Prevalence of Mental Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba

Transcription

1 Prevalence of tal Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba Patricia J Martens, PhD 1, Randall Fransoo, MSc 2, The Need To Know Team 3, Elaine Burland, MSc 4, Charles Burchill, MSc 5, Heather J Prior, MSc 6, Okechukwu Ekuma, MSc 7 Objectives: To determine the prevalence of mental illness in older adults and its effect on home care and personal care home (PCH) use. Methods: Using nonidentifying administrative records (fiscal years to ) from the Population Health Research Data Repository housed at the Manitoba Centre for Health Policy, we determined the 5-year period prevalence for individuals aged 55 years and over ( men and women) for 3 mental illness categories: cumulative mental disorders (those having a diagnosis of depression, anxiety disorder, personality disorder, schizophrenia, and [or] substance abuse), any mental illness, and dementia. We calculated age-specific and age-adjusted rates of home care and PCH use and the prevalence of mental illness in PCH residents. Results: From the group aged 55 to 59 years to the group aged 90 years or older, the prevalence of mental illness increased with the population s age. The prevalence of any mental illness rose from 32.4% to 45.0% in men and from 42.6% to 51.9% in women, and dementia prevalence rose from 2.0% to 33.6% in men and from 1.3% to 40.3% in women. The age-adjusted annual rates of open home care cases per 1000 population aged 55 and older varied by mental illness grouping (no mental disorder, 57 for men and 91 for women; cumulative mental disorders, 162 for men and 191 for women; dementia, 300 for men and 338 for women). The age-adjusted rates of PCH use per 1000 population aged 75 years and older also varied by mental illness grouping (no mental disorder, 53 for men and 78 for women; cumulative mental disorders, 305 for men and 373 for women; dementia, 542 for men and 669 for women). Among patients admitted to (or resident in) a PCH in , 74.6% (87.1%) had a mental illness, and 46.0% (69.0%) had dementia. Conclusions: tal illness affects the use of home care and nursing homes profoundly. Individuals with dementia used home care at 3 times the rate of those having no mental illness diagnosis, and they used PCHs at 8 times the rate. (Can J Psychiatry 2007;52: ) Information on funding and support and author affiliations appears at the end of the article. Clinical Implications The high 5-year period prevalence (one-third to one-half) of any mental disorder in older adults indicates the need for appropriate and accessible mental health care resources. Dementia is associated with very high use of community services, and the aging population will continue to affect this use. Health care providers and lay staff working in community settings (both home care and PCH settings) need to receive training in mental disorders and to offer appropriate services. The Canadian Journal of Psychiatry, Vol 52, No 9, September

2 Limitations This study is based on administrative claims data and not on chart audit or direct diagnosis of the population. The comparative rates of use of community services are based on cross-sectional data, which carry the limitation of association rather than causation. The study could only be replicated in jurisdictions having population-based data files similar to those in Manitoba. Key Words: mental illness, home care, nursing homes, institutional care, dementia, sex differences, health services research, administrative claims data, population-based rates, Manitoba Centre for Health Policy tal illness is a profound problem worldwide, yet population-based information on its prevalence and resource use by those affected is lacking. This research study focuses on community service use patterns of older adults with mental illness specifically, the use of home care and personal care (nursing) homes. A 2002 Canadian study on mental illness highlighted the need to understand the use of community services beyond the acute care settings. 1 Senator Michael Kirby subsequently chaired a Canada-wide Senate consultation on mental illness, producing an extensive report in 2006, entitled Out of the Shadows at Last. 2 The report recommendations included the necessity of providing home care to people with mental illness and making this service equivalent to that received by people with physical ailments. In addition, the issue of appropriate housing was addressed, with recommendations including a shift from acute care to long-term care facilities and appropriate housing options, as well as staffing competencies and training programs to ensure clinically appropriate care for seniors experiencing mental illness. The importance of helping aged couples to live together or in close proximity, regardless of the level of services required, was also underscored. Manitoba is fortunate to have access to many data sources (specifically, community, hospital, and mental health Abbreviations used in this article CI CIHR FY ICD-9-CM MCHP MHMIS PCH RHA confidence interval Canadian Institutes of Health Research fiscal year International Classification of Diseases, 9th revision, clinical modification