Professor Tom Burns in this edition of the journal has helpfully summarised changes and progress made in community psychiatry over at least the last Half-Century. This commentary from an author in Melbourne, Australia will pick up some key points for discussion, suggesting comparisons and contrasts with mental health service developments in Australia.
The decline in inpatient numbers that Burns notes is even more dramatic when considered as a population rate. Between the 1950s and 1994 the population of the USA increased from about 165 million to about 265 million (United States Census Bureau, 2014), and between the 1950s and now, that of the UK increased from about 52 million to about 63 million (Office for National Statistics, 2014). Hence the change in bed numbers considered as a population rate is approximately a tenfold reduction in the USA (from 302 per hundred thousand to 27 per hundred thousand); a little less in the UK (from 290 per hundred thousand to 43 per hundred thousand). Australia's bed count peaked about 30 000 in 1965 (Commonwealth of Australia, 2007) (260 per hundred thousand) and by 2010–2011 was 6755 (30 per hundred thousand) (Department of Health and Ageing, 2013). Such reduction is not uniform across the world, there is considerable variation in remaining bed numbers even within Europe (Priebe et al.
Reference Priebe, Badesconyi, Fioritti, Hansson, Kilian, Torres-Gonzales, Turner and Wiersma2005) and more in low- and middle-income countries (Saxena et al.
Reference Saxena, Lora, Morris, Berrino, Esparza, Barrett, Van Ommeren and Saraceno2011). It has been argued that there is some evidence of a degree of re-institutionalisation occurring in parts of the European continent (Priebe et al.
Reference Priebe, Badesconyi, Fioritti, Hansson, Kilian, Torres-Gonzales, Turner and Wiersma2005) and that more people with mental illness may be finding themselves in custody or forensic psychiatric settings – a contemporary example of operation of Penrose's law (‘the population size of prisons and psychiatric hospitals are inversely related’ see Penrose, Reference Penrose1939). Rates of psychosis in the UK prisons at least are well above those in the community (Brugha et al.
Reference Brugha, Singleton, Meltzer, Bebbington, Farrell, Jenkins, Coid, Fryers, Melzer and Lewis2005). This is a possible adverse consequence of increased care in communities and one against which constant vigilance is needed.
In his review, Burns describes the development of the evidence base through to the 1980s then forward from that decade. From an Australian perspective, a dramatic punctuation in Australian mental health care policy and service delivery around 1993 probably illustrates well a response to some challenges described in Burns's summary (Gerrand et al.
Reference Gerrand, Singh, Nagel, Ash, Turnbull, Meadows, Fossey, Grigg, Rosen, Weir, Benson, Meadows, Singh and Grigg2012). In 1992, Australia adopted the National Mental Health Policy and the first 5-year national plan for mental health services commenced in 1993. The plan included a drive towards accelerating deinstitutionalisation along with development of community services. Australia, then, entered a phase that lasted some 15 years during which strong Federal investment in innovation and promoting restructuring in mental health care led introduction and dissemination of novel ideas and interventions. Initiatives such as substantial national incentive funding, the forum for interchange between service providers, consumers and carers provided by The Mental Health Services (TheMHS) conferences and the recognition and encouragement provided by National Mental Health Services Achievement awards (Gerrand et al.
Reference Gerrand, Singh, Nagel, Ash, Turnbull, Meadows, Fossey, Grigg, Rosen, Weir, Benson, Meadows, Singh and Grigg2012) played their part then and still do. While variable in its effects in the States and Territories, and followed more recently by a pivot back towards developments led from this lower governmental level (Ash et al.
Reference Ash, Bland, Brown, Browne, Burvill, Davies, Grigg, Gurr, Hughson, Meadows, Nagle, Rosen, Scott, Singh, Smith, Weir, Meadows, Singh and Grigg2012), the Australian National Mental Health Strategy has represented an example of a strongly translational approach to developing mental health care nationally and one that can be credited with some substantial successes.
