Background
There is a growing body of research about the use of formal and informal support services for mental health problems. For example, recent national surveys from Australia [
1], USA [
2], England [
3] and Canada [
4] have documented prevalence rates for mental disorders, professional and non-professional service usage patterns [
5‐
9], perceived need profiles [
10‐
12], and associated linkages to sources of mental health information [
13]. However, evidence about the prevalence of mental disorders, symptoms, and patterns of service use in rural and remote communities is in short supply for a variety of reasons. Surveying residents of these regions is expensive and difficult. Indeed, the 2007 Australian National Survey of Mental Health and Wellbeing (NSMHWB) acknowledged that it was ‘… limited in its representation of people from remote and very remote areas’ ([
1], p. 597) while in some Canadian provinces oversampling has been used in the national survey to provide reliable sub-provincial prevalence estimates [
4].
The meaning of rurality varies and there are difficulties in assessing evidence from countries with substantial rural populations such as Scotland, Australia, Canada or the United States. Some authors [
14] have concluded that the definition and official classifications of rurality are not helpful in making meaningful distinctions between communities and that we should either change the classifications [
15] or stop using them; for example Rost et al. proposed ‘… that investigators no longer employ any of the multiple definitions of rurality proposed in the literature [
14] (p. 232).
There are a number of reports that conclude that the prevalence of mental health symptoms is similar between rural and urban areas [
16‐
18]. Likewise, Smith [
19] reports that ‘there is little evidence linking prevalence of mental health disorders with rurality’ (p. 58). Some authors, such as Ziller [
20], found that self reported need for mental health services was slightly (but statistically significantly) higher in rural US. However, there are a number of reasons to believe that the use of services for mental health problems may differ between rural/remote residents and their urban peers including: specialist workforce shortages in rural areas [
21]; generalist workforce shortages in remote and very remote areas [
21,
22]; costs, travel and waiting times for rural and remote patients [
23]; culture and attitudinal factors including “rural stoicism”, concerns about privacy and confidentiality in small communities [
24]; and, lower rates of engagement in and continuation with treatment for mental illness in rural areas [
14].
In a study of 78,000 employees in Australia, Hilton et al. [
25] found, perhaps unsurprisingly, that those in very remote regions were less likely to access mental health treatments. Parslow et al. [
7] compared service use in Australia using the 1997 and 2007 NSMHWB and concluded that use of any mental health service had risen significantly but that the proportion seeing General Practitioners (GPs) for mental health services had not increased (p. 206). This increase in mental health service use has been influenced by Federal programs to improve access to mental health practitioners, including the Access to Allied Psychological Services and the Better Access to Psychiatrists, Psychologists and General Practitioners initiatives introduced in 2001 and 2006 respectively [
26]. GPs are key providers in rural areas with specialist shortages and they are the first choice for many people seeking help for a mental health problem [
26].
While Mechanic [
27] argues that we cannot use prevalence of mental disorders as a proxy for service need, the aggregate pattern of service use in Australia and elsewhere by those with mental health disorders would appear to be one of underuse [
28]. The 2007 Australian NSMHWB concluded that ‘rates of service use for people with a mental health condition are less than optimal [
5] (p. 615). About one-third (34.9%) of those meeting the Composite International Diagnostic Interview (CIDI) criteria for a mental disorder reported that they received services in the previous 12 months. There was evidence that those with severe disorders made more use of community based mental health services (63.5%) than those with mild disorders (17.7%). The extent of this “underuse” is not clear, since no-one is suggesting that all those with a mental health disorder identified in a research interview would benefit from treatment from a generalist or a mental health specialist. In another Australian study, Short et al. [
29] concluded that public mental health services care for people with psychotic disorders while ‘those with high prevalence disorders … seek psychiatric treatment elsewhere’ (p. 475).
It is important to acknowledge that the Australian health care system differs from unitary health care systems, such as those in the UK and New Zealand, having a state-federal separation of specialist and primary mental health care funding and responsibilities. Unitary systems are better able to integrate specialist and primary care based services since they have population based planning and funding arrangements. In Australia, access to services can be compromised since there is no effective process for joint or shared planning [
30].
There is some international evidence about non-specialist sources of care such as religious advisors [
31] or self care [
32]. Verhaak et al. [
33] note that patients may not recognise that their problems have a mental health origin or may question the effectiveness of their providers. In a cross-European survey of adults from six nations, ten Have et al. [
34] reported a low perceived effectiveness of professional care and about a third of respondents believed it was worse than no care for those with serious emotional problems. In the Australian context, Olesen et al. [
6] concluded that there is widespread use of non-professional services and self-management strategies (e.g., 33.5% of adults with an affective or anxiety disorder in the last 12 months sought help from family or friends). These findings are consistent with those from studies investigating community attitudes to mental illness in general, with low rates of confidence in professional treatment, as a component of the prevailing stigma of mental disorders [
35].
