Evaluating measurement equivalence using the item response theory log-likelihood ratio (IRTLR) method to assess differential item functioning (DIF): applications (with illustrations) to measures of physical functioning ability and general distress

BACKGROUND

Methods based on item response theory (IRT) that can be used to examine differential item functioning (DIF) are illustrated. An IRT-based approach to the detection of DIF was applied to physical function and general distress item sets. DIF was examined with respect to gender, age and race. The method used for DIF detection was the item response theory log-likelihood ratio (IRTLR) approach. DIF magnitude was measured using the differences in the expected item scores, expressed as the unsigned probability differences, and calculated using the non-compensatory DIF index (NCDIF). Finally, impact was assessed using expected scale scores, expressed as group differences in the total test (measure) response functions.

METHODS

The example for the illustration of the methods came from a study of 1,714 patients with cancer or HIV/AIDS. The measure contained 23 items measuring physical functioning ability and 15 items addressing general distress, scored in the positive direction.

RESULTS

The substantive findings were of relatively small magnitude DIF. In total, six items showed relatively larger magnitude (expected item score differences greater than the cutoff) of DIF with respect to physical function across the three comparisons: "trouble with a long walk" (race), "vigorous activities" (race, age), "bending, kneeling stooping" (age), "lifting or carrying groceries" (race), "limited in hobbies, leisure" (age), "lack of energy" (race). None of the general distress items evidenced high magnitude DIF; although "worrying about dying" showed some DIF with respect to both age and race, after adjustment.

CONCLUSIONS

The fact that many physical function items showed DIF with respect to age, even after adjustment for multiple comparisons, indicates that the instrument may be performing differently for these groups. While the magnitude and impact of DIF at the item and scale level was minimal, caution should be exercised in the use of subsets of these items, as might occur with selection for clinical decisions or computerized adaptive testing. The issues of selection of anchor items, and of criteria for DIF detection, including the integration of significance and magnitude measures remain as issues requiring investigation. Further research is needed regarding the criteria and guidelines appropriate for DIF detection in the context of health-related items.

IRT health outcomes data analysis project: an overview and summary

BACKGROUND

In June 2004, the National Cancer Institute and the Drug Information Association co-sponsored the conference, "Improving the Measurement of Health Outcomes through the Applications of Item Response Theory (IRT) Modeling: Exploration of Item Banks and Computer-Adaptive Assessment." A component of the conference was presentation of a psychometric and content analysis of a secondary dataset.

OBJECTIVES

A thorough psychometric and content analysis was conducted of two primary domains within a cancer health-related quality of life (HRQOL) dataset.

RESEARCH DESIGN

HRQOL scales were evaluated using factor analysis for categorical data, IRT modeling, and differential item functioning analyses. In addition, computerized adaptive administration of HRQOL item banks was simulated, and various IRT models were applied and compared.

SUBJECTS

The original data were collected as part of the NCI-funded Quality of Life Evaluation in Oncology (Q-Score) Project. A total of 1,714 patients with cancer or HIV/AIDS were recruited from 5 clinical sites.

MEASURES

Items from 4 HRQOL instruments were evaluated: Cancer Rehabilitation Evaluation System-Short Form, European Organization for Research and Treatment of Cancer Quality of Life Questionnaire, Functional Assessment of Cancer Therapy and Medical Outcomes Study Short-Form Health Survey.

RESULTS AND CONCLUSIONS

Four lessons learned from the project are discussed: the importance of good developmental item banks, the ambiguity of model fit results, the limits of our knowledge regarding the practical implications of model misfit, and the importance in the measurement of HRQOL of construct definition. With respect to these lessons, areas for future research are suggested. The feasibility of developing item banks for broad definitions of health is discussed.

Identification of differential item functioning using item response theory and the likelihood-based model comparison approach. Application to the Mini-Mental State Examination

BACKGROUND

An important part of examining the adequacy of measures for use in ethnically diverse populations is the evaluation of differential item functioning (DIF) among subpopulations such as those administered the measure in different languages. A number of methods exist for this purpose.

