CMPH Journal

Introduction

Over 50% of healthcare providers experience symptoms related to burnout, which is significantly higher than the general population [1,2]. One indication of burnout is high turnover. Researchers reported that professionals working in jobs requiring in-home visits tend to demonstrate a high turnover rate [3-5]. Additionally, professionals with high turnover are those who experience secondary traumatic stress or compassion fatigue when working with families with a history of significant trauma [6,7]. Gill and colleagues, and Lee and colleagues demonstrated high turnover rates in home visitors in early childhood education and prevention programs [4,5], while West et al. associated a high turnover rate due to burnout resulting from emotional exhaustion associated with feelings of hopelessness, anger, frustration, and difficulties in coping with or performing one’s job effectively [8].

In paediatrics, providers in early intervention (E.I.) programs in the United States work in clients’ homes as defined by Part C of the Individuals with Disabilities Education Act (IDEA) [9]. The Part C program outlines the services provided to young children who face various risk factors to their health or well-being, such as developmental delays and disabilities and their families [9]. E.I. practitioners focus on empowering and educating the family to be experts on their child and make decisions according to the child’s needs and characteristics [10]. Due to the work environment the E.I. provides may be at risk for burnout and experiencing compassion fatigue due to the family’s life and events [8].

Burnout is essential to investigate due to its link with physical health impairments including muscle pain, headache, insomnia, respiratory illnesses, and gastrointestinal disorders [11]. Although burnout may occur in any job, it is most notable in the healthcare/ helping professions [12-14]. The World Health Organization in 2019 added burnout to the International Classification of Diseases, characterizing it as a “syndrome of three dimensions: feelings of energy depletion or exhaustion, increased mental distance from one’s job or feelings of cynicism or negativism about one’s job, and reduced professional efficacy” [15].

Burnout consists of three core aspects: emotional exhaustion, depersonalization, and feelings of reduced personal accomplishment. Burnout is a complex process, developing gradually, with symptoms ranging from almost none to severe [16]. As E.I.  providers work in families’ homes, they often take on the stress and concern of the families adding to the provider’s emotional exhaustion.

Considering E.I. providers engage with families in their natural environment, they often encounter families experiencing multiple stressors like families dealing with children challenged by disabilities, mental health issues, substance abuse, and poverty [17]. Alitz, et al. documented that the delivery of evidence-based home visiting models facilitates positive outcomes among families and their children, however home visitors experienced increased stress when they were unable to meet clients’ needs [3]. These home environments increase E.I. providers’ risk of developing burnout and compassion fatigue due to the prolonged exposure of witnessing and empathizing with the experiences of children and their families [18].

Sorenson et al. associated compassion fatigue in providers with behavioral, physical, and emotional changes such as loss of empathy and objectivity, sadness and depression, anxiety, sleeplessness, nightmares, chronic lateness, substance abuse, sleep and eating disturbances, disinterest, helplessness, exhaustion, recurring unwanted thoughts, and avoidance [18]. Additional risk factors for the development of compassion fatigue include traumas and Adverse Childhood Experiences (ACEs), role ambiguity, inability to establish boundaries between work and personal life, and job demands [8,19]. Dauber et al. and Harden et al. reported home providers in Head Start and nursing demonstrated elevated levels of compassion fatigue that can compromise the quality-ofcare families receive due to disruptions in relationships between the providers and families [20,21].

In addition to compassion fatigue, Begic et al. reported several other job-related factors to exacerbate the risk for burnout among home visitors [19]. These include low job satisfaction and pay levels, lack of organizational structure, supervisor and co-worker support, and lack of autonomy and flexibility, leading to unhealthy work environments. Compassion fatigue occurs due to prolonged exposure to witnessing and empathizing with the traumatic experiences of clients and their families [22].

Sacco & Copel determined those providers in helping professions also experience compassion satisfaction [23]. Compassion satisfaction occurs when the care provider feels a sense of connection with his/her patients and a sense of achievement in his/her work [24]. This satisfaction in helping children and their families is often why providers work in E.I. Burnett & Wahl reported compassion satisfaction is inversely related to compassion fatigue and may effectively reduce provider burnout [25]. Ross et al. published findings indicating that home visitors reporting higher satisfaction levels are more likely to experience higher levels of compassion satisfaction and lower levels of burnout and compassion fatigue [22].

