To date, there is very little in the way of empirical data on pharmacologic, non-pharmacologic, or service models specifically designed and adapted for older age bipolar disorder [3]. This is in contrast to late-life depression and to late-life schizophrenia, for which several randomized clinical trials of medications or psychotherapies have been conducted. This dearth of research leads to the necessity for clinicians to extrapolate evidence from younger people with bipolar disorder or from studies of older adults with other diagnoses, leading to what has been dubbed the practice of “evidence-free” medicine [4]. As Bartels and coworkers note, NIH-funded research applications require that investigators provide scientific justification for excluding children from research, which may increase the number of studies that incorporate children. However, no such justification is required for excluding older adults [2]. Similarly, services research in older age bipolar disorder has largely been descriptive and retrospective and only a handful of studies have employed prospective approaches such as clinical trials to systematically address which service models best accommodate the unique features of older age bipolar disorder.

In the USA, the number of geriatric psychiatrists and geriatric mental health specialized allied health professionals has remained largely static, while the number of older adults is expected to double between the years 2000 and 2030 [5]. There were approximately 1.5 geriatric psychiatrists per 10,000 people over the age of 75 in the year 2000, and projections for the year 2050 indicate that there will be only 0.3 geriatric psychiatrists per 10,000 older adults [6]. During the 2011–2012 academic year, there were only 58 geriatric psychiatry fellows in the country [7]. Despite geriatricians’ higher job satisfaction than many other specialties, roughly 40 % of geriatric medicine fellowship slots are not filled. This dearth of providers is similar for allied health professions, with reports from psychology, nursing, and social work fields indicating severe gaps in developing the workforce for older adults with mental health problems. For example, only 4 % of psychologists specialize in geropsychology [8]. To date, remedies for addressing this shortage, including financial incentives for geriatricians, have not resulted in increases in participation in geriatrics specialty training over time. As a result, an alternative approach may be to promote the use of geriatric specialty providers to educate and support geriatric mental health services delivered by general adult practitioners. Even so, general adult mental health practitioners may not meet the needs of older adults with bipolar disorder. In 2010, only 54 % of outpatient psychiatrists accepted Medicare, which is generally attributed to the disadvantaged rate of reimbursement [9]. As such, limited availability of mental health care for older adults with bipolar disorder is evident in both general adult and specialty-trained geriatrics providers.

Medical comorbidity accounts for a large proportion of the staggeringly high rate of years of life lost to bipolar disorder [10]. At least some of the excess morbidity from medical problems is due to diminished access to care, suboptimal quality of medical care, and primary care–mental healthcare fragmentation. People with bipolar disorder are less likely to receive care for diabetes or cardiovascular disease than are people without mental illnesses [11]. In the broader population of adults with serious mental illnesses, the prevalence of untreated chronic illnesses is staggering, estimated at 30 % for diabetes, 62 % for hypertension, and 88 % for hyperlipidemia [12]. In addition to undertreatment, individuals with serious mental illnesses also appear to experience worse outcomes following hospitalization for medical problems. In the Veterans Health System, patients with serious mental illnesses are less likely to receive preventative care [13].

Many factors likely contribute to higher rates of medical comorbidities in adults with serious mental illnesses. Patient-level factors include diminished care-seeking for chronic medical problems and lower rates of adherence to prescribed regimens or healthy lifestyles. Pragmatic patient-level barriers include poor access to transportation and poverty, which reduces access to some forms of care. Provider-level barriers include therapeutic nihilism (assumptions of limited potential to benefit from medical care given the patient-level barriers described above) and the often-limited participation of psychiatric providers in monitoring and treating medical comorbidities. System-level factors include fragmentation of medical and behavioral health services. Indeed, Bartels and colleagues demonstrated that one of the strongest predictors of engagement in mental and physical health services in a mixed-diagnosis sample of chronically mentally ill older adults was the physical distance between clinic settings. When medical and mental health clinics were located far apart from each other, the rate of successful referral to mental health services was lower [14].

In addition to the burden on the health system, bipolar disorder is associated with substantial participation of informal caregivers in the day-to-day management of the illness. Tasks of caregiving in bipolar disorder are diverse and include assistance with managing medications and finances. In addition, caregivers must take an active role in care when the person with bipolar disorder is incapacitated by the illness. A small but compelling number of reports indicate that bipolar disorder is associated with substantial caregiver distress and experienced burden. The level of caregiver burden in bipolar disorder appears to be equivalent to that of caregivers of people with schizophrenia [15]. Indeed, caregivers themselves frequently require mental health services (29 % were current users of mental health services in one study) [16], and there is a strong association between caregiver burden and increased caregiver mental health service use among caregivers of people with bipolar disorder.

