Maternal Immigrant Status and Readiness to Transition to Home From the NICU

This prospective cohort study includes participants of a larger study called Partnering With Parents, the Medical Home, and Community Providers to Improve Transition Services for High-Risk Preterm Infants in Rhode Island (THP).²³ The study population included mothers of preterm infants (<37 weeks’ gestation) cared for in a level 3 to 4, single-family-room NICU at Women and Infants Hospital (WIH) for >5 days. The study was approved by the WIH Institutional Review Board, and informed consent was obtained. The exclusion criteria were mothers not being able to read English or Spanish and mothers and/or infants having a life-threatening diagnosis. Between October 2012 and September 2015, 776 mothers consented and enrolled when their infants were medically stable and early in the discharge-preparation stage.

All THP families received the WIH-standard NICU-discharge process, including teaching and/or demonstration of infant care, formula mixing, and cardiopulmonary resuscitation; follow-up appointment review; and, if needed, a trial of home equipment. However, as part of the THP, mothers received additional support and guidance facilitated by social workers (SWs) and family resource specialists (FRSs). The FRSs were former NICU parents who spoke English, Spanish, and/or Portuguese and were matched with the parents to maximize culturally sensitive care. The THP SWs provided oversight of the FRSs and collaborated with the NICU-staff SWs with a focus on preparedness and transitioning to home. The THP SWs and FRSs facilitated transitions to mental health support, community providers, and resources (ie, housing, food, and heat) as indicated. Discharge education was provided in both verbal and written format in the preferred and/or native language specified by parents. The FRSs and/or SWs reviewed detailed education binders with parents, including topics of infant safety, infection control, smoking, housing, and depression. The number of in-person meetings and phone contacts were dependent on family needs and requests.

Data collection of maternal characteristics included primary language, race, and level of education. Insurance type was recorded as Medicaid or private and/or nonpublic; there were no uninsured patients in this cohort. Data on history of substance abuse, involvement with child protective services (CPS), domestic violence, and mental health disorders (MHDs) were collected by study SWs either by maternal report or chart review (admission notes, social work assessments, and/or psychiatric consultations). The study definition of history of MHD has been described in detail in previous publications and consists of a documented diagnosis and/or mental health treatment of anxiety, depression, bipolar disorder, posttraumatic stress disorder, or obsessive-compulsive disorder.^21,24 Infant variables included birth weight; gestational age; neonatal morbidities, including intraventricular hemorrhage grades III and IV, necrotizing enterocolitis, sepsis, and bronchopulmonary dysplasia (oxygen at 36 weeks postmenstrual age); oxygen at discharge; length of hospitalization; and breast milk use.

The Fragile Infant Parental Readiness Evaluation (FIPRE) was developed by HealthActCHQ (healthactchq.com/survey/fipre) as a quality measure of parent outcomes and perceptions of NICU-discharge readiness.²⁵ Demographic questions of country of birth and length of time in the United States are included. Immigrant status was defined as mother responding “no” to, “Were you born in the United States?” The FIPRE does not collect details pertaining to refugee or asylum-seeking status.

The FIPRE measures how the parents feel about the NICU care provided to them and their infants and how emotionally prepared they believe they are to care for their infants at home. There are 4 response options ranging from “not at all” to “a lot” for multiitem scales of perception of NICU support, infant well-being, maternal well-being, and maternal comfort. Family cohesion and time impact are single-item scales, with response options ranging from “excellent” to “poor” and “a lot” to “none,” respectively.

NICU support indicates if the parents had a positive NICU experience and if they felt included and well-informed in their infants’ care. Infant well-being indicates how concerned a mother is regarding her infant’s condition and survival. Maternal well-being reflects feelings of self-competency and emotional confidence in caring for herself as well as her infant. Maternal comfort indicates a mother’s degree of worry and/or distress pertaining to her infant’s sleeping, eating, development, and medical needs.

For all scales, excluding the time-impact scale, scores were converted to a standard 0-to-100 metric, with higher scores being more favorable. Scores <75 were evaluated for each scale and represent responses in the lower–three-quarters range of possible scores. Mean scores were compared between immigrant and native mothers.

Maternal and infant characteristics were compared on the basis of immigrant versus native status by using Student’s t tests and Wilcoxon rank tests for continuous data and χ² tests for categorical data. For infant analyses, mixed models were used for the continuous data, and bivariate regressions with generalizing estimating equations adjustment were used for the categorical data to adjust for multiple births. Limited subanalyses were done by regions of interest (developed regions, Africa, Latin America, the Caribbean, and Asia and Oceania) and years in the United States (<5, 5–9, 10–14, and ≥15 years).

