Caregiver Engagement During Pediatric Post-Acute Care Hospitalization

A retrospective review of medical records to examine caregiver engagement was conducted at a 49 bed pediatric pPACH in the Northeastern United States. This pPACH serves a multinational patient population with complex medical and/or rehabilitative needs referred from acute care hospitals primarily from the Northeastern United States. Service providers include physicians, nurse practitioners, psychologists, physical therapists, occupational therapists, speech-language pathologists, respiratory therapists, clinical nutritionists, child life specialists, social workers, and interpreter services (available 24 hours/day through hospital-based interpreters and/or contracted phone and tablet services allowing for expansive range of languages). Patient diagnoses may include complications related to premature birth, genetic/chromosomal disorders, acquired illnesses, and traumatic injuries, which may necessitate invasive and noninvasive respiratory support, enteral feeding and/or total parenteral nutrition, and polypharmacy. Family members (>18 years) may sleep at bedside in a mix of private and 2 patient rooms.

Patient demographics were retrieved by a data analyst from nursing and social work documentation in the hospital’s electronic health record (EHR; Medical Information Technology, Incorporated Westwood, MA). Demographic data included patient: sex, payer, Child Protective Services (CPS) involvement, race/ethnicity, age at first admission to pPACH, total number of pPACH admissions during study timeframe, residence distance from the hospital, length of stay, and caregiver visits and calls.

Caregivers were defined as adults who accepted responsibility for the patient by providing a home and care for and learn the patient’s medical care. This definition encompassed the EHR options: parent, legal guardian, foster parent, and caregiver. The number of caregiver visits and telephone calls were retrieved by electronic data query from the nursing daily shift assessment in the EHR for patients admitted from January 1, 2013, to December 31, 2017. These dates reflect the first full 5 years of EHR documentation of caregiver visits and calls. Caregiver engagement was defined as occurring in the following 2 ways: (1) in-person visits to the patient and (2) telephone calls made to the patient’s unit. Caregivers may visit their child or call their nurse 24 hours per day.

Dichotomous variables included sex, payer, and CPS involvement. Sex was identified as male/female. Payer, which for the purposes of this study was classified as the patient’s primary insurance, was recoded to 2 variables, private (commercial, foreign embassy, or self-pay) and public (Medicaid programs). CPS is a state-run federally mandated agency that receives, investigates, and responds to reports of child abuse or neglect. Family involvement with CPS (before and/or during pPACH admissions) was retrieved from the nursing staff’s daily shift assessment, supported by social work documentation, and was coded as yes (involved) or no (not involved). CPS involvement was included to describe the pPACH population and understand the potential relationship to caregiver visits and/or calls.

Multilevel variables included race/ethnicity, age groups, and number of pPACH admissions groups. Race/ethnicity (Asian, Hispanic, white, African American/Black, Middle Eastern Indian/Arabic, other, unknown, and missing), as listed in the EHR, was reported by caregivers. The largest race/ethnic group was white and were used as the reference group. Age at first admission (date of first admission – date of birth) was coded into 5 age groups: infant (<1 year), toddler (1–3 years), school age (4–11 years), adolescent (12–17 years), and young adult (≥18 years). Number of admissions was coded into 3 admission groups: single, 2, and 3 or more pPACH admissions. Residence distance from hospital (mileage difference between pPACH address and approximate centroid of home ZIP code) was coded into 4 groups based on quartiles: 0 to 8.4, 8.5 to 21.4, 21.5 to 45.8, and ≥45.9 miles.

Continuous variables included length of stay (= sum of pPACH days for all admissions), total number of visits, and total number of calls. Both visits and calls were individually coded as 1 = any visit(s)/call(s) and 0 = no visit/call in a 24-hour period. The cumulative count was calculated separately for both variables.

Unique variables were calculated to determine proportions of success with missed opportunities for visits and calls (see Supplemental Fig 1 for variable calculations).

Deidentified data from patient records were uploaded to SPSS 22 (IBM) and R (version 3.6.2) for statistical analyses. Descriptive statistics (percentages) were used to describe the sample demographics. The current study used multivariable logistic regression models to estimate the relationship between sociodemographic variables and caregiver engagement, measured as the odds of caregiver visit and odds of caregiver call (Supplemental Fig 1). Probabilities were extracted from the model using the R package emmeans.¹² 

Odds ratios with 95% confidence intervals were used to measure the association between demographic characteristics and caregiver engagement outcomes (caregiver visit and call ratios). P values < .05 were considered statistically significant.

Approval for this study was granted by the hospital’s institutional review board.