Burns in the paper describes Europe adopting a highly sectorised approach to planning. The idea of organising mental health care services in defined geographical catchment areas or sectors has indeed been influential internationally but implemented in very varying ways. As an example, in Victoria, Australia, in the 1990s a sectorised approach to mental healthcare delivery was introduced, with typical sector sizes of 200 000 to 215 000, less in rural or remote areas (Meadows & Singh, Reference Meadows and Singh2003). In Australia, State funded services only constitute part of the mental health care delivery funded from the national tax pool, since a parallel fee-for-service system is Federally administered. This may tend to drive sector-based planning towards larger population bases to assure critical mass for specific service components. Here, then the continuity provided by clinicians working across inpatient and community settings has less typically been the model of care provision. A key advantage of the sectorised service principal is the ability to apply systematic measurement of population health needs to guide resource distribution. In passing it may be noted that this is much easier in capitation-funded healthcare systems such as the British National Health Service; in Australia where, as introduced earlier, part of government funded mental health care provision is on a fee-for-service basis allowing co-payment, equity is more difficult to engineer (Meadows et al.
Reference Meadows, Singh, Burgess and Bobevski2002; Meadows & Tylee, Reference Meadows and Tylee2013). In 2010, a review of formulae used to guide resource distribution in the UK could identify six such indices (Tulloch & Priebe, Reference Tulloch and Priebe2010). An extended amount of work was also done on this in the USA including systematic formal evaluation of the function of six such indicators in the Colorado Social Health Survey. (Ciarlo et al.
Reference Ciarlo, Shern, Tweed, Kirkpatrick and Sachs-Ericsson1992; Tweed et al.
Reference Tweed, Ciarlo, Kirkpatrick and Shern1992). Here in Victoria, Australia, there was also development of formal resource distribution process along these lines and this was perhaps critically important in the reasonably effective transition from institutionalised to deinstitutionalised care (Meadows, Reference Meadows1997). However, in Victoria, it is at least 15 years since such a formula approach was explicitly and transparently used for allocation or distribution of mental health resources. It could be argued that in planning terms the incorporation of systematic assessment of the geographic dimension (Thornicroft & Tansella, Reference Thornicroft and Tansella2009) has been another significant achievement of community psychiatry. Developing services across a national or regional planning environment where there is considerable inequity in resource distribution will make it hard or impossible to put in place a standardised model of care delivery operating with comparable boundary conditions. Neglecting such systematic planning processes could invite a progressive creep away from equitable service delivery as demography and associated needs for care change.
As Burns moves on to talk more about developing research evidence, perhaps it is germane to note challenges for research in this area that have played some part in limiting development of the evidence base. Many relevant studies are by necessity Cluster Randomised Controlled Trials (RCTs); these have their particular design challenges – statistical and methodological. It is now widely understood that statistically there are implications for power calculations and that there is a need for appropriate and relatively advanced analytic approaches. Considering some design issues however, there are several additional threats to internal and external validity of findings. Usually the control condition involves no intervention – so there is typically no analogue to the placebo control often used in individual RCT designs. In a psychotherapeutic intervention trial, this might be carefully matched for duration and treatment expectation. In contrast in a cluster RCT, teams may not receive any intervention in an analogue of a Treatment as Usual comparison group, with limited opportunity for blinding. While possible effects on the intervention group, often considered as Hawthorne effects (performance can change in response to a change in the environment rather than in response to the nature of the change itself), are often considered, there also are possible effects on the non-intervention teams or services. These may experience demoralisation and function at a lower level – with a risk of bias favouring the intervention condition. Alternatively they may develop a stronger motivation to outcompete the teams selected for the intervention. More generally this is described as the John Henry effect, in which members of a control group may actively work harder to overcome the ‘disadvantage’ of being in the control group (Colman, Reference Colman2008). In that case, the advantage of the intervention may be obscured and this source of bias may have contributed to some less than encouraging findings from studies in this field. There a need progressively to refine design approaches here. Stepped wedge designs (Hussey & Hughes, Reference Hussey and Hughes2007) may provide opportunities better to manage the expectations related to team or service-level interventions since all participating service elements will receive intervention at some point. Where issues of demoralisation or instances of the John Henry effect are threats to validity then some variant of a stepped wedge design may represent an advance over a simple parallel-groups cluster RCT
Burns notes that the State of Victoria is an outlier in terms use of community treatment orders. At the time of writing, Victoria as a state is about to embark on major transitions in how these orders are managed, with a new mental health act (Parliament of Victoria, 2014). This act lays much more premium on the notion of recovery and of presumed capacity as well as allowing for advanced statements on promoting greater involvement of patients in decision-making. Whether this results in a change in how community treatment orders are implemented in this State is, along with so much in this field, very much a work in progress.