Maslach, Jackson, and Leiter reported burnout symptoms seem more prevalent in the helping professions, such as physicians and teachers and linked burnout to emotional exhaustion [14]. Individuals experiencing emotional exhaustion feel overextended and are unable to offer emotional support to others. Individuals with emotional exhaustion often transition to a cycle of depersonalization, or cynicism, referring to the negative and cynical attitudes toward work, leading to impersonal response toward recipients of one’s care and a reduced sense of personal accomplishment [26-28]. Maslach defined a reduced sense of personal accomplishment as a negative view of oneself, mainly related to one’s work [26]. As providers get to the very low point of a negative view of oneself at work, emotional exhaustion, burnout, and compassion fatigue can overwhelm the provider placing them at risk of leaving their work and placing the need to support these essential healthcare workers on administrators.

The burnout rate among those in the health care/helping professions is reported at medium to high levels, and the burnout risk in health care providers is higher than in the general working population [29]. Researchers investigated burnout, depression, compassion satisfaction, and compassion fatigue in healthcare fields, including physicians, nurses, and trauma therapists [8,18]. However, more evidence of burnout and compassion fatigue needs to be documented concerning home visitors. Krabn, et al. determined that E.I. workers were at risk of burnout; however, no existing studies documented current burnout trends, depression, compassion satisfaction, and compassion fatigue in E.I. providers [30]. Therefore, this study seeks to determine the current state of burnout, depression, compassion satisfaction, and compassion fatigue among E.I. providers.

Study Design

This study was an exploratory cross-sectional survey to determine current levels of burnout, depression, compassion satisfaction, and compassion fatigue in E.I. providers in the United States. Institutional review board approval from the University of Oklahoma Health Sciences was obtained.

Participants

Early Intervention providers in the United States. The inclusion criteria were E.I. providers working in their state’s Part C program and completing the survey. Exclusion criteria include providers not explicitly working in their state’s E.I. program.

Measures

Participants completed a demographic survey exploring their training, years of experience, and workloads.

Maslach Burnout Inventory – Human Services Survey (MBIHSS)

The MBI-HSS is a self-administered, validated 22-question survey that assesses the three significant areas of burnout, used extensively in healthcare fields [31,32]. The questionnaire uses a 0 to 6 Likert-type scale to assess each burnout area. An elevated level of burnout in each area consists of a score of 27 or higher in the emotional exhaustion subscale, a score of 10 or higher in the depersonalization subscale, and a score of 33 or lower in the personal accomplishment subscale [14,33].

Reliability of the MBI-HSS using Cronbach’s coefficient alpha resulted in reliability for the total scale as 0.87; 0.85 for emotional exhaustion, 0.80 for depersonalization, and 0.75 for personal accomplishment [34]. Pérez-Mármol and Brown demonstrated scalability, structural validity, and reliability of the MBI-HSS in conjunction with Rasch Measurement Model requirements [35].

Patient Health Questionnaire (PHQ-2)

The PHQ-2 is a depression screen with good construct and criterion validity, given that its discriminant validity was excellent for the PHQ-2 compared to the PHQ-9, a more extended survey [36]. The PHQ-2 is a valid tool to assess depression diagnosis, severity, and outcome [37]. Comprehensive assessment established the reliability, construct and criterion validity, and sensitivity to change of the PHQ-2 [38].

Professional Quality of Life Survey for Health Care Workers (ProQOL)

ProQOL is a 30-item measure to determine the negative and positive effects of helping others who experience suffering and trauma. ProQOL has subscales for burnout, compassion satisfaction, and compassion fatigue [24]. The measures Cronbach’s alpha of 0.738, meaning the questionnaire demonstrates internal reliability and good construct validity. The ProQOL assesses three subscales: burnout, compassion satisfaction, and compassion fatigue. Cronbach’s alpha revealed good internal reliability for the subscales: burnout α=0.805; compassion satisfaction α=0.909; and compassion fatigue α=0.797 [39-41]. Regarding validity, the ProQOL demonstrated excellent person- and item-fit characteristics and fulfilled the strict requirements of Rasch measurement models [42].

Procedures

We used snowball convenience sampling to gather participants through recruitment using survey distribution via Facebook groups,

E.I. listservs from the American Physical Therapy Association (APTA), the American Occupational Therapy Association (AOTA), the American Speech and Hearing Association (ASHA), and emails to E.I. leaders and providers across the nation. The invitations to participate included a link and Q.R. code to the survey. The survey included demographics and MBI-HSS, PHQ2, and ProQOL. Participant completion of the electronic survey served as the participants’ consent.

