| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ALL ISSUES | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Special Article |
Received June 5, 2002; revised August 29, September 10, 2002; accepted September 16, 2002. From the University of California, San Francisco (PAA,BLC), Stanford University (DG-T), Dartmouth Medical School (MTH), Cornell University (HCS), and University of Pittsburgh (RS). Address correspondence to Dr. Areán, Department of Psychiatry, University of California, San Francisco, 401 Parnassus Avenue, Box F-0984, San Francisco, CA 94143-0984; e-mail: pata{at}lppi.ucsf.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGUIDELINES FOR GEROPSYCHOTHERAPY...CONCLUSIONREFERENCES |
|---|
Key Words: Outcome Studies • Psychotherapies
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGUIDELINES FOR GEROPSYCHOTHERAPY...CONCLUSIONREFERENCES |
|---|
|
First, the psychotherapy under investigation must have initial pilot data demonstrating the intervention's feasibility, acceptability to the patient, and potential effect on the intended outcome. Because psychotherapy is influenced by the patient's sociodemographic variables, the outcomes achieved in one sociodemographic population may not generalize to another. Although there are no established recommendations regarding pilot sample size, we recommend a sample size of 10 to 20 subjects per study group.
Second, psychotherapists must be trained to competently administer the treatment. A manual or set of guidelines for conducting the specified psychotherapy must exist, and a reliable method of instruction developed. Competency must be demonstrated through audio- or videotaped reviews of pilot cases. In the IMPACT and PROSPECT studies, for instance, each therapist participated in workshops to learn how to deliver the intervention and then did so with five pilot cases. The videotaped treatment sessions were reviewed and rated for competency by expert psychotherapists. Therapists did not treat research subjects until they displayed a satisfactory competency.
Third, researchers should assess all three aspects of treatment implementation: delivery, receipt, and enactment.7 Treatment delivery refers to the implementation of specific treatment components by providers to patients. Methods of assessment include ongoing internal and independent monitoring and provision of feedback to the interventionist to guarantee that the treatment is being delivered as intended. The internal supervision, generally conducted by a member of the research team, ensures that therapists are delivering the intervention at an intended quality level; the independent evaluation, generally done by a consultant who is not a member of the research team, is an audit that confirms the therapists are delivering treatment as specified by the manual. In trials comparing two or more treatments, proponents for each treatment model should be equally represented so as to avoid the outcome's possibly being dominated by an "allegiance" effect to one treatment alone.8
Treatment receipt refers to the degree to which the client actually received the intervention. It is typically assessed by records of contact between client and therapist, interventionist documentation of the session, and feedback from the client indicating understanding of the treatment components. Treatment enactment focuses on the degree to which clients demonstrate in their environment the changes sought by the intervention. Thus, skills learned from cognitive–behavioral therapy (CBT) would be demonstrated by completion of homework assignments and reductions in negative thoughts as recorded in a Negative Thoughts Diary. Measures that tap into treatment receipt and enactment are typically assessed at post-treatment and follow-up time-points in order to determine their status as active ingredients of psychotherapy that may have directly influenced outcomes. As Lichstein and colleagues7 suggest, the best way to determine whether a treatment has been successfully delivered is to note whether the participants actually act on what has been delivered. Thus, skill acquisition is not sufficient; there must also be evidence of utilization post-treatment.
Fourth, outcome assessments should include manipulation checks. For example, a measure of skill acquisition is necessary for behavioral interventions to determine whether changes in symptoms reflect changes in skills. This requirement resembles blood levels for medication adherence in pharmacology trials. Because approximately nine percent of the variance accounting for treatment effects in psychotherapy is due to nonspecific factors (e.g., genuineness, warmth, and empathy), it is important to determine whether symptom change is related to acquisition of the treatment goal (e.g., resolution of role-transition impairments for interpersonal therapy).8
Finally, the psychotherapy must follow a logical and sequential course.9 Once pilot data have demonstrated that the psychotherapy is feasible, acceptable, and potentially efficacious, the investigator may initiate a randomized controlled trial comparing the new intervention to either a placebo-control (typically, supportive therapy or another form of credible but nonessential interaction) or to an intervention that is considered the standard of treatment, typically medication or another psychotherapy. After this trial shows efficacy in treating the targeted disorder, it can be investigated with an effectiveness trial that compares the study intervention with usual care. Although patients in effectiveness trials are more clinically complex than those in efficacy trials (because of relaxed inclusion and exclusion criteria), they are also more representative of the target population. Finally, once effectiveness is established, the psychotherapy can be studied in a dissemination trial that investigates whether the typical therapist can master the intervention and continue to use all of its components.
