Impact of the Family-Centered Early Intervention Program Good Start on Psychosocial Constructs Related to Quality of Life in Parents of Children with Cerebral Palsy: A Pre-Post-Follow-Up Study

Aim

The aim of the current study was to evaluate the impact of the intensive four-day, family-centered early intervention program Good Start on psychosocial constructs related to QoL, including well-being, life satisfaction, sense of coherence, efficacy in parenting, and family empowerment, among parents of children with CP.

Study design

A pre-post-follow-up intervention study was used to evaluate the impact over time of the early intervention program Good Start among parents of children with CP, from baseline (before participation) to post-program at three weeks and follow-up at six months after the program. The overall evaluation employed a combination of quantitative and qualitative methods, including surveys, observations, and interviews. Results from the observations and interviews are published elsewhere (Skovgaard et al., submitted). A timeline of the evaluation process is presented in Figure 1. The study is reported with reference to the STROBE checklist for cohort studies (STrengthening the Reporting of OBservational studies in Epidemiology) (Von Elm et al. 2007), although with adaptations of the headings to suit the reporting of this study.

Figure 1

Context – Study setting

Good Start is a four-day, family-centered intervention program that targets families of children aged six months to four years diagnosed with CP. It has been developed and is administered by the Elsass Foundation, a privately funded nonprofit organization in Charlottenlund, Denmark, which provides the entire program at no cost to the families. The intensive nature of the program is unique, and families can find and enroll in the program themselves without assistance from the public healthcare system. During the four-day program, families reside in housing provided by the Elsass Foundation. This includes parents, the child with CP, siblings, and extra babysitter(s), who are typically grandparents. Prior to residency, a telephone conversation takes place between the Elsass Foundation and one or both parents to identify and assess activity-related challenges and goals. Approximately three months after the program, the Elsass Foundation initiates a follow-up telephone contact with the family. The entire intervention therefore encompasses the initial call to the follow-up contact three months later, with the primary focus being the four-day program at the Elsass Foundation (Figure 1). Families are assigned to an interdisciplinary team comprising a physiotherapist, an occupational therapist, and a psychologist. Together with the parents, this team tailors specific interventions for each family. The intervention includes presentations, individually tailored training activities, and networking events. The intervention is conducted alongside up to five other families approximately matched based on their children’s levels of gross motor function. Any training activities emphasize the integration of training tasks into the family’s daily routines through play and the child’s everyday activities. The activities are designed to address the specific desires and needs of each family, which are either communicated before the program or emerge during the course of the program (Johansen et al. 2023; Præst and Larsen 2023).

Evaluation design

The design of the evaluation was based on a logic model that was collaboratively developed with the personnel of the Elsass Foundation during two workshops in the spring of 2021. A logic model serves as a methodological tool that provides a structured overview of how an intervention is expected to lead to a sequence of outcomes for a specific target group (Funnell and Rogers 2011). The model illustrates the proposed interrelatedness of psychosocial constructs of quality of life, including well-being, life satisfaction, sense of coherence, parenting efficacy, and family empowerment within the family. These constructs capture both the internal (emotional, psychological) and relational (intra-family social) aspects of parental experience (Irwin, Jesmont and Basu 2019), but do not include the physical health domain. During the formulation of the logic model, discussions and identifications were undertaken regarding the concrete activities of the intervention, the expected immediate, intermediate, and ultimate outcomes, as well as the target audience for the intervention. In addition, a set of key assumptions concerning the relationships between activities and presumed outcomes in the logic model were outlined. These provided guidance for the specific selection of the means for outcome measurement. Further information regarding the logic model can be found in Appendix 1 and elsewhere (Johansen et al. 2023; Præst and Larsen 2023).

Participants, data collection, and variables

This pre-post-follow-up study recruited participants from four Good Start programs between January 2021 and December 2022. The Elsass Foundation personnel randomly selected four out of eight possible Good Start programs to be included in the evaluation. All potential participants (38 parents) who were to take part in one of these four Good Start programs received a cover letter containing information about the evaluation study and a link to a web-based questionnaire, which was sent out along with the information about the program (with the option for inquiries in case of questions). If the web-based questionnaire had not been filled out prior to the start of the program, parents were also allowed to complete a paper-based questionnaire upon arriving at the four-day program. Three weeks and six months after attending the program, the parents received the same web-based questionnaire, excluding questions related to sociodemographic factors and information about the child. If the parents had not returned their questionnaire within approximately two weeks of the initial posting, up to two reminders were sent at intervals of approximately seven to 10 days.

Outcome measures

Statistics

Continuous variables were tested for normality (Shapiro-Wilk test, histogram, and quantile-quantile (Q-Q) plots). Descriptive statistics (mean and standard deviation) were calculated for all variables (Table 2). A linear mixed model was used to detect any overall differences between related means for each dependent variable at baseline, after three weeks, and after six months, with ‘subject’ and ‘family cluster’ as random effects. This model takes into consideration the interdependence between measurements and the unequal number of data points across time points. The dependent variables were treated as continuous variables. The independent variable ‘time’ consisted of three categorical, related groups. These analyses were performed and reported as unadjusted differences (beta coefficients with 95% confidence intervals (CIs)) at each time point (Table 3). The residuals of the linear mixed models were checked for normal distribution. The significance level was considered to be p < 0.05. To illustrate the predicted outcomes, we used margin plots for each outcome (Figures 2 and Appendix 1). All analyses were performed in the statistical software STATA BE 17 (StataCorp 2021).

