Parenting Self-Efficacy in Fathers of Medically Complex Infants: A Longitudinal Study

Abstract

Introduction:

Infants with medical complexity are have multiple chronic conditions and require specialized intensive care. One important factor in optimizing infant health and development is parenting self-efficacy (PSE). The purpose of this study was to examine parental self-efficacy in fathers over time.

Methods:

A longitudinal survey study was conducted with fathers of medically complex infants. We used the validated Karitane Parent Confidence Scale to assess PSE and multivariable linear regression examined the associations between father and infant characteristics on PSE.

Results:

Fathers (n=27) were white (74%), married (85%), high school educated (37%), with incomes ≥ $US50,000 (66%). Father’s mean PSE score was 39.28 (±3.9). Hispanic ethnicity and total number of chronic conditions were significant predictors of lower PSE in fathers (p < .03).

Conclusions:

Fathers of medically complex infants reported low PSE. More strategic interventions need to focus on self-efficacy and creating opportunities for connection between fathers and infants.

Introduction

A successful transition to parenthood has long-term implications for the health and wellbeing of parents and the health and development of their child (Belsky & Jaffee, 2015). This successful transition to parenting includes mastery of new skills, formation of new identity, and development of a caregiving bond with the child (Tremblay & Pierce, 2011). In the first few months of their child’s life, men and women engage in these parenting behaviors at their own pace and fathers are becoming more involved in childcare (Sethna et al., 2017). The amount and the quality of father-child interactions is crucial for child well-being.

Compared to motherhood, fatherhood is a distinct transition period contributing differential effects on child development. Paternal involvement improves cognitive development for full-term (Sarkadi et al., 2008; Sethna et al., 2017) and preterm infants (Yogman et al., 1995), encourages good sleep behaviors (Tikotzky et al., 2010) and supports the development of pro-social behaviors (Cabrera et al., 2007; Trahan, 2018). Father’s play with their infant encourages independence and exploration, whereas mother’s play offers safety and balance (Feldman, 2003; Yogman et al., 1995). Furthermore, mothers have higher parenting self-efficacy (Shorey et al., 2019) and greater marital satisfaction (Elek et al., 2003) when a father is engaged in the care of their infant. These important contributions from paternal involvement are necessary to create supportive experiences that positively impact infant health and development. Even so, there are barriers that may limit paternal involvement during infancy, such as work schedule, traditional gender roles, or lack of inclusion in decision making.

Fathers of neonatal intensive care unit (NICU) infants or those with complex care needs face distinctive challenges in transitioning to their paternal role. Mothers often take a lead role in caregiving and decision making when a child has complex care needs, which encourages confidence. Thus, fathers may struggle in developing confidence as they often feel excluded from care or are hesitant to physically interact with their infant given their medical fragility (Loewenstein et al., 2019; Provenzi & Santoro, 2015). Fathers want to be involved even if they are apprehensive due to perceptions of their infant’s fragility (Fegran et al., 2008). Fathers also experience intense emotional reactions to hospitalization, similar to mothers (Bright et al., 2013) but often feel unnecessary or unable to voice concern, reactions, or emotions due to long-standing, traditional masculine gender roles and common perceptions that mother’s provide more competent care (Deeney et al., 2012). Fathers are torn between work and family responsibilities as the weight of social norms and expectations compel them to provide financially for their family (Hearn et al., 2019), while also managing role identity expectations as a member of the family. The expectations for coparenting, the quality of the partnered relationship and degree of confidence as a parent influence father - child dynamics.