Despite a prevalence of <1 % in community-based epidemiologic surveys of older adults, older age bipolar disorder is not an uncommon diagnosis in outpatient, inpatient, and residential treatment settings. Much of what is known about the prevalence of bipolar disorder in treatment settings is from chart reviews of selected sites and so exact estimates of proportions are not available. As reviewed previously, bipolar disorder accounts for approximately 8–10 % of inpatient geropsychiatry admissions, 6 % of outpatient geriatric psychiatry cases, and 14–17 % of older adults seen in psychiatric emergency settings [17]. In a handful of studies, about 3 % of patients in skilled nursing facilities have a diagnosis of bipolar disorder. It is clear that although bipolar disorder is relatively rare among community-dwelling older adults, it is commonly seen in treatment settings where psychiatric care is administered to older adults. At the same time, it is important to note that a sizable proportion of older adults with bipolar do not use any mental health services. Byers and coworkers evaluated the National Comorbidity Survey and found that 41 % of patients with bipolar disorder older than the age of 55 surveyed were not using any mental health services in the prior year [18].

Older adults with bipolar disorder, just as their younger counterparts, seem to consume more mental health services than patients with major depression. Bartels and coworkers compared service use patterns of older patients with bipolar disorder to those of older patients with major depression and found that the group with bipolar disorder used approximately four times the amount of mental health services than the group with major depression [19]. In particular, higher rates of mental health service utilization were observed for inpatient, partial hospitalization, and case management services, although the use of outpatient psychotherapy was comparatively lower for older adults with bipolar disorder. In comparison with older people with schizophrenia, the rate of hospitalization for older adults with bipolar disorder was slightly lower although the mean length of stay was longer.

Older adults with bipolar disorder have been reported to use a different array of mental health services compared to younger or middle-aged patients. Using administrative data from a sample of 40,000 patients who were clients of a large public mental health system, it seems that older adults (people over age 60) with bipolar disorder used substantially fewer crisis residential and emergency psychiatry services than did their younger counterparts [20]. However, older adults with bipolar disorder used more case management services. These data are consistent with the few studies of the long-term course of bipolar disorder, which suggest the severity of manic symptoms may decline with age [21], yet at the same time, cognitive and functional impairment becomes less distinguishable from that in schizophrenia, increasing the need for functional assistance such as case management [22].

Inpatient hospitalization is a relatively frequent occurrence among older adults with bipolar disorder. In the current era of increased attention to rising healthcare costs, there has been substantial attention regarding the fiscal impact of inpatient psychiatric hospitalization. As of 2006, the average length of stay for adults with bipolar disorder was 9.4 days [23]. Examining inpatient treatment in community hospitals, the rate of hospitalization among older adults with bipolar disorder compared to younger adults with bipolar disorder was somewhat lower and the mean cost of hospitalization was approximately $8000 [23]. In a sample of 65,556 veterans with for bipolar disorder seen in the Veterans Health Administration setting, patients with bipolar disorder who were older than age 60 experienced psychiatric hospitalization approximately 6 times over a three-year period, which was equivalent to that of the middle-aged patients and slightly more than that of adults younger than age 30 [24]. However, adults over the age of 60 had average lengths of stay that were three times longer than that of younger adults, with the mean of 98 hospital days per patient compared to 30 days for younger adults. About 60 % of the inpatient hospitalizations in this sample were for mania and 40 % were for depression.

In attempting to contain costs associated with the “revolving door” of rehospitalization, attention has been paid to population factors that predict greater risk for hospitalization. In two separate retrospective chart review studies analyzing both diagnostic and non-diagnostic variation factors and risk of rehospitalization in separate geriatric inpatient psychiatric settings, bipolar disorder emerged as a strong predictor of risk of repeated hospitalization [25, 26]. Other factors that emerged in these studies in conjunction with bipolar disorder that predicted rehospitalization included male gender, living alone, and residing in supported housing. Within the population of older patients with bipolar disorder hospitalized for mania, Lehmann and Rabins [27] compared prehospitalization profiles of older adults with bipolar disorder in early-onset (onset <45 years) to late-onset (onset >45 years) subgroups and found that early-onset patients were (1) more likely to have had a recent dose change by a provider; (2) more likely to have been non-adherent to medications; and (3) more likely to have been exhibiting aggressive/agitated behavior. Other factors, such as the rate of comorbidities, and demographic variables were similar between early- and late-onset groups. Taken together, studies of the rate of inpatient hospitalization for bipolar disorder in later life are somewhat inconsistent as to whether risk of psychiatric hospitalization increases or decreases with age in older adults with bipolar disorder. There does appear to more consistent indication that older adults experience longer stays than their younger counterparts. Available data described above indicate that in geriatric inpatient psychiatry settings, bipolar disorder is a risk factor for rehospitalization compared to other diagnoses.

