This study was approved by the Research Ethics Boards of the Université du Québec en Outaouais (#2021-1167), the CIUSSS de la Capitale-Nationale (#MP-13-2021-2135). A cross-sectional design was used to recruit a convenience sample of French-speaking Canadian adult mental health workers across Quebec (Canada). To be eligible, professionals had to meet the following criteria: 1) work at the time of the survey with adult with mental illness (under 65 years old), 2) have direct contact with clients, and 3) be fluent in French. Those working exclusively with children were excluded.

From March to December 2021, all eligible professionals were invited to read a detailed description of the study, sign a consent form and respond to an online questionnaire via a LimeSurvey platform. Following Fan and Yan (54) recommendations, various recruitment strategies were employed to ensure a satisfactory participation rate: 1) emails to professional groups and orders, 2) emails sent through managers of all integrated health and social services centers (CISSS) from each of the 17 administrative regions of Quebec, medical clinics and community organizations offering adult mental health services and 3) promotion on social networks and local newspapers. Presentations conducted by the principal investigator were delivered at targeted workplaces (13 CISSS) to explain the project and to seek assistance with the recruitment of professionals. To encourage participation, compensation prizes were drawn among study participants, with one $25 prize per 50 participants. A total of 512 French-speaking adult mental health workers completed the questionnaire.

The study collected sociodemographic and occupational information such as gender, age, ethnicity, years of experience in mental health, workplace and location of services, employment status (full-time or part-time), percentage of parents with mental illness on caseload (low: ≥ 20%, moderate: 21-50%, high: >50%), level of education, personal experience of mental illness (yes or no), level of comfort around children.

The French version of the Family-Focused Mental Health Practice Questionnaire (FFMHPQ-FR) was used to collect self-report data on mental health professional’s FFP and related factors. The original questionnaire consisted of 45 items rated on a seven-point Likert scale (1=strongly disagree to 7=strongly agree) and demonstrated good psychometric properties for 13 of the 16 subscales (Cronbach’s alphas = 0.61 to 0.89) (55). The development of the FFMHPQ-FR underwent a rigorous process, as detailed in Piché et al. (56). It was initially translated into French through a back-translation procedure (57) and was subsequently adapted to improve its psychometric proprieties (i.e. seven items added to the subscales with limited consistency) and to consider recent literature (i.e. three item added on attitudes and beliefs toward FFP) (36). Finally, it was validated through both exploratory and confirmatory factor analysis, resulting in 42 items grouped into five subscales (Family-Focused Practices; Workplace support for FFP; Skills, knowledge and confidence; Openness to improve practice; Attitudes and beliefs toward FFP), with good psychometric properties (α = 0.61 to 0.89) ( Table 1 ) (56). Scores are calculated by averaging the items included within each subscale for every participant.

Across the 42 items of the FFMHPQ-FR, the rate of valid data amounted to 90.8%. Participants were required to answer each question in the questionnaire, while also having the option to select the responses “refusal to answer” or “does not apply.” These response choices were coded as missing data and treated using mean imputation per item (58).

All statistical analyses were performed using SPSS Version 27.0 software. As a preliminary analysis, the following assumptions were tested: sample size, multicollinearity, outliers and normality. Outliers were treated by winsorizing 2^nd and 98^th percentile. Descriptives (means, standard deviations) were calculated for all variables. Analyses of variance (ANOVA) and chi-square tests were conducted aiming to verify the relationship between mental health worker’s FFP and variables of interest in this study. Post-hoc analyses allowed to specify the nature of the differences between the groups.

