Determinants of Service Use Among Young Canadians With Mental Disorders

Objective: To identify the determinants of service use by young Canadians with mental health problems.

Methods: Data were drawn from a recent large Canadian mental health survey. The analyses were conducted on a subsample of 1092 Canadians aged 15 to 24 years and identified as presenting a mood disorder, an anxiety disorder, or a substance-related disorder in the 12 months preceding the survey. We classified variables potentially associated with any type of service use for a mental health problem over a 12-month period according to predisposing, enabling, and need factors. We conducted weighted multivariate logistic regressions to determine the association of each factor with service use.

Results: In the final model, being female and living alone were the predisposing factors associated with service use. None of the enabling factors predicted help seeking. In regard to the perceived need factors, those who had difficulties with social situations were more likely to use services. Having a mood disorder and (or) having a diagnosed chronic illness were the evaluated need factors associated with service use.

Conclusion: Certain groups of young Canadians are less likely to seek help for mental health problems and could be the target of interventions aimed at increasing service use.

(Can J Psychiatry 2005;50:629-636)

Information on funding and support and author affiliations appears at the end of the article.

Clinical Implications

* Only 25% of young Canadians with mental health problems seek help. Therefore, it is essential to initiate interventions that will aim at increasing their use of services.

* Sex, living arrangement, reaction to social situations, diagnosis of mood disorders, and chronic physical disorder influence the likelihood of service use.

* Interventions to encourage service use should be targeted toward people who are less likely to use the services, namely, young men, young people living with their parents or with unrelated others, and people diagnosed with an anxiety or a substance-related disorder.

Limitations

* The survey, which is not specifically designed for young people, does not contain information on familial context.

* The study does not account for the impact of income on service use.

* No distinction is made between types of service, disorder, age, and sex.

Key Words: adolescents, young adults, mental health, mood disorders, anxiety disorders, substance-related disorders, service use, help seeking, Canada

National surveys across developed countries indicate that, when compared with older age groups, people aged 15 to 24 years have the highest rates of mental disorders (1,2). This trend is also observed in the CCHS 1.2 (3), which specifies that 18% of Canadians aged 15 to 24 years meet the criteria for a mood disorder, an anxiety disorder, or a substance-related disorder, compared with almost 12% of those aged 25 to 44 years, 8% of those aged 45 to 64 years, and 3% of those aged 65 years and over. Although young people are particularly at risk for mental disorders, they underuse services for these problems (1,3). Therefore, it is important to investigate factors associated with service use in this particular subgroup.

Prior research has examined potential determinants of help seeking by young people with mental disorders. However, these studies have overlooked some of the factors examined in studies targeting adults in general. For example, factors associated with service use, such as the size of the social network (4) and certain types of living arrangements (5), may also influence young people's help-seeking behaviour. Moreover, few studies based their analyses on an established model of service use. Finally, the only Canadian study that explored the determinants of help seeking by young people with mental disorders only considered a few variables (6). Considering these limitations, this study aims to identify the factors associated with service use for mental health problems by young Canadians. We used Andersen's Behavioural Model of Health Care Use as a theoretical framework (7,8).

Method

Data were drawn from the CCHS 1.2 (n = 36 984), which is a nationally representative, computer-assisted, face-to-face household survey. The sampling procedure used was a multistage stratified cluster design. The response rate was 77% (3). A more detailed description of the methodology used for this survey is provided elsewhere in this issue (9).

Sample

For the present study, the sample (n = 1092) included people aged 15 to 24 years meeting the criteria for mood disorders (major depressive episode or mania), anxiety disorders (panic disorders, social phobia, or agoraphobia), or substancerelated disorders (alcohol or illicit drug dependence) during the 12 months preceding the survey. We assessed mental disorders, using a modified version of the WMH-CIDI according to the DSM-IV (3).

Variables

This study looked at the determinants of service use for mental health problems over a 12-month period. The services considered included hospitalization, as well as consultation with psychiatrists, psychologists, GPs, nurses, social workers, support groups, help lines, alternative health care providers, and clergy.