Data Analysis

MBI-HSS burnout, PHQ-2, and ProQOL were scored by summing all scores across subjects. Initially, there were n=244 responses to the survey, but some individuals had single items from each measure missing, so this effectively reduced the total sample size per outcome to MBI burnout (n=235), PHQ-2 (n=243), and ProQOL (n=233). No imputation was considered, and missingness was not assumed to be random.

Descriptive statistics were computed using E.I. provider discipline for each subscale scores from the MBI-HSS, including emotional exhaustion, depersonalization, personal accomplishment, and from the ProQOL including burnout, compassion satisfaction, and compassion fatigue. A backward selection process was used to arrive at the ProQOL final subscale scores to determine burnout, compassion satisfaction, and compassion fatigue. Normality was assessed for each subscale score among professions. None of the subscale scores were normally distributed among the professions. Nonparametric Kruskal Wallis tests were used to determine if each subscale score differed among the professions. Correlation analysis was conducted among all subscale scores using the entire sample to determine whether subscale scores for MBI-HSS correlated to subscale scores of the ProQOL. Each subscale score was tested for normality among the entire sample. Even among the sample, all subscale scores were not normally distributed. A Spearman correlation coefficient matrix was created for all six subscale scores to conduct correlation analysis.

Results

Participants

Two hundred forty-four E.I. providers completed the survey from October 23, 2023, to December 31, 2023. The responses from participants included 16.6% Physical Therapists (PT), 12% Occupational Therapists (OT), 33.2% Speech-Language Pathologists (SLP), 5.8% Service Coordinators (SC), 18.7% Special Instructors/early interventionists/developmentalists (SI), and 13.7% vision specialists and nutritionists (Other). One participant identified themselves as an occupational therapist and a service coordinator. Since our number of service coordinators was small, we used the occupational therapist as the provider for this individual. Providers from 41 states responded that Montana, Wyoming, Vermont, West Virginia, Alaska, Arkansas, Hawaii, Mississippi, and New Hampshire were not represented. Nineteen percent of the E.I. providers had a fixed schedule, and 81% had a flexible schedule and covariate variables. The mean years working in their profession was 18.4 years (SD=11.5). The range of providers years in E.I. was 0-51 years, with the average working 35.6 hours per week and a range of 2-100 children on their caseload. Table 1 provides descriptive statistics, variables, frequency, and percentages for the participants.

Table 1: Descriptive statistics of Categorical Variables, Frequency and Percentages.

There was no significance in the relationships between the MBI-HSS with years in profession, years in early intervention, provider discipline, regions, or whether the provider had a fixed or flexible schedule. There was no significant relationship to the PHQ-2 to the years in profession, years in early intervention, number of hours in a typical week, number of children in workload, provider discipline, regions, or whether the provider had a fixed or flexible schedule. The ProQOL showed no significance for years in profession, years in early intervention, number of days worked per week, number of hours in a typical week, number of children in workload, provider discipline, regions, or whether the provider had a fixed or flexible schedule. Table 2 provides the descriptive statistics for the outcome measures, including the total score data for the MBI-HSS, PDQ-2, and the ProQOL measures.

Table 2: Descriptive Statistics for Outcome Measures.

The MBI-HSS demonstrated significance with the number of days worked per week and number of hours in a typical week. Number of hours worked each week (R=0.27, p<0.0001) and the number of days worked (R=0.23, p=0.0003) were associated with MBI-HSS burnout score univariately. The only outcome with multiple variables associated with the outcomes was the MBI-HSS burnout score. On the PDQ-2, the number of hours worked each week (Spearman =0.18, p=0.0042) was the only variable associated with the measure. Provider type was the only variable associated with ProQOL (F=2.48, p=0.0330) measure in the univariate associations. Table 3 illustrates these relationships.

Table 3: Covariant relationships to the outcome measures of MBI-HSS, PDQ-2 and ProQOL (Bold=P>.05).

MBI-HSS Average Subscale Scores

The participants' average scores for each of the subscales for the MBI-HSS are in Figure 1. Frequency scores from a general population of 11,000+ people in the human services professions are included for comparison [43]. Higher emotional exhaustion and depersonalization contribute to burnout, while higher personal accomplishment reduces burnout. Our participants scored higher average emotional exhaustion than the general population. High emotional exhaustion is indicated by 46.38% of participants, and high depersonalization was scored by 16.6%, which would increase the burnout rate, compared to 20.85% of participants who scored high personal accomplishment.