A multinomial logistic regression analysis was performed to assess prediction of mental health worker’s FFP on the basis of ten factors: Workplace support for FFP subscale, employment status, percentage of parents with mental illness on caseload, years of experience in mental health, level of education, Skills, knowledge and confidence subscale, Openness to improve practice subscale, Attitudes and beliefs toward FFP subscale, level of comfort around children, and personal experience of mental illness. To perform this, scores from the FFP subscale were recoded into three categories: low (1.00-4.77), moderate (4.78-5.55) and high (5.56-7.00) score of FFP, based on percentile-based cut-points (respectively on the 33^rd and 67^th percentiles). Using a three-categorical variable will provide interpretable coefficients to quantify the relationship between predictors and the outcome variable. Thus, the high score category allows to discriminate professionals that provide more FFP to clients and their families. The results are presented as inverted odds ratio (IOR), easier to interpret with negative beta. The significance level in all analyses was .05.

The majority of participants were female (87.1%) and the average age was 40.34 years (range 19-75 years). Most were Caucasians (95.9%), born in Quebec (92.0%). The majority (85.7%) were university graduates, from various disciplines including social work (27.9%), nursing (20.9%), psychology (13.9%), psychoeducation (13.1%) and special education (8.0%). The average number of years’ of experience in mental health was 11.77 (range 1-40 years). Among the participants, 27% reported that more than half of their caseload included clients with mental illness who have at least one minor child. In the past five years, 43.3% had received family-focused training, while 34.8% had attended child-focused training.

Table 1 presents the average scores of participants on the FFMHPQ-FR’s subscales and items. The average score on the FFP subscale (M=5.13; SD=1.06) indicates a moderate level of FFP among mental health professionals in this sample. The highest item scores were reported on Identification of children, Evaluating parenting competencies and Information to parents, while the lowest were attributed to Information to children, Family meetings and Referrals to support services. Among the subscales, the participants scored the lowest on Workplace support for FFP (M=3.70; SD=1.13), with most items scoring under 4. The highest subscale scores were reported on the Attitudes and beliefs toward FFP (M=5.58; SD=1.04) and the Openness to improve FFP (M=6.00; SD=0.85).

Table 2 presents descriptive statistics of all variables, according to the three percentile-based categories of FFP score (low, 34.8%; moderate, 32.4%; high, 32.8%). Differences between groups for the following eight independent variables have been identified through post-hoc analyses: Workplace support for FFP, employment status, proportion of parental clientele, Skills, knowledge and confidence in FFP, Openness to improve FFP, Attitudes and beliefs toward FFP, highest academic degree, and years of experience in mental health.

The model including all ten predictors, when compared to a constant-only model, was statistically significant, χ2 (5, N=482) = 145.22, p<.001. This indicates that the predictors, as a set, significantly discriminate between low, moderate, and high score participants. The model explained 29.3% of the variance in FFP score, according to Nagelkerke R².

This model classified 53.3% of cases correctly, which is greater than the proportional by chance criterion of 41.7% ( Table 3 ). However, specific classification rates (low score, 66.1%; moderate score, 28.4%; high score, 64.2%) underline the model’s tendency to under classify the moderate FFP scores compared to low and high score categories.

Comparing low score FFP with high score FFP categories, it was found that the strongest predictors of reporting a high FFP score as a professional are employment status (full versus part time) (IOR=2.34), and skills, knowledge, and confidence toward FFP (high versus low score) (IOR=2.26) ( Table 4 ). This indicates that professionals employed full-time are 2.34 times more likely to report a high FFP score rather than a low FPP one, controlling for all other factors in the model. Likewise, for every point increase on the Skills, knowledge and confidence subscale, professionals are 2.26 times more likely to present a high FFP score. The perceived level of workplace support for FFP (high versus low) (IOR = 1.90), openness to improve FFP (high versus low) (IOR = 1.52), and attitudes and beliefs towards FFP (high versus low) (IOR =1.61) were also found as important predictors of FFP. Weaker predictions were found between high FFP and the proportion of caseloads with a parental status (high versus low) (IOR = 1.49) and the professional’s years of experience working in mental health (high versus low) (IOR = 1.04). The professional’s education (highest diploma), being comfortable with children, and personal experience with mental illness were not found to predict a high score of FFP compared with a low FFP score.