We looked at the determinants that were found to be associated with service use in previous studies and classified them according to Andersen's Behavioural Model of Health Care Use (7,8), an established framework used to study service use for mental health problems (4,10,11 ). The model suggests that use of health services is a function of predisposing, enabling, and need factors. Predisposing factors are present before illness onset. They are related to the individual's propensity to service use. In the present study, the predisposing factors included age, sex, country of origin (Canada or other), type of living arrangement, and occupation (that is, school attendance and employment status). The enabling factors are related to the means by which individuals might access mental health services, for example, factors influencing knowledge about where to seek help. The enabling factors considered in this study were social support, determined by the global score on the MOS social support survey form (12); size of the social network; mental disorders among relatives; province of residence; and MIZ. MIZ typology devises areas outside CMAs and CAs according to the percentage of people commuting to and the distance from a CMA or a CA. The MIZ classification better reflects the urban-rural continuum than is possible with the urban-rural dichotomy (13).

Originally, a third category in Andersen's model comprised perceived and evaluated needs. However, in recent studies using Andersen's model, perceived and evaluated needs were considered separately (10,11). This study follows this trend. The perceived need factors included perception of mental and physical health, ability to face day-to-day demands, reaction to social situations, and psychological distress. A global measure of psychological distress was assessed by the KlO symptom scale (14). The evaluated need factors category comprised each assessed disorder (that is, mood disorders, anxiety disorders, or substance-related disorders) and their cooccurrence, as well as each chronic physical disorder.

Analyses

We conducted frequency analyses to assess the 12-month prevalence of any service use. We used multivariate logistic regressions to determine the association of each of the predisposing, enabling, perceived, and evaluated need factors with service use. According to Andersen's model, variables were entered in blocks: predisposing, enabling, perceived need, and evaluated need. To account for the complexity of the CCHS 1.2 sampling plan, we conducted analyses using the Wesvar software (15) and applied bootstrap weights. The strength of the associations is reported in ORs, with a 95%CI.

Results

Over a period of 12 months, only 25% of people with a mental disorder consulted for mental health problems. The results of the regressions showed that, when only predisposing factors were entered into the model, being female increased the likelihood of service use for mental health problems. Young people who were unemployed, whether attending school or not, were more likely to use the services than were those who were employed and (or) attending school. Moreover, people living with someone were less likely to seek help than were those living alone.

With the enabling factors added to the model, being female remained associated with service use. People living with someone were still less likely to seek help than were people living alone, whereas occupation ceased to be a significant predictor. Regarding the enabling factors, having relatives experiencing the same disorder and having more social support predicted service use.

When perceived need factors were added to the model, being female, living alone, and having relatives with the same disorder remained associated with service use. Social support ceased to be a significant predictor. In addition, those who perceived their mental health as poor or fair were 4 times more likely to use any service during the 1-year period than were those who perceived it as excellent. Having difficulties with social situations was also a predictor of help seeking.

The final model was obtained with the inclusion of evaluated need factors. Being female remained a predictor of service use. Canadians who were living with someone were less likely to use the services than were those living alone, with the exception of those who were living with a partner and (or) a child. The association between having difficulties with social situations and service use persisted, whereas the perception of mental health ceased to be a significant predictor. Having a mood disorder and having a diagnosed chronic physical disability were associated with an increased likelihood of service use.

Discussion

Results from the CCHS 1.2 indicate that few Canadians aged 15 to 24 years use services for their mental health problems. This suggests a need for interventions.

We explored various potential determinants of service use to identify those that predicted help seeking. A final regression model emerged from the simultaneous inclusion of predisposing, enabling, perceived need, and evaluated need factors. Regarding the predisposing factors, young women were more likely to use the services than were young men. The direction of this finding is supported by previous studies that have found an association between sex and service use (16-19). Our study was also interested in the association between types of living arrangement and help seeking. Living alone was also found to be a significant predictor of frequent service use in a study of adults of all ages (5). However, of the studies concerned with young people, this is the first to examine the association between this type of living arrangement and help seeking. We also found that being unemployed, whether attending school or not, increased the likelihood of service use. However, this association disappeared when enabling factors were entered into the model. Biddle and others also studied the impact of occupation on consultation of GPs for mental health problems and found no association (20).

No enabling factors had a significant impact on the final model. Social support was a predictor of service use until perceived need factors were entered into the model. In the literature, findings concerning the association between social support and service use for mental health problems are inconsistent. Biddle and others found that social support was not a predictor of consultation with a GP (20), whereas Saunders and others found that use of informal support was associated with service use (21). These inconsistencies could result from conceptual differences in the definition of social support or from the type of services studied. We also found that having relatives who suffered from the same disorder was associated with an increased likelihood of service use, but the association disappeared when evaluated need factors were added to the model. Cunningham and others (22) and Wu and others (23) have demonstrated that, when relatives received mental health care, young people were significantly more likely to use services. Unfortunately, information concerning the use of services by relatives was not available in the CCHS 1.2.