Manitoba Centre for Health Policy tal Health Management Information System personal care home Regional Health Authority facilities) that can assist in providing a more accurate picture of the extent and implications of mental illness. The collaborative researcher planner group known as The Need To Know Team identified the need for an overview of population-based indicators of mental illness (that is, prevalence as well as patterns of health care use) to assist Manitoba s RHAs in planning and decision making. The Need to Know Team comprises researchers from the University of Manitoba s health services and population health research unit (the MCHP), planners from Manitoba Health, and high-level planners from each RHA. The Need To Know project strengthens the capacity of the academics to carry out research that is relevant to rural and northern RHAs and also strengthens the capacity of team members to understand, interpret, and apply research at the planning and decision-making level. 3 5 The objective of the study reported here was to investigate how mental illness affects older adults use of home care and PCHs, with an a priori hypothesis that having a mental illness increases the likelihood of use. The Manitoba Home Care Program, established in 1974, is the oldest program of its kind in Canada. It is a comprehensive, provincially funded universal program with 2 objectives. First, it provides services to individuals with inadequate informal resources to return home from hospital or to remain in the community. Second, it assesses and places individuals in long-term care facilities (such as nursing homes) if and when home care services cannot maintain them safely and (or) economically at home (home care services are provided until they are placed). 6 Reassessments at predetermined intervals provide the basis for decisions by case managers to discharge individuals or to change the type or amount of service delivered. Available Canadian data on the use of home care are currently limited to self-reported rates of home care use for selected provinces, derived from the Canadian Community Health Survey. 7 Statistics Canada estimated that 2.7% of the population aged 18 years and older used home care in the fiscal year (2.2% of men and 3.4% of women). These 582 La Revue canadienne de psychiatrie, vol 52, no 9, septembre 2007

3 Prevalence of tal Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba estimates were based on data from Prince Edward Island, Nova Scotia, New Brunswick, and Quebec. There are no data specifically comparing case rates for individuals with and without mental illness. The rate of home care use in Manitoba for all ages was 2.7% in Among those aged 10 years and older, 7% of Manitobans (n = ) received home care from to PCHs, also known as long-term institutions or nursing homes, are residential facilities for individuals with chronic illness or disability particularly those who have mobility and eating problems. 10 In , there were 9105 licensed PCH beds in Manitoba. 11 There were (3040 male and 7252 female) Manitoba residents aged 75 years and older who were newly admitted to a PCH during to , with a total of residents during the 5 years. 9 The prevalence of mental illness in PCHs is high. In a US study of 454 PCHs, 34.9% of the residents experienced senile dementias. 12 As well, dementia and other psychiatric disorders were present in about 80% of the new admissions. In another US study, 51% of the residents in nursing homes had dementia, and 4% had schizophrenia. 13 The Canadian Study of Health and Aging and its provincial sample, the Manitoba Study of Health and Aging, began in 1991 and followed a cohort of community-dwelling individuals aged 65 years and older, collecting data through interviews, medical examinations, and PCH records. 14 Several studies derived from this cohort have determined factors that were independently associated with admission to PCHs, including depressive symptoms, being female, being unmarried, having dementia and other cognitive impairments, having functional impairment or certain physical diseases (such as diabetes, stroke, and Parkinson disease), and the state of caregiving (such as a caregiver s presence or absence, age, burden, and state of depression) Factors associated with home care use differed, with one study showing only a weak association between cognitive status and the use of home care services by older adults and another demonstrating that dementia diagnoses increased the use of home care, whereas other cognitive impairments did not. 18,19 Methods This study used nonidentifying administrative claims data from the Population Health Research Data Repository (the Repository) housed at the MCHP. The study included a cohort of all Manitoba residents aged 55 years and older ( men and women) who had lived in Manitoba for at least 1 year over the 5 fiscal years from to For the purposes of home care and PCH use analyses, we determined crude age-specific rates for age categories above 54 years as well as an aggregate rate for home care (ages 55 and older) and for PCH (ages 75 and older), reflecting the respective ages at which the use of home care and nursing homes increases rapidly. We used SAS