Six-Month Utilization of Psychotherapy by Older Adults with Depressive Symptoms

Abstract

We explored psychotherapy utilization patterns for community-dwelling older adults with depressive symptoms identified during a statewide initiative designed for identifying risk of substance misuse. Individuals screening negative for substance misuse, but positive for depressive symptoms, agreed to participate in monthly interviews conducted over 6 months (n = 144). Results showed that 39 (27 %) received psychotherapy, of which nearly two-thirds reported four or fewer visits. Mental health counselors were the most frequently reported service providers (50–62.5 %). Location of service varied considerably. This study documents low rates of psychotherapy use and few visits. Current efforts to increase psychotherapy access are discussed.

Introduction

National epidemiological studies are documenting a decrease in psychotherapy utilization in recent years. This trend is concerning for several reasons and may be especially harmful for older adults, who experience disparities in access to mental health services and have complex comorbid conditions often warranting a multi-modal approach. In this paper, we examine psychotherapy utilization in detail over a 6-month period for community-dwelling older adults (psychotherapy providers, locations, frequency on a monthly basis) and discuss implications and potential strategies to enhance access to psychotherapy.

In what has been described as a “sea change in the provision of mental health services” (Druss 2010, p. 1419), use of psychotherapy has declined in recent years, and use of psychotropic medications has increased in the same time frame. As such, “the range of options available to patients for mental health treatment is declining” (Druss 2010, p. 1420). In the largest national study of health service utilization comparing trends from 1998 to 2007 (Marcus and Olfson 2010), while the proportion of adults with depression who receive psychotropic medication has remained stable (80 % in 1998, 82 % in 2007), the proportion receiving psychotherapy alone or psychotherapy with psychotropic medications has declined dramatically (54–43, and 42–35 %, respectively). In 2007, depressed older adults were least likely of all age groups to receive psychotherapy alone (29 vs. 39–66 % for younger age groups) or psychotherapy and antidepressant medication combined (21 vs. 33–36 %), but they had similar or higher rates for antidepressant medications alone (74 vs. 52–82 %; Marcus and Olfson 2010). This striking change has several possible causes, including the advent of newer generation psychotropic medications, a shift in how psychiatrists are trained and practice (Mojtabai and Olfson 2008), movement toward managing mental health conditions using psychotropic medications (Olfson and Marcus 2009) in primary care settings (Wang et al. 2006), and advertising and insurance reimbursement policies that favor medication management (Druss 2010).

These trends are troubling for several reasons, particularly for some groups of underserved populations including older adults. Compared to other age groups, older adults are less likely to use mental health services at all, and if they do receive services, they are less likely to receive services of minimally adequate treatment quality (Wang et al. 2005). Older adults usually present with mental health issues in primary care settings, where medications are much more readily available than psychotherapy and where quality of care is usually lower compared to specialty mental health settings (Wang et al. 2005). Older adults do not adhere to psychotropic medication regimens (Bosworth et al. 2008), and most prefer psychotherapy (Gum et al. 2006). Meta-analyses have demonstrated that psychotherapy is effective (Pinquart et al. 2006), and that older adults benefit from psychotherapy just as well as younger adults (Cuijpers et al. 2009). Access to psychotherapy is also advantageous for older adults because they commonly experience comorbid physical and mental disorders. Psychotherapy and other behavioral interventions (e.g., focused on adherence to medical regimens, health behavior change, pain management, sleep disturbance) are particularly suited to individuals with physical comorbidity when combined with other medical treatments, as these interventions can benefit both mental and physical symptoms and may even result in cost offset (Blount et al. 2007).

To inform future efforts to improve access to psychotherapy for older adults, we believed it was first important to conduct a needs assessment in order to gain a detailed understanding of current psychotherapy utilization. Existing large-scale studies (Marcus and Olfson 2010; Olfson and Marcus 2010; Wang et al. 2005) possess very large rigorous sampling designs but are cross-sectional and do not provide detailed information about psychotherapy delivery for older adults (e.g., who is delivering it, where, and how often?). Psychotherapy can be delivered by a variety of professionals in a variety of settings and for various amounts. The most efficient allocation of resources requires careful planning. Detailed information about psychotherapy utilization by older adults in a large geographical area could be used to identify resources upon which to capitalize. For example, if professionals from a particular discipline or setting were delivering psychotherapy to large proportions of older adults, this group of professionals or setting could be used to expand psychotherapy availability or their methods could be replicated by other professionals. Moreover, this information could be used to identify gaps in psychotherapy access that might present opportunities for personnel training or policy changes. For example, we know that integration of psychotherapy into primary care settings improves psychotherapy access and outcomes for older adults (Areaán et al. 2008); if low psychotherapy use was observed in primary care, this suggests an opportunity.

