graph produced from R code used to determine how to differentiate several groups. Three of the clusters (3, 4, and 5) had very small sample sizes but were distinct from clusters 1 and 2 based on infant and parental variables. These clusters were combined to represent the group at the highest risk. Most participants fell in the first two clusters (lowest and moderate risk). The characteristics of the three main groups of clusters are presented in Table 3.
      TABLE 3Description of clusters
      Parents at lowest risk (n = 77)Parents at moderate risk (n = 68)Parents at highest risk (n = 8)
      Least severe infant health during NICU and postdischargeModerately severe infant during NICU and postdischargeMost severe infant health during NICU and postdischarge
      Moderate parental stressModerate parental stressHigh parental stress
      Low family burdenModerate family burdenHigh family burden

      Few family resources
      Note. NICU, newborn intensive care unit.
      Descriptive statistics were conducted on all clusters to examine characteristics (Table 4). Tests of differences were examined for lowest-risk and moderate-risk groups only because the sample size for highest-risk groups was too small to have meaningful results. Independent t tests were conducted to determine differences between low-risk and moderate-risk groups. Results revealed that these clusters significantly differed on all variables except medications, family resources, and parental stress (Table 5). The lowest-risk group had significantly lower means on all variables than the moderate-risk group.
      TABLE 4Cluster descriptive statistics (N = 153)
      Variable
      All variables were standardized.
      ClusterMeanSDMin–max
      NICU medical devicesLow risk1.701.030–5
      Moderate risk2.870.751–5
      High risk2.750.712–4
      Discharge medical devicesLow risk0.210.550–3
      Moderate risk0.530.700–2
      High risk3.622.261–8
      Current medical devicesLow risk000
      Moderate risk0.180.380–1
      high risk1.501.201–3
      SpecialistsLow risk1.301.410–6
      Moderate risk4.402.520–10
      High risk6.383.023–11
      RehospitalizationsLow risk0.140.420–2
      Moderate risk0.600.930–4
      High risk1.753.060–9
      MedicationsLow risk0.551.230–6
      Moderate risk0.681.070–3
      High risk4.882.472–7
      Length of NICU stayLow risk2.882.140.29–42
      Moderate risk12.224.854–30
      High risk16.626.3211–30
      Gestational age (weeks)Low risk34.883.2729–42
      Moderate risk28.083.3223–40
      High risk27.892.3124–30
      Birth weight (grams)Low risk2,4758731,051–4,990
      Moderate risk1,188581480–3,207
      High risk876239614–1,354
      Parental stressLow risk72.1720.8840–128
      Moderate risk71.4319.8436–143
      High risk101.5028.2373–126
      Family burdenLow risk9.8613.900–52
      Moderate risk21.6817.040–57
      High risk43.1212.9024–59
      Family resourcesLow risk80.2716.2516–105
      Moderate risk82.6614.5823–105
      igh risk66.6212.2243–82
      Note. NICU, newborn intensive care unit.
      a All variables were standardized.
      TABLE 5Differences between the low-risk and moderate-risk groups
      VariableLow-risk group (n = 77)Moderate-risk group (n = 68)tP
      NICU medical devices1.702.87−7.87<.001
      Discharge medical devices0.210.53−3.05.002
      Current medical devices0.000.17−3.79<.001
      Specialists1.304.40−8.96<.001
      Rehospitalizations0.140.60−3.75<.001
      Medications0.550.67−0.69.494
      Length NICU stay2.8812.22−14.66<.001
      Gestational age (weeks)34.8828.0812.38<.001
      Birth weight (grams)2,4751,18810.56<.001
      Family resources80.2882.66−0.94.352
      Parental stress72.1771.430.22.837
      Family burden9.8621.68−4.54<.001
      Note. NICU, newborn intensive care unit.