Version 8.0 (SAS Inc, Cary, NC, 1999) to perform the statistical testing and calculate rates and 95%CIs. The Repository data files for this research included hospital claims (records of hospital admissions), medical claims (records of visits to physicians apart from those occurring to a hospital inpatient), home care (records of the use of provincial home care services), PCHs (records of the use of nursing homes), registry files (records of the time a person was registered as a resident of Manitoba, as well as his or her age, sex, and area of residence), vital statistics (records of births and deaths), pharmaceutical claims (pharmaceutical use from the Drug Program Information Network), and the MHMIS, which is a record of mental health community services (including psychiatrist and psychologist diagnoses) and institutional services for Selkirk and Eden tal Health Services. Although records had been made anonymous, all files were linkable through the use of an encrypted personal health number and crosswalks between databases. We derived the diagnoses from physician, hospital, MHMIS, and pharmaceutical files and used home care and PCH files to determine the use of these services. Ethical approval was obtained from the University of Manitoba s Faculty of Medicine Human Research Ethics Board and the provincial Health Information Privacy Committee. We obtained permission to use the MHMIS from Medical Directors for the psychiatric facilities in Manitoba and from the Chief Provincial Psychiatrist. All data came from nonidentifying secondary administrative files, with no primary data collected from any individual in the study. We defined 5-year period prevalence of mental illness as the percentage of the population diagnosed with a certain condition at any time during fiscal years to , according to ICD-9-CM coding. This was actually treatment prevalence because, to be included, an individual must have contacted the health care system and received a diagnostic coding of a mental illness disorder within the 5-year period. Therefore, undiagnosed mental illness would not have been included in the Repository data. However, this may not present as great a problem in the population admitted to either home care or PCH services because admittance is normally preceded by a physician visit and referral. We considered it important to make this research comparable to similar Canadian reports using administrative data. 1,20 However, other reports are often limited to hospital records only, whereas our research included nonhospital claims to derive more population-based estimates. Physician claims are limited to the use of 3-digit coding without the additional decimals, making it difficult to detect some specific diagnostic categories. For example, bipolar could not be distinguished The Canadian Journal of Psychiatry, Vol 52, No 9, September

4 from depression, and pharmaceutical use patterns could not assist in specifying diagnoses because many drugs are used for both conditions. Consultations with pharmaceutical experts helped determine the appropriate drug choices to identify such conditions as depression on the basis of drug patterns for the 5-year study period. When we compared treatment prevalence estimates obtained through our definitions with other study estimates, we found that all were similar to other studies except for a possible underestimate of personality disorders. For a complete discussion of validity issues for the diagnostic categories, see Appendices 3 and 6 of the web-accessible MCHP report. 9 To assist in a broad understanding of the use of health care services by individuals with mental illness, we defined 4 groupings of the cohort: those with any mental disorder, those with no mental disorder, those with dementia, and those with cumulative mental disorders (see definitions below). The group having any mental illness was defined as those individuals having at least one diagnosis of ICD-9-CM 290 to 319 between and The group with no mental disorders had been assigned none of these diagnostic codings in the 5 years. The dementia group was defined as individuals having at least one diagnosis of ICD-9-CM 290 to 292 (organic psychotic conditions), 294 (other organic psychotic conditions), 331 (cerebral degenerations), or 797 (senility) within the study period. Dementia is difficult to code with administrative claims data, and the possibility of substantial undercoding is well known We consulted with geriatricians and epidemiologists specializing in research on the older adult to define dementia for our study. This definition is meant to be inclusive (higher sensitivity) rather than exclusive (specificity). ICD-9-CM codes 291 and 292 have some decimal codings that are clearly dementia (291.1, 291.2, ) but others that are not. Codings 291 and 292 only contribute 5% of the total dementia cases to the cohort; thus, some may be false positives, whereas others will be true positives. However, this would contribute in only a minor way to the overall numbers, and the benefit lies in not underestimating the prevalence. The fourth grouping, referred to as the cumulative mental disorders group, had been assigned one or more diagnostic codings in the 5-year