Figure 1: Average subscale scores on MBI-HSS compared to the general population.

MBI-HSS Profiles

Maslach and Leiter compiled profile analyses with the MBI-HSS scales to determine patterns in MBHSS scores [44]. Figure 2 illustrates the profiles based on cut-off scores of the MBI-HSS subscale scores. Using the subscales' scores, the profiles categorize exhaustion, cynicism, and efficacy and their relationship to one another. The profiles are engaged, ineffective, overextended, disengaged, and burnout. People with different burnout profiles have different workplace experiences. Twenty percent of participants were categorized by the MBI-HSS profiles as engaged, 29% were categorized as ineffective profile, 31% fell into the overextended profile, less than .02% were disengaged, and 14% were profiled as burnout. Depression was present in 19.9%, or 48 out of 241 early intervention providers. Table 4 contains subscales and descriptive statistics by MBI-HSS and participants’ discipline.

Figure 2: Profiles and the Impact of the MBI-HSS Subscale Scores.

Table 4: Subscales Descriptive Statistics by MBI-HSS and Professions.

Figure 3: MBI-HSS Subscales of Depersonalization, Emotional exhaustion, and Personal accomplishment.

PHQ-2

On the PHQ-2 Depression scale, only the number of days worked per week indicated significance for this measure (p= 0.0042. Table 3). Our results did indicate that depression (Total PHQ2 equal to or greater than 3) was present in 19.9% or 48 out of 241 E.I. providers.

ProQOL Average Subscale Scores

The ProQOL showed no significant difference for participants' years in the profession, years in E.I., number of days worked per week, number of hours in a typical week, number of children in workload, provider discipline, regions, or whether the provider had a fixed or flexible schedule. Speech language pathologists (SLP) showed the lowest mean ProQOL scores. Special instructors, physical therapists and the other category of E.I. providers were not significantly different from speech language pathologists in ProQOL scores. Service coordinators were 5.8 (95% CI: 1.0, 10.6) points higher than speech language pathologists (t=2.36, p=0.0191). Occupational therapists were 5.3 (95% CI: 1.8, 8.9) points higher on the ProQOL than SLPs (t=2.93, p=0.0038). Table 5 indicates the ProQOL descriptive statistics by subscale and discipline.

Table 5: ProQOL Descriptive Statistics by Subscales and Profession.

The results indicate medium burnout and medium compassion satisfaction scores for all E.I. providers based on cut-off scores according to the ProQOL manual [24]. The ProQOL measure determined that 83% of the providers showed medium levels of burnout, and 17% showed low levels of burnout. The providers in this study showed a medium-elevated level of compassion satisfaction. None of our providers scored a low level of compassion satisfaction. Forty percent of the providers showed high levels of compassion satisfaction, and 60% of our participants scored medium levels of compassion satisfaction. The ProQOL measure provided evidence that 44% of our participants had low levels of compassion fatigue, while 55% had medium levels of compassion fatigue, and less than 1% had high levels of compassion fatigue.

Correlation of MBI-HSS and ProQOL

Subscale scores among the same measure (i.e., subscale scores within the MBI or subscale scores within the ProQOL) were correlated. However, subscale scores across measures (i.e., MBI vs ProQOL) were not correlated (Figure 4).

Figure 4: ProQOL Subscales of Compassion Satisfaction, Burnout and Compassion Fatigue.

Discussion

This study aimed to explore the current state of burnout, depression, compassion satisfaction, and compassion fatigue among E.I. providers. The results indicate many E.I. providers identified with traits of burnout or had traits that lead to burnout and almost twenty percent of the participants were depressed. Compassion satisfaction, which acts in a protective nature to decrease burnout, was rated as medium or high for everyone in the study, while compassion fatigue, which can exacerbate burnout, was rated as low or medium by study participants. While our two primary measures of burnout did not correlate, the results became more similar when accounting for burnout risk factors on the MBI-HSS summary (Figure 5).

Figure 5: Comparison of current study data to Mukherjee et al. results.

The E.I. provider profiles on the MBI-HSS profiles were similar to Mukherjee et al. [45], recent investigation of health care workers in underserved areas during COVID-19. Mukherjee et al. study included 655 primarily Caucasian women health care providers, aged 50 years and younger, who worked in inpatient wards, emergency departments, or intensive care units. Even though the work environments were different, the need for compassion and caring are similar between the two studies [45].