Abstract
The planning of mental health services in Australia, as in most countries, is greatly influenced by the needs of individuals who present for health care. The demand from these individuals is what health professionals see in their daily work and is what places political pressure on governments, especially when it becomes the focus of media attention. The unmet need for services and queuing of patients waiting for various types of health care is endemic in many places and a highly sensitive topic in medico-political debates, including in mental health [1, 2].
What is much less visible, especially to the average clinician, is the large number of individuals who have injuries and illness but who do not, or have not yet, sought and received treatment. The personal suffering, lost productivity, lost opportunity for early intervention and poorer clinical outcomes arising from untreated disorder and the delay in treating disorders is well recognized. Only by undertaking population surveys is it possible to estimate the prevalence of mental disorders in the community, understand how these disorders impact on the individual and their family and carers, gain some insight into who gets treatment and who does not, and learn where people go for treatment.
The Australian Bureau of Statistics undertook Australia's first mental health population survey, the National Survey of Mental Health and Wellbeing (NSMHWB), in 1997 [3]. The data produced from this survey, in conjunction with the conclusions of the 1998 Australian Burden of Disease Study, were profoundly influential. The Burden of Disease study found that mental disorders and substance abuse accounted for nearly 30% of all health-related disability and that depression was the leading cause of disability for male and female subjects among all health conditions [4]. The 1997 NSMHWB confirmed that nearly one in five Australians met criteria for a mental disorder in that year, that most people who met criteria did not get treatment from a health professional, and those who did, received it from a general practitioner [5].
The policy response to these findings was to expand the population health scope of the National Mental Health Strategy, which until that time had been primarily focused on the structural reform of the public specialist mental health service system. The target population for this service system was the lower prevalence disorders, especially psychotic disorders and mood, anxiety and personality disorders with high levels of acuity and risk. There were policy initiatives to treat more people with common mental disorders. These initiatives included a national primary mental health-care initiative that commenced in June 1999 and provided education and skills training in mental health for general practitioners. The Better Outcomes in Mental Health Care programme was introduced in 2001 [6] and the Better Access to Psychiatrists, Psychologists and General Practitioners (Better Access) programme in November 2006 [7]; both of which expanded access to allied mental health professionals.
Ten years after the first NSMHWB, the Australian Bureau of Statistics carried out a second national mental health survey. The second National Survey of Mental Health and Wellbeing (2007 NSMHWB) is a nationally representative household survey of adults aged between 16 and 85 years. Papers in this issue of the Journal describe the methodology and some findings. The findings have a considerable number of policy implications, which are briefly discussed.
Mental disorders are highly prevalent
The NSMHWB 2007 showed that one in five people experienced a mental disorder in the past 12 months. This finding, consistent with other international research and the 1997 survey, highlights again that mental disorders are among the most common illnesses in society. As in 1997, anxiety disorders were the most common class of mental disorders. Among the anxiety disorders, post-traumatic stress disorder and social phobia are the most prevalent. As a class, anxiety disorders are more than twice as common as affective disorders and almost three times as common as substance use disorders.
The Survey also found that 45% of the Australian population have experienced mental disorder at some point in their life. Even more than the 20% 12 month and 10% 30 day prevalence figures, such a high lifetime prevalence shows the extent to which mental disorders afflict the population. They are not restricted to a minority whose conditions remit and relapse. They are, as Andrews et al. argue, ‘waiting for most’ of us [8]. This finding reinforces what must still be the message for the broader community: that mental disorders are not conditions that happen only to someone else.
At a national level, there has been a major effort to educate Australians about mental disorders, especially depression, alcohol and illicit drug use, during the past decade. There is some evidence that these efforts have resulted in better awareness about depression and reduced discrimination against Australians with depression [9]. There is also some evidence of a reduction in harmful drinking and illicit drug usage, although rates of substance use disorders have not fallen and treatment rates remain very low [10]. Less effort has been made in the area of anxiety disorders. The high prevalence and chronicity of this class of disorders suggests that specific action is also needed for this group.