As mentioned earlier, bipolar disorder is diagnosed in 2–3 % of patients in skilled nursing facilities. In a study conducted by the Department of Veterans Affairs, patients with bipolar disorder were approximately 28 % more likely to be admitted to skilled nursing facilities (SNFs) than people without mental illnesses [28]. Not surprisingly, patients with bipolar disorder in such facilities, compared to individuals with OABD in community settings, are more likely to have impairment in instrumental activities of daily living, greater cognitive impairment, fewer social supports, and more severe psychotic symptoms [29]. Within the population of residents of such facilities, people with serious mental illnesses are also more likely than individuals without serious mental illnesses to reside in skilled nursing facilities for greater than 90 days [30]. People with serious mental illnesses residing in skilled nursing faculties are also comparatively younger than residents without serious mental illnesses. Finally, patients with serious mental illnesses in SNFs are also more likely to be hospitalized than patients without serious mental illnesses, with high rates of hospitalizations for ambulatory care-sensitive conditions (i.e., ones that are seen as being likely to be treatable as an outpatient) [31].

The Olmstead Act (1999) required that “qualified” individuals with serious mental illnesses be placed in community settings rather than skilled nursing facilities whenever possible. Of note, while the general preference for consumers and for clinicians is for community placement, a study of older adults in New Hampshire residing in skilling nursing settings by Bartels et al. [32] revealed that there was a high degree of discordance between patients’ preferred level of appropriateness for community placement and clinician-assigned level of appropriateness. In this study, the majority of clinicians believed that ideal transition settings were in community group homes, whereas patient consumers almost uniformly preferred apartment/independent living.

Partial hospitalization, according to Medicare’s definition, is “a structured program of outpatient psychiatric services provided to patients as an alternative to inpatient psychiatric care.” The term “partial hospitalization” is used interchangeably with day treatment. These services can be delivered in a community mental health or hospital setting and do not involve overnight stays. Medicare Part B provides coverage for partial hospitalization. Little is known about the prevalence of use or impact of partial hospitalization in bipolar disorder, although at least in one study older patients with bipolar disorder were more likely to use this service than were older patients with major depression [29]. Partial hospitalization programs vary in terms of structure and content, but typically sessions are held 2–4 times per week for 4–6 h per day, combining group and individual therapy as well as pharmacotherapy. Therapeutic targets include both symptom management and functional rehabilitation, with the involvement of multidisciplinary team care. One focus for partial hospitalization is for transitional care as a “step-down” from inpatient care. In light of recent attention to rapid rehospitalization in a variety of chronic medical illnesses, the case for partial hospitalization programs may build.

The National Comorbidity Survey [33] found that older adults with mental health diagnoses are less likely to participate in outpatient treatment than younger cohorts. The reasons for this lower rate of participation are unknown but generally attributed to patient-driven factors such as lower acceptance of mental health care among older adults and attendant stigmatization along with provider/system-driven factors such as diminished screening and recognition resulting in reduced referral to such services. As previously stated, 41 % of patients with OABD are not receiving mental health care, although this is a lower rate of non-participation than in older adults with other mood or anxiety diagnoses [18].

Additional national trends are for mental health services to be delivered by primary care providers rather than specialty mental health providers, although these trends are less prominent for bipolar disorder than other mood or anxiety disorders [34]. A recent study reported that older adults use a comparatively lower rate of psychotherapy than younger people, and only 25 % of older adults with mental health diagnoses use psychotherapy [35]. Older adults frequently report a preference to discuss mental health problems with family practitioners rather than specialty mental health providers [36]. In addition, a number of studies have reported that older adults with depressive symptoms prefer psychotherapy to antidepressant medication [37, 38], yet depression is typically managed in the family practice setting with medication treatment and not psychotherapy. As such, there is a disconnect between the site of preferred treatment and the preference for treatment modality in geriatric depression. The treatment preferences of older patients with bipolar disorder have received scant research attention. At least one study indicated that group psychotherapy was feasible and acceptable in a small sample of older patients with bipolar disorder [39].