period, representing at least one of the following: depression, anxiety disorder, personality disorder, substance abuse, or schizophrenia. Depression was defined in the hospital or MHMIS files as either any of ICD-9-CM codes to (affective psychoses), (neurotic depression), 309 (adjustment reaction), or 311 (depressive disorder); or an ICD-9-CM code 300 (neurotic disorders) plus a prescription for an antidepressant or mood stabilizer (excluding the antianxiety drugs paroxetine, citalopram, and venflaxamine, because of their possible use for mental illnesses apart from depression). In the physician files, depression was defined as either any of ICD-9-CM codes 296, 309, or 311; or ICD-9-CM code 300 plus a prescription for an antidepressant or mood stabilizer (excluding the antianxiety drugs paroxetine, citalopram, and venflaxamine). Anxiety disorders were determined by the presence of one or more ICD-9-CM codes (anxiety states), (phobic disorders), or (obsessive compulsive disorders) in hospital abstracts or MHMIS files or by a physician coding of 300 at least 3 times in the 5-year span. Personality disorders were defined as ICD-9-CM code 301 (personality disorders). Substance abuse was defined as any of ICD-9-CM codes 291 (alcoholic psychoses), 292 (drug psychoses), 303 (alcohol dependence), 304 (drug dependence), or 305 (nondependent abuse of drugs). Schizophrenia was defined as ICD-9-CM code 295 (schizophrenic disorders). Many people diagnosed with mental illness experience comorbid conditions. For Manitobans aged 10 years and over (n = from to ), 24.01% of the population were in the cumulative mental disorders group. Of those in the cumulative group, 70.8% (n = ) had only a single diagnosis of the 5 (that is, depression, anxiety disorder, substance abuse, schizophrenia, or personality disorder), and the remaining 29.2% (n = ) had 2 or more such diagnoses. Moreover, of the 36.89% of the population with any mental illness, 71.0% (n = ) had only a single diagnosis, and the remaining 29.0% (n = ) had 2 or more mental illness diagnoses. 9 We defined home care use as the average annual number of open home care cases per 1000 residents over the 5-year period. A case may remain open when home care is provided intermittently or when the individual is hospitalized for a period. Also, a resident may have had more than one episode of home care in the study period; these were counted as separate cases. The age-adjusted rate reflects the population of Manitoba on December 31, One limitation was that, in certain large buildings housing older adults in Winnipeg, block home care services were provided but not individually recorded. Therefore, although the service counts were available, it was not possible to assign specific home care service use to specific people within the building. We defined PCH use as the number of individuals living in a PCH for any portion of the 5-year period ( to ) per 1000 Manitobans in that age category. Once admitted, almost all PCH residents remain in a PCH until death (although there will be transfers between PCHs). We calculated age-specific rates for those aged 55 years and older and an overall rate for those aged 75 years and older. Individuals aged 75 years and older make up 87% of all PCH 584 La Revue canadienne de psychiatrie, vol 52, no 9, septembre 2007

5 Prevalence of tal Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba Table 1 Five-year period prevalence (FY to FY ) of cumulative mental illness disorders, any mental disorders, and dementia for adults aged 55 years or older (n = men; n = women) a Cumulative mental disorders Any mental disorders Dementia Age categories, years % % % % % % 55 to ( ) 60 to ( ) 65 to ( ) 70 to ( ) 75 to ( ) 80 to ( ) 90 and older ( ) ( ) ( ) ( ) 2.03 ( ) 1.30 ( ) ( ) ( ) ( ) 2.78 ( ) 2.32 ( ) ( ) ( ) ( ) 4.83 ( ) 4.14 ( ) ( ) ( ) ( ) 8.61 ( ) 8.04 ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) a For further explanation of the diagnostic codings by category, refer to the Methods section residents. 24 Resident and admission data are not available for federal nursing homes. Results As illustrated in Table 1, about 20% of the men and 29% of the women aged 55 years and older were classified in the cumulative mental disorders group during the 5-year study period. That is, they received a diagnostic coding of at least one of the following: depression, anxiety disorder, personality disorder, substance abuse, or schizophrenia. Correspondingly, the approximate proportion of the population having any mental disorder diagnosis increased from one-third of the population aged 55 to 59 years to one-half the population aged 90 years and older. The most dramatic changes with increasing age appeared for the prevalence of dementia, which rose rapidly from 2% or less in the group aged 55 to 59 years, to over one-quarter of the group aged 80 to 89 years and to over one-third in the group aged 90 years and older. The age-specific rates of home care use among those aged 55 years and older showed substantial differences by mental disorder category. These differences decreased as age increased (Table 2). In Figure 1, we see that the lowest age-adjusted annual rate of open home care cases was for