Figure 2

Ethics

According to Danish legislation (LBK no. 1338 of September 1, 2020) (LBK 2020), formal ethical review was not mandated, as the study did not involve the collection of human biological material. Nonetheless, institutional approval for the study and its data management procedures was obtained from the Research and Innovation Organization (RIO) at the University of Southern Denmark (journal nr. 11.260). Participants gave informed consent to participate in accordance with the Declaration of Helsinki (Vollmann and Winau 1996; World Medical 2013). Expert and parent representatives were actively involved in the preparation of the information materials for the participating parents and in the pre-testing of the questionnaire. Participation in the study was voluntary, with informed consent obtained in electronic form from all participants prior to data collection. Participants and their families were fully informed about the study’s objectives, procedures, and their rights, including the option to withdraw at any stage without repercussions. Personal identifiers were removed or anonymized to ensure participant confidentiality, and all data were handled with strict attention to privacy and security.

Participant characteristics

In the context of our study, a total of 19 families, comprising 38 parents, participated in the four separate Good Start programs. The response rates for outcome measures at pre-program, post-program, and six-month follow-up assessments were 100% (n = 38), 97% (n = 37), and 92% (n = 35), respectively. Half of the parents who filled in the questionnaire were female, and all parent couples cohabited. In addition, 12 out of 19 families (63% of the sample) included siblings living at home with the child diagnosed with CP (not shown in table). The parents’ ages ranged from 24 to 50 years, with the majority being between 29 and 40 years old (84%). Furthermore, more than half of the parents (63%) reported having attained either a medium (bachelor-level) or long-term higher education (master’s-level or above), and 84% were employed, with 66% holding full-time positions (Table 1). Parents’ education was categorized according to the International Standard Classification of Education (ISCED) as short-cycle tertiary (e.g., post-secondary vocational programs), medium-cycle (bachelor’s level and professional bachelor’s level such as nursing or teaching), and long-cycle (master’s or doctoral level) education (Statistics 2012).

In the families, eight children with CP were girls (42%) and 11 were boys (58%), and the majority were aged 1 to 2 years (68%). A significant proportion of the children (58%) exhibited a moderate level of gross motor functional impairment due to their CP diagnosis, resulting in compromised balance and coordination and necessitating the use of assistive devices to support mobility (Table 1).

Changes in outcome measures

Linear mixed model analyses revealed significant effects of time for all outcome measures except for parenting efficacy and family empowerment in post-program compared to pre-program measures. Table 2 displays means, standard deviations, and range for outcome measures at pre- and post-program and at the six-month follow-up assessment.

At the six-month follow-up, a statistically significant effect of time was observed for all outcomes compared to pre-program measures, apart from family empowerment (Table 3). Among all outcomes, the percentages of change ranged from 0.3% to 13.9%, with three outcome measures exhibiting changes exceeding 10% (Table 3).

The margins plots presented in Figure 2 visually represent the results of the linear mixed model analysis by displaying the predicted mean scores for participants at each time point across the six outcome measures. As illustrated in Figure 2, there is a notable upward trend in the predicted mean scores for parenting well-being, life satisfaction, sense of coherence (downward trend), and the family situation. These trends indicate significant improvements from baseline to post-program, and these gains are sustained at the six-month follow-up. For parenting efficacy, while the improvement was not statistically significant from pre- to post-program, it became statistically significant at the six-month follow-up. However, the Family Subscale shows no significant change, which indicates that family empowerment remained stable throughout the intervention program.

Well-being and life satisfaction

In one randomized early intervention trial for infants at very high risk of CP (Hielkema et al. 2020), the family-centered intervention program (COPCA) and the typical infant physiotherapy resulted in similar family outcomes. However, specific elements of the COPCA intervention, such as ‘caregiver coaching’, were associated with increased family empowerment and improved QoL.

Similar findings were reported in another intervention trial involving families with children with CP aged three to six years (Myrhaug et al. 2018). In this study, parents received three weeks of conductive education in addition to conventional practice, and this was compared to conventional practice alone. Although there were no significant differences in QoL scores between the groups, the intervention group showed notable improvements. Specifically, the percentage change in QoL scores was ~12%, which indicates positive outcomes consistent with those observed in the ‘Good Start’ program. The results from the Good Start program ranged from 6% to 14%, depending on which aspect of QoL was being measured (well-being, life satisfaction), which is likely to be of clinical relevance. One meta-analysis of a limited number of randomized controlled trials (Irwin, Jesmont and Basu 2019) aiming to improve caregiver well-being (anxiety, competence, stress, etc.) through educational, psychological, or supportive interventions indicates that such interventions can enhance the well-being of parents of children with CP, which supports the results of the present study.