Thus, we designed a longitudinal study to assess detailed information about service utilization (including psychotherapy), by capitalizing upon an existing statewide program in Florida that conducted screening and referral for depressive symptoms in community settings. The Florida BRief Intervention and Treatment for Elders (BRITE) Project (Schonfeld et al. 2010) focused on screening and brief interventions for older adults with substance misuse. BRITE providers also screened for depressive symptoms and referred depressed older adults to mental health services. In an effort to work toward expanding BRITE to directly address depression, we conducted the current study of older adults who screened positive for depressive symptoms. We aimed to explore psychotherapy utilization patterns in depth for a sample of community-dwelling older adults with identified depressive symptoms. On a monthly basis for 6 months, we assessed psychotherapy utilization, including providers, locations, and frequency of visits.

Methods

Sample

Data for this report were derived from a 6-month longitudinal study of mental health service utilization patterns for older adults with depressive symptoms, described in detail elsewhere (Gum et al. 2011). Participants were recruited from outreach and screening efforts in six counties of Florida by partnering with the Florida BRITE Project (Schonfeld et al. in press, 2010). BRITE is a statewide program of community outreach, screening, and brief interventions. BRITE was funded to screen adults aged 60 and older for misuse of alcohol, medications, or illicit drugs; BRITE offered brief interventions to those demonstrating moderate to high risk. Employed by agencies that contracted with the state of Florida, BRITE personnel screened older adults in various sites, such as hospital emergency departments, walk-in clinics, senior centers, retirement communities, health fairs, and other community sites within their counties.

BRITE also screened individuals for depressive symptoms using the Short Geriatric Depression Scale (S-GDS; Sheikh and Yesavage 1986). When BRITE personnel identified individuals with clinically significant depressive symptoms (i.e., S-GDS C 5), they referred those individuals to existing treatment options for depressive symptoms (e.g., mental health centers, primary care physicians), because BRITE offered brief interventions for substance misuse only, not depression. Providers were trained to deliver referrals in a standardized manner, using a script for participants with S-GDS scores C5 or who requested a referral. Providers selected the referral sources based on their knowledge of their local communities, and providers were not monitored during the referral process; thus, there were likely differences in how referrals were given and discussed.

In addition to offering referrals to individuals who screened positive for depressive symptoms and negative for substance misuse, BRITE personnel invited them to participate in the present study, a naturalistic study designed to describe service use patterns over 6 months following BRITE screening and referral. Participants also were recruited locally at the university site using the same outreach methods as BRITE, including screenings and soliciting referrals from a wide variety of community agencies and senior housing agencies. In addition to depressive symptoms, other eligibility criteria for the present study were: (a) age C65; (b) English-speaking; (c) no substance misuse (i.e., C3 drinks/week; C2 drinks/1 day; or any illicit substance use); (d) no current specialty mental health service utilization (i.e., psychotropic medication from psychiatrist, psychotherapy); and (e) passed a six-item cognitive screener (i.e., score C3, Callahan et al. 2002). During recruitment, 5,963 screenings were recorded from the sites involved in the present study, with 313 recorded as screening positive for depressive symptoms but negative for substance misuse. Detailed data regarding these individuals are not available (e.g., numbers of these individuals who were excluded, who were invited to participate but refused, who were not invited to participate).

As described previously (Gum et al. 2011), 144 individuals participated in the current study. The 144 participants were 75.5 years old on average (SD = 7.6), predominantly female (n = 114, 79.2 %), and white (n = 104, 72.2 %) or black (n = 34, 23.6 %; Asian n = 2, 1.4 %; multi-racial n = 4, 2.8 %; Hispanic ethnicity n = 13, 9.0 %). Most participants had high school education or greater (n = 107, 74.3 %), and almost half lived in poverty (n = 61, 42.4 %; income was missing for 22 participants). Few participants were married (n = 23, 16.0 %) and lived with others (n = 50, 34.7 %). Most perceived themselves to be in fair health (n = 65, 45.1 %). The average S-GDS score was 8.3 (SD = 2.7), and over a third met criteria for a current major depressive episode (n = 55, 38.2 %). Compared to the overall sample of individuals screened by BRITE (n = 85,001), the overall BRITE sample contained more males (44.9 %), Hispanic individuals (20.2 %), and was younger (M = 69.0 years, but BRITE screened adults C55 years old; Schonfeld et al. in press).

Measures

Utilization of Psychotherapy

At each follow-up interview, participants were presented with a list of professionals (psychologist/therapist/counselor, social worker/case manager, psychiatrist, other physician, nurse, faith leader, or other). They were asked, “Since last month, have you talked to any of these professionals about feelings of stress or sadness?” For each professional to whom they responded “yes,” they were asked follow-up questions regarding type of service, frequency of visits, and location of visits: (a) “what type of service did you receive from this professional—counseling, case management, specialist referral, over the counter medication, prescribed medication, or other (specify)?” (b) “how often did you see this professional—not at all, once a month, twice a month, once a week, or nearly every day?” and (c) “Where did you see this professional—primary care, other medical outpatient, emergency room, other medical inpatient, mental health or substance abuse outpatient, mental health or substance abuse inpatient, senior center or senior housing, religious setting, home, telephone, or other (specify)?”