those with no mental disorder (57/1000 men and 91/1000 women), followed by the cumulative mental disorders group (162/1000 men and 191/1000 women) and then the dementia group (300/1000 men and 338/1000 women). In other words, about 6% to 9% of those not diagnosed with a mental disorder in the 5-year study period received home care. In contrast, 16% to 19% of those in the cumulative mental disorders group received home care (for women, this represents 2.1 times more than the group with no mental disorders, and for men, 2.8 times more). At 30% to 34%, the highest prevalence of home care use was for people with dementia diagnoses (for women, this was 3.7 times higher than in the group with no mental disorder, and for men, it was 5.2 times higher). Table 3 illustrates the crude age-specific rates of PCH use by mental disorder category. Once again, the differences between these categories decreased as age increased. Figure 2 shows the age-adjusted overall rates for people aged 75 years or older. In this broad age group, less than 10% of those with no mental disorder (53/1000 men and 78/1000 women), about one-third of those with cumulative mental disorders (305/1000 men and 373/1000 women), and one-half to two-thirds of the dementia group (542/1000 men and 669/1000 women) were residents of a PCH. A woman in the cumulative mental disorders group was 4.8 times more likely to be a PCH resident than a woman with no mental disorder, and a man in the cumulative group was 5.7 times more likely to be a PCH resident than a man having no mental disorder. In the group diagnosed with dementia, women were 8.6 times more likely, and men were 10.2 times more likely, to be in a PCH than were their counterparts with no mental disorder. The Canadian Journal of Psychiatry, Vol 52, No 9, September

6 Table 2 Population prevalence of open home care cases by sex, age category, and mental disorder category (FY to FY ) Open home care cases per 1000 population (95%CIs) No mental illness Cumulative mental disorder Dementia Age categories, years 55 to ( ) ( ) ( ) ( ) ( ) ( ) 60 to ( ) ( ) ( ) ( ) ( ) ( ) 65 to ( ) ( ) ( ) ( ) ( ) ( ) 70 to ( ) ( ) ( ) ( ) ( ) ( ) 75 to ( ) ( ) ( ) ( ) ( ) ( ) 80 to ( ) ( ) ( ) ( ) ( ) ( ) 90 and older ( ) ( ) ( ) ( ) ( ) ( ) Table 3 Population prevalence of PCH residency by sex, age category, and mental disorder category (FY to FY ) PCH residency rate per 1000 population (95%CIs) No mental illness Cumulative mental disorder Dementia Age categories, years 55 to ( ) 0.94 ( ) ( ) 7.19 ( ) ( ) ( ) 60 to ( ) 1.99 ( ) ( ) ( ) ( ) ( ) 65 to ( ) 3.03 ( ) ( ) ( ) ( ) ) 70 to ( ) 7.46 ( ) ( ) ( ) ( ) ( ) 75 to ( ) ) ( ) ( ) ( ) ( ) 80 to ( ) ( ) ( ) ( ) ( ) ( ) 90 and older ( ) ( ) ( ) ( ) ( ) ( ) 586 La Revue canadienne de psychiatrie, vol 52, no 9, septembre 2007

7 Prevalence of tal Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba Figure 1 Age-adjusted open home care cases per 1000 by sex and by mental disorder category, age 55+ years (FY to FY ) Open Open home home care care cases cases per per thousand none cumulative mental illness dementia Figure 2 Population prevalence of PCH residency by sex and by mental disorder category, age 75+ years (FY to FY ) PCH cases PCH cases per per thousand none cumulative mental illness dementia were consistently more likely to reside in a PCH than men 1.5 times as likely in the group with no mental disorder and 1.2 times as likely in both the group with cumulative mental disorders and the group with dementia. There is a strong association between having a dementia diagnosis and residing in a PCH. Of the Manitoba residents aged 75 years and older with dementia, were living in a PCH at some time during the 5-year study period. The 3733 men represented 54.2% of all men with dementia, and the 8915 women represented 66.9% of all women with dementia. To give an overall picture of the mental disorder diagnosis profile of PCH residents, Table 4 shows the prevalence of mental disorder diagnoses in the 5 years before , by 2 categories: those admitted to a PCH during the fiscal year (n = 2252) and those residing in a PCH during that fiscal year (n = ). Within the previous 5 The Canadian Journal of Psychiatry, Vol 52, No 9, September

8 Table 4 Proportion of all residents admitted to a PCH in , and proportion of all residents living in a PCH in FY , within each mental disorder category in the previous 5 years (FY to FY ) tal disorder category Percentage admitted to PCH within each category Percentage of PCH residents within each category Total cumulative mental disorders 38.7 Separate diagnoses a ( ) Depression 33.8 ( ) Anxiety disorder 11.4 ( ) Substance abuse 7.1 ( ) Schizophrenia 5.2 ( ) Personality disorder 2.0 ( ) Any mental disorder 74.6 ( ) Dementia 46.0 ( ) 51.6 ( ) 42.2 ( ) 10.7 ( ) 7.4 ( ) 8.4 ( ) 4.6 ( ) 87.1 ( ) 69.0 ( ) a The percentage of residents in the separate diagnostic categories of "cumulative mental disorders" does not add to 100% because residents may have more than one mental illness and may be identified in more than one illness category. years, about one-half of those admitted to a PCH had a diagnosis of dementia, and three-quarters had some mental disorder diagnosis prior to admission. Of those residing in a PCH, 69% had dementia and 87.1% had at least one mental disorder diagnosis within the previous 5 years. Discussion Along the spectrum of community care, there are many seniors who are healthy and able to live independently in the community, whereas others may require home supports or the more intensive nursing care that a PCH can provide. With the increasing senior population (aged 65 years and older) growing to an anticipated 17.8% of the Manitoba population by 2020, it will be necessary to identify and plan for their health service needs. 