      DISCUSSION

      The current study, which was part of
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      thesis, expanded on the current literature by using a novel analytic technique (i.e., hierarchical cluster analysis) to identify which parents may be at greatest risk for negative psychosocial sequelae after their infants’ NICU hospitalization. Three main clusters were identified based on infant health characteristics, parental stress, family burden, and family resources. Clusters represented parents with the least severe infants and least family burden, parents with moderately ill infants and moderate levels of stress and family burden, and parents with most severely ill infants and greatest levels of parental stress and family burden and fewest family resources. However, results for the highest-risk group should be interpreted with caution given the small sample size.
      Most participants (n = 145, 94.77%) were categorized into low- and moderate-risk groups. Although low-risk and moderate-risk groups differed on nearly all infant medical indices and family burden, they did not differ on the number of infant medications, family resources, or parental stress. Parental stress did not differ between these groups. It is hypothesized that this finding could be related to the fact that longer NICU stays and more frequent NICU rehospitalizations may result in increased face time and communication between parents and members of the medical and psychosocial teams in the NICU. This additional interaction, in turn, may increase parental confidence when managing their child's illness postdischarge. Using the current data set, it was reported elsewhere that a multivariable regression analysis revealed that more rehospitalizations were associated with less parental stress (
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      ). As a result, there may be confounding factors that influence stress that was not addressed in the current study. Another hypothesis to explain this finding could be that the current sample included more low to moderate medically at-risk infants and therefore limit the ability to detect differences in parental stress given the small number in the high-risk group. For example, the high-risk group had higher means for parental stress, which suggests that increased stress could be a function of infants with more severe medical outcomes, particularly those born extremely premature. It has been reported that parents of extremely low-birth-weight infants experience high levels of stress compared with parents of infants born full term (
      • Singer L.T.
      • Salvator A.
      • Guo S.
      • Collin M.
      • Lilien L.
      • Baley J.
      Maternal psychological distress and parenting stress after the birth of a very low-birth-weight infant.
      ). In the current study, low- and moderate-risk groups represent more mild-moderately ill infants. Therefore, it is possible that stress is a more useful predictor of parental functioning among those with the most severely ill infants in the NICU.
      Parents of infants with increased health issues during a NICU hospitalization (e.g., birth weight, medical devices, length of time in NICU) and after discharge (e.g., number of specialists, rehospitalizations, medical devices) reported increased family burden. These findings are in agreement with the thesis by
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      , which highlighted that family burden, as compared to parental stress and couple functioning, was most influenced by infant medical risk (i.e., infant use of ECMO, the longer length of stay, more medical devices). These findings are also consistent with reports that infants who are born with lower gestational ages experience increased medical complications during and after hospitalization and require increased attention and support from parents, which can take a significant toll on the family (
      • Geller P.A.
      • Bonacquistic A.
      • Patterson C.A.
      Maternal expereince of neonatal intensive care unit hospitalization: Trauma exposure and psychosocial responses.
      ,
      • Stephens B.E.
      • Bann C.M.
      • Poole W.K.
      • Vohr B.R.
      Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
      ). Specific infant medical indices and levels of family burden can be used to identify parents who may need additional psychosocial support. The present findings indicate that screening techniques should focus on infant medical severity and family burden within the NICU setting to identify parents at risk for negative psychosocial outcomes. Currently, clinicians often assess parental mental health or stress in the NICU but do not consider the burden of the child's health on family functioning. Assessment of family burden may be valuable for health care professionals to connect families to increased support in the NICU, as well as to appropriate support services after discharge.
      The findings also indicate that parental report of infant medical severity is feasible given that they were consistent with previous literature that suggests that ongoing medical issues influence parental outcomes (
      • Stephens B.E.
      • Bann C.M.
      • Poole W.K.
      • Vohr B.R.
      Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
      ,
      • Treyvaud K.
      Parent and family outcomes following very preterm or very low birth weight birth: A review.
      ). In addition, these indicators of infant health could be used to sort parents into low-, medium-, and high-risk groups depending on stress and family burden. Given that all data were checked for face validity before analyses, it appears that a psychometrically sound parental report measure of infant medial severity is warranted.

      Clinical Implications

      The transition from the NICU to home can be difficult for parents, especially among those with infants with medical complexities. Parents must learn to manage their child's illness without medical team assistance, maintain family and work, and manage their self-care and relationship in the context of having increased responsibilities for their child's care. Identifying at-risk families may help to inform more effective screening tools during and after their child's hospitalization, as well as provide valuable information for the development of psychosocial interventions that continue throughout this adjustment period. Parents of infants at a more severe medical risk, during and after discharge, should be targeted with early intervention aimed to improve family and child outcomes. Intervention may promote family coping by improving communication with the medical team, addressing the value of self-care, and increasing visibility regarding the influence of parental mental health on child development. The current study also indicates that parents appear to be reliable reporters of child health. Therefore, the development of a parental report measure of infant medical risk would help to increase dissemination of a classification system that identifies families at greater risk for negative psychosocial sequelae.