In the final model, having difficulties with social situations was associated with an increased likelihood of help seeking. In previous studies focusing on young people, this specific measure of impairment was not explored, although some studies examined the impact of other impairment variables. Even if the definition of these variables varied considerably between studies, impairment was found to be associated with service use (23-25). The perception of poor or fair mental health was also found to be a predictor of service use, but its influence disappeared in the final model, with the inclusion of evaluated need factors. Biddle and others did not include evaluated need factors in their final model (20) and found that perception of mental health was not a predictor of consultation. Bearing in mind that they only investigated help seeking from a GP, these diverging results could indicate that the impact of this variable may vary according to the type of professional considered.

Finally, evaluated need factors were entered into the model. We found that having a diagnosed chronic physical disability increased the likelihood of service use. Wu and others have comparable findings (23), whereas Haarasilta and others found no significant association (26). A possible explanation for this association is that regular contact with health care professionals for physical problems may give more opportunities to discuss mental health issues and may increase the likelihood of being identified as having a mental health problem. We also found that young people with a mood disorder were more likely to use services than were those with either an anxiety disorder or a substance-related disorder. This conclusion can be reached because our sample only includes people diagnosed with one of these 3 disorders; thus not having one disorder implies having one of the others. This result cannot be directly compared with previous findings, since the population (for example, the general population vs people with mental disorders) and the disorders studied differ between studies.

Some variables that were not associated with service use in the present study found empirical support in prior studies targeting young people. For instance, some studies identified age as a predictor of service use for mental health problems, but the direction of the association is inconsistent (16,19,23,24,27). Moreover, some studies indicated that age is not a significant predictor of service use (17,19,25,26). These inconsistencies may result from an interest in different age groups, different countries, and differences in the setting surveyed (for example, school or community). Finally, the size of the social network was only considered by studies targeting adults (4). Therefore, the absence of association for young people could be specific to this population.

This study has limitations that should be noted. Considering that the survey was not specifically designed for young people, data on specific determinants of service use by youth were not available for the analyses. For instance, there is a lack of information on other family members and on family functioning. Previous studies have found that determinants related to familial context, such as the impact of symptoms on family functioning (24) and parents' marital status (21), are associated with help seeking. Therefore, this study may overlook predictors of service use.

Further, when considering young people, it is difficult to correctly engage some concepts. One problematic variable is income. With the information collected by the CCHS 1.2, it is impossible to establish the extent of the monetary support provided by parents. Thus the declared personal income does not necessarily reflect the amount of money that a young person may benefit from. Further, many young people are unaware of their parents' income; thus missing data for family income are substantial.

Future research could examine determinants of health care use for particular types of services, disorders, and ages, and for both sexes. It has been demonstrated that specific determinants may be associated with different types of services (23,24,28). Certain determinants may be associated with service use only for specific types of disorders or may vary for adolescents and young adults as well as for men and women.

Conclusions

Despite the high prevalence of mental disorders among young Canadians, this study revealed that relatively few youth consulted a professional for their mental health problem. It is therefore essential to initiate interventions that will aim at increasing young Canadians' use of services for such problems.

This study also provides unique information about the determinants of service use by young Canadians. The results indicate that certain groups of people are less likely to use the services for mental health problems. These groups can be targeted to increase service use. Efforts should be made to reach young men, people living with their parents, and those living with unrelated others. The results indicate that people with anxiety disorders and substance-related disorders are less likely to seek help. Consequently, interventions aimed at increasing the detection of these disorders may be required. For example, interventions could aim at increasing parents' capacity to detect mental health disorders and to incite their children to seek professional support.

Funding and Support

This study was supported by a grant from the CIHR. This research was also supported in part by a scholarship to EB from the Groupe de recherche sur l'�quit� d'acc�s et l'organisation des services de 1�re ligne, funded by the CIHR (2003-2005), by an award from the Quebec Inter-University Center for Social Statistics (2004), and by a RAMHPS award, funded by the CIHR and Quebec Mental Health and Neuroscience Network (2003-2005). This research was also supported in part by scholarships from the Strategic Training Program in public and population health research of Quebec, a partnership of the CIHR, the Quebec Population Health Research Network, and the AnEIS strategic program, funded by the CIHR and the Fonds de la recherche en sant� du Qu�bec.