OBJECTIVE

The objective of this study was to introduce and demonstrate the identification of DIF using item response theory (IRT) and the likelihood-based model comparison approach.

METHODS

Data come from a sample of community-residing elderly who were part of a dementia case registry. A total of 1578 participants were administered either an English (n = 913) or Spanish (n = 665) version of the 21-item Mini-Mental State Examination. IRT was used to identify language DIF in these items with the likelihood-based model comparison approach.

RESULTS

: Fourteen of the 21 items exhibited significant DIF according to language of administration. However, because the direction of the identified DIF was not consistent for one language version over the other, the impact at the scale level was negligible.

CONCLUSIONS

IRT and the likelihood-based model comparison approach comprise a powerful tool for DIF detection that can aid in the development, refinement, and evaluation of measures for use in ethnically diverse populations.

A program to reduce the disparity in the rate of biennial lipid profiles between African-American and white Medicare beneficiaries with diabetes mellitus in New York City

The burden of diabetes is more severe for minority populations than for the white population in the United States. Analysis of Medicare fee-for-service claims data revealed a significant disparity regarding screening for lipids between white and African-American Medicare beneficiaries with diabetes as measured by the biennial lipid profile indicator in New York State and the disparity was even greater in New York City. Thus IPRO, the Medicare New York State Quality Improvement Organization, launched a multifaceted quality improvement project to reduce the disparity, targeting African-American Medicare beneficiaries with diabetes in New York City and the providers who serve them. There was an absolute increase of 16.7% in the proportion of African-American beneficiaries with diabetes receiving a biennial lipid profile in the intervention areas and the disparity reduced by 9.8% between African-American and all eligible white beneficiaries in the intervention areas. Although it was not feasible to determine the direct impact of selected interventions on reducing the disparity, the interventions collectively appeared to be effective.

Effect of inpatient quality of care on functional outcomes in patients with hip fracture

OBJECTIVES

We sought to examine the relationship between functional outcome and process of care for patients with hip fracture.

RESEARCH DESIGN AND PARTICIPANTS

We undertook a prospective cohort study in 4 hospitals of 554 patients treated with surgery for hip fracture.

MEASUREMENTS

Information on patient characteristics and processes of hospital care collected from the medical record, interviews, and bedside observations. Follow-up information obtained at 6 months on function (using the Functional Independence Measure [FIM]), survival, and readmission.

RESULTS

Individual processes of care were generally not associated with adjusted outcomes. A scale of 9 processes related to mobilization was associated with improved adjusted locomotion (P = 0.006), self care (P = 0.022), and transferring (P = 0.007) at 2 months, but the benefits were smaller and not significant by 6 months. These processes were not associated with mortality. The predicted value for the FIM locomotion measure (range, 2-14) at 2 months was 5.9 (95% confidence interval 5.4-6.4) for patients at the 10th percentile of performance on these processes compared with 7.1 (95% confidence interval 6.6, 7.6) at the 90th percentile. Patients who experienced no hospital complications and no readmissions retained the benefits in locomotion at 6 months. Anticoagulation processes were associated with improved transferring at 2 months (P = 0.046) but anticoagulation and other processes of care were not otherwise associated with improved function.

DISCUSSION

Our findings indicate the need to attend to all steps in the care of patients with hip fracture. Additionally, functional outcomes were more sensitive markers of improved process of care, compared with 6-month mortality, in the case of hip fracture.

The development and psychometric analyses of ADEPT: an instrument for assessing the interactions between doctors and their elderly patients

BACKGROUND

The lack of instruments and methodologies designed specifically for assessing doctor-elderly patient interactions has constricted research on effective communication in the medical care of older adults.

PURPOSE

This article reports on the development, qualitative analyses, and psychometric testing of the Assessment of Doctor-Elderly Patient Transactions (ADEPT), an instrument for assessing interactions between doctors and their elderly patients.

METHODS

The ADEPT was based on the recommendations of an expert panel and designed around the three-function model of the medical interview. The ADEPT is meant to operationalize the research findings of interactional analysis studies of doctor-patient interaction. Following preliminary testing with standardized patients, the ADEPT was applied to videotaped visits of 433 patients 65 years of age and older to the doctor (n = 40) identified as their primary source of care.