The degree of confidence one has in their parental role (i.e., parenting self-efficacy (PSE)) was first described by Bandura (1997) in social cognitive theory. PSE is a dynamic process influenced by circumstances, experiences, and exposure to new demands (Bandura, 1997; Jones & Prinz, 2005). PSE is defined as an individual’s belief about their own ability to be successful in their role as parent (Vance & Brandon, 2017) and their participation in medical care is linked to improved health status and better adherence to treatment (Wysocki & Gavin, 2004). Co-parenting encourages greater feelings of confidence (Pinto et al., 2016) and confidence helps fathers develop positive role identity (Feeley et al., 2013; Hearn et al., 2019) and increases involvement in care of their medically complex infant, which may lead to fewer hospital days (Melnyk et al., 2006). However, there is limited data about paternal self-efficacy with medically complex infants or how efficacious they feel compared to their partners. Findings related to maternal self-efficacy has been published (Vance, Pan, et al., 2020). Thus, the purpose of this study was to examine parental self-efficacy in fathers, specifically by 1) comparing PSE in mothers and fathers, 2) determining what factors may be associated with PSE in fathers, and 3) examining trajectories of PSE in fathers.

Methods

Data Collection and Participants

We conducted an exploratory, longitudinal study to examine the development parental self-efficacy over time. Participants were recruited from an academic medical center in the southeastern United States who had an infant born with a complex chronic condition and admitted to intensive care. Initial eligibility was determined by whether the infant had a medically complex diagnosis using the Complex Chronic Condition Classification system (Feudtner et al., 2014). Self-identified parents (i.e., mothers and fathers) were eligible if they spoke English, were older than 18 years of age, and anticipated the infant being discharged into their care. Adoptive parents and parents of multiples were ineligible. This study was approved by the hospital Institutional Review Board (similar to Ethics Boards in international settings).

Individual informed consent was obtained from all participants. After consenting, each parent was emailed a secure link using the online database, Research Electronic Data Capture (REDCap), an online research application. All fathers completed the survey via the internet except for one father who requested a paper version. Surveys were completed within three weeks of birth (T1), at discharge (T2), and three months after discharge (T3). Two fathers were lost to follow up and every father enrolled was associated with a mother and infant (e.g., no single fathers).

Measures

The Karitane Parenting Confidence Scale (KPCS) (Crncec et al., 2008) was used to measure task-specific parenting self-efficacy (PSE). The KPCS is a validated 15-item measure used to assess confidence in the parenting role for parents of infants 0 – 12 months in age. The total score ranges from 0 – 45, on a 4-point Likert scale (0–3) with higher scores reflecting greater confidence (i.e., self-efficacy). Scale anchors are, ‘no, hardly ever’, ‘no, not very often’, ‘yes, some of the time’, and ‘yes, most of the time.’ This scale was selected as it is sensitive to detecting changes over time and had a clinical cut off index. Scores ≤ 39 indicate clinically low confidence, identifying those who may be at risk and in need of supportive resources. The scale demonstrates high reliability across study time points (T1 a = 0.84; T2 a = 0.79; T3 a = 0.76).

The Family Assessment Device (FAD) (Epstein et al., 1983) was used to measure family functioning. The FAD is a validated, 12-item scale designed to measure family communication, roles and involvement. Items are rated on a 4-point Likert scale, averaged across responses, with higher scores indicative of poorer family functioning. Cronbach’s alpha for this study demonstrated high reliability (T1 a = 0.84; T2 a = 0.88; T3 a = 0.87)

The Psychological General Well-Being Index (PGWBI) (Gaston & Vogl, 2005) was used to measure psychological health and well-being. This 22-item validated measure generates a total score from a 6-point Likert scale (0–5). Scores above 73 indicate positive well-being. Scores from fathers at 3 months post discharge were used in this analysis. Cronbach’s alpha for this study demonstrated high reliability (T3 a = 0.80).

The Technology Dependence Scale (TDS)(Docherty et al., 2011) was used to measure infant illness severity. The TDS is scored cumulatively based on technology requirements in the following care domains: level of care, parental lines, nutrition, monitors, blood draws, respiratory assistance, skin and oral care, external drains and/or catheters, mobility, transfusions, and medications. The frequency of monitoring or amount of technological dependence generates higher values. Each point increment indicates increasing severity with higher scores representing a greater dependence on technology, or illness severity.