|
GUIDELINES FOR GEROPSYCHOTHERAPY RESEARCH |
|---|
|
TOPABSTRACTINTRODUCTIONGUIDELINES FOR GEROPSYCHOTHERAPY...CONCLUSIONREFERENCES |
|---|
Guideline 1: Select an Age-Appropriate Psychotherapy
During the development of a geropsychotherapy research study, the chosen psychotherapeutic intervention should be adapted to the needs of older adults without substantially changing the therapeutic content. Age-appropriate adaptations of psychotherapy include adding a patient-education component to treatment; changing treatment components to reflect developmental differences between older and younger adults; adjusting the pace of the psychotherapy to account for age-related changes in information-processing; and allowing flexibility in the delivery of psychotherapy to overcome cognitive, medical, and physical barriers to care. We refer the reader to several papers and manuals that describe how psychotherapy should be adapted for older adults and briefly describe the adaptations here.10–12
Patient education.
Older people tend to have misconceptions about the content and process of psychotherapy that may lead to difficulty focusing on the aims of therapy and to early treatment dropout.13 To address this, initial sessions must include socialization to psychotherapy. In the PEPUP study, patients were asked to relate their knowledge about psychotherapy and the concerns they had about starting treatment. The provider then addressed patient concerns, at times with family present, and provided patients with written information about the treatments provided in the study. Patients also had the opportunity to participate in an information group about services in the clinic, which included a question-and-answer session about psychotherapy.
Adjusting for changes in information-processing.
Although research shows that older people can learn new tasks and do maintain mental flexibility, this research also shows that there are a number of cognitive changes associated with aging that should be attended to when providing psychotherapy to an older patient.11 Age-associated cognitive slowing decreases use of inferential reasoning, and declines in working memory affect how psychotherapeutic material is processed and retained. The adage "start low and go slow" not only applies to the use of psychotropic medication in older patients, it also applies to the delivery of psychotherapy.
One method for accommodating changes in information-processing and inferential reasoning is to adjust the length and number of psychotherapy sessions. In the PROSPECT study, interpersonal therapy (IPT) sessions were decreased from the usual 50 minutes to 30 minutes for patients cognitively unable to maintain the therapeutic focus established with the therapist for a particular session. Another strategy for accommodating age-related changes in information-processing is to discuss new therapeutic material in a variety of modalities. As an example, the IMPACT investigators created a strategy called "cue and review." The therapist explained each step of the problem-solving skill, assisted the patient as necessary to complete the step, and then reviewed what the patient just accomplished. At the end of the session, the entire process was briefly reviewed. Therapists were also taught the importance of allowing the older person time to explain problems in the context of their own history without losing the here-and-now focus of treatment.
Adjusting treatment on the basis of social-developmental stages.
An important point in adapting psychotherapy is to make sure the strategies being used match the developmental process of older adults. Familiarity with "normal aging" is crucial, in that encouraging older people to engage in psychosocial activities that counter the developmental process can result in more harm than good.14 For instance, in CBT for depression, patients are encouraged to increase their social network of friends and take up new activities to increase exposure to pleasant events. Research on older adults' social adaptation indicates, however, that, as we age, it is adaptive to become more selective about whom we socialize with, and thus older adults have smaller, yet closer, social networks.15 Encouraging older people to seek out new social contacts may not be a successful approach for improving mood. In the PEPUP and IMPACT studies, for example, therapists encouraged patients to increase their pleasant activities by increasing the frequency of social contacts with existing friends and family whom the patient felt emotionally connected with, rather than encouraging new contacts. In cases where patients lost all social contacts through death or migration, they were encouraged to engage in activities where they could meet people similar to their previous friends.
Increasing the flexibility of treatment to accommodate social and functional constraints.
Older adults are faced with familial, medical, and time constraints that can influence participation in psychotherapy.16 In order for patients to benefit from psychotherapy, the treatment must be flexible with regard to structure, location, and presentation of the treatment. To accommodate time, resource, and family constraints, it is common for psychotherapy to incorporate case-management strategies such as helping the older person with caregiving issues and finding respite care so that the older patient can attend therapy sessions. To accommodate visual and fine-motor disabilities, the IMPACT study adapted the Problem-Solving Therapy for Primary Care (PST-PC) forms with larger print and larger writing spaces. Audiotaped recording of sessions and the details of homework assignments can help to overcome problems associated with visual or memory impairments; this was done in the PEPUP study. Finally, stigma and mobility issues that may deter the older person from participating can be overcome by changing the setting in which treatment is delivered. In REACH, treatment was provided in non-mental health settings, such as primary care, patients' homes, senior centers, and churches. Often, further adaptations have to be made to the psychotherapy so that it can be delivered in the non-mental health setting. For instance, in both the IMPACT and PROSPECT programs, the psychotherapies were selected, in part, on their ability to match the pace of primary-care practices.