Procedures

The baseline interview was conducted in-person, and the six monthly follow-up interviews were conducted by telephone. Interviewers were counselors in the BRITE Project who conducted all BRITE services (screening, referrals, brief interventions for substance misuse) or research interviewers at the university. All interviews were audio recorded unless the participant refused. All interviewers underwent the same 1-day training led by the principal investigator. The project coordinator reviewed audio recordings of all baseline interviews and each interviewer’s first seven follow-up interviews, providing additional training as needed. The study was approved by the University of South Florida Institutional Review Board, and all participants provided written informed consent. There are no known conflicts of interest for any author. All authors certify responsibility for the manuscript.

Data Analysis

Descriptive statistics were investigated for all variables related to psychotherapy, including frequencies and crosstabs of reported counseling by follow-up period, professional, frequency, and location. Total number of visits was estimated based on frequency for each monthly period (e.g., “once a week” was estimated as four visits per month). Bivariate analyses (Chi square, t test) were conducted to examine receipt of at least one psychotherapy visit over the study period (yes/no) in relation to gender, minority status, marital status, education level, poverty status, having a current MDE at baseline, or baseline S-GDS score. Due to the small number of participants who received psychotherapy services, additional analyses of predictors of psychotherapy utilization were not conducted. All models were tested using SAS 9.2 (Cary, NC).

Results

Psychotherapy Providers and Locations

Of the 144 study participants, 39 (27 %) reported that they engaged in at least one psychotherapy visit with a professional about feelings of stress or sadness during the 6-month study period. The majority of the reported psychotherapy was delivered by psychologists/therapists/ counselors at each time point, with the total number of psychologists/therapists/counselors seen across all participants each month ranging from 7 (50.0 %; percentages are based on total number of professionals seen for psychotherapy each follow-up period) to 10 (62.5 %) across follow-up periods. Social workers/case managers delivered the next highest proportion of psychotherapy at each follow-up period, except for follow-up 5, when three psychiatrists (20.0 %) were seen compared to two social workers/case managers (13.3 %). Numbers of social workers/case managers seen ranged from 2 (13.3 %) to 5 (31.3 %) across months. Numbers of psychiatrists seen ranged from one (6.3 %) to three (20.0 %) across months. Fewer nurses [0–3 (18.8 %)], faith leaders [0–2 (12.5 %)], and other MD/health professionals [0–1 (7.1 %)] were seen each month for psychotherapy.

The most common location for psychotherapy was in participants’ homes. Of professionals seen each month for psychotherapy, the number providing psychotherapy in home ranged from three (21.4 %) to seven (43.8 %) across follow-up periods. There was more variability across other locations: range of 0–4 (25 %) for aging service or senior center locations, 1–2 (6.3–14.3 %) for religious setting, 0–3 (18.8 %) for mental health outpatient setting, 0–1 (7.1 %) for mental health inpatient setting, 0–1 (7.1 %) for primary care, 0–2 (12.6 %) for other outpatient setting, 0–1 (7.1 %) for other inpatient setting, 0–1 (6.3 %) for telephone, and 0–2 (13.3 %) for other setting (location was missing for 10 professionals over all follow-up periods).

Psychologists/therapists/counselors saw participants in their homes (range of 1–5 home visits across follow-up periods), aging service or senior center settings (1–2), religious settings (0–3), mental health outpatient settings (0–1), primary care (0–1), and other outpatient settings (0–1). Social workers and case managers saw participants in their homes (1–2), aging service or senior center settings (0–2), other outpatient settings (0–1), and other inpatient settings (0-1). Psychiatrists saw participants in their homes (0–1), mental health outpatient settings (0–2), and mental health inpatient settings (0–1). Nurses saw participants in their homes (0-2) or other locations (0–1). Religious leaders saw participants once at home, in a religious setting, and by telephone. At two follow-up periods, other MD/health professionals delivered psychotherapy to one participant in primary care.

Number and Frequency of Psychotherapy Visits

Participants reported a total of 91 psychotherapy visits Study participants who reported receiving any psychotherapy during the 6-month study period averaged a total of 5.6 visits, with the number of reported visits ranging from one to 24. The median and mode were both four visits. Twenty-five (64.1 % of 39) individuals reported four or fewer total visits (n = 8 with 1 visit, n = 4 with 2 visits, n = 1 with 3 visits, n = 12 with 4 visits). Twelve individuals (30.8 % of 39) reported more than four visits.