Over the 5 year study period, 619 patients were admitted to the pPACH. Five outliers were removed because of extreme values in number of admissions (9-18 vs 1-8 admissions) representing <1% of the patient population. Cumulative total of caregiver visits = 32 224, whereas the cumulative total of caregiver calls = 14 006; caregiver visitations occurred approximately twice as frequently as phone calls. Among the 614 patients, more than one-half were male (n = 337, 54.9%), almost two-thirds had public payer (n = 402, 65.5%), less than one-quarter were involved with CPS at some time during their admission(s) (n = 136, 22.1%), and the largest racial/ethnic group was white (n = 221, 36.0%). Infants composed the largest age group (n = 271, 44.1%) and most patients had a single admission (n = 453, 73.8%). Table 1 presents patient demographics and unadjusted probabilities for caregiver visits and calls; Table 2 presents the adjusted probabilities and odds ratios for caregiver visits and calls.

Patients had higher odds of being visited by a caregiver if they had private insurance (compared with public insurance); their caregiver identified the patient as Asian, Arabic/Middle Eastern, or Hispanic (compared with those identified as white); were between the ages of 1 and 17 years (compared with those less than 1 year old); and those who lived between 8.5 and 21.4 miles from the hospital (compared with those who lived within 8.4 miles of the hospital).

Patients had lower odds of being visited by a caregiver if they had CPS involvement (compared with no CPS involvement), were identified by their caregiver as being African American/Black, other, or unknown race/ethnicity (compared with those identified as white), were 18 years old or older (compared with those less than 1 year old), were admitted 2 or more times (compared with having only a single admission), or lived ≥21.5 miles from the hospital (compared with those who lived within 8.4 miles of the hospital).

Patients had higher odds of having a caregiver call if the patient had 2 or more admissions (compared with having only a single admission).

Patients had lower odds of having a caregiver call if they were male (compared with female), had private insurance (compared with public insurance), had CPS involvement (compared with no CPS involvement), were identified by their caregiver as African American/Black, Asian, Arabic/Middle Eastern, Hispanic, other, or unknown (compared with those identified as white), and were ≥1 year old (compared with those less than 1 year old).

Prior research has described parental/caregiver engagement for children who are acutely hospitalized,^1,2,4  although there is a lack of evidence describing caregiver engagement in a pPACH. In the current study, caregiver engagement in a pPACH was explored by examining caregiver visits and phone calls during patients’ admission(s). Based on cumulative total visits and calls, visitation is the primary mode of caregiver engagement at this pPACH, where caregiver engagement through visitation is encouraged and supported through financial assistance, interpreter services, planned family teaching and events, and dedicated private visit space. Although calls are another potential point of contact with health care providers, calls may provide only limited opportunity for attachment between caregiver and patient or for caregivers to learn their child’s care. In addition, calls directly to patients were not expected for this population because of functional communication and cognitive limitations, and were not included in the data collection.

Using private versus public insurance payer as socioeconomic proxy, odds of caregiver visits were higher for patients with private than those with public payers. This may suggest that families with private payers may have resources allowing them to visit their child more frequently. It should be noted, however, that payers may not always reflect economic resources; children with medical complexity may access public health payer systems because of the substantial cost of extensive health care needs.^7,13 

The odds of visitation by caregivers with CPS involvement was lower than those without CPS involvement. The number of children with CPS involvement at this pPACH (22.1%) is relatively consistent with findings in Canadian NICUs/PICUs, where 30% of children experiencing hospitalizations >6 months had Child and Family Services involvement (the Canadian equivalent of CPS).¹⁴  Catlin evaluated children with lengthy hospitalizations (>6 months) in US NICUs and PICUs and reported only 5% of patients had CPS involvement; however, 23% of all extended-stay patients had parents stop their relationship with their child (either gradually diminished calls/visits, relinquish custody, or had custody removed).¹⁵  In the present analysis, the relationship between low caregiver visits as a result of restrictions placed or reasons for CPS involvement was not examined.

Some groups (Asian, Arabic/Middle Eastern, Hispanic, and white) demonstrated higher odds of caregiver visits than the remaining groups (African American/Black, Other, and Unknown). These results are supported by a recent study examining factors which may predict parent’s presence in NICU settings. Among 71 patients, race/ethnicity was significantly correlated with parental presence.¹⁶  Social variables, including race/ethnicity were included to observe potential systemic racism and some known consequences, including poverty, segregation, and limited access to quality health care.¹⁷ 

Patients who were young adults (≥18 years) had lower odds of caregiver visitation than infants; however, many do not have the same developmental needs for in-person caregiver contact as infants and thus this may not present as a clinically pertinent finding. Odds of caregiver visitation within the infant group were lower than all other groups. This is concerning because of potential issues related to attachment and critical periods of development.^18,19  In a meta-analysis examining parenting in the NICU, Vazquez and Cong noted that parents often feel unprepared and emotionally challenged when becoming the parent of an infant requiring NICU care.²⁰  Being engaged by proximity and having physical contact with the child allows parents to gain knowledge and skills and eventually become experts in the care of their child. Lower visitation among caregivers of pPACH infants may delay knowledge and skill development needed to learn the complex care of their child.