Guideline 2: Select an Age-Appropriate Comparison Group
Special attention should also be given to selection of the comparison condition against which the intervention will be tested, particularly when the trial is comparing a new intervention to an existing one or to usual care. When comparing interventions, it is important that both undergo the adaptations described above. If the comparison intervention is not adapted, then differences in outcome between interventions may simply be due to one intervention not being amenable to the needs of older adults, and, thus, outcomes will be difficult to interpret. In comparing an intervention to usual care, the researcher must first be careful to detail what constitutes "usual care," and then make sure usual care is not a potentially toxic alternative. In the PROSPECT trial, usual care was theoretically the best comparison for the intervention, but was found to be an unethical option for the target population. Because older adults with suicidal ideation are more likely to kill themselves than those in any other age-group, the investigators modified usual care by informing primary-care physicians about their patients' suicidality.
Guideline 3: Use Consumer-Based Methods for Recruiting and Retaining Research Participants
One of the biggest challenges of conducting any geriatric treatment study is recruiting and retaining an adequate number of older people into the project. This is particularly true for geropsychotherapy research, where misconceptions about psychotherapy can prevent older people from volunteering and result in the participation of only motivated and psychologically sophisticated people. Because the usefulness of psychotherapy should be demonstrated for all older people who encounter psychotherapy as a treatment option, recruitment and retention of broader samples of older adults is necessary. Several methods for recruiting older adults into geropsychotherapy research exist, and are all based on the premise that to attract volunteers, the target population must be involved in the design of recruitment methods.17 The REACH project, one of the largest cohorts of dementia caregivers achieved to date (N=1,222), used social marketing as its recruitment approach. In social marketing, the product (i.e., therapy), price (i.e., participant burden), place (i.e., location of recruitment), and promotion techniques (i.e., recruitment strategies) are set in collaboration with representatives of the target population. Using this method, the REACH investigators from multiple sites developed general and site-specific recruitment and retention strategies by "partnering" community organizations such as physicians' practices, the Visiting Nursing Association, and leaders in the local minority communities. These partners helped determine the best ways to recruit for the study (e.g., via the radio, in Latino populations; face-to-face recruitment, in medicine), and the best location for treatment (churches and senior centers), and they helped determine the most acceptable length of the treatment and research assessments.
Guideline 4: Evaluate Age-Related Treatment-Process Variables
In addition to measuring processes related to treatment delivery, receipt, and enactment, a number of age-related process variables are important to measure in order to ascertain when the intervention is most successful and under what circumstances it may need to be altered. For instance, stigma concerns regarding mental health are important to measure because they may negatively influence patients' engagement regardless of a priori expectations they may have about the effectiveness of psychotherapy. It is also common practice to measure therapeutic expectancies before and after the treatment experience. Although not used as a primary determinant of outcome, these measures can help to clarify whether treatment expectancies influence outcomes and/or change over time. This information can be used to modify the process of psychotherapy. Examples of common measures are the Lorr Scale18 and the Treatment Rationale Scale.19
These process measures are best incorporated in conjunction with process measures of attendance and utilization of related services. In traditional psychotherapy research, participants are told not to engage in other forms of treatment, so that the pure effect of psychotherapy can be measured. For older adults, this requirement is unrealistic because they often have other supportive encounters with a personal physician, case manager, caregiver support groups, medical illness support groups, and social activities through senior centers. Thus, geropsychotherapy researchers must also record utilization of those services so that both the distinct and combined effects of psychotherapy can be measured.20 In PEPUP, the investigators used a modified version of Lehman's case-management questionnaire to record both the number of social service contacts the patients had and how satisfied they were with those services.21 These measures of related services can serve as covariates in assessing primary outcomes of therapeutic interventions or mediators of therapeutic effects. They can also be viewed as direct outcomes of treatment, inasmuch as some interventions enable clients to access appropriate services.
Additional measures that should be collected to help inform the process of therapy are life events, cognitive status, and pain/illness measures. Life events, such as onset of acute illness or loss of a loved one, can affect how older people report their outcomes at follow-up assessments. In PEPUP, researchers tracked life events at each assessment period to qualify/quantify outcomes in mood.2 Unless a study is explicitly assessing a psychotherapy's usefulness for older adults with cognitive impairment, people with frank impairment are usually screened out of the study either through use of a cognitive screening exam, such as the Folstein Mini-Mental State Exam (MMSE)22 or through more formal assessment. However, cognitive status can change over time, potentially influencing the outcomes of the study. Thus, in the PEPUP project, the MMSE was administered at the follow-up assessment periods. Chronic pain and illness is also common in older adults and has been known to influence mood and activity level. The McGill Pain Questionnaire23 is one scale that has been used extensively in older, disabled populations. As with use of services, changes in pain, cognition, and resiliency to life events over time can be used both as a covariate on psychotherapy effects and as an outcome.