You have
Access
Professor Tom Burns in this edition of the journal has helpfully summarised changes and progress made in community psychiatry over at least the last Half-Century. This commentary from an author in Melbourne, Australia will pick up some key points for discussion, suggesting comparisons and contrasts with mental health service developments in Australia.
The decline in inpatient numbers that Burns notes is even more dramatic when considered as a population rate. Between the 1950s and 1994 the population of the USA increased from about 165 million to about 265 million (United States Census Bureau, 2014), and between the 1950s and now, that of the UK increased from about 52 million to about 63 million (Office for National Statistics, 2014). Hence the change in bed numbers considered as a population rate is approximately a tenfold reduction in the USA (from 302 per hundred thousand to 27 per hundred thousand); a little less in the UK (from 290 per hundred thousand to 43 per hundred thousand). Australia's bed count peaked about 30 000 in 1965 (Commonwealth of Australia, 2007) (260 per hundred thousand) and by 2010–2011 was 6755 (30 per hundred thousand) (Department of Health and Ageing, 2013). Such reduction is not uniform across the world, there is considerable variation in remaining bed numbers even within Europe (Priebe et al. Reference Priebe, Badesconyi, Fioritti, Hansson, Kilian, Torres-Gonzales, Turner and Wiersma2005) and more in low- and middle-income countries (Saxena et al. Reference Saxena, Lora, Morris, Berrino, Esparza, Barrett, Van Ommeren and Saraceno2011). It has been argued that there is some evidence of a degree of re-institutionalisation occurring in parts of the European continent (Priebe et al. Reference Priebe, Badesconyi, Fioritti, Hansson, Kilian, Torres-Gonzales, Turner and Wiersma2005) and that more people with mental illness may be finding themselves in custody or forensic psychiatric settings – a contemporary example of operation of Penrose's law (‘the population size of prisons and psychiatric hospitals are inversely related’ see Penrose, Reference Penrose1939). Rates of psychosis in the UK prisons at least are well above those in the community (Brugha et al. Reference Brugha, Singleton, Meltzer, Bebbington, Farrell, Jenkins, Coid, Fryers, Melzer and Lewis2005). This is a possible adverse consequence of increased care in communities and one against which constant vigilance is needed.
In his review, Burns describes the development of the evidence base through to the 1980s then forward from that decade. From an Australian perspective, a dramatic punctuation in Australian mental health care policy and service delivery around 1993 probably illustrates well a response to some challenges described in Burns's summary (Gerrand et al. Reference Gerrand, Singh, Nagel, Ash, Turnbull, Meadows, Fossey, Grigg, Rosen, Weir, Benson, Meadows, Singh and Grigg2012). In 1992, Australia adopted the National Mental Health Policy and the first 5-year national plan for mental health services commenced in 1993. The plan included a drive towards accelerating deinstitutionalisation along with development of community services. Australia, then, entered a phase that lasted some 15 years during which strong Federal investment in innovation and promoting restructuring in mental health care led introduction and dissemination of novel ideas and interventions. Initiatives such as substantial national incentive funding, the forum for interchange between service providers, consumers and carers provided by The Mental Health Services (TheMHS) conferences and the recognition and encouragement provided by National Mental Health Services Achievement awards (Gerrand et al. Reference Gerrand, Singh, Nagel, Ash, Turnbull, Meadows, Fossey, Grigg, Rosen, Weir, Benson, Meadows, Singh and Grigg2012) played their part then and still do. While variable in its effects in the States and Territories, and followed more recently by a pivot back towards developments led from this lower governmental level (Ash et al. Reference Ash, Bland, Brown, Browne, Burvill, Davies, Grigg, Gurr, Hughson, Meadows, Nagle, Rosen, Scott, Singh, Smith, Weir, Meadows, Singh and Grigg2012), the Australian National Mental Health Strategy has represented an example of a strongly translational approach to developing mental health care nationally and one that can be credited with some substantial successes.