Previous research demonstrated that challenges faced by parents of infants in a NICU may influence their presence at the child’s bedside.²¹  One notable theme included being torn between home and hospital, especially if there were siblings at home. In a qualitative study examining acute care hospital to home transitions for children with medical complexity, parents endorsed the challenge of balancing hospital, home, job, and other responsibilities.²²  This dilemma may be exacerbated for caregivers of children with multiple readmissions and/or increasing residence distance. Both factors showed lower odds of visiting by caregivers.

The majority of the pPACH population had a single admission; however, some patients experienced multiple admissions. Readmissions reflect admission to acute care rather than being discharged to home. This readmission cycle often indicates increased medical fragility and/or complications requiring the patient to experience multiple admissions to the acute care setting for higher level of care, surgeries, and/or procedures. Children with medical complexity have the highest rates of readmissions and use a large portion of the health care expenditure in pediatrics.¹³  This level of medical complexity is often associated with prolonged hospitalizations. Prolonged hospitalization has been shown to reduce caregiver engagement over time.¹⁴ 

Residence distance was found to be a significant variable associated with caregiver visits. The odds of visitation were lower for caregivers living more than 21.5 miles from the pPACH. Caregivers coming from the inner radius of 8.5 to 21.4 miles, however, showed increased odds of visiting compared with those within 8.4 miles of the pPACH; the reason for this phenomenon is unclear. Two studies examining accommodations for parents across multiple acute care hospitals in the United States and Canada found most parents stayed at the child’s hospital bedside, at least initially, whether they lived 25 miles or more than 100 miles from the hospital.^23,24 

Caregiver calls placed to the patient’s unit nurse indicates a level of engagement but not physical contact with the patient. There may, however, be unique experiences occurring prompted by recent changes in technology. A call to the nurse may prompt a digital interaction between caregiver and patient (videoconferencing); thus, a caregiver may interact with his or her child even if the child has limited ability to communicate.

Males had significantly lower odds of caregiver calls than females, although the reasons and clinical implications are unclear. Patients with private insurance had lower odds of caregiver calls than patients with public insurance. Children with CPS involvement had lower odds of caregiver calls, which may reflect either limitations imposed by CPS and/or rationale for CPS involvement. All non-white racial/ethnic groups had lower odds of calling than white, perhaps related to language barriers. Infants had the highest odds of caregiver calls than all other age groups, possibly related to caregivers concern for more information about their infant in general or diagnosis/treatment because this experience is new to these caregivers. Patients with multiple admissions (≥2) showed increased odds of calls, which may be associated with prolonged hospitalizations.

An interplay between caregiver visits and calls is notable in certain racial/ethnic groups, for instance the odds of visits in Asian and Middle Eastern families is high, whereas the odds of calls is low. This could be interpreted as families who visit frequently find it unnecessary to call. Alternatively, some variables, such as CPS involvement, are associated with lower odds of both call and visits. This may indicate that CPS involvement is associated with lower overall caregiver engagement as opposed to only impacting 1 mode of engagement.

A paucity of literature has examined caregiver calls and combined caregiver calls and visits as measures of engagement in the context of a pediatric hospitalizations. Although statistically significant relationships were identified in the current study, the authors offer these only as initial findings. Additional research would be needed to understand the role that calls, as well as video-based communication, play within the construct of caregiver engagement.

In this retrospective study, a patient cohort from a single pPACH was examined. Inherent in retrospective research using EHRs, threats to reliability and validity may exist. For example, the limited EHR options for caregiver race/ethnicity may not accurately reflect families’ preference and/or may have been entered erroneously by medical staff. Caregiver visits and calls are required fields in nursing documentation; each shift in the EHR lists visitor type or no family contact; however, there may be instances of missed visits or phone calls ifs the nurse did not receive the call or witness a visit. The authors calculated caregiver visit and call rates based on at least 1 instance in a single 24-hour period, which does not indicate quality or length of the visit/call. Phone calls to nurses were the only caregiver phone calls considered in the study and caregivers may have called other disciplines or patients directly. Last, family primary language was not included as a variable in this study but is planned to be incorporated in future work.

Caregiver engagement in a pPACH is not well understood. This is the first study to describe caregiver engagement at a pPACH and identify related factors. Patients at the pPACH were visited more frequently by their caregivers than nurses were contacted via phone. Factors associated with lower odds of caregiver visits as well as higher odds of caregiver calls included: younger age (infancy), multiple (≥2) pPACH readmissions, and using a public payer. Acute care and pPACH providers are encouraged to evaluate caregiver engagement for all pediatric patients and consider factors that promote engagement as well as factors that may be preventing caregivers from visiting and calling. Future research may include identifying barriers to caregiver visitation as well as assessment of caregivers’ stress and coping.

The authors thank Helene Dumas for invaluable guidance and mentoring throughout this project. The authors also thank Child Life Specialists, Courtney DiStefano and Kristen Hildreth, for bringing this important topic to our attention.