Guideline 5: Measure Age-Specific Treatment Outcomes
Often, the greatest changes in older adults may not be symptom resolution, but how they negotiate their environment. For this reason, geropsychotherapy researchers assess a broad range of clinical outcomes, including activity engagement, disability, life satisfaction, and psychosocial functioning. Therefore, the key outcomes of geropsychotherapy are not solely symptom resolution, but functional change and improvement, as well. In the IMPACT, PEPUP, and PROSPECT studies, baseline and follow-up observation included standard assessments of disability and functioning. Both PEPUP and IMPACT used the SF–36,24 whereas the PROSPECT study used the WHODAS–II.25 PEPUP also measured outcomes such as life satisfaction, using the Life Satisfaction Index, a problem checklist that measured resolution of target problems, and activity engagement, using the Older Adult Activity Schedule.26,27
Multiple outcome measures increase the complexity of analyzing treatment outcomes. Researchers must distinguish between what they consider to be the primary outcome of a treatment from related secondary outcomes, or alternatively, adopt a multivariate-outcomes analytic strategy to assess treatment effectiveness. In either case, it will be important to define the key outcomes of a study before treatment is initiated and to ensure a conceptual link between the treatment and its expected outcomes. The power analysis performed for the PROSPECT study to determine its needed sample size considered a reduction in suicidal ideation as the primary outcome, but it also incorporated the reduction of depressive symptomatology as an additional outcome.
Guideline 6: Adjust the Length and Timing of Research Evaluations
As Guidelines 4 and 5 imply, participant interviews in geropsychotherapy research have the potential for being lengthy and redundant. These are two key problems that have been found to increase refusal to continue protocol participation.17 The geropsychotherapy researcher should design the assessment battery with the physical and cognitive needs of the older adult in mind, particularly the older adult with significant or multiple medical and/or psychiatric problems. For example, fatigue is a factor that may cause older adults to exhibit decrements in attention and is a factor to be considered when selecting individual measures and considering the length of the entire battery of measures. In this regard, although the entire battery should encompass a wide range of non-overlapping late-life process variables and outcomes, it nevertheless should be as brief as possible and should be administered by an assessor experienced in testing older adults. Examples of measures selected with the needs of older patients in mind include the Beck Depression Inventory-Primary Care (BDI–PC), rather than the standard BDI, to measure depressive symptoms, the use of the Mini-International Neuropsychiatric Interview (MINI), rather than the Structured Clinical Interview for DSM-IV (SCID-IV) for clinical diagnostic purposes, and the 12-item Short-Form Health Survey (SF–12), rather than the SF–36, with 36 items for measuring health status.28–31 Also, special care should be given to the timetable for measuring particular outcomes. Not all outcomes need be measured simultaneously. For instance, IMPACT process measures to determine application of the therapy skills were not part of the outcome assessment, but rather were collected after critical periods during the intervention (i.e., after Sessions 1, 3, and 6 of a four- to eight-session treatment protocol).
We recognize that there are circumstances when the demands of the research preclude streamlining the research interview. For example, case management was a crucial component of treatment in PEPUP and the addition of case-management questions to the 6- and 18-month follow-up interviews resulted in 2-hour assessments. In order to make the interview experience less demanding, the 6- and 18-month interviews were divided into two separate meetings, and were often administered in the participants' homes.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONGUIDELINES FOR GEROPSYCHOTHERAPY...CONCLUSIONREFERENCES |
|---|
Psychotherapy with older populations is a growing area of research. Following these guidelines will serve to improve the quality of future geropsychotherapy research, and this improvement can serve as a springboard for continued methodological refinements and quality research.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGUIDELINES FOR GEROPSYCHOTHERAPY...CONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. P. Vitaliano, W. Katon, and J. Unutzer Making the Case for Caregiver Research in Geriatric Psychiatry Am J Geriatr Psychiatry, October 1, 2005; 13(10): 834 - 843. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Granholm, J. R. McQuaid, F. S. McClure, L. A. Auslander, D. Perivoliotis, P. Pedrelli, T. Patterson, and D. V. Jeste A Randomized, Controlled Trial of Cognitive Behavioral Social Skills Training for Middle-Aged and Older Outpatients With Chronic Schizophrenia Am J Psychiatry, March 1, 2005; 162(3): 520 - 529. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. D. Lebowitz Clinical Trials in Late Life: New Science in Old Paradigms Gerontologist, August 1, 2004; 44(4): 452 - 458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Arean Advances in Psychotherapy for Mental Illness in Late Life Am J Geriatr Psychiatry, February 1, 2003; 11(1): 4 - 6. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ALL ISSUES | SEARCH | TABLE OF CONTENTS |
