Technology Assessment: Non-Pharmacological Interventions for Post-Disc

Technology assessment on the effectiveness of non-pharmacological interventions for post-discharge care in heart failure patients.

Table of Contents


Living with HF is a challenge to the sufferers of this condition. It also represents a significant burden for the caregivers. The effective management of HF is achieved through optimal medical therapy. In addition, current emphasis has been geared toward a comprehensive and proactive support for self-management utilizing delivery models before, during, and after hospitalizations. These self-management models use multiple intervention components of comprehensive patient education, self care behaviors, medication review, proactive nurse followup, and increased access to providers.

This report identified 49 randomized controlled trials that evaluated interventions utilizing various combinations of individual components initiated in three different settings. We included studies published since 1990 — the time point associated with rapid advances in the medical management of HF, and changing health needs of the patients and changes in the practice patterns. The majority of the included studies reported readmissions as the primary outcome of interest. A few studies also evaluated combined endpoint of readmissions or death as the primary outcome of interest. In general, studies were not adequately powered to evaluate the clinical outcome of mortality. We performed meta-analysis and subgroup analyses to address the key question of the effectiveness of interventions of post-discharge support to prevent readmissions. Our report used lax inclusion criteria, thereby reviewing a large body of literature. We identified those interventions that utilized increased clinic visits, home visits, and multidisciplinary care to reduce the risk of readmissions in the intervention group compared with the usual care group. Often these interventions utilized a combination of secondary components, with telephone followup being the most common across the studies; however, when telephone followup was utilized alone, it did not demonstrate a significant difference in all cause readmissions compared with usual care.

In our subgroup analyses, the characteristics of intermediate- to long-term followup (>6 month and >12 month), interventions initiated in the inpatient setting, and patient age greater than 75 years had a statistically significant impact on the all cause readmissions in the intervention group. Inpatient setting benefits may be explained due to better transition care from the inpatient to home care in the intervention group. Generally studies did not report adequate information about the transition of care coordination in the usual care group. Improvements with longer followup suggest increased compliance and motivation.

Compared to usual care, one-quarter of the 20 studies reported a significant reduction in all-cause mortality when interventions were initiated during an index hospitalization. Similar significantly decreased length of stay during readmissions was reported when interventions were initiated during an index hospitalization. Individual studies were not adequately powered to ascertain meaningful differences between the interventions and usual care groups for mortality outcomes; however, there was no distinct combination of intervention components and improved outcomes of QOL changes, and miscellaneous clinical outcomes (costs, and composite endpoint of mortality or readmission).

Eighteen randomized controlled trials compared self-management delivery models beginning after discharge from an index hospitalization. Only one of 18 studies suggested decreased event rates in mortality, readmissions, and reduction in length of hospital stay during readmissions with interventions. Also, the evidence was sparse for interventions beginning in the outpatient clinics.

The studies were heterogeneous with regard to intervention components, intervention duration, followup duration, and components of usual care. In addition, some studies utilized components of usual care along with their intervention components. The studies were performed in a range of settings, in patients with a wide range of HF severity, across various countries, over a long time period during which the standards of HF care have changed considerably, and in patients with different underlying risks of clinical events. Studies often utilized several different combinations of intervention components, resulting in considerable heterogeneity that was difficult to dissect and ascertain the effects of individual components.

The consistency of results in the settings of diverse studies adds credibility to the conclusion that patient outcomes can be improved with optimized transition of care coordination between the inpatient settings to home care, and a combination of interventions that increase access to providers. However, there was no easily discernable pattern based on a particular intervention component, time period (year), intervention duration, and specific coexisting medical illnesses. Across studies, similar rates of coexisting medical illnesses, and proportions of patients who received ACE inhibitors, beta-blockers, and/or diuretics were reported at baseline among patients in the intervention and control group. Of note, no studies evaluated or provided data on adverse effects due to the interventions.

Additional limitations of the studies include small sample size and short followup durations, and a substantial number of the studies reported a composite endpoint of mortality or readmission as their primary outcome. It is well-acknowledged that the use of a composite endpoint can erroneously attribute reductions in mortality to interventions that do not actually reduce deaths.72 The relative effectiveness of the individual components of interventions remains unknown, since none of the studies compared one intervention component with each other. This lack of clarity on the necessary combination of components of a HF management program may be answered by the future publications. Future research with long-term followup is needed to determine which individual components, if any, in what settings and circumstances, may benefit. A determination should be made as to which patients are most likely to benefit from which combinations of intervention components, and then studies focused on these patients should be conducted. Any future studies should continue to focus primarily on clinical benefits, as short-term intermediate outcomes or composite endpoints are inadequate surrogates for clinically important outcomes.


Current as of February 2009
Internet Citation: Conclusion: Technology Assessment: Non-Pharmacological Interventions for Post-Disc. February 2009. Agency for Healthcare Research and Quality, Rockville, MD.