25,26 This study points out major trends for both present and future needs of community care. There is a high prevalence of mental illness in the older adult population, with particular increases among those aged 80 and older about one-half having some kind of mental disorder diagnosis and over one-quarter having a diagnosis of dementia. Further, the likelihood of using community care resources is strongly associated with a diagnosis of dementia: compared with those having no mental disorder, those aged 55 years and older with dementia are 4 to 5 times more likely to use home care, and those aged 75 years and older are 8 to 10 times more likely to use PCHs. Given the aging population, the inevitable conclusion is that there will be increased demand for community support such as home care or PCH support that is sensitive to the needs of those experiencing mental illness. In Canada, the home care sector is currently dealing with such issues as developing standards for home care, measures of quality of care, training and human resource management, and developing electronic information systems. 7 Embedded in these discussions must be the realization that home care clients have a high likelihood of mental illness, which may affect home care staffing standards, quality of care, and training. During discussions with home care coordinators and workers in Manitoba, it was pointed out that the training and mandate of home care services does not necessarily meet the needs of those with mental illness. Home care is designed more to deal with physical issues of disability or limited physical functioning than with mental illness issues such as supervision of tasks of daily living. Mirroring the home care situation, the prevalence of mental disorders is extremely high among individuals being admitted to, and those resident in, a PCH; 75% of those being admitted to a PCH and 87% of those already resident had one or more mental disorder diagnoses within the previous 5 years. Thus the burden of clients with mental illness begs the question of 588 La Revue canadienne de psychiatrie, vol 52, no 9, septembre 2007

9 Prevalence of tal Illness and Its Impact on the Use of Home Care and Nursing Homes: A Population-Based Study of Older Adults in Manitoba the appropriateness and training of PCH staff. How well-trained are PCH staff members to address issues of mental illness? In Manitoba, very few PCHs actually have psychiatric nurses as part of the staff mix. PCH residents have a notably high prevalence of dementia 46% of those admitted and 69% of those already resident. Moreover, those over age 85 years are particularly vulnerable to developing dementia. 25 Dementia patients require about 36% more nursing care than patients without the disease, so the pressures caused by an increasingly vulnerable population, together with the high costs of care, will make this a significant health care issue. 27 It will become increasingly important to ensure that our PCH system is able to deal with the high burden of mental illnesses and not just with clients physical illnesses. In addressing mental illness issus, further studies will need to examine the qualifications of PCH staff and determine how they relate to quality of care for those clients. One interesting finding in this study is the sex difference in the use of home care and PCHs. were more likely than men to use these resources, especially in the group with no mental disorder. This difference may be driven by 2 factors. First, women may take on the role of primary caregiver more often than men, thus compensating for the use of home care if the man is the one in need of extra care. In contrast, if the woman is the one in need, the man may be more likely to receive help with the caregiver role through home care or a PCH. Second, and somewhat related, women have a higher life expectancy than men and thus require external community resources if there is no spousal caregiver within the home. However, what is also striking about the findings is that this observed sex difference was smaller in situations where the client had a mental disorder, especially in the case of dementia. The current study was unable to incorporate marital status as a variable with any degree of validity. This could be an area for further study to understand why the sex difference in use of community services was lower among those with dementia. Conclusion The results of this research reflect the concerns of Senator Kirby s Report. 