RESULTS

Four final scales derived from exploratory and confirmatory factor analyses were scored: Supporting, comprised of the 12 items from the first factor; Eliciting Needs, containing the 5 items from the second factor; and Informing, based on the final 6-item factor. Individual Cronbach's alphas across raters for this sample ranged from .71 to .79 for the first scale, from .83 to .88 for the second scale, and from .64 to .81 for the third scale. The reliability estimates for the total scale (23 items) ranged from .80 to .86 across raters. A fifth summed index composed of 46 binary checklist items also was computed.

CONCLUSIONS

The findings indicate that credible scales can be developed for assessing communication behaviors.

The relationship between organizational systems and clinical quality in diabetes care

OBJECTIVE

To assess the clinical quality of diabetes care and the systems of care in place in Medicare managed care organizations (MCOs) to determine which systems are associated with the quality of care.

STUDY DESIGN

Cross-sectional, observational study that included a retrospective review of 2001 diabetes Health Plan Employer and Data Information Set (HEDIS) measures and a mailed survey to MCOs.

METHODS

One hundred and thirty-four plans received systems surveys. Data on clinical quality were obtained from HEDIS reports of diabetes measures.

RESULTS

Ninety plans returned the survey. Composite diabetes quality scores (CDSs) were based on averaging scores for the 6 HEDIS diabetes measures. For the upper quartile of responding plans, the average score was 77.6. The average score for the bottom quartile was 53.9 (P < .001). The mean number of systems or interventions for the upper-quartile group and the bottom-quartile group was 17.5 and 12.5 (P < .01), respectively. There were significant differences in the 2 groups in the following areas: computer-generated reminders, physician champions, practitioner quality-improvement work groups, clinical guidelines, academic detailing, self-management education, availability of laboratory results, and registry use. After adjusting for structural and geographic variables, practitioner input and use of clinical-guidelines software remained as independent predictors of CDS. Structural variables that were independent predictors were nonprofit status and increasing number of Medicare beneficiaries in the MCO.

CONCLUSIONS

MCO structure and greater use of systems/interventions are associated with higher-quality diabetes care. These relationships require further exploration.

Modern psychometric methods for detection of differential item functioning: application to cognitive assessment measures

Cognitive screening tests and items have been found to perform differently across groups that differ in terms of education, ethnicity and race. Despite the profound implications that such bias holds for studies in the epidemiology of dementia, little research has been conducted in this area. Using the methods of modern psychometric theory (in addition to those of classical test theory), we examined the performance of the Attention subscale of the Mattis Dementia Rating Scale. Several item response theory models, including the two- and three-parameter dichotomous response logistic model, as well as a polytomous response model were compared. (Log-likelihood ratio tests showed that the three-parameter model was not an improvement over the two-parameter model.) Data were collected as part of the ten-study National Institute on Aging Collaborative investigation of special dementia care in institutional settings. The subscale KR-20 estimate for this sample was 0.92. IRT model-based reliability estimates, provided at several points along the latent attribute, ranged from 0.65 to 0.97; the measure was least precise at the less disabled tail of the distribution. Most items performed in similar fashion across education groups; the item characteristic curves were almost identical, indicating little or no differential item functioning (DIF). However, four items were problematic. One item (digit span backwards) demonstrated a large error term in the confirmatory factor analysis; item-fit chi-square statistics developed using BIMAIN confirm this result for the IRT models. Further, the discrimination parameter for that item was low for all education subgroups. Generally, persons with the highest education had a greater probability of passing the item for most levels of theta. Model-based tests of DIF using MULTILOG identified three other items with significant, albeit small, DIF. One item, for example, showed non-uniform DIF in that at the impaired tail of the latent distribution, persons with higher education had a higher probability of correctly responding to the item than did lower education groups, but at less impaired levels, they had a lower probability of a correct response than did lower education groups. Another method of detection identified this item as having DIF (unsigned area statistic=3.05, p<0.01, and 2.96, p<0.01). On average, across the entire score range, the lower education group's probability of answering the item correctly was 0.11 higher than the higher education group's probability. A cross-validation with larger subgroups confirmed the overall result of little DIF for this measure. The methods used for detecting differential item functioning (which may, in turn, be indicative of bias) were applied to a neuropsychological subtest. These methods have been used previously to examine bias in screening measures across education and ethnic and racial subgroups. In addition to the important epidemiological applications of ensuring that screening measures and neuropsychological tests used in diagnoses are free of bias so that more culture-fair classifications will result, these methods are also useful for the examination of site differences in large multi-site clinical trials. It is recommended that these methods receive wider attention in the medical statistical literature.