There has also been a major effort in the past decade to promote youth mental health services. This has seen the development of national programmes such as headspace. One argument for this development has been that the prevalence of mental disorders is highest in young people aged 18–24 years [11]. Overall the survey found that mental disorders were most common among those aged 16–24, especially substance use disorders. As Slade et al. report, treatment rates for mental disorders are low in this age group [12]. The New Zealand Mental Health Survey and other international surveys show that the age of onset for affective disorders and anxiety disorders is in the teenage years [13, 14]. Services for younger Australians must be a priority and youth services must target substance use disorders assertively.
In conjunction with the advocacy for youth mental health, there has been a major effort over the past decade to promote early intervention in the delivery of services [11]. The evidence of poorer outcomes with longer duration of untreated symptoms applies to common mental disorders, such as anxiety and depression, as well as for psychosis [15]. Although debate continues about the long-term effectiveness of early intervention programmes [16, 17], services that target this younger age group remain critical. This should be supported by research efforts to clarify what interventions produce the best long-term outcomes.
The survey also found that the prevalence for any affective disorder is now lower than the corresponding rates from the New Zealand and US prevalence studies [12]. It is unclear whether better awareness campaigns, treatment approaches or some other unidentified factors have led to this difference.
The NSMHWB 2007 found that the 12 month prevalence of suicidal thoughts, plans or actions was, as expected, much higher among those with mental disorders than among the general population. As Johnson et al. argue in their paper in this issue [18], this is particularly important because suicidal ideation is predictive of suicide and a major public health issue. Furthermore, they noted that the 12 month prevalence of suicide attempts was similar in 2007 and 1997, despite a slightly lower prevalence of suicidal ideation. Service use for mental health problems was higher among suicidal individuals than it was among the general population, but significant numbers of those experiencing suicidal thoughts and behaviours did not receive treatment.
Suicide prevention in Australia is being progressed by the renewed Living is for Everyone (LiFE) framework [19], which calls for further initiatives aimed at known risk groups in the general community. It is clear that one of those major risk groups is people with mental disorders, especially those who do not receive services. A better coordination of the activities of the National Mental Health Strategy and the National Suicide Prevention Strategy could help achieve the aim of getting more people with mental disorders and suicidal behaviours into appropriate treatment.
Mental disorders are often comorbid with each other and physical disorders
The survey confirmed that comorbidity between mental disorders is common [20], a finding that has been replicated in many countries. In the NSMHWB 2007 one-quarter of people with a 12 month disorder also had another mental disorder from a different class (anxiety, affective or substance use). Anxiety and affective disorders were the most common comorbidity. Comorbidity was associated with increased severity, days out of role and treatment rates.
Nearly half of female subjects and more than one-third of male subjects with substance use disorders also had an anxiety or affective disorder (or both). Given that comorbidity is so common, the continued separation of mental health services from drug and alcohol services does not make policy or programme delivery sense. This is highlighted by the finding from the survey that the treatment rate for comorbid anxiety and affective disorders is nearly 70%, whereas for comorbid anxiety and substance use it is 30% and for comorbid affective disorders and substance use it is 28%. Policy development should address the issue that parallel service streams have been created, whereby those people comorbid for mental illness and substance use disorders do not access services at a rate that would be suggested by the level of disability they experience. While the National Comorbidity Initiative has provided a focus on this area [21], more effort will be needed to improve the outcomes for people with comorbid mental and substance use disorders.
It is well established that people with mental disorders also have increased numbers of physical health problems. This has been especially recognized in the low-prevalence disorders such as schizophrenia. Lawrence et al. in Western Australia demonstrated that psychiatric patients have elevated mortality rates for every major cause of death [22], and a systematic review of the literature by Saha et al. found that the differential mortality gap between those with schizophrenia and the general community has worsened in recent decades [23]. The impact of mental disorders on outcomes for physical health problems has also been demonstrated, for example, the presence of an affective disorder in those with stroke may impede rehabilitation.
