Effect of a community-based nursing intervention on mortality in chronically ill older adults: a randomized controlled trial

Abstract

Background: Improving the health of chronically ill older adults is a major challenge facing modern health care systems. A community-based nursing intervention developed by Health Quality Partners (HQP) was one of 15 different models of care coordination tested in randomized controlled trials within the Medicare Coordinated Care Demonstration (MCCD), a national US study. Evaluation of the HQP program began in 2002. The study reported here was designed to evaluate the survival impact of the HQP program versus usual care up to five years post-enrollment.

Methods and findings: HQP enrolled 1,736 adults aged 65 and over, with one or more eligible chronic conditions (coronary artery disease, heart failure, diabetes, asthma, hypertension, or hyperlipidemia) during the first six years of the study. The intervention group (n = 873) was offered a comprehensive, integrated, and tightly managed system of care coordination, disease management, and preventive services provided by community-based nurse care managers working collaboratively with primary care providers. The control group (n = 863) received usual care. Overall, a 25% lower relative risk of death (hazard ratio [HR] 0.75 [95% CI 0.57-1.00], p = 0.047) was observed among intervention participants with 86 (9.9%) deaths in the intervention group and 111 (12.9%) deaths in the control group during a mean follow-up of 4.2 years. When covariates for sex, age group, primary diagnosis, perceived health, number of medications taken, hospital stays in the past 6 months, and tobacco use were included, the adjusted HR was 0.73 (95% CI 0.55-0.98, p = 0.033). Subgroup analyses did not demonstrate statistically significant interaction effects for any subgroup. No suspected program-related adverse events were identified.

Conclusions: The HQP model of community-based nurse care management appeared to reduce all-cause mortality in chronically ill older adults. Limitations of the study are that few low-income and non-white individuals were enrolled and implementation was in a single geographic region of the US. Additional research to confirm these findings and determine the model's scalability and generalizability is warranted.

Trial registration: ClinicalTrials.gov NCT01071967. Please see later in the article for the Editors' Summary.

Figure 1. CONSORT flow diagram.

Figure 2. Subgroup analyses.

Deaths and tests for interaction by subgroup. HRs and adjusted HRs (aHR) along with 95% CIs are represented by forest plots with x-axis in log 2 scale. The regression model used for the aHR includes covariates for sex, age group, primary diagnosis, perceived health rating, number of medications taken, hospital stays in the past 6 mo, and tobacco use. CHD, coronary heart disease.

Figure 3. Kaplan-Meier estimate of cumulative mortality up to 5 y from enrollment.

The plot includes results for all participants randomized into the study, (unadjusted data), with p-value calculated using the log-rank test.