Placebo run-in period in studies of depressive disorders. Clinical, heuristic and research implications

BACKGROUND

In spite of the virtually ubiquitous nature of the initial 10-day placebo run-in period (IPR) in drug trials, there is little empirical data establishing its relevance.

METHOD

Data from 593 subjects were examined retrospectively to determine whether or not the prognosis of subjects minimally improved during the IPR was different to those who were unimproved. The IPR period was single-blind and was followed by a six-week double-blind phase in all studies.

RESULTS

Twenty-six per cent of the subjects were minimally improved and 74% were unimproved. Approximately 10% of the subjects who were much improved were not followed systematically. Across a range of diagnosis, severity and chronicity subjects minimally improved (versus unimproved) after IPR had a more favourable prognosis whether assigned to drug or placebo.

CONCLUSIONS

Change during IPR appears to be a meaningful predictor. Stratification should be considered in future antidepressant studies.

Imipramine treatment of opiate-dependent patients with depressive disorders. A placebo-controlled trial

BACKGROUND

The literature is inconclusive on the role of antidepressant medications in treating drug dependence. Studies have either not focused on depressed patients or have selected patients with depressive disorders based on cross-sectional symptoms rather than a syndromal diagnosis. A clinical trial of an antidepressant was, therefore, conducted on drug-dependent patients with syndromal depression.

METHODS

Patients receiving methadone hydrochloride maintenance treatment were selected if they met the criteria for a DSM-III-R depressive disorder that was chronologically primary, had persisted during a past abstinent period or was long-standing, and persisted during at least 1 month of stable methadone treatment. Subjects were randomized to a 12-week, double-blind, placebo-controlled trial of imipramine hydrochloride. A favorable response was defined as a Clinical Global Impression scale score for depression of 2 ("much improved") or 1 ("very much improved") and at least a 75% reduction in self-reported drug or alcohol use or abstinence.

RESULTS

One hundred thirty-seven patients were randomized, and 84 completed a minimum adequate trial of at least 6 weeks. Among the 84 adequately treated patients, 57% (24/42) receiving imipramine were rated as responders compared with 7% (3/42) receiving placebo (P < .001). On measures of mood, there was a robust effect of imipramine. Imipramine was superior to placebo on some self-reported measures of substance use and craving, and mood improvement was associated with improvement in self-reported substance use. However, few patients achieved urine-confirmed abstinence.

CONCLUSIONS

Imipramine was an effective antidepressant in patients receiving methadone who were selected via syndromal criteria for depressive illness. Imipramine may reduce substance abuse among patients whose mood improves; however, this effect was less robust.

The efficacy of imipramine and psychotherapy in early-onset chronic depression: a reanalysis of the National Institute of Mental health Treatment of Depression Collaborative Research Program

The authors compared the effectiveness of Cognitive Behavioral Therapy (CBT), Interpersonal Psychotherapy (IPT), Imipramine Clinical Management (ICM) to Placebo Clinical Management (PCM) for outpatients with early-onset chronic depression (N = 65) in the National Institute of Mental Health (NIMH) Treatment of Depression Collaborative Research Program (TDRP). The post-treatment depression scores of the CBT. IPT, and ICM groups were not significantly different from the PCM group. We did not find a relationship between the duration of Major Depression and response to a specific treatment. Studies are needed to determine if combining psychotherapy with medication improves social functioning and enhances the quality of life for patients with chronic depression.