A striking finding of the NSMHWB 2007 is that there are higher rates of mental disorders in those with a National Health Priority Area (NHPA; i.e. diabetes, cancer, asthma, coronary heart disease, stroke or arthritis) compared to those without these disorders [20]. A total of 28% of those with at least one NHPA comorbidity (i.e. diabetes, cancer, asthma, coronary heart disease, stroke or arthritis) had a mental disorder, compared to only 17% of those without an NHPA. The policy imperative is to create systems in which people with chronic physical health disorders also have their mental disorders properly addressed, while also continuing our efforts to ensure better physical health care for those with mental disorders. This would require closer coordination between medical care and mental health care. One way of achieving this is to ensure the involvement of general practitioners in the management of patients; another is for mental health services to include physical health screening as a routine part of their service delivery, and finally, by ensuring that clinicians address both mental and physical conditions in order to get adequate treatment for this important comorbidity.
Mental disorders are often severe and disabling
What should governments and funders make of such a high population prevalence of mental disorders? Do all those meeting criteria need treatment? The survey provides some guide by assessing the severity and disability caused by mental disorders.
In 46% of those with a mental disorder, their condition was mild. There will be debate about the policy significance of this group and in the absence of longitudinal follow up we do not know how many of these cases may be transient. But mild disorders do not equate with trivial disorders. Diagnostic criteria must still be met and many mild disorders at the time of the survey may go on to become more severe. They should be a focus for early intervention.
People with mental disorders experience nearly 4 days in the previous 30 days when they are unable to carry out their usual activities. In comparison, people without a mental disorder report 1 day out of role in the previous 30 days. These findings emphasize not only the personal but also the social and economic impact of mental disorders. The loss of productivity caused by common mental disorders is increasingly recognized in Australian and international research [24]. Mental disorders make the largest contribution of all major health conditions to health-related labour force non-participation rates. Productivity Commission modelling has shown that of interventions to improve the two major components of human capital, health and education, averting the impact of mental disorders has the greatest potential to lift labour force participation rates [25].
Given that effective treatments exist for common mental disorders, much lost productivity can be averted [26]. Policy reform needs to be directed toward ensuring that people with mental disorders are given the greatest opportunity to re-enter the workforce and function at their highest level. In addition for those already in the workforce, screening is possible with established, well-validated tools that have scoring algorithms that correlate well with the presence of a mental disorder. Examples include instruments such as the Kessler-6 or Kessler-10, which do not require the time and expense of instruments such as the World Mental Health (WMH)-CIDI [27]. High-risk individuals can be identified by these instruments and early intervention initiated. Occupational health programmes should be aware that affective disorders were found in the NSMHWB 2007 to be more severe, cause more days out of role and interference with life, and may therefore constitute a priority for action. In the current global economic turmoil, increased levels of psychological distress among employees can be expected and this may, in some, eventually manifest as mental disorders, resulting in an even greater impact on productivity, as well as personal well-being. Employers should understand that untreated mental health problems among their employees will have an economic impact and there are programmes in which they can invest to reduce this impact [26].
Treatment rates are low
Despite the impact that mental disorders have on the individual, only one in three people with a disorder sought professional health for their condition. This is essentially the same as found in the NSMHWB 1997 and is a consistent finding in international studies. As Burgess et al. report in their paper in this issue, Australia's treatment rates are approximately in the middle of an international group of countries in which similar surveys have been conducted [28]. There is a very large variation in treatment rates among the different classes of mental disorder. Nearly half of those with an affective disorder alone, access treatment compared to <12% of those with a substance use disorder alone.
The lowest rates of treatment seeking were in those with substance use disorders. This is in stark contrast to the fact that, according to the Australian Institute of Health and Welfare, hospital separations for alcoholic liver disease have increased by 61% between 1998 and 2007 [29]. We are clearly failing to provide timely and effective interventions for this disorder. Australia-wide, alcohol abuse alone is estimated to result in an annual economic burden exceeding $7.5bn [30], and current estimates are that only 2% of the burden attributable to alcohol use disorders is averted by current treatment because treatment rates are so low [31].