      Limitations and Future Directions

      The current study's self-report, cross-sectional, retrospective research design included potential bias. In addition, parents were retrospectively asked to report about their infant's health during NICU admission, which may have been up to 3 years earlier. As such, recall about the experiences and severity of the child's health may differ from what transpired. However, retrospective reports of childhood health are reasonably reliable to warrant their use in population research (
      • Haas S.A.
      The long-term effects of poor childhood health: An assessment and application of retrospective reports.
      ).
      • Haas S.A.
      The long-term effects of poor childhood health: An assessment and application of retrospective reports.
      reported that >16 years after childhood health issues, parents were able to accurately identify illness >50% of the time and when identifying dichotomous data, they were accurate 91% to 95% of the time.
      The highest-risk group was very small, and as a result, tests of differences could not be used to distinguish among all groups. Infant medical severity ranged from mild to severe; however, most parents had infants who were mild to moderately ill. Recruiting more parents may help to increase variance and allow for more accurate differentiation among groups of parents.
      It is important to note that the current sample only included parents who were in an intimate relationship with their partners’ since their child's hospitalization. Although this approach helps to increase internal validity, this sample does not reflect the range of families who have a child in the NICU. For example, single parents, including those for whom the stress of the NICU experience might have contributed to separation or divorce, may report less parental coping and greater parental stress and family burden. A better understanding of the experiences of parents with a broader range of family circumstances may help determine additional factors that place families at increased risk for stress and family burden.
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      thesis research indicated that parental stress was inversely related to couple functioning (r = −.40, p < .001), which highlights the need to explore various relationships among parents in future work.
      Furthermore, including nontraditional families would also provide valuable information regarding diverse family experiences following the child's hospitalization. Most of the participants were biological mothers and may not accurately represent the entire family experience. It would be valuable to figure out how to successfully recruit larger numbers of men for these studies because sex may be an important component to consider when understanding the impact of the NICU experience on the family.
      Despite these limitations, the extant literature is the first to use cluster analysis to group parents on distress levels as a function of their infant's medical severity. Given this analytic technique, it provides additional information to the findings of
      • Grunberg V.A.
      Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
      thesis because it demonstrates how all of these variables could be used to distinguish parents at increased risk for distress. This information provides key information on which specific medical indices of infant health may be particularly challenging for parents. These findings may help inform clinicians and researchers how indices of infant health can be used to identify parents who may benefit from additional support during and after infant hospitalization. Clinicians should consider using these specific factors to develop and modify psychosocial interventions for both parents (regardless of sex) and examine treatment efficacy between groups.

      CONCLUSIONS

      The current study used hierarchical cluster analysis to distinguish how NICU parents group together based on specific infant health characteristics and parental functioning variables. Three clusters were identified, and parents were categorized into low-, moderate-, or high-risk groups for negative psychosocial adjustment. The low and moderate groups significantly differed on all outcomes except medications, family resources, and parental stress. Future research should consider collecting more data to increase the sample size to more accurately differentiate among the high-risk groups. These findings might also help the health care workforce (e.g., physicians, nurse practitioners, nurses, psychologists) identify which parents to target for additional psychosocial support during their NICU admission and after discharge. Psychosocial interventions should be tailored to each group to improve biopsychosocial outcomes and development for NICU parents, families, and children.

      ACKNOWLEDGMENTS

      We thank John Chuo, MD, of the Division of Neonatology at the Children's Hospital of Philadelphia for his valuable input on the development of this measure.