Demographic information and infant characteristics were collected from the enrollment survey and included: parent age (years), race/ethnicity (Asian, African American, Caucasian, Hispanic, Other), partnered status (single, married, or living with partner), education (high school, college, graduate), income (categorical), health insurance (private, public, self-pay), family structure (number of people in home), prenatal care (yes/no), infant gestation age at birth (weeks), total hospital days (length of stay), number of chronic conditions, and total number of medications at discharge.

Statistical Analysis

Statistical analysis was performed using STATA Version 16 (College Station, TX: StataCorp LLC.). Descriptive statistics were conducted on demographic information. A one-way repeated measures ANOVA was used to examine differences in PSE scores among mothers and fathers. To evaluate factors associated with PSE in fathers, multivariable linear regression was used, clustering by family. Regression analysis was used to examine paternal demographic and social characteristics (age, race, education, marital status, family functioning, and psychological well-being) and infant characteristics (illness severity, chronic conditions, medications) as predictors of PSE and evaluate change in scores over time. Due to the small sample size of this study, two separate analyses were conducted to decrease the likelihood of Type II error. Length of hospitalization was a covariate in both models to control for amount of time an infant was in the hospital. Significance was set at the 0.05 level.

Using visual analysis in Tableau Desktop (2019), we categorized PSE scores into 3 main patterns using the clinical cut off index (scores < 39) from the KPCS scale from enrollment to 3 months after discharge. The same process was used to categorize mother’s PSE patterns (Vance, Knafl, et al., 2020).

Results

Sample Description

Twenty-nine fathers consented to participate but two fathers were list-wise deleted from the analysis for incomplete and missing data, resulting in a final sample of 27 fathers (89% retention rate). The majority of fathers were white (74%), married (85%), high school educated (37%), had an income ≥ $US50,000 (66%) and private insurance (78%). Fathers ranged in age from 19 to 48 years (M = 33 years, SD = 6.3) and mothers ranged in age from 19 to 43 (M = 29 years, SD = 5.4). Infant’s mean gestational age at birth was 35 weeks (SD = 4.8) with an average of length of stay at 45 hospital days (SD = 39.5). The majority of infants were the first child (55%) and had a prenatal diagnosis of a complex chronic condition (70%) (Table 1).

Table 1.

Father/Infant demographics and illness characteristics (N=27)

	M	SD
Age	33.8	6.3
Number of Children	1.9	1.2
Gestational Age (weeks)	35.3	4.8
Length of Stay (days)	45.5	39.5
Chronic Conditions	3.5	0.19
Medications	3.4	2.9
Illness severity (TDS score)	18.3	9.3
	n (%)	%
Race
Non-Hispanic White	20	74
Non-Hispanic Black or Other	3	11
Hispanic	4	15
Marital Status
Married/Living together	23	85
Single/Divorced	4	15
Education
Some HS	2	7
High school	10	37
College Graduate	7	26
Graduate/Professional degree	8	30
Income
≤ 25,000	5	18.5
25,001 – 50,000	4	15.5
50,001 – 100,000	9	33
≥ 100,001	9	33
Health Insurance
Medicaid	4	15
Private	21	78
None/Self-Pay	2	7
Infant Diagnosis Category
Birth defects & Neurological issues	15	55
Complex Heart	6	20.6
Extreme Prematurity	6	20.6

Descriptive statistics and regression results

The mean PSE score for fathers across all time points was 39.28 (±3.9), indicating low PSE (i.e., scores ≤ 39 for clinical cut off index). Father PSE scores significantly increased over time (β = 1.58, p < .00). ANOVA analysis showed PSE mean scores between mothers and fathers were significantly different, with fathers having lower scores than mothers at each time point (F= 11.27, p < 0.00) (Table 2).

Table 2.