Burns in the paper describes Europe adopting a highly sectorised approach to planning. The idea of organising mental health care services in defined geographical catchment areas or sectors has indeed been influential internationally but implemented in very varying ways. As an example, in Victoria, Australia, in the 1990s a sectorised approach to mental healthcare delivery was introduced, with typical sector sizes of 200 000 to 215 000, less in rural or remote areas (Meadows & Singh, Reference Meadows and Singh2003). In Australia, State funded services only constitute part of the mental health care delivery funded from the national tax pool, since a parallel fee-for-service system is Federally administered. This may tend to drive sector-based planning towards larger population bases to assure critical mass for specific service components. Here, then the continuity provided by clinicians working across inpatient and community settings has less typically been the model of care provision. A key advantage of the sectorised service principal is the ability to apply systematic measurement of population health needs to guide resource distribution. In passing it may be noted that this is much easier in capitation-funded healthcare systems such as the British National Health Service; in Australia where, as introduced earlier, part of government funded mental health care provision is on a fee-for-service basis allowing co-payment, equity is more difficult to engineer (Meadows et al. Reference Meadows, Singh, Burgess and Bobevski2002; Meadows & Tylee, Reference Meadows and Tylee2013). In 2010, a review of formulae used to guide resource distribution in the UK could identify six such indices (Tulloch & Priebe, Reference Tulloch and Priebe2010). An extended amount of work was also done on this in the USA including systematic formal evaluation of the function of six such indicators in the Colorado Social Health Survey. (Ciarlo et al. Reference Ciarlo, Shern, Tweed, Kirkpatrick and Sachs-Ericsson1992; Tweed et al. Reference Tweed, Ciarlo, Kirkpatrick and Shern1992). Here in Victoria, Australia, there was also development of formal resource distribution process along these lines and this was perhaps critically important in the reasonably effective transition from institutionalised to deinstitutionalised care (Meadows, Reference Meadows1997). However, in Victoria, it is at least 15 years since such a formula approach was explicitly and transparently used for allocation or distribution of mental health resources. It could be argued that in planning terms the incorporation of systematic assessment of the geographic dimension (Thornicroft & Tansella, Reference Thornicroft and Tansella2009) has been another significant achievement of community psychiatry. Developing services across a national or regional planning environment where there is considerable inequity in resource distribution will make it hard or impossible to put in place a standardised model of care delivery operating with comparable boundary conditions. Neglecting such systematic planning processes could invite a progressive creep away from equitable service delivery as demography and associated needs for care change.
As Burns moves on to talk more about developing research evidence, perhaps it is germane to note challenges for research in this area that have played some part in limiting development of the evidence base. Many relevant studies are by necessity Cluster Randomised Controlled Trials (RCTs); these have their particular design challenges – statistical and methodological. It is now widely understood that statistically there are implications for power calculations and that there is a need for appropriate and relatively advanced analytic approaches. Considering some design issues however, there are several additional threats to internal and external validity of findings. Usually the control condition involves no intervention – so there is typically no analogue to the placebo control often used in individual RCT designs. In a psychotherapeutic intervention trial, this might be carefully matched for duration and treatment expectation. In contrast in a cluster RCT, teams may not receive any intervention in an analogue of a Treatment as Usual comparison group, with limited opportunity for blinding. While possible effects on the intervention group, often considered as Hawthorne effects (performance can change in response to a change in the environment rather than in response to the nature of the change itself), are often considered, there also are possible effects on the non-intervention teams or services. These may experience demoralisation and function at a lower level – with a risk of bias favouring the intervention condition. Alternatively they may develop a stronger motivation to outcompete the teams selected for the intervention. More generally this is described as the John Henry effect, in which members of a control group may actively work harder to overcome the ‘disadvantage’ of being in the control group (Colman, Reference Colman2008). In that case, the advantage of the intervention may be obscured and this source of bias may have contributed to some less than encouraging findings from studies in this field. There a need progressively to refine design approaches here. Stepped wedge designs (Hussey & Hughes, Reference Hussey and Hughes2007) may provide opportunities better to manage the expectations related to team or service-level interventions since all participating service elements will receive intervention at some point. Where issues of demoralisation or instances of the John Henry effect are threats to validity then some variant of a stepped wedge design may represent an advance over a simple parallel-groups cluster RCT
Burns notes that the State of Victoria is an outlier in terms use of community treatment orders. At the time of writing, Victoria as a state is about to embark on major transitions in how these orders are managed, with a new mental health act (Parliament of Victoria, 2014). This act lays much more premium on the notion of recovery and of presumed capacity as well as allowing for advanced statements on promoting greater involvement of patients in decision-making. Whether this results in a change in how community treatment orders are implemented in this State is, along with so much in this field, very much a work in progress.