More than 70% of those with any disorder who sought treatment consulted a general practitioner, with more than one-quarter being treated only by their general practitioner. This emphasizes the importance of primary mental health care and the critical role that general practice plays in the treatment of common mental disorders. Mental health professionals, however, are increasingly important in the delivery of services for Australians with a mental disorder.
Patterns of service use are changing. Compared with the 1997 survey there is an almost twofold increase in the number of persons consulting psychologists. In 2007 around 13% of people with a mental disorder had consulted a psychologist for a mental disorder, compared to around 6% in 1997. This is likely to be due to the dramatic uptake by allied mental health professionals of the Medicare Benefit Schedule (MBS) rebates under the Better Access to Psychiatrists, Psychologists and General Practitioners (Better Access) initiative introduced in November 2006 [7]. One could conclude that more psychological therapies are now being used in the treatment of mental disorders. While we need information on the quality of care being delivered by these providers, increasing access to clinicians such as psychologists who can deliver psychological treatments for anxiety and affective disorders (such as cognitive therapy and interpersonal therapy) is a positive development.
Young people (both male and female) were less likely to receive treatment. Young women were more likely than men to receive treatment. Young men (those aged 16–24) and older men (those aged 75–85) were the least likely to receive treatment. This argues for an ongoing emphasis to improve service utilization by young and elderly men. As would be expected, service use was highest in people with more severe disorders, greater comorbidity and in people who reported suicidal behaviours. It is clear, however, that a concerted effort is required to address treatment rates. In doing this we need to better understand the relationship between diagnosis and need for services.
Diagnosis and the need for services are very different
Understanding of why people with disorders do not access treatment and how to address this is needed to better target our services. Some results from the survey, discussed by Meadows and Burgess [32], provide information that helps understand this conundrum.
The perspective of the person with the mental disorder is important in gaining an understanding as to whether treatment is successful, and why the treatment rates are much lower than the prevalence of the disorders. A total of 78% of people with a 12 month disorder and a perceived need for care considered that their needs were wholly (37%) or partially (41%) met. So, and this is no surprise, our services can and should do better for those in treatment.
Among those with a 12 month mental disorder, however, more than half (57%) did not perceive a need for care. Those with a mental disorder who were not receiving treatment were asked whether they needed certain forms of help, such as counselling or medications. A total of 86% said that they did not need any type of help (information, counselling, medication, social intervention or skills training). Further analysis of the data from the survey, supplemented by data from other sources, is required to better understand what is behind people's responses and how to best improve access to services.
The fact that 86% of those with a disorder not receiving care do not think they need any help, has major implications for policy and service development. Simply improving access to more providers (e.g. through the MBS) will not address the broader unmet need. One could argue that much more needs to be done in providing information to improve mental health literacy, including the recognition of symptoms that constitute a mental disorder and knowledge about available and effective treatments. As Meadows and Burgess also argue, the priority for improving treatment rates should be those with disorder and a perceived need for care, because a demand for services would be expected from this group [32].
In contrast, nearly 5.5% of those receiving mental health services have no lifetime diagnosis. Caution should be exercised in automatically assuming that this group are inappropriately consuming scarce resources. Some of these individuals are likely to have one of the mental disorders not covered in the survey (personality disorders, eating disorders and psychosis). Others may also have psychiatric symptoms that fall short of meeting full diagnostic criteria, but are accessing treatment, and this could constitute early intervention that prevents the progression to a mental disorder.
Be careful making comparisons with the NSMHWB 1997
There has been a tendency in some quarters to rush into a comparison between the findings of the 1997 and 2007 surveys. Comparison of the two surveys will need to be done, by academics, policy makers and others. But it is important that when this is done, the limitations are known and like is being compared with like.