When comparing moderate FFP score with high FFP score categories, the strongest predictor was found to be the employment status of professionals (full-time versus part-time) (IOR=2.26). Thus, professionals who work full-time are 2.26 times more likely to have a high FFP score than a moderate one, controlling for all other factors in the model. The perceived skills, knowledge, and confidence in FFP (IOR=1.65) and attitudes and beliefs towards FFP (IOR=1.59) were also found to be important predictors of FFP. This indicates that for every point increase on the Perceived skills, knowledge and confidence and the Attitudes and beliefs toward FFP subscale, professionals are respectively 1.65 times and 1.59 times more likely to have a high FFP score rather than a moderate one. None of the other factors were found to be associated with a high score of FFP compared with a moderate FFP score.

The study’s findings shed light on the state of FFP among mental health professionals in Quebec, revealing a generally moderate level of engagement with families. This involves actions such as checking if clients have minor children, providing written materials on mental health and parenting, and discussing parenting strategies. Moreover, professionals reported a moderate level of perceived skills, knowledge and confidence around FFP. Results also suggest that most workers understand the relevance of offering support to children and families whose parent has a mental illness, and that they are generally open to improve their practice to better support families. Interestingly, the results seem to contrast with international studies that show lower self-reported FFP levels among adult mental health workers than FFP scores in our study. For example, in Skogøy et al. (43), the average scores for the FFP subscales are respectively 3.85 (Family support) and 4.08 (Referrals). This could suggest a unique context in Quebec where professionals appear more inclined towards FFP practices.

It is however worth noting that even though the participants in this study reported using moderate FFP levels, there were variations in responses across specific FFP activities. While a majority of the sample (84%) report that they routinely inquired about clients having minor children, discussing parenting strategies and providing mental health information directly to clients’ children were not frequently reported. These findings emphasize the need for improvement in mental health care practices concerning support for children of parents with mental illness. Professionals should not only identify such children but also play a proactive role, including offering age-appropriate information on mental illness, offering help around parenting issues and referring children and parents to appropriate support services. The results underscore the importance of expanding professionals’ awareness of their role in supporting families affected by mental illness, urging them to go beyond identification and incorporate further comprehensive strategies.

Simultaneously, results indicate that workplace support scored the lowest among all the other worker or workplace factors reported by participants. This finding generally converges with the literature, underlining that few organizations have clear guidelines supporting a family-focused approach and that most professionals feel that there are still major obstacles in their workplace for them to adopt FFP with their caseload, such as lack of necessary space and resources to invite their clients’ children for family meetings, as well as limited time to do so (36, 59). Our findings thus show that Quebec professionals feel unsupported by their work environment when it comes to offering support to families of parents receiving mental health services. Most professionals report not having sufficient time to work with families, indicated that their workload is too high, and that there are no specific parenting or family programs to refer clients’ families to when needed.

The study revealed significant predictors of FFP, encompassing both organizational and professional factors. Surprisingly, neither of the two personal factors examined—comfort with children or personal experience with mental illness—emerged as significant predictors of professionals’ FFP scores. This finding contrasts with previous research findings (44), thus enriching the existing literature. Additionally, while some predictors were consistent across high versus low FFP scores and high versus moderate FFP scores, a closer examination reveals a greater number of significant associations between low and high FFP levels.

Results underscore the significance of professionals’ attitudes toward FFP, and their perceived competence, knowledge, and confidence in predicting higher levels of self-reported engagement in FFP. These findings align with prior research, emphasizing that professionals who exhibit confidence in using FFP and acknowledge its importance are more likely to integrate it into their daily practice (39, 43, 52). Recognizing the unique parenting challenges faced by parents that have a mental illness, understanding the needs of their children, and feeling confident in engaging with and offering them support, are identified as crucial elements for the successful adoption of FFP with parents and their families. Notably, participants reported moderate to high scores on these factors, indicating a positive trend toward FFP.