Acknowledgements

We thank the Quebec Inter-University Center for Social Statistics for their support during the analyses. The research and analyses are based on data produced by Statistics Canada; the opinions expressed do not represent the views of Statistics Canada.

R�sum� : Les d�terminants de l'utilisation des services chez les jeunes Canadiens souffrant de troubles mentaux

Objectif : Identifier les d�terminants de l'utilisation des services par les jeunes Canadiens ayant des probl�mes de sant� mentale.

M�thodes : Les donn�es �taient tir�es d'une vaste enqu�te canadienne r�cente sur la sant� mentale. Les analyses ont �t� men�es aupr�s d'un sous-�chantillon de 1 092 Canadiens �g�s de 15 � 24 ans et reconnus pr�senter un trouble de l'humeur, un trouble anxieux ou un trouble li� � une substance dans les 12 mois pr�c�dant l'enqu�te. Les variables potentiellement associ�es � un type d'utilisation des services quelconque pour des raisons de sant� mentale, sur une p�riode de 12 mois, �taient class�es selon des facteurs de pr�disposition, d'habilitation et de besoin. Nous avons men� des r�gressions logistiques multivari�es pond�r�es pour d�terminer l'association de chaque facteur � l'utilisation des services.

R�sultats : Dans le mod�le final, �tre de sexe f�minin et vivre seul �taient les facteurs de pr�disposition associ�s � l'utilisation des services. Aucun des facteurs d'habilitation ne pr�disait la recherche d'aide. Pour ce qui est des facteurs de besoin per�u, ceux qui �prouvaient des difficult�s avec les situations sociales �taient plus susceptibles d'utiliser les services. Avoir un trouble de l'humeur et/ou avoir une maladie chronique diagnostiqu�e �taient les facteurs de besoin �valu� associ�s � l'utilisation des services.

Conclusion : Certains groupes de jeunes Canadiens sont moins susceptibles de rechercher de l'aide pour des probl�mes de sant� mentale et pourraient �tre la cible d'interventions visant � accro�tre l'utilisation des services. L'information sur le financement et le soutien ainsi que sur les affiliations de l'auteur figure � la fin de l'article.

[Reference]

References

1. Andrews G, Hendcrson S, Hall W. Prevalence, comorbidity, disability and service utilisation. BrJ Psychiatry 2001;178:145-53.

2. Kessler RC1 McGonagle KA, Zhao S. Lifetime and 12 month prevalence of DSM-III-R psychiatric disorders in United States. Arch Gen Psychiatry 1994;51:8-19.

3. Statistique Canada. Enqu�te sur la sant� dans les collectivit�s canadiennes : sant� mentale et bien-�tre [homepage on the Internet], [updated 2004 Sept 09]. Available: www.statcan.ca/francais/freepub/82-617-XIF/def_f.htm. Accessed 2005 Feb 23.

4. Albizu-Garcia CE, Alegria, M, Freeman D, Vera M. Gender and health services use for a mental health problem. Soc Sci Med 2001;53:865-78.

5. ten Have M, Vollebergh R, Bijl RV, de Graaf R. Predictors of incident care service utilisation for mental health problems in the Dutch general population. Soc Psychiatry Psychiatr Epidcmiol 2001;36:141-9.

6. Schonert-Reichl K, Muller JR. Correlates of help-seeking in adolescence. J Youth Adolesc 1996;25:705-31.

7. Andersen RM. Revisiting the behavioural model and access to medical care: does it matter? J Health Soc Behav 1995;36:1-10.

8. Andersen RM, Ncwman JF. Societal and individual determinants of medical care utilization in the United States. Milbank Mem Fund Q I973;51:95-124.

9. Gravel R, B�land Y. The Canadian Community Health Survey: Mental Health and Well-Being. Can J Psychiatry 2005;50:573-9.

10. Goodwin R, Andersen RM. Use of the Behavioral Model of Health Care Use to identify correlates of use of treatment for panic attacks in the community. Soc Psychiatry Psychiatr Epidemiol 2002;37:212-9.

11. Koenen KC, Goodwin R, Slrucning E, Hellman F, Guardino M. Postlraumatic stress disorder and treatment seeking in a national screening sample. J Trauma Stress 2003;16:5-16.