Chronological milestones to guide drug change. When should clinicians switch antidepressants?

BACKGROUND

We attempt to identify the time when patients whose conditions are unimproved while receiving antidepressants are unlikely to respond and should have their treatment changed.

METHODS

A total of 593 patients were studied. The course of treatment for patients was examined to determine the weeks at which patients who received drug therapy had a better chance of being rated as responders at the study end (week 6) vs patients who received placebo.

RESULTS

At the end of week 3, 19 (32%) of the 59 patients who received drug therapy and 6 (10%) of the 57 patients who received placebo and who never minimally improved were rated as responders at week 6. For those who showed no improvement by week 4, the effects of drug therapy and the placebo were equal. Patients who received drug therapy and whose conditions were unimproved but who had been minimally improved at some point had a superior prognosis with drug therapy vs placebo until week 4. Of those unimproved at week 4 but minimally improved at some point previously, 20 (39%) of the 51 patients who received drug therapy vs 3 (8%) of the 36 patients who received placebo were rated as responders at week 6. Of the 75 patients who minimally improved while receiving drug therapy at the end of week 5, 33 (44%) had a chance of being rated a responder at the end of week 6 vs 9 (26%) of the 35 patients receiving placebo.

CONCLUSIONS

Patients tolerant of an adequate dose, whose conditions have never been at least minimally improved by the end of week 4, should have their treatment regimen altered. These patients represented a minority of drug-treated patients in the sample studied (ie, 39/392 [10%]). Patients whose conditions minimally improve at some prior week but not after week 5 should have their treatment changed. Patients whose conditions minimally improve in week 5 should continue treatment until week 6.

Imipramine treatment of alcoholics with primary depression: A placebo-controlled clinical trial

BACKGROUND

Depressive disorders are commonly comorbid with alcoholism, particularly in treatment-seeking samples. If antidepressant treatment were safe and improved the treatment outcome in the subset of actively drinking alcoholics with depression, this would be of clinical importance.

METHODS

We conducted a randomized, 12-week placebo-controlled trial of imipramine hydrochloride combined with weekly relapse prevention psychotherapy. The subjects were 69 actively drinking alcoholic outpatients with current depressive disorders. The first onset of depression was either antecedent to the abuse of alcohol or occurred during prolonged periods of sobriety. Depression and drinking outcomes at 12 weeks, as well as their relationship, were measured.

RESULTS

Imipramine treatment was safe and associated with improvement in depression in both adequately treated and intention-to-treat samples. While there was no overall effect on drinking outcome, patients whose mood improved showed decreased alcohol consumption that was more marked in those treated with imipramine.

CONCLUSIONS

Imipramine treatment is effective for primary depression among actively drinking alcoholic outpatients, and may improve alcoholic outcome for those whose depression responds to treatment.

Imipramine treatment of cocaine abuse: possible boundaries of efficacy

A 12-week placebo-controlled, randomized clinical trial was undertaken to evaluate imipramine as a treatment for cocaine abuse, and to examine whether its effect may be limited to subgroups defined by route of use or by diagnosis of depression. One-hundred thirteen patients were randomized, stratified by route of use and depression. All patients received weekly individual counseling. Compared to placebo the imipramine group showed greater reductions in cocaine craving, cocaine euphoria, and depression, but the effect of imipramine on cocaine use was less clear. A favorable response, defined as at least 3 consecutive, urine-confirmed, cocaine-free weeks was achieved by 19% (11/59) of patients on imipramine compared to 7% (4/54) on placebo (P < 0.09). The imipramine effect was greater among nasal users--33% (9/27) response on imipramine vs. 5% (1/22) on placebo (P < 0.02). Response was also more frequent, but not significantly so, among depressed users on imipramine (26%, 10/38) than on placebo (13%, 4/31) (P < 0.19). Response rates were low in intravenous and freebase users and those without depression. Considered together with the literature on desipramine, these data suggest tricyclic antidepressants are not promising as a mainstay of treatment for unselected cocaine abusers. However, tricyclics may be useful for selected cocaine abusers with comorbid depression or intranasal use, or in conjunction with a more potent psychosocial intervention.