The majority (57.1 %) of the psychotherapy recipients who received psychotherapy in the first month of follow-up received just one visit during that time. Few individuals received psychotherapy once per week across follow-up periods (25.0–35.7 %), except for the last follow-up, when eight (50 %) reported visits once per week.

Covariates of Psychotherapy

Bivariate analyses revealed nonsignificant associations between receiving at least one psychotherapy visit and all other variables considered, i.e., gender, minority status, marital status, education level, poverty status, having a current MDE at baseline, or baseline S-GDS score (p [ 0.10). Detailed results for all variables are available from the first author.

Discussion

Not surprisingly, this study documents low rates of psychotherapy use compared to national estimates for adults across ages. Here, 27 % of participants reported at least one psychotherapy visit with a mean number of 5.6 visits, compared to 43 % receiving a mean number of 8.16 visits in a representative national study of outpatient treatment for individuals with depression (Marcus and Olfson 2010). In the current sample, psychotherapists delivered most psychotherapy visits in various locations, including home. There was greater variability for other professional types and locations, with no clear patterns of high psychotherapy delivery by any other professional type or location. Given that most participants received four or fewer psychotherapy visits, it is unlikely they received evidence-based psychotherapy. Recommended number of sessions in evidence-based interventions is 7–12, based on a meta-analysis that examined treatment effectiveness and dropout rates (Pinquart et al. 2007).

Before discussing potential implications, it is useful to consider the findings within the context of the study’s limitations. First, this was a sample recruited from a statewide screening program in multiple sites with slightly different recruitment and referral strategies across sites. While reaching large numbers of older adults, it is not a randomly selected, representative sample. Also related to the sampling frame, we do not have detailed data regarding individuals screened who might have been eligible but were not recruited or who refused to participate. The sample size is somewhat small, particularly given the low numbers who received psychotherapy. Thus, there is possibility for sampling bias as well as possible differences based on referral methods by different providers. Moreover, this was not an intervention study. These limitations preclude more detailed analyses of factors associated with psychotherapy use or changes in depressive symptoms.

The primary strength of the study is the degree of detail regarding psychotherapy utilization assessed in monthly increments for 6 months. We designed this study to capture a great deal of detail in order to inform possible opportunities to improve access. First, regarding professionals, we observed that psychotherapy was delivered by a variety of professionals in a variety of settings. This observation suggests that efforts to increase access to psychotherapy for older adults will need to span multiple disciplines and settings. Psychologists and mental health counselors delivered most psychotherapy visits, so these providers are two groups that should be included in any change efforts. Several other professionals were involved in delivering some psychotherapy as well, and psychotherapy is often part of other health and social services, so a transdisciplinary effort is necessary.

Second, regarding settings, the largest proportion of psychotherapy visits occurred in participants’ homes. Home-based delivery can improve access for older adults who cannot or are not willing to travel to outpatient service settings, although many barriers exist to delivering home-based psychotherapy, such as the provider’s time and reimbursement policies for travel policies. The expansion of telehealth may help to further improve access to psychotherapy without traveling from one’s home. Primary care was rarely accessed for psychotherapy; although research studies now demonstrate the value of integrating brief behavioral interventions into primary care settings (Areaán et al. 2008), this trend does not seem to have led to broad utilization yet. This presents an opportunity for expansion and also for research on facilitating the adoption and implementation of behavioral interventions in primary care.

Third, the vast majority of participants received very few psychotherapy visits. This observation indicates that efforts are needed to retain older adults in psychotherapy. There are likely several reasons for the low number of visits (e.g., financial, insurance, transportation, quality of services, client preferences) that require further research. For providers, this finding suggests the need to regularly assess an older client’s engagement and possible barriers to continuing services and possibly the provision of social services to facilitate access (Choi 2009).

While this study points to possible opportunities to improve psychotherapy access, the study’s descriptive nature precludes drawing firm conclusions about precisely what strategies are needed to improve access. Future intervention and policy research is needed to determine whether various changes (e.g., professionals’ training, reimbursement policies, delivery methods such as telehealth and integration into medical and social service settings) will actually improve access. It is likely that the engagement of multiple stakeholders, including state policymakers, will be key to improving psychotherapy access for depression—as it was for improving behavioral intervention access for older adults with substance misuse (Schonfeld et al. in press, 2010; Schonfeld and Hedgecock 2008). Policymakers in state legislature and state agencies (such as state units on aging, mental health, and substance abuse) could work together to develop protocols and pool resources for training, implementation, and evaluation at multiple provider agencies. Their leadership could provide the necessary initiative for provider agencies interested in incorporating depression interventions. The availability of funding mechanisms through Medicare and the Older Americans Act (Administration on Aging and Substance Abuse and Mental Health Services Administration 2012) for evidence-based behavioral interventions provide new opportunities for such collaborative research.