12. Sherbourne CD, Stewart A. The MOS social support survey. Soc Sci Med 1991;32:705-14.

13. McNiven C, Puderer H, Janes D. Census metropolitan area and census agglomeration influenced zones (MIZ): a description of the methodology. Ottawa: Geography Division of Statistics Canada; 2000. Report 92F0138MIF.

14. Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand, SLT, and others. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med 2002;32:959-76.

15. Westat statistical software. Wesvar 4.0. Rockville (MD); 2000.

16. Gasquet I, Chavance M, Ledoux S, Choquet M. Psychosocial factors associated with help-seeking behaviour among depressive adolescents. Eur Child Adolesc Psychiatry 1997;6:151-9.

17. Pumariega AJ, Glover S, Holzer CE, Nguyen H. Utilization of mental health services in a tri-ethnic sample of adolescents. Comm Ment Health J 1998;34:145-56.

18. Gasquet I, Ledoux S, Chavance M, Choquet M. Consultation of mental health professionals by French adolescents with probable psychiatric problems. Acta Psychiatr Scand 1999;99:126-34.

19. Garland AF, Ziegler EF. Psychological correlates of help-seeking attitudes among children and adolescents. Am J Orthopsychiat 1994;64:586-93.

20. Biddle L, Gunnell D, Sharp D, Donovan JL. Factors influencing help seeking in mentally distressed young adults: a cross-sectional survey. Br J Gen Pract 2004;54:248-53.

21. Saunders SM, Resnick MD, Hoberman HM, Slum RW. Formal help-seeking behaviour of adolescents identifying themselves as having mental health problems. J Am Acad Child Adolesc Psychiatry 1994;83:49-52.

22. Cunningliam PJ, Freiman MP. Determinants of ambulatory mental health services use for school-age children and adolescents. Health Serv Res 1996;31:409-27.

23. Wu P, Hoven CW, Bird HR, Moore RE, Cohen P, Alegria M, and others. Depressive and disruptive disorders and mental health service utilization in children and adolescents. J Am Acad Child Adolesc Psychiatry 1999;38:108I-90.

24. Angold A, Erkanli A, Fanner EMZ, Fairbank JA, Burns, BJ, and others. Psychiatric disorder, impairment, and service use in rural African American and white youth. Arch Gen Psychiatry 2002;59:893-901.

25. Wu P, Hoven CW, Cohen P, Liu X, Moore RE, Tiet Q, and others. Factors associated with use of mental health services for depression by children and adolescents. PsychiatrScrv 2001 ;52:189-95.

26. Haarasilta L, Marttunen M, Kaprio J, Aro H. Major depressive episode and health care use among adolescents and young adults. Soc Psychiatry Psychiatr Epidemiol 2003;38:366-72.

27. Cohen P, Hesselbart CS. Demographic factors in the use of children's mental health services. Am J Public Health 1993;83:49-52.

28. Fournier E, Lemoine O, Poulin C, Poirier LR. Enqu�te sur la sant� mentale des Montr�alais. Volume 1: La sant� mentale et les besoins de soins des adultes. Montreal (QC): Direction de la sant� publique, R�gie r�gionale de la sant� et des services sociaux de Montr�al-Centre; 2002. Report ISBN 2-89494-342-3.

[Author Affiliation]

Emilie Bergeron, MSc1, L�o-Roch Poirier, MSc2, Louise Fournier, PhD3, Pasquale Roberge PhD4, Genevi�ve Barrette, PhD5

[Author Affiliation]

Manuscript received and accepted May 2005.

Previously presented at the Canadian Association of Psychiatric Epidemiology; 2004 October 14; Montreal (QC).

1 PhD Candidate, Faculty of Medicine, Department of Social and Preventive Medicine, University of Montreal, Montreal, Quebec.

2 Researcher, Institut National de Sant� Publique, Montreal, Quebec.

3 Researcher, Institut National de Sant� Publique, Montreal, Quebec; Associate Professor, Faculty of Medicine, Department of Social and Preventive Medicine, University of Montreal, Montreal, Quebec.

4 Post Doctoral Fellow, Groupe de Recherche Interdisciplinaire en Sant�, University of Montreal, Montreal, Quebec; Researcher Institut National de Sant� Publique, Montreal, Quebec.

5 PhD, Department of Psychology, Universit� du Qu�bec � Montr�al, Montreal, Quebec.

Corresponding address: Dr L Fournier, Institut national de sant� publique, 4835 Christophe-Colomb, Montreal, QC H2J 3G8

e-mail: louise.fournier@inspq.qc.ca