2 tal illness is a driving force behind the use of home care and PCHs by older adults. It is therefore critical to understand the service mix necessary to provide support for those experiencing mental illness, in addition to supporting those experiencing physical illness. With the high prevalence of mental illness in the elderly and the changing demographics of Canada, options that include transitions from independent housing to assisted living to PCHs all in close proximity will become more in demand. Moreover, along with housing options comes the issue of staff competencies to deal with a growing number of clientele experiencing mental illness. We need to examine not only how our society views and cares for people with mental illness but also how our older adult population can live in settings that optimize their quality of life. Further, we must support them whether their illness takes the form of physical or mental disability, or more likely, both. Funding and Support Dr Martens has received funding from the CIHR, which supports her research through a CIHR New Investigators Award (2003 to 2008) and a CIHR Community Alliances for Health Research grant (2001 to 2006) that helped fund The Need To Know Team (directed by Dr Martens). Acknowledgments This work formed part of a project on mental illness completed in 2004, one of several projects undertaken each year by the MCHP under contract to Manitoba Health. The results and conclusions are those of the authors and no official endorsement by Manitoba Health was intended or should be inferred. The authors are indebted to the Decision Support Services of Manitoba Health and the Office of Vital Statistics in the Agency of Consumer and Corporate Affairs for the provision of data. Thanks to Dr Phillip St John and Dr Verena ec and to the Working Group for the MCHP report who worked alongside The Need To Know Team to provide clinical guidance and research expertise: Eckhard Goerz, Dr Renée Robinson, Christine Ogaranko, Dr John Walker, and Dr Marni Brownell. References 1. Health Canada. A report on mental illnesses in Canada. Ottawa (ON): Health Canada; The Standing Senate Committee on Social Affairs, Science and Technology. Out of the shadows at last. Final report of the Standing Senate Committee on Social Affairs, Science and Technology. Ottawa (ON): Government of Canada; Bowen S, Martens PJ. A model for collaborative evaluation of university community partnerships. J Epidemiol Community Health. 2006;60: Bowen S, Martens PJ, The Need to Know Team. Demystifying knowledge translation : learning from the community. J Health Serv Res Pol. 2005;10(4): Martens PJ, Roos NP. When health services researchers and policy-makers interact: tales from the tectonic plates. Healthc Pol. 2005;1(1): Manitoba Health. Annual statistics Part 2 Health Programs, Section 2 Home Care [Internet]. Manitoba Health; [cited 2007 Jun 20]. Available from: 7. Statistics Canada. Home care utilization, by sex, household population aged 18 and over, selected provinces, territories and health regions, 2000/01. Canadian Community Health Survey optional content and related tables, cat no XIE. Ottawa (ON): Statistics Canada; Roos NP, Stranc L, Peterson S, et al. A look at home care in Manitoba. Winnipeg (MB): Manitoba Centre for Health Policy; Martens P, Fransoo R, McKeen N, et al. Patterns of regional mental illness disorder diagnoses and service use in Manitoba: a population-based study [Internet]. Winnipeg (MB): Manitoba Centre for Health Policy; 2004 Sep [cited 2007 Jan 3]. Available from: reports.htm. 10. Manitoba Centre for Health Policy. MCHP s concept dictionary (Types of personal care homes >PCH supply and utilization >Personal care home data >Waiting times, admissions, days used, number of residents) [Internet] Jul [cited 2006 Oct 4]. Available from: concept/concept.frame.shtml. 11. Manitoba Health. Annual statistics Part 3 Health Services Insurance Fund. Section 3 Personal Care Home Program. Table 1 Personal care homes by Regional Health Authority: licensed beds and resident days 1997/98 [Internet]. Manitoba Health; 2004 Jul [cited 2006 Oct 4]. Available from: The Canadian Journal of Psychiatry, Vol 52, No 9, September

10 12. Rovner BW, German PS, Broadhead J, et al. The prevalence and management of dementia and other psychiatric disorders in nursing homes. Int Psychogeriatr. 1990;2: Burns BJ, Wagner HR, Taube JE, et al. tal health service use by the elderly in nursing homes. Am J Public Health. 1993;83: Canadian Study of Health and Aging Working Group. Canadian Study of Health and Aging: study methods and prevalence of dementia. CMAJ. 1994;150(6): Rockwood K, Stolee P, McDowell I. Factors associated with institutionalization of older people in Canada: testing a multifactorial definition of frailty. J Am Geriatr Soc. 1996;44(5): Hebert R, Dubois M-F, Wolfson C, et al. Factors associated with long-term institutionalization of older people with dementia: data from the Canadian Study of Health and Aging. J Gerontology A Biol Sci Med Sci. 2001;56(11):M693 M St John PD, Montgomery PR. Depressive symptoms in older people predict nursing home admission. J Am Geriatr Soc. 2006;54(11): Hawranik P. The role of cognitive status in the use of inhome services: implications for nursing assessment. Can J Nurs Res. 1998;30(2): Shapiro E, Tate RB. The use and cost of community care services by elders with unimpaired cognitive function, with cognitive impairment/no dementia and with dementia. Can J Aging. 