Two additional professional characteristics emerged as significant predictors of FFP adoption: years of experience in adult mental health and professionals’ openness to improving their practice. This aligns with previous findings, showing that professionals with more than five years of experience were more likely to report higher levels of family and parenting support than their less-experienced counterparts (41). Hence, these worker-related characteristics play a pivotal role in ensuring the adoption of FFP in their interactions with clients. On this basis, it is recommended to offer targeted FFP training to professionals, to empower them to feel confident and prepared to use FFP with parents and families (41). Specifically, such training should include informing professionals about the impact of mental illness on parenting and equipping them with practical skills for engaging, communicating, and offering support to families (60). Best practices for interacting with children and families in the context of adult mental health should be systematically integrated into treatment protocols.

Results show that three workplace-related factors significantly predict a high FFP score: perceived workplace support for FFP, the professional’s employment status and the percentage of their clients with children. These findings converge with earlier research (43, 52), underlining the crucial role of workplace factors, including time constraints, workload, and access to supervision, as important predictors of FFP engagement among adult mental health professionals. These findings support the importance of raising awareness among responsible for clinical programs in mental health services and community organizations, about professionals’ roles and responsibilities in FFP and the need to provide adequate resources, particularly time.

Moreover, our finding regarding employment status, revealing that full-time work basis predicts a higher level of FFP, has not been yet addressed in the literature. It is possible to hypothesize that professionals employed on a full-time basis may have more time to work with families and may benefit from enhanced training or supervision, compared to those working part-time. This is an important finding, especially considering that more and more professionals work part-time in industrialized countries (61, 62). While this could account for their higher level of FFP, future research should investigate the differences between professionals working full-time and part-time, specifically around workload, training, and supervision. Furthermore, recognizing that working with families is complex and may require more time, organizations might adjust caseload expectations when screening and assessing the needs of clients with children, as well as ensure that all professionals, regardless of their employment status, receive comprehensive support in FFP.

To further support professionals in adopting FFP, the implementation of clear clinical guidelines and procedures is recommended. Additionally, providing access to specific FFP supervision or facilitating multi-disciplinary team discussions, might enhance understanding and acceptance of FFP as an integral part of all professions within the team (43). These measures may not only contribute to a shared sense of responsibility within the team but also help professionals feel actively involved and supported in embracing FFP practices (36). Overall, fostering a supportive work environment with the necessary resources and infrastructure is crucial to encouraging widespread adoption of FFP among mental health professionals.

Findings underscore the importance of ongoing monitoring and benchmarking of FFP within every workplace. To cultivate a work environment conducive to FFP, it is crucial for managers responsible for clinical programs to not only be aware of professionals’ roles and responsibilities but also to continually monitor the integration of FFP practices. Regular assessments and benchmarking can provide valuable insights into the effectiveness of existing support systems, allowing for necessary adjustments and improvements (63).

Although the sample size is adequate and the profile of participating professionals is comparable to that of other studies conducted in Quebec (e.g., 64), it is nevertheless a convenience sample and may not be representative of the population of professionals working with adults receiving mental health services in Quebec. It is possible that those who participated were more concerned about children who have a parent with a mental illness, which could have led to a bias in the representativeness of the results, resulting in a tendency to overestimate their use of FFP, as evoked by Gregg et al. (36). Future studies should be conducted with a larger and more diverse sample that is representative of all professionals working with adults receiving mental health services, to enhance the generalizability of findings. It is also important to remember that the data are self-reported, questionnaire-based, and may not reflect the actual practice of professionals. More studies including direct observations or audits of professionals’ interactions with families and children (32), or qualitative interviews, should be conducted to ensure the accuracy of results, as well as longitudinal studies to track professionals’ practices over time. Finally, this study is cross-sectional, indicating that the relationships identified in our analysis may not be causative. Based on our results, it is not possible to determine whether the consideration of these predictors in the care and services provided to adults with a mental illness and their families might impact their recovery and well-being.