Duration of antidepressant trials: clinical and research implications

The objective of our study was to demonstrate that additional antidepressant benefit occurs between weeks 4 and 6 in adult outpatients, even when dose is not increased. Response between weeks 4 and 6 was studied among depressed outpatients randomly assigned to imipramine, phenelzine, or placebo under double-blind conditions. Patients were selected for analysis only if they did not have a dose increase after the start of the fourth week of treatment (day 22). Eighty-eight patients met this condition. Conditional probability analysis was performed. Nonresponders to 4 weeks (28 days) of treatment had a significantly greater likelihood of responding by week 6 if they were on phenelzine rather than placebo. The same is probably true for patients on imipramine. In research and clinical care, 4 weeks is too short a trial of phenelzine to conclude a lack of efficacy. Four weeks is probably also too short a trial of imipramine.

A double-blind placebo-controlled comparison of phenelzine and imipramine in the treatment of bulimia in atypical depressives

Although antidepressants have been found to be superior to placebo in 12 of 14 studies, the relationship between improvement in the depressive diathesis and bulimia is unclear. In this study, the efficacy of placebo, imipramine, and phenelzine is examined in patients comorbid for atypical depression and bulimia. Greater improvement was observed for both depressive and bulimic symptoms with phenelzine than with either imipramine or placebo. Consistent with its poor antidepressant effects in atypical depression, imipramine seemed to have minimal efficacy for the bulimic symptoms of atypical depressives. These data suggest that the presence of bulimia does not alter the treatment response of atypically depressed patients. Furthermore, the data may suggest a link between depression and bulimia in atypical depressives. Demonstrating a statistical difference with a small sample suggests the effect size is robust, however conclusions are limited by a small sample size.

Chronic depression: response to placebo, imipramine, and phenelzine

We reanalyzed data from a larger, previously published study in order to directly address whether very chronically depressed patients could benefit from antidepressant medications. This study entered 598 depressed patients into a study randomizing patients to 6 weeks of double-blind treatment with imipramine, phenelzine, or placebo. Patients were assessed for chronicity on a four-point scale from "mostly well" to "virtually always depressed." The current analyses include only the 153 study completers who were rated as "virtually always depressed." In these patients, imipramine was effective for significantly more patients than was placebo (22 [46%] of 48 responding to imipramine vs. 9 [17%] of 52 responding to placebo; chi 2 = 9.50; p = 0.002), whereas phenelzine was significantly more effective than imipramine (37 [70%] of 53 responding to phenelzine; chi 2 = 5.96; p = .015). Patients with mild depression, early onset, or histories of panic attacks did not have substantially different outcomes than patients without these characteristics. These findings suggest that some chronically depressed patients may be good candidates for treatment with antidepressant medication. Because the majority (80%) of the sample met Columbia criteria for definite or probable atypical depression, too few chronic depressives were available to evaluate separately antidepressant efficacy in chronically depressed outpatients who did not have atypical depression. Hence, these results may be applicable only to patients with atypical depression.

Columbia atypical depression. A subgroup of depressives with better response to MAOI than to tricyclic antidepressants or placebo

We summarise a series of studies using a MAOI to help establish the validity of a subgroup of depressives referred to as atypical depressives. Patients with reactive mood meeting DSM-III criteria for depressive illness who had associated atypical features (which include hyperphagia, hypersomnolence, leaden paralysis, and rejection sensitivity) were randomised to imipramine, phenelzine and placebo. Non-responders were crossed over, and in all there were over 400 patient trials. Phenelzine consistently was found to be superior to imipramine. Only in trials which included patients lacking atypical, vegetative symptoms was imipramine found to equal phenelzine. We conclude that the researcher and the clinician should consider the relevance of the atypical depressive syndrome.

Imipramine treatment of alcoholism with comorbid depression

Of 60 depressed alcoholics who completed an open trial of imipramine, 27 (45%) responded with improvement in both mood and drinking behavior, and eight (13%) responded after further dosage increases or treatment with disulfiram. In a subsequent 6-month, randomized discontinuation trial, four of 13 subjects (31%) relapsed during imipramine treatment and seven of 10 (70%) relapsed while taking placebo. This suggests a potential treatment approach for a high-risk subgroup of alcoholics.

How symptomatic do depressed patients remain after benefiting from medication treatment?

Symptom Check List-90 (SCL-90) scores of 318 moderately depressed, nonmelancholic outpatients treated with phenelzine, imipramine, or placebo were compared with scores obtained from a demographically matched control group of 69 "normal" subjects. Pretreatment scores of depressed patients were significantly higher than those of the control group; after 6 weeks of treatment, they were indistinguishable from scores of the control group. These results suggest that symptoms of moderately depressed patients who benefit from treatment not only decrease from baseline, but are reduced to a level of symptomatology that is psychiatrically normal.

Loss of drug effects during continuation therapy

OBJECTIVE

This study sought to determine what proportion of relapses during continuation therapy with antidepressants can be attributed to loss of nonspecific placebo effects while the patients are taking the drugs.

METHOD

Depressed patients were studied over a 12-week period. One hundred sixty-four patients were randomly assigned to placebo, 174 to imipramine, and 169 to phenelzine. At 6 weeks 35 were judged to be responders to placebo, 70 to imipramine, and 96 to phenelzine. These patients continued their double-blind treatment for weeks 7-12.

RESULTS

Thirty-one percent of the patients who were taking placebo, approximately 12% who were taking imipramine, and approximately 9% who were taking phenelzine relapsed in the 7- to 12-week phase. Two different methods of estimating relapses suggested that during the first 3 months of treatment, a large percentage of the relapses of patients taking drugs was attributable to the loss of nonspecific placebo effects rather than true drug effects.

CONCLUSIONS

A considerable proportion of relapses in the first 3 months of treatment with antidepressants appears to be due to loss of placebo effects. These clinically relevant data may be used to encourage patients who relapse during this period, and who erroneously conclude that anti-depressant effects are temporary, to try another medication.

Further evidence that a placebo response to antidepressants can be identified

OBJECTIVE

The authors' goal was to analyze the acute phase of antidepressant drug treatment to identify placebo responses.

METHOD

Patients rated as improved after 6 weeks of double-blind treatment with imipramine or phenelzine were followed for an additional 6 weeks of double-blind treatment. Initial responses were classified according to the speed of improvement (abrupt or gradual), the persistence or nonpersistence of improvement, and the timing of improvement (early or late onset).

RESULTS

It was predicted that patients with nonpersistent, abrupt responses to the drugs were actually experiencing a placebo response and would have the worst prognosis. In fact, this group accounted for a disproportionate number of the relapses. Nonpersistent responders to a drug had a 23.7% relapse rate, but persistent responders had only a 9.0% relapse rate, a significant difference.

CONCLUSIONS

The authors conclude that a significant proportion of relapses within the first 6 weeks of treatment with an active drug are not related to loss of a true drug effect. Rather, some are related to loss of nonspecific placebo effects, and abrupt nonpersistent responses during drug treatment are most likely the result of placebo effects.

Demoralization Predicts Nonresponse to Cognitive Therapy in Depressed Outpatients

Thirty-nine depressed outpatients meeting DSM-III criteria for nonmelancholic major depression or dysthymic disorder were treated with 16 weekly individual cognitive therapy sessions. Prior to treatment, they completed the Beck Depression Inventory, the Hopelessness Scale, and the Dysfunctional Attitudes Scale. Independent of knowledge of outcome, the authors chose from these scales items indicating demoralization, that is, that patients perceived their ability to positively affect their own future as too likely to be ineffectual to warrant efforts at change. After cognitive therapy, 20 patients were considered responders (51%) although three quickly relapsed (44% responded and maintained). Nonresponders had significantly higher pretreatment demoralization scores than did responders. These results suggest that high levels of demoralization may predict poor response of depression to cognitive therapy, although the small sample size precluded differentiation of demoralization from hopelessness.