PSE Mean Scores by Parent (n=27)

			Parent
	Event		Mothers		Fathers
	M	SD	M	SD	M	SD
Enrollment (n=27)	38.0	4.06	38.2	6.1	37.8	3.5
Discharge (n=26)	39.5	4.03	39.6	3.5	39.5	4.2
3 Months (n=24)	41.3	3.25	42.0	3.2	40.6	3.6

In the first regression model of paternal and social characteristics (Table 3), Hispanic ethnicity predicted lower PSE scores (p = 0.01). Fathers who were partnered (e.g., married or living with partner) (p = 0.02) and those fathers with better psychological well-being scores (p = 0.01) were predictive of higher PSE scores, with father demographic and social characteristics accounting for 48% of the variance in PSE scores.

Table 3.

Multivariable regression analysis of Father PSE, paternal and infant characteristics

Model 1: Paternal	Coef	p
Intercept	36.7	.000*
Ethnicity (Hispanic vs. Non-Hispanic)	−5.16	.01*
Partner status (single vs. partnered)	4.1	.02*
Psychological well-being	.13	.04*
Model 2: Infant	Coef	p
Intercept	39.8	.01*
Chronic Conditions (total number)	−1.21	.01*
Medications (total number)	−0.30	.02*
Illness severity (TDS score)	−0.16	.00*

^*Significant at p<.05

In the second regression model of infant characteristics (Table 3), total number of chronic conditions (p = 0.01), medications at discharge (p = 0.02), and greater illness severity (p = 0.01) was predictive of lower PSE scores in fathers. Infant characteristics accounted for 33% of the variance in PSE scores.

Father PSE Patterns

Each father’s PSE trajectory line was graphed and categorized into the following patterns: (1) Increasing, (2) Stable (high or low), (3) Mixed. Categorization was determined by using the clinical cut off index score of 39. Our definitions for each pattern were as follows: increasing pattern had an increase in scores crossing the cut off, (2) Stable patterns had either high or low scores that were consistently above or below the cut off, (3) mixed patterns had scores that were not consistent overtime. The majority of fathers had an increasing PSE trajectory over time (n=13). Several PSE trajectories were stable over time (stable high, n = 6 vs. stable low, n=4). The remaining fathers had a mixed trajectory (n = 4). Figure 1 provides example trajectories identified for each pattern.

Fig. 1.

Example patterns of paternal self-efficacy.

Discussion

Our findings revealed that Father’s PSE scores increased overtime but when compared to their partners, fathers reported significantly lower PSE scores. Most fathers had scores in the clinical risk range, indicating the need for greater support in their parenting roles. Factors associated with higher PSE in fathers were greater psychological well-being and living with or married to a partner. Fathers of ethnic minority backgrounds demonstrated lower confidence as it relates to caring for their infant with complex care needs. To date, this is the first study using the KPCS measure to document PSE in fathers.

From available data on PSE in fathers and mothers, our results are consistent with previous findings that fathers report an increase in confidence over time, but their scores are often lower than mothers (Hudson et al., 2001). Additionally, the factors associated with higher PSE is different for mothers and fathers. Compared to previously published findings of mothers from the same study, ethnicity and infant characteristics such as greater illness severity and number of chronic conditions was predictive of lower PSE in fathers, but not for mothers (Vance, Pan, et al., 2020). However, better psychological well-being contributed to higher PSE in both fathers and mothers (Vance, Pan, et al., 2020). Of note, our previous findings indicated that marital status (e.g., being married or living with a partner) was predictive of lower confidence in mothers but for fathers it was predictive of better confidence, which is a new finding not previously identified (Murdock, 2013; Trahan, 2018). This finding could suggest that mothers may be more sensitive to dissatisfaction or lack of support perceived in the marital relationship as compared to fathers. Fathers may also benefit from the marital relationship as they watch the mother engage in competent caregiving, which contributes to their confidence. There is evidence that married fathers are healthier and report fewer mental health issues (Meadows, 2009); thus, fathers may experience additional or protective benefits from a marital relationship that is different from mothers. More evidence is needed regarding the quality of the martial relationship and satisfaction with partner support to elucidate how partnered status contributes to overall confidence in the parenting role.