Sense of coherence

In our study, we found short- and long-term improvements in relation to coherence (12–13%) and to the family situation (9–12%) after attending Good Start. The positive direction of these results is in line with those of a single one-year intensified habilitation program that included parents of preschool children with CP (Damgård et al. 2016). Although the sense of coherence and optimism scores did not change significantly during the intervention period in the study, many parents reported clinically relevant changes, with a minimum of 5% change. Specifically, 67% of the mothers who reported clinically relevant changes experienced increased optimism (Damgård et al. 2016).

Parenting efficacy

The findings of sustained improvement in parenting efficacy are consistent with previous research, as shown by pre- and post-intervention measures from an early childhood intervention program lasting more than one year (Hughes-Scholes and Gavidia-Payne 2019) and one meta-analysis of intervention studies on parent training programs, which identified a moderate positive effect size in enhancing parenting efficacy (Hohlfeld, Harty and Engel 2018). Although interventions and measures of efficacy differed between these studies, both included caregivers of young children with various neurodevelopmental disabilities, with a minor focus specifically on CP. As no established minimal clinically important difference (MCID) exists for the Efficacy Subscale, we cannot determine whether the observed changes are clinically meaningful.

Family empowerment

Only limited knowledge exists regarding the trajectories of parenting empowerment over time among parents of children with CP who attend early intervention programs. A cohort study of 58 families (Kalleson, Jahnsen and Ostensjo 2019) with children aged four years and under, who were offered follow-up by the Norwegian Cerebral Palsy Follow-up Program over three years and received support from a multidisciplinary team, a coordinator, training, and/or an intensive rehabilitation program, found consistently high scores (with mean scores above 4 out of 5) in the Family Subscale measuring empowerment during the daily management of home activities. This result is similar to the findings of the present study, which indicates robust family capacity, with most parents having further education, being employed, and living together as couples. However, previous intervention studies have reported modest increases in family empowerment: ~8% in a single-group early family-centered intervention study conducted over a two-year period in India involving children with developmental delays (Muthukaruppan et al. 2022), and ~5–8% in a larger feasibility study of the ENVISAGE program, which involved continuous workshops over a five-week period for caregivers of children with neurodisabilities (Miller et al. 2023). Notably, there was variation in baseline levels of empowerment among participants across studies.

The logic model underpinning the Good Start program provides a theoretical foundation for understanding how various psychosocial constructs, such as parental well-being, life satisfaction, sense of coherence, parenting efficacy, and family empowerment, are interrelated in parents of children with CP. While definitive causal pathways cannot be claimed, the model suggests plausible connections supported by prior research. As parents gain knowledge, emotional support, and confidence, their perceived sense of coherence, parenting efficacy, and sense of empowerment are strengthened (Damgård et al. 2016; Hohlfeld, Harty and Engel 2018; Miller et al. 2023). These factors, in turn, have been linked to improvements in both individual well-being and family quality of life (Eriksson and Mittelmark 2017; Irwin, Jesmont and Basu 2019; Kalleson, Jahnsen and Ostensjo 2019). Importantly, these constructs appear to reinforce one another in a dynamic process, rather than operate in isolation.

The overall assumption of a perceived improvement in participating parents’ sense of competence and QoL after attending the Good Start program, which was stated in the logic model, is supported by qualitative findings (Skovgaard et al., submitted).

Study comparability and implications

The comparability of studies investigating the changes in psychosocial constructs related to QoL arising from family-centered care interventions in families with children with CP is limited due to variations in intervention durations, study designs, measurement instruments, follow-up periods, and cultural contexts. Despite these differences, the findings reviewed above demonstrate similar trends. The addition of a short yet intensive intervention to this body of research, such as the Good Start program, adds to the growing evidence of the positive impacts of a family-centered care approach on parents of young children with disabilities. This underscores its relevance in clinical practice and supports its consideration in informing best practices and policy decisions in early intervention for families of children with CP.

Strengths and limitations

Several limitations must be acknowledged that may affect the interpretation of the results of the current study. Firstly, potential influences of parental gender and age were not investigated, which leaves open the possibility that these factors may impact the perceived effects of the intervention. Secondly, the lack of a control group comparison prevents us from ruling out that the observed changes may be attributed to general developmental trends or external factors; however, it is also conceivable that in the absence of intervention, parenting outcomes, such as well-being or efficacy, could have worsened over time due to ongoing caregiving challenges. Thirdly, despite utilizing standardized outcome measures previously employed with parents of children with various disabilities, these measures have not been specifically validated for the population under study, and no statistical power calculations were performed as part of the study design. Fourth, the relatively small sample size combined with the number of outcomes tested is a limitation of this study, increasing the risk of spurious positive findings. The strengths of this study lie in its prospective design, which captures observations and reduces bias within a usual practice context, thereby enhancing the real-world relevance of the findings (Concato, Shah and Horwitz 2000). Furthermore, the development of a logic model to guide the selection of outcome assessment instruments enhances transparency and ensures that they align with the objectives of the Good Start program.