As emphasized by Slade et al., there are important differences between the surveys [12]. The surveys used different versions of the CIDI. The change in versions changed the number and content of questions used to tap the diagnostic criteria, the structure of the interview (specifically with regard to the placement of diagnostic screener questions in a separate early module) and changed the sequencing of questions within some diagnostic modules. Even small changes to the wording of a questionnaire can result in large differences in the extent and type of information elicited from respondents. Importantly the 2007 survey estimates lifetime prevalence, introducing problems in relation to recall bias.
In addition the response rate of the 2007 survey was considerably lower (60%) than that for the 1997 survey (78%). People who did not participate in the survey may have had a higher (or lower) likelihood of meeting criteria for a mental disorder. Although a non-response follow-up study was conducted that suggested that errors in estimation are likely to be small at an aggregate level, the 40% non-response rate means that caution should be exercised in generalizing the findings to the entire Australian population.
Conclusion
The 2007 NSMHWB provides unique data, particularly with regard to the complex issues of prevalence, disability and service use for mental disorders. But it does not tell the whole story. The 1997 NSMHWB was accompanied by two other surveys: a low-prevalence survey [33] and a child and adolescent survey of those aged 4–17 years [34]. The ages covered in the 2007 NSMHWB was lowered from 18 years in 1997 to 16 years in 2007 to ensure that there was some cross-over in the data.
Both of these additional 1997 surveys added considerably to the evidence base for mental health reform. Low-prevalence disorders, such as schizophrenia, consume much of the resources of public sector mental health services. They also require considerable input from other social services including housing, disability support services and income support. The treatment of mental disorders in children and adolescents remains an area of considerable under-investment. Most mental disorders begin in childhood or adolescence [14], and the evidence base for prevention of mental disorders is strongest in children [35].
Up-to-date information on the prevalence, disability and service utilization of low-prevalence disorders and child and adolescent disorders is urgently needed to support service development and reform, including that being undertaken through the Council of Australian Governments (COAG) National Action Plan for Mental Health, as well as informing the way in which the Fourth National Mental Health Plan is implemented. Until such information is available, information on one-quarter of the population, covering years in which mental disorders first emerge for many people, remains deficient.
The 2007 NSMHWB also provides us with epidemiological data that can be used to inform burden of disease estimates and link these with cost-effectiveness of health sector interventions to develop essential packages of clinical and preventative care [36]. To significantly increase treatment rates for common mental disorders it will be necessary to consider non-face-to-face treatments, including Internet- and telephone-delivered therapy, where this is shown to be cost-effective [37, 38]
Preliminary analysis of the NSMHWB 2007 suggests that mental disorders are more common in those who were unemployed and in those who had been homeless and in jail. Given that the survey was of households, this means that homeless people, people resident in nursing homes, hostels, and hospices and those in prison or other corrective service facilities were not included. The prevalence of mental disorders is higher in these groups and they are heavy users of health and social services. In addition, people who did not speak English were excluded, and their needs may be different from the surveyed population. More information is needed about these subpopulations to assist service planning.
Nevertheless, for improving the outcome of Australian adults with common mental disorders, the 2007 NSMHWB is a major achievement and will be of enormous benefit to the sector. As Insel and Fenton point out, psychiatric epidemiology is not just about counting [39]. The breadth of the content of the survey provides scope for further analyses, which will provide invaluable information on the complex relationship between symptomatology, diagnosis, comorbidity, the experience of mental disorder, perceived needs for care and use of services in rural Australia and by other underserved populations. These analyses will need to take into account the technical issues involved in the use of survey instruments such as the CIDI and how this impacts on the interpretation of the data [40, 41]. The results of these analyses will nevertheless provide vital information on the impact of mental disorders in our society and guide the delivery of services in the next decade and beyond.
Footnotes
Acknowledgements
The authors thank Philip Burgess, Jane Pirkis, Maree Teesson, Tim Slade, Suzy Saw, Brian Kelly and Graham Meadows for their comments on earlier drafts of this paper. The views expressed in this paper are those of the authors and do not necessarily reflect the position of the Commonwealth Department of Health and Ageing (where Professor Whiteford is the Principal Mental Health Advisor), Queensland Health or the Australian National Mental Health Standing Committee.