Some American studies have examined whether rural residents have access to an appropriate range and “dose” of services. Fortney reported that about 60% of a rural sample with self-reported depression received some formal treatment, with more than 50% receiving pharmacotherapy and 25% receiving psychotherapy [
36,
37]. The rates of minimally adequate treatment were measured as the number of scripts filled and the number of psychotherapy sessions attended [
36]. He noted that if these rural residents do not respond to medication they may not have access to psychotherapy as an alternative treatment. In a study of American veterans with depression, anxiety or post traumatic stress disorder, Cully et al. [
38] found that rural veterans were less likely to receive psychotherapy than their urban peers and that the “dose” was less likely to be adequate. While shortages of specialist mental health staff characterise rural and remote communities in Australia, it is not clear if these findings hold true.
In a study of depression and remoteness in South Australia, Goldney [
17] found that rates of treatment of depression with antidepressants were similar in less accessible and more accessible areas and that a higher proportion of the rural depressed had contact with a health service, social worker or counsellor than their more accessible peers. Diagnostic interviews were not used in this study; hence, a limitation was the reliance on self-identification of past “depression”. Jackson et al. [
39] have identified factors contributing to mental health usage in a rural setting, including gender, age, marital status, mental disorder, co-morbidity and psychological distress. Similarly, in the 2007 Australian NSMHWB, service usage among adults with a mental health problem was lowest among males and the youngest and oldest age groups [
5,
6].
In summary, the use of services for mental health disorders presents a complex picture. A focus on rural and remote settings complicates matters and international evidence is hard to compare. Given that access to services for mental health problems is limited in rural and remote regions, questions remain about from whom rural residents in Australia seek help, how frequently this occurs and what characteristics are associated with this help-seeking behaviour. This paper examines the impact of remoteness on the patterns of service use for mental health problems in a rural community sample, how they compare with national data for urban residents from the 2007 Australian NSMHWB, and the relationship between self-reported service use and estimates of the potential need for services.
Heath service context
Australia is a federation of 6 states and 2 territories, where public hospitals are funded by state departments of health and admission to public hospitals is free. The majority of GPs and medical specialists, including psychiatrists, are self-employed and their fees are subsidised or paid in full by the Federal Medical Insurance system. GPs act as gatekeepers for accessing specialist services, and patients can be referred to psychiatrists, psychologists and other allied health professionals, including social workers, mental health nurses and occupational therapists [
26].
While the number of full-time equivalent GPs per 100,000 members of the population is reasonably similar across all categories of remoteness in Australia, the number of medical specialists decreases with increasing remoteness. For example, there are approximately 18 full-time equivalent psychiatrists per 100,000 people comprising 23 (per 100,000) in major cities, compared to 7 for inner regional, 5 for outer regional, and 3 for remote and very remote areas [
40]. In rural and remote areas, access to GPs is also complicated by distance and sometimes by cost, through the use of co-payments.
Methods
Complete details of the methodology and measures used in the Australian Rural Mental Health Study (ARMHS), a longitudinal population-based cohort, are provided elsewhere [
41,
42]. This project was approved by the Human Research Ethics Committees of the University of Newcastle, University of Sydney, Greater Western Area Health Service, Hunter New England Area Health Service and the North Coast Area Health Service.
Recruitment
Sixty Local Government Areas were identified from three Australian rural health service regions of the state of New South Wales (NSW) using the Rural, Remote and Metropolitan Areas classification, representing approximately 70% of the geographic region of non-metropolitan NSW. Over-sampling of the remote and very remote regions ensured sufficient sample size from these regions. The baseline sample was recruited between 2007 and 2009 and comprised residents aged 18 years or older living in private dwellings. Households were identified from Australian electoral rolls and sent a letter informing them of the study. Matching telephone numbers were found from an electronic directory and households were called requesting consent for the study and to identify a contact person. Consent papers and questionnaires were sent to those who agreed to participate.
Participants aged 65 years or over were screened for cognitive impairment using the modified Telephone Interview for Cognitive Status (TICS-M) [
43] and those with a TICS-M total score < 17 were excluded. Non-English speaking members of a household, those with significant hearing impairment that impeded consent and/or interview, and those with no identifiable telephone contact number (after directory and electronic database search) were also excluded. The baseline survey used self report measures, administered by post in two parts (survey A and B) mailed two weeks apart and excluded special dwellings (such as hospitals, nursing homes, prisons, hotels and hostels) and overseas visitors usually resident outside Australia.
Variables and instruments
Socio-demographic characteristics
Basic socio-demographic questions were included in Survey A and the following characteristics were coded for the current analyses: age in years (grouped into five categories 18–34, 35–44, 45–54, 55–64 and 65+); gender (male, female); education (partial schooling, completed high school or above, and unknown); and marital status (married or de facto, divorced or separated, widowed, and never married). Financial position was assessed using the perceived prosperity item from the Household, Income and Labour Dynamics in Australia (HILDA) study (“Given your current needs and financial responsibilities, would you say that you and your family are: prosperous, very comfortable, reasonably comfortable, just getting along, poor, or very poor?”) [
44]. Rurality/remoteness was categorised using the Australian Standard Geographic Classification (ASGC) system, which allocates classes of remoteness to localities, based on the Accessibility/Remoteness Index of Australia Plus (ARIA+) values: major cities, inner regional, outer regional, remote and very remote. The current study selected people residing in non-metropolitan areas, with the remaining categories used for classification of district level remoteness. ARIA+, developed by the National Centre for Social Applications of Geographic Information Systems (
http://www.gisca.adelaide.edu.au), describes remoteness from goods and services for any part of Australia relying on road distance as a surrogate for remoteness and on the population size of a service centre as a surrogate for the availability of services [
45].