A series of logistic regression models was done to estimate the effect of immigrant status on FIPRE scores based on the literature. For the total population, the dependent variables were FIPRE scores (NICU support, infant well-being, maternal well-being, maternal comfort, and family cohesion) <75, and the independent variables were immigrant-mother status, days in the NICU, non-English primary language, Medicaid, MHD, and social risk score (a simple count of the number of characteristics previously identified as increasing the risk of negative outcomes, including having less than a high school education, being single, CPS involvement, domestic abuse, and substance abuse). For the immigrant mothers–only subanalysis, the dependent variables were the FIPRE scores, and the independent variables were years in the United States, Medicaid, and MHD. The final model included those variables that were found to be significant in the bivariate analysis and the literature.

Table 1 presents the maternal characteristics of 176 immigrant and 556 native mothers. Immigrant mothers were more likely to be older, gravida >1, and multiracial or people of color and have a non-English primary language, Medicaid, and less than a high school education. Immigrant mothers were less likely to have social adversities, which included CPS involvement, substance abuse, and a history of MHD compared with native mothers. During the study period, the number of SW and/or FRS contact points were similar between groups.

Thirty-five different countries were reported by mothers as their birthplace and were grouped by regions: 36% Caribbean, 25% Latin America, 20% Africa, 14% Asia or Oceania, and 5% developed regions. Fourteen of 16 (88%) mothers who reported being born in Puerto Rico identified themselves as not being born in the United States, so all 16 mothers are included in the Caribbean category. Thus, native born reflects being born on the US mainland. Among immigrant mothers, 51% reported Spanish as their primary language, 32% reported English, and 17% reported another language.

Table 2 presents maternal characteristics of immigrant mothers by years in the United States. Of the 152 mothers (99%) who answered this question, 22% lived in the United States for <5 years, 26% lived in the United States for 5 to 9 years, 19% lived in the United States for 10 to 14 years, and 33% lived in the United States for ≥15 years. Immigrant mothers who lived in the United States for longer time periods were more likely to be high school graduates, consider English their primary language, have private health insurance, and report a history of MHD. Infant characteristics (Table 3) were similar, except infants of immigrant mothers were more likely to have severe intraventricular hemorrhage and receive breast milk at discharge.

The FIPRE scales of immigrant and native mothers are presented in Table 4. Immigrant mothers reported significantly lower mean scores for infant well-being, maternal well-being, and comfort pertaining to infant medical needs when compared with native mothers. Additionally, immigrant mothers were more likely to score in the lower–three-quarters range for the infant and maternal well-being scales (Fig 1). Immigrant mothers also felt their newborn infants would limit their free time.

Table 5 presents a logistic regression model for the total population to predict FIPRE scores <75. Immigrant-mother status, non-English primary language, and MHD predicted unfavorable FIPRE scores <75 for infant and maternal well-being, whereas Medicaid insurance predicted a more favorable perception of maternal well-being, maternal comfort, and family cohesion. Increased number of NICU days was associated with an unfavorable perception of maternal comfort and family cohesion. The social risk score was not significantly associated with any FIPRE scales.

The logistic model to predict low FIPRE scores among immigrant mothers (Table 5) revealed that increased years in the United States (odds ratio [OR] 1.11; confidence interval [CI] 1.02–1.20) predicted an unfavorable score for perception of NICU support, with the odds increasing by 11% for each year spent in the United States. Immigrant mothers with a history of MHD were 4 times more likely to have a less favorable perception of their infants’ well-being (OR 4.08; CI 1.67–9.94).

Our hypothesis that immigrant mothers would perceive less discharge readiness compared with native mothers is supported by the study findings. Significant differences were observed between study groups in multiple domains of NICU-discharge readiness. Immigrant mothers were more likely to have increased worry about their infants’ survival, perceive themselves as less competent or emotionally ready to care for their infants, and have more concern about their infants’ ongoing medical needs. However, immigrant and native mothers had similar perceptions of NICU support and family cohesion. We propose that the enhanced SW and peer involvement provided by the THP may have contributed to similar perceptions of NICU support.

The importance of exploring the family and home environment and providing multidisciplinary NICU support during discharge-readiness assessment is recommended.²⁶ Peer parents have been used in a variety of settings with demonstrated benefit.^27,–29 Matching NICU parents with peers who have a similar language and culture has helped non-English–speaking mothers gain confidence,¹⁴ and peer support is a recognized, integral component of NICU family-centered care.²⁶ 

In this immigrant-mother cohort, only 32% reported English as their primary language. The language barriers faced by limited-English–speaking families have been well described, including difficulties with information gathering and joint decision-making, which may add to increased distress and anxiety during the discharge process.^14,–16,30 In the study, non-English speaking was associated with maternal unfavorable perception of both infant and maternal well-being. Potential language barriers should continually be assessed because 40% of the immigrant mothers in our cohort who had been living in the United States for >15 years did not consider English their primary language.

