Comparison of evidence-based interventions on outcomes of hospitalized, cognitively impaired older adults

Abstract

Aim: This article reports the effects of three evidence-based interventions of varying intensity, each designed to improve outcomes of hospitalized cognitively impaired older adults.

Materials & methods: In this comparative effectiveness study, 202 older adults with cognitive impairment (assessed within 24 h of index hospitalization) were enrolled at one of three hospitals within an academic health system. Each hospital was randomly assigned one of the following interventions: Augmented Standard Care (ASC; lower dose: n = 65), Resource Nurse Care (RNC; medium dose: n = 71) or the Transitional Care Model (TCM; higher dose: n = 66). Since randomization at the patient level was not feasible due to potential contamination, generalized boosted modeling that estimated multigroup propensity score weights was used to balance baseline patient characteristics between groups. Analyses compared the three groups on time with first rehospitalization or death, the number and days of all-cause rehospitalizations per patient and functional status through 6-month postindex hospitalization.

Results: In total, 25% of the ASC group were rehospitalized or died by day 33 compared with day 58 for the RNC group versus day 83 for the TCM group. The largest differences between the three groups on time to rehospitalization or death were observed early in the Kaplan-Meier curve (at 30 days: ASC = 22% vs RNC = 19% vs TCM = 9%). The TCM group also demonstrated lower mean rehospitalization rates per patient compared with the RNC (p < 0.001) and ASC groups (p = 0.06) at 30 days. At 90-day postindex hospitalization, the TCM group continued to demonstrate lower mean rehospitalization rates per patient only when compared with the ASC group (p = 0.02). No significant group differences in functional status were observed.

Conclusion: Findings suggest that the TCM intervention, compared with interventions of lower intensity, has the potential to decrease costly resource use outcomes in the immediate postindex hospitalization period among cognitively impaired older adults.

Keywords: cognitive impairment; functional status; rehospitalizations; transitional care.

Figures

Figure 1. Consolidated Standards of Reporting Trials diagram of eligibility, enrollment and attrition from 2006 to 2008

ASC: Augmented Standard Care; RNC: Resource Nurse Care; TCM: Transitional Care Model.

Figure 2. Time to first all-cause rehospitalization or death for three intervention groups using propensity score-weighted Kaplan–Meier curves (n = 202)

ASC: Augmented Standard Care (lowest dose) intervention; RNC: Resource Nurse Care (medium dose) intervention; TCM: Transitional Care Model (highest dose) intervention group.

Figure 3. Cumulative percentage of the number of rehospitalizations over time by group

Demonstrates the cumulative percentage of all-cause rehospitalizations over time for each intervention group using raw data. These are not adjusted using propensity score weighting. Percentages were calculated by dividing the raw cumulative number of all-cause rehospitalizations by the total sample size for each interventions group. ASC: Augmented Standard Care; RNC: Resource Nurse Care; TCM: Transitional Care Model.

Figure 4. Propensity score-weighted predicted mean number of rehospitalizations per patient over time

ASC: Augmented Standard Care; RNC: Resource Nurse Care; TCM: Transitional Care Model.

Figure 5. Propensity score-weighted predicted mean number of rehospitalization days per patient over time

ASC: Augmented Standard Care; RNC: Resource Nurse Care; TCM: Transitional Care Model.

Figure 6. Propensity score-weighted mean basic and instrumental activities of daily living score over time

(A) BADL. (B) IADL. Means were adjusted for death and hospital/skilled nursing facility stay and for presentation we subtracted 2 for interpretation of BADL scores ranging 0–6 [35] and IADL scores ranging 0–8 [36]. Higher scores indicate fewer functional deficits. ASC: Augmented Standard Care; BADL: Basic activities of daily living; IADL: Instrumental activity of daily living; RNC: Resource Nurse Care; TCM: Transitional Care Model.