1997;16: Yuen, EJ, Gerdes JL, Gonzales JJ. Patterns of rural mental health care: an exploratory study. Gen Hosp Pyschiatr. 1996;18: Taylor DH, Fillenbaum GG, Ezell ME. The accuracy of medicare claims data in identifying Alzheimer s disease. J Clin Epidemiology. 2002;55: Fillit H, Geldmacher DS, Welter RT, et al. Optimizing coding and reimbursement to improve management of Alzheimer s disease and related dementias. J Am Geriatr Soc. 2002;50: Tyas SL, Tate RB, Wooldrage K, et al. Estimating the incidence of dementia: the impact of adjusting for subject attrition using health care utilization data. Ann Epidemiol. 2006;16: Doupe M, Brownell M, Kozyrskyj A, et al. Using administrative data to develop indicators of quality care in personal care homes. Winnipeg (MB): Manitoba Centre for Health Policy; ec V, MacWilliam L, Soodeen RA, et al. The health and health care use of Manitoba s seniors: have they changed over time? [Internet]. Winnipeg (MB): Manitoba Centre for Health Policy; 2002 Sep [cited 2006 Oct 5]. Available from: Frohlich N, De Coster C, Dik N. Estimating personal care home bed requirements [Internet]. Winnipeg (MB): Manitoba Centre for Health Policy; 2002 Dec [cited 2006 Oct 5]. Available from: O Brien JA, Caro JJ. Alzheimer s disease and other dementia in nursing homes: levels of management and cost. Int Psychogeriatr. 2001;13: Manuscript received October 2006, revised, and accepted January Director and Senior Researcher, Manitoba Centre for Health Policy, Winnipeg, Manitoba; Associate Professor, Department of Community Health Sciences, Faculty of Medicine, University of Manitoba, Winnipeg, Manitoba. 2 Researcher, Manitoba Centre for Health Policy, Winnipeg, Manitoba. 3 A collaboration of the rural and northern Regional Health Authorities of Manitoba, Manitoba Health, and the Manitoba Centre for Health Policy. 4 Research Assistant, Manitoba Centre for Health Policy, Winnipeg, Manitoba. 5 Senior Systems Analyst, Manitoba Centre for Health Policy, Winnipeg, Manitoba. 6 Systems Analyst/Programmer, Manitoba Centre for Health Policy, Winnipeg, Manitoba. 7 Systems Analyst, Manitoba Centre for Health Policy, Winnipeg, Manitoba. Address for correspondence: Dr PJ Martens, Director, Manitoba Centre for Health Policy, University of Manitoba, Room 408, 727 McDermot Avenue, Winnipeg, MB R3E 3P5; Pat_Martens@cpe.umanitoba.ca Résumé : La prévalence de la maladie mentale et son effet sur l utilisation des soins à domicile et des établissements de soins prolongés : une étude de la population des personnes âgées du Manitoba Objectifs : Déterminer la prévalence de la maladie mentale chez les personnes âgées et son effet sur l utilisation des soins à domicile et des foyers de soins personnels (FSP). Méthodes : À l aide des dossiers administratifs anonymes (exercices financiers à ) du dépôt de données de recherche sur la santé de la population du centre de politiques en santé du Manitoba, nous avons déterminé la prévalence sur 5 ans, pour les personnes de 55 ans et plus ( hommes, femmes), de 3 catégories de maladie mentale : les troubles mentaux cumulatifs (ceux qui ont un diagnostic de dépression, de trouble anxieux, de trouble de la personnalité, de schizophrénie, et [ou] d abus de substance), toute maladie mentale, et la démence. Nous avons calculé les taux par âge et les taux rectifiés selon l âge d utilisation des soins à domicile et des FSP ainsi que la prévalence de la maladie mentale chez les résidents des FSP. Résultats : La prévalence de la maladie mentale augmentait avec l âge de la population, depuis le groupe des 55 à 59 ans jusqu au groupe des 90 ans et plus. La prévalence de toute maladie mentale passait de 32,4 % à 45,0 % chez les hommes et de 42,6 % à 51,9 % chez les femmes, et la prévalence de la démence passait de 2,0 % à 33,6 % chez les hommes, et de 1,3 % à 40,3 % chez les femmes. Les taux annuels rectifiées selon l âge des cas ouverts de soins à domicile par tranche de de population de 55 ans et plus variaient selon le regroupement de maladies mentales (aucun trouble mental, 57 pour les hommes et 91 pour les femmes; troubles mentaux cumulatifs, 162 pour les hommes et 191 pour les femmes; démence, 300 pour les hommes et 338 pour les femmes). Les taux rectifiées selon l âge d utilisation des FSP par tranche de de population de 75 ans et plus variaient aussi selon le regroupement de maladies mentales (aucun trouble mental, 53 pour les hommes et 78 pour les femmes; troubles mentaux cumulatifs, 305 pour les hommes et 373 pour les femmes; démence, 542 pour les hommes et 699 pour les femmes). Parmi les patients hospitalisés (ou résidents) des FSP en , 74,6 % (87,1 %) avaient une maladie mentale, et 46,0 % (69,0 %) souffraient de démence. Conclusions : La maladie mentale affecte profondément l utilisation des soins à domicile et des établissements de soins prolongés. Les sujets souffrant de démence utilisaient les soins à domicile à 3 fois le taux de ceux qui n avaient pas de diagnostic de maladie mentale, et les FPS, à 8 fois le taux des personnes sans diagnostic. 590 La Revue canadienne de psychiatrie, vol 52, no 9, septembre 2007