Fathers of ethnic minority backgrounds demonstrated lower confidence as it relates to caring for their infant with complex care needs. Yet while confidence is associated with paternal involvement (Trahan, 2018), paternal involvement has not been linked to any demographic variables (e.g. race/ethnicity, income) (Shorey et al., 2019; Trahan, 2018). However, the samples from these studies were racially and ethnically homogenous. Consequently, fathers reporting less confidence in the care of their child may be due to fewer opportunities to engaged with the healthcare team and master skills to care for their child in the hospital setting related to increased job or family.

In an effort to support fathers and bridge the gap in providing opportunities for interaction, clinicians can focus on ways fathers can engage in decision making and care planning. Positive interaction with healthcare providers is associated with paternal involvement in care which promotes child wellbeing (Garfield & Isacco Iii, 2012). In one study of paternal confidence, higher scores were associated with control, meaning that fathers may perceive their ability to control their child’s behavior and engage with their child as an indication of their success as a parent (Murdock, 2013). Within the hospital setting, often parents feel as though they have limited control in the care of their infant. Nurses are uniquely positioned to help fathers feel welcome and validated in providing care in the hospital environment, which may contribute to a sense of control. Engagement of fathers can also help to reduce feelings of uncertainty and develop confidence through involvement in care (Hearn et al., 2019; Koliouli & Zaouche Gaudron, 2018). Future research must focus on parenting competence throughout infancy, starting in the hospital setting, as their infant’s illness trajectory may be uncertain.

Overall, paternal PSE scores did increase over time as commonly seen in mothers during the transition to parenthood (Pinto et al., 2016). The majority of fathers (70%) had increasing or stable high PSE trajectories, which is encouraging given the medical complexity of their infants. Even so, lower scores were associated with greater caregiving complexity, suggesting fathers may need more time and multiple opportunities for interaction to support paternal confidence. Parents often rise to meet the challenges presented and can provide the necessary care for their infant should their needs demand it. In light of this, more evidence related to the quality of parenting behavior (e.g., competence) would help identify how parents actually provide care for their child and their ability to create an environment that fosters growth and development, for their child and for themselves as a parent.

Our analysis was limited by the small sample size, decreasing power needed to detect moderate to large population effects. While the father’s perspective is needed in future research, fathers are often difficult to recruit for research given multiple competing demands. Recruitment of fathers for this study was impacted by whether the mother provided contact information. Mothers were approached at the hospital bedside and often fathers were not present during the consenting process. Of those fathers who were approached to consent for the study, all but one agreed to participate, signifying fathers are willing to participate in research. The hurdle in recruiting and contacting fathers directly to introduce the study and consent remains. Nevertheless, the literature is scant on large samples of fathers; small sample sizes are common in studies of fathers (Fong et al., 2018; Murdock, 2013). Future studies should look to oversample fathers to provide adequate power to detect changes. The lack of diversity within this sample limits the generalizability of the findings. Most fathers sample were married, white couples, even though partnered status was not considered for study eligibility. This may also be the result of selection bias, as those who were available to participate had more flexibility regarding job demand and ability to be present in the hospital. Future studies should consider recruitment strategies within the community and/or via alternative methods such as social media and father support groups to enhance our understanding of the diverse experiences of parental self-efficacy for fathers and mothers.

Conclusion

This study offers clinical implications for fathers who have an infant with complex care needs. Fathers self-efficacy scores were consistently at or below the clinical risk range indicating low confidence in their parental role. Thus, supportive parenting resources and opportunities for engagement need to be provided during hospitalization and fathers welcomed into care management and discharge planning. Specifically, if nurses and other clinicians engage fathers during a healthcare interaction, it may support the development of confidence in their parenting role. Paternal involvement has positive impacts on the health of the family, especially those of infants with complex care needs and identifying ways to support parental involvement and parental self-efficacy is critical.