Health status characteristics
For the current analyses, seven health characteristics were selected, which comprised direct global ratings about health, standardised measures frequently used in mental health epidemiological research, risk behaviours or circumstances linked with mental health (e.g., substance use, smoking, recent adversity). Specifically: single item self report measures were devised (for Survey A) to assess overall physical and overall mental health, using 5-point Likert scales ranging from poor to excellent; the Kessler K-10 (K-10) was used (in Survey A) to assess psychological distress [
18,
46]; the World Health Organisation’s ten item Alcohol Use Disorders Identification Test (AUDIT) was used (in Survey B) to measure current hazardous alcohol use, dichotomised as low risk (scores of 0–7) or at-risk (scores ≥ 8) [
47,
48]; the nine item Patient Health Questionnaire (PHQ-9) was used (in Survey B) to assess current depressive symptoms (at least one symptom for more than half the time) [
49]; and a 12 item adverse life events scale was used (in Survey A) for events within the preceding 12 months [
50], which was categorised as: 0–2, 3–5, or more than 5 events.
Service utilisation: self-reported professional and non-professional contacts for mental health
Recent service use for mental health problems was investigated primarily using items from the 2007 Australian NSMHWB [
1,
5,
10], which included: questions about the frequency of visits to a range of mental health professionals “… in the past 12 months regarding your own mental health problems, such as stress, anxiety or depression, or worries about alcohol or drugs”; and comparable questions about other contacts (e.g., with other professionals, friends or family, clergy, etc.). Questions about the forms of help received, their overall adequacy, and potential barriers were asked, including a measure of inadequate help – “In general, did you get as much help as you needed (yes or no)?” and factors that might have stopped or delayed access to health care when it was needed: distance to travel; cost; transport; limited choices; time taken to get an appointment; confidentiality; time away from work, carer responsibilities, stress and other. Also included in Survey B were a range of questions about other ways of accessing support for mental health problems during the past 12 months, including the frequency of using internet support groups, self-help groups, and telephone counselling.
Data analysis
Data entry, cleaning, aggregation and analysis techniques primarily involved the Statistical Package for Social Sciences (SPSS version 17.0; Chicago, IL, USA) and SAS (SAS V9.2; SAS Institute Inc., Cary, NC, USA) statistical software. The initial analyses reported below are descriptive in nature, detailing the socio-demographic and health status characteristics of the participants who completed the service utilisation questions, followed by patterns of self-reported professional and non-professional contacts during the past 12 months (including the percentages with any contacts, and the number and distribution of contacts by service users). For the major analyses, hierarchical logistic regressions were used for the outcome variables of professional contact and non-professional contact, each coded as no contact (0) versus any contact (1). All predictor variables were categorical in nature, with socio-demographic characteristics entered simultaneously at step 1, followed by each of the individual health status characteristics entered separately at step 2 (as they tended to be interdependent) thereby assessing relationships between the various health status characteristics and service use, while controlling for any underlying socio-demographic influences. For the sub-group of participants who reported having contact with a professional or non-professional during the past 12 months for their own mental health problem, factors associated with inadequate help, defined as answering no to the question about receiving as much help as needed, were explored using logistic regression techniques. Participants were only included in these analyses if they had the full set of socio-demographic characteristics; and, as a partial control for the number of statistical tests, the threshold for significance was set at p < 0.01. These associations are reported as Odds Ratios (OR) with 99% Confidence Intervals (CI).
Acknowledgements
The study was funded by the National Health and Medical Research Council (Project Grants #401241 and #631061), and also supported by a Research Capacity Building Grant to the Australian Rural Health Research Collaboration. The authors wish to acknowledge the other investigators on this study: Prof John Beard, Prof Vaughan Carr, A/Prof Lyn Fragar, Prof David Lyle, Prof Prasuna Reddy, and Dr Helen Stain. We also wish to acknowledge the research site coordinators: Jan Sidford, John Ogle (Broken Hill), Trim Munro, Amy Strachan (Moree), Louise Holdsworth, Kath O’Driscoll (Lismore), Cheryl Bennett, Jannelle Bowler (Orange), along with Fleur Hourihan, Dr Gina Sartore, Denika Novello.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
DP led the writing of the paper, TL devised the Provisional Service Need Index and led the data analysis with MF, CC managed the data collection and custodianship, BK had the original idea for the study and is chief investigator, while the remaining authors JF, KI, JA, DA, RR and RB took a full part in the drafting and development of the paper. All authors read and approved the final manucripts.