In our primary analyses, immigrant mothers were more likely to be older and multiparous and to have Medicaid and less than a high school education, findings that are consistent with other US-immigrant cohorts of similar race and/or ethnicity.^7,31 Additionally, psychosocial risk factors, including CPS involvement, substance abuse, and MHDs, were less likely in this population. Although these reports may be accurate, we recognize that cultural differences and barriers to health care, in either native countries or the United States, may lead to underreporting and/or underdiagnosis, particularly with regard to mental health.^32,–35 

When exploring the socioeconomic changes observed over time spent in the United States, increased report of English as the primary language and a higher level of education for immigrant mothers were not unexpected. Assimilation into US culture and more opportunities for education could facilitate more employment and thus less Medicaid dependence. Reports of MHDs also increased, with most notable changes being seen after the first 5 years of time spent in the United States. This may be due to increased access to or use of psychiatric care over time, which may be partially a result of increased private insurance coverage. However, first-generation immigrants have been reported to have better health outcomes, including mental health, when compared with their US-born counterparts. This phenomenon, known as the Healthy Immigrant Paradox, may not last because the health advantages of subsequent generations begin to decline and eventually resemble those of the US-born population.^36,37 In our cohort, MHD rates increased by eightfold after 5 years in the United States.

The regression analysis of immigrant mothers revealed that increased years in the United States was an independent risk factor for perception of poor NICU support. For each year in the United States, the odds of scoring poorly on this scale increased by 11%. It could be hypothesized that as English improves, less time is needed for the translation of acute medical issues, the health care system, and resources. As families assimilate, the intensity of support services lessens, which may be perceived by some as overall less time spent with the family. In the current study, immigrant and native mothers received similar services from the THP SWs and FRSs.

Despite receiving enhanced THP services, differences in mean FIPRE scores were identified on the well-being and comfort scales. A low FIPRE score for infant well-being indicates that mothers were more likely to fear for their infants’ lives and feel that their infants were not receiving enough medical attention. Immigrant mothers were nearly twice as likely to score lower for infant well-being. A non-English primary language was also an independent risk factor for lower perception of infant well-being. Because the majority of immigrant mothers were non-English speaking, this finding suggests a need for culturally competent providers who are able to communicate with families.

Similarly, immigrant mothers reported lower overall scores for the maternal well-being compared with native mothers. This well-being scale captures perceptions of feeling overwhelmed and decreased parenting self-confidence. In regression analyses, immigrant mothers were 1.7 times more likely to have less favorable well-being scores compared with native mothers. In part, this may be due to the natural challenges of transitioning to motherhood while balancing complex issues of migration and acculturation and thus leading to inadequate maternal coping methods.^38,39 Similarly, mothers who reported a non-English primary language were more likely to report low infant well-being scores. Regression analysis also revealed that mothers with Medicaid were less likely to score poorly on the maternal well-being and comfort scales. We have reported previously that this finding may be secondary to perceptions of Medicaid as a safeguard or reporting bias when in socially undesirable situations.²¹ 

Immigrant mothers’ mean scores for the maternal comfort scale were significantly lower than native mothers’ scores, with approximately half of the mothers reporting scores <75, indicating increased concern and/or anxiety pertaining to infant development, sleep, nutrition, and overall medical needs. These are all common worries previously reported by NICU parents.⁴⁰ Although the THP provided all mothers with enhanced support and education, immigrant mothers had poorer perceptions of maternal comfort, infant well-being, and maternal well-being. These findings are an indicator of the complex and unique challenges of nonnative families.

We cannot overemphasize the importance of recognizing maternal mental wellness. History of MHD was an independent risk factor of poor perception of infant and maternal well-being among all mothers. Additionally, immigrant mothers who reported an MHD were 4 times more likely to have lower infant well-being scores. This is of importance because increased rates of MHDs were associated with increased years of US residency and are consistent with our previous reports of low FIPRE scores among mothers with adverse mental health.^21,24 

Strengths of this study include the diverse cohort of mothers, reporting of maternal psychosocioeconomic characteristics, and relatively large cohort of immigrant mothers. Capturing length of stay in the United States offered opportunities to examine trends over time. By using a discharge-readiness tool targeted to a high-risk NICU population, the FIPRE provided insight into multiple domains of preparedness for transitioning to home. Because all families enrolled in the THP received SW and peer-matched support, we were able to uncover parent-reported unmet needs. Limitations include the potential biases of maternal reporting, including mental health diagnoses. Another limitation is the challenge of examining multiple ethnicities and/or races as 1 cohort because differences may be obscured.

Nearly one-quarter of mothers of preterm infants admitted to the NICU were immigrants. Immigrant mothers perceived themselves to be less ready for discharge in multiple domains pertaining to the well-being and needs of both their infants and themselves. Analysis of psychologic and socioeconomic changes over time provides insight into the changing needs of immigrant mothers with preterm infants. The provision of culturally competent care and the importance of establishing effective communication with immigrant families cannot be overestimated.

CI

confidence interval

CPS

child protective services

FIPRE

Fragile Infant Parental Readiness Evaluation

FRS

family resource specialist

MHD

mental health disorder

OR

odds ratio

SW

social worker

THP

transition home program

WIH

Women and Infants Hospital