      REFERENCES

        • Abidin R.R.
        Parenting Stress Index: Professional manual.
        3rd ed. Psychological Assessment Resources, Lutz, FL1995
        • Blashfield R.K.
        Propositions regarding the use of cluster analysis in clinical research.
        Journal of Consulting and Clinical Psychology. 1980; 48: 456-459
        • Boykova M.
        • Kenner C.
        Transition from hospital to home for parents of preterm infants.
        Journal of Perinatal and Neonatal Nursing. 2012; 26: 81-87
        • Doucette J.
        • Pinelli J.
        The effects of family resources, coping, and strains on family adjustment 18 to 24 months after the NICU experience.
        Advances in Neonatal Care. 2004; 4: 92-104
        • Feeley N.
        • Gottlieb L.
        • Zelkowitz P.
        Infant, mother, and contextual predictors of mother-very low birth weight infant interaction at 9 months of age.
        Journal of Developmental and Behavioral Pediatrics. 2005; 26: 24-33
        • Geller P.A.
        • Bonacquistic A.
        • Patterson C.A.
        Maternal expereince of neonatal intensive care unit hospitalization: Trauma exposure and psychosocial responses.
        in: Muzik M. Rosenblum K.L. Integrating Psychiatry and Primary Care. Springer, Switzerland2018: 227-247
        • Gray P.H.
        • Edwards D.M.
        • O'Callaghan M.J.
        • Cuskelly M.
        Parenting stress in mothers of preterm infants during early infancy.
        Early Human Development. 2012; 88: 45-49
        • Grunberg V.A.
        Neonatal intensive care unit infants, parental stress, couple and family impact: How family resources may attenuate the stress.
        2016 (Retrieved from)
        • Grunberg V.A.
        • Geller P.A.
        • Bonacquisti A.
        • Patterson C.A.
        NICU infant health severity and family outcomes: A systematic review of assessments and findings in psychosocial research.
        Journal of Perinatology. 2019; 39: 156-172
        • Haas S.A.
        The long-term effects of poor childhood health: An assessment and application of retrospective reports.
        Demography. 2007; 44: 113-135
        • Holditch-Davis D.
        • Miles M.S.
        • Weaver M.A.
        • Black B.
        • Beeber L.
        • Thoyre S.
        • Engelke S.
        Patterns of distress in African-American mothers of preterm infants.
        Journal of Developmental and Behavioral Pediatrics. 2009; 30: 193-205
        • Lefkowitz D.S.
        • Baxt C.
        • Evans J.R.
        Prevalence and correlates of posttraumatic stress and postpartum depression in parents of infants in the neonatal intensive care unit (NICU).
        Journal of Clinical Psychology in Medical Settings. 2010; 17: 230-237
        • Lorenz J.M.
        • Wooliever D.E.
        • Jetton J.R.
        • Paneth N.
        A quantitative review of mortality and developmental disability in extremely premature newborns.
        Archives of Pediatrics and Adolescent Medicine. 1998; 152: 425-435
        • March of Dimes Perinatal Data Center
        Special care nursery admissions.
        2011 (Retrieved from)
        • Nosarti C.
        • Murray R.M.
        • Hack M.
        Neurodevelopmental outcomes of preterm birth: From childhood to adult life.
        Cambridge University Press, New York, NY2010
        • Obeidat H.M.
        • Bond E.A.
        • Callister L.C.
        The parental experience of having an infant in the newborn intensive care unit.
        Journal of Perinatal Education. 2009; 18: 23-29
        • Shaw R.J.
        • Bernard R.S.
        • DeBlois T
        • Ikuta L.M.
        • Ginzburg K.
        • Koopman C.
        The relationship between acute stress disorder and posttraumatic stress disorder in the neonatal intensive care unit.
        Psychosomatics. 2009; 50: 131-137
        • Singer L.T.
        • Salvator A.
        • Guo S.
        • Collin M.
        • Lilien L.
        • Baley J.
        Maternal psychological distress and parenting stress after the birth of a very low-birth-weight infant.
        JAMA. 1999; 281: 799-805
        • Stein R.E.
        • Jessop D.J.
        The impact on family scale revisited: Further psychometric data.
        Journal of Developmental and Behavioral Pediatrics. 2003; 24: 9-16
        • Stephens B.E.
        • Bann C.M.
        • Poole W.K.
        • Vohr B.R.
        Neurodevelopmental impairment: Predictors of its impact on the families of extremely low birth weight infants at 18 months.
        Infant Mental Health Journal. 2008; 29: 570-587
        • Treyvaud K.
        Parent and family outcomes following very preterm or very low birth weight birth: A review.
        Seminars in Fetal and Neonatal Medicine. 2014; 19: 131-135
        • Treyvaud K.
        • Anderson V.A.
        • Howard K.
        • Bear M.
        • Hunt R.W.
        • Doyle L.W.
        • Anderson P.J.
        Parenting behavior is associated with the early neurobehavioral development of very preterm children.
        Pediatrics. 2009; 123: 555-561
        • Van Horn M.L.
        • Bellis J.M.
        • Snyder S.W.
        Family Resource Scale-Revised: Psychometrics and validation of a measure of family resources in a sample of low- income families.
        Journal of Psychoeducational Assessment. 2001; 19: 54-68

      Biography

      Victoria A. Grunberg, Clinical Psychology Ph.D. Candidate, Department of Psychology, Drexel University, Philadelphia, PA.
      Pamela A. Geller, Associate Professor of Psychology, Ob/Gyn, and Public Health, and Interim Director of Clinical Training, Clinical Psychology PhD program, Drexel University College of Medicine, Philadelphia, PA.
      Chavis A. Patterson, Director of Psychosocial Services, Division of Neonatology, The Children's Hospital of Philadelphia, and Assistant Professor of Clinical Psychology in Psychiatry, Department of Psychiatry, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA.