Annotated Abstracts of Journal Articles
2014, 3rd Quarter
Annotations by Jeff Huffman, MD, FAPM, Scott R. Beach, MD, and Christopher Celano, MD
September 2014
ANNOTATION (Jeff Huffman)
The Finding: Examined from numerous angles, a telephone-based collaborative care depression intervention for CABG patients was generally associated with no significant increase in healthcare costs, and on almost all analyses met accepted criteria for being a highly cost-effective intervention.
Strengths and Weaknesses: Strengths included a comprehensive and sophisticated cost analysis that included actual costs, actual costs outside of one high-cost outlier, and estimated costs using complex statistical models to estimate/extrapolate projected costs (to account for big differences in mean/median and the small sample size). Weaknesses included a substantial proportion of participants (~40%) without full claims data and the variability mentioned above that required removal of outliers and use of multiple analyses to examine costs, resulting in some uncertainty in the reader about which analysis is “best” or most accurate. Indeed, simply using all patients and actual costs, the collaborative care intervention was much more expensive, but with the removal of a single outlier, it became much less costly than usual care.
Relevance: Taken together, the analyses here do appear to support the notion that collaborative care, provided by telephone, in a high-risk cohort of depressed post-CABG patients, is cost-effective and may even be cost-saving.
Objective: To determine the 12-month cost-effectiveness of a collaborative care (CC) program for treating depression following coronary artery bypass graft (CABG) surgery versus physicians’ usual care (UC).
Methods: We obtained 12 continuous months of Medicare and private medical insurance claims data on 189 patients who screened positive for depression following CABG surgery, met criteria for depression when reassessed by telephone 2 weeks following hospitalization (nine-item Patient Health Questionnaire ≥10) and were randomized to either an 8-month centralized, nurse-provided and telephone-delivered CC intervention for depression or to their physicians’ UC.
Results: At 12 months following randomization, CC patients had $2068 lower but statistically similar estimated median costs compared to UC (P=.30) and a variety of sensitivity analyses produced no significant changes. The incremental cost-effectiveness ratio of CC was -$9889 (-$11,940 to -$7838) per additional quality-adjusted life-year (QALY), and there was 90% probability it would be cost-effective at the willingness to pay threshold of $20,000 per additional QALY. A bootstrapped cost-effectiveness plane also demonstrated a 68% probability of CC “dominating” UC (more QALYs at lower cost).
Conclusions: Centralized, nurse-provided and telephone-delivered CC for post-CABG depression is a quality-improving and cost-effective treatment that meets generally accepted criteria for high-value care.
ANNOTATION (Scott Beach)
The Finding: Enhanced depression treatment in patients with acute coronary syndrome (ACS) reduced rates of a composite outcome of death or rehospitalization for ACS during the 6-month study period, but these effects did not persist during the subsequent year after treatment ended.
Strengths and Weaknesses: Strengths of the study included a thorough method of evaluating for the composite outcome including the use of three cardiologists, sound statistical methods, and relatively long follow-up period for this type of study. Weaknesses included relatively small sample size, a small total number of cardiac readmissions of deaths, and post-hoc nature of the analysis. Furthermore, the relatively high rates of depression persisting at the end of the treatment period in both groups makes it difficult to draw conclusions regarding the findings.
Relevance: The authors suggest that one reason for the trend reversal following the treatment period may be a “catch up” phenomenon, in which patients in the enhanced treatment group experienced worsening medical outcomes once treatment ceased, essentially “catching up” to the comparison group. This may suggest a need for ongoing enhanced treatment in order to maintain improved cardiac outcomes.
Background: The Coronary Psychosocial Evaluation Studies (COPES) trial demonstrated promising results for enhanced depression treatment to reduce cardiovascular risk of patients with acute coronary syndrome and comorbid depression, but the long-term effectiveness of this intervention is unclear.
Methods: 157 participants with persistent depression after hospitalization for acute coronary syndromes were enrolled in the COPES trial. 80 participants were allocated to 6-months of enhanced depression treatment, and 77 were allocated to usual care. We report on an additional 12 months of observational follow-up for the composite outcome of death or first hospitalization for myocardial infarction or unstable angina.
Results: Although the intervention was previously shown to have favorable cardiovascular effects during the treatment period, we observed a significant time-by-treatment group interaction during extended follow-up (p=0.008). Specifically, during the 6-month treatment period, death or hospitalization for myocardial infarction / unstable angina occurred in 3 (4%) participants in the treatment group, compared with 11 (14%) in the usual care group (HR=0.25, 95% confidence interval [CI] 0.07-0.90, p=0.03). In contrast, during 12 months of additional observational follow-up, 11 (14%) participants in the treatment group experienced the composite outcome, as compared to 3 (4%) in the usual care group (HR=2.91, 95% CI 0.80-10.56, p=0.10)
Conclusions: Enhanced depression treatment was associated with a reduced risk of death or hospitalization for myocardial infarction / unstable angina during active treatment, but this effect did not persist after treatment ceased. Future research is needed to confirm our findings, and to determine the optimal duration of depression treatment in patients with depression after acute coronary syndromes.
ANNOTATION (Christopher Celano)
The Finding:
In patients with (generally mild to moderate) heart failure, anhedonia (as measured by a four-item subscale on the HADS) was strongly associated with poor self-care and a lower likelihood of consultative behavior (i.e., talking to physicians about symptoms), even after controlling for positive affect and depressive symptoms. While positive affect (as measured by the Global Mood Scale and PANAS) was associated with improved self-care in unadjusted analyses, these associations became non-significant when controlling for anhedonia. This suggests that anhedonia may play a unique role in affecting health behaviors in patients with heart failure and argues for further investigation into the links between this construct and health behaviors/outcomes.
Strengths and Weaknesses: The main strengths of the study were its prospective nature, its sample size, and its inclusion of several different psychological scales measuring depression (CES-D, HADS) and positive affect (PANAS, GMS). Its main limitations were that the self-care outcomes were not objectively measured and that anhedonia was measured via a subscale from the HADS, a depression and anxiety instrument.
Relevance: This is one of the first manuscripts examining the associations between positive and negative psychological states and health behaviors in heart failure. As such, it raises several questions: What kind of psychological states (positive or negative) play the biggest role in health behaviors in heart failure? Is anhedonia a separate construct from positive affect, or are they opposite ends of a spectrum? Similarly, is anhedonia a symptom of depression, or is it a separate entity altogether? Finally, to what degree are divergent findings in this study reflective of differences in scales rather than differences in psychological states? This manuscript argues for the need for further investigation into the links between psychological states and health behaviors and outcomes in heart failure, as well as an evaluation of the scales used to measure psychological states in medical populations.
Objective: Optimal self-care is crucial in patients with chronic heart failure (HF). While the focus of research has been on negative mood states, adequate psychological resources may be required to successfully engage in HF self-care. Therefore, the longitudinal associations of multiple positive affect measures in explaining HF self-care including consulting behavior were examined while adjusting for depressive symptoms and potential covariates (e.g., disease severity).
Methods: In this prospective cohort study, 238 patients (mean age: 66.9±8.6years, 78% men), with chronic HF completed questionnaires at baseline and 1-year follow-up. Positive affect was assessed with the Positive and Negative Affect Schedule (PANAS) and the Global Mood Scale (GMS). Anhedonia, i.e. diminished interest or pleasure, was assessed with a subscale of the Hospital Anxiety and Depression Scale (HADS). The 9-item European Heart Failure Self-care Behaviour scale was completed to assess HF self-care including consultation behavior.
Results: Linear mixed modeling results showed that anhedonia was most strongly associated with both poor self-care (estimate=-.72, P<.001) and consulting behavior (estimate=-.44, P<.001) over time, after adjustment for covariates and depressive symptoms. GMS positive affect was related to better HF self-care adjusting for standard depressive symptoms but not when adjusting for anhedonia. PANAS positive affect was not independently related to self-care.
Conclusion: Anhedonia was associated with worse compliance with self-care among chronic HF patients over time, irrespective of disease severity and depression. Associations between positive affect and self-care were dependent on the measures used in multivariable analyses.
