A transitional care model for low-income older adults does not reduce readmission rates or emergency department visits during care transitions
- Correspondence to: Dr Pamela Nadash, Department of Gerontology, University of Massachusetts Boston, 100 Morrissey Blvd, Boston, MA 02125, USA;
Implications for practice and research
A relatively simple, low-cost care transition intervention can encourage low-income, predominantly African-American elders to better utilise primary care, posthospital discharge.
An adequately powered study is needed in order to enable stronger conclusions to be drawn, particularly regarding the model's impact on hospital re-admission and emergency department (ED) visits.
Studies benefiting from high levels of minority participation should exploit this intervention through study design.
Improving management of care transitions is an important goal in health systems change, and various attempts have been made to minimise the negative outcomes and avoidable costs resulting from poorly managed transitions, particularly in vulnerable, high-cost populations. The efforts range from wholesale systems change to smaller, less resource-intensive interventions, such as that tested in this study. This study also addresses the relevance of such interventions among an older, largely low-income sample of African-Americans with chronic conditions, a group that is typically under-represented in this area of research.
A treatment group of 69 individuals was compared with 52 controls, identified from hospital records, on outcomes of hospital re-admissions, ED admissions and primary care utilisation. While the control group received standard care, the treatment group members received a modified Care Transitions Intervention,1 which incorporates tools to promote cross-site communication, patient self-management training and a ‘transition coach’, who provided guidance and postdischarge follow-up, often by telephone. All participants were aged 60 and over and admitted to an urban safety net hospital in Atlanta (USA), with one or more of the following comorbidities: hypertension, stroke, heart condition, diabetes mellitus, dementia (with a proxy) or taking four or more medications. The treatment and control groups were matched by age, sex, comorbidities and age-adjusted Charlson Comorbidity Index.
The sample was 90% African-American, with average incomes of less than US$20 000/annum and an average age of 70. Nearly 90% of the participants had diabetes, and nearly 75% had hypertension. No significant differences in hospital re-admissions or ED use were found. However, primary care usage was significantly higher among the treatment group participants at the 30, 90 and 180 days of follow-up points, but only marginally so after a year, even though the postdischarge intervention lasted only 30 days. A preintervention and postintervention analysis of the treatment group also yielded null findings on all outcomes.
Improving care for low-income, vulnerable older people is important, due to their high healthcare costs. If served inadequately, they may consume considerable resources and experience poor outcomes as they bounce from setting to setting, due to poorly managed postacute transitions. In addition, the support needs of low-income African-Americans might also differ in some important ways from the needs of others.
Unfortunately, the study's small sample size and overall design limit its usefulness. An encouraging finding was the increase in primary care visits among the treatment group participants. Such visits have been associated with positive health outcomes.2–4 However, it is unclear whether the intervention's null impact on other outcomes results from its small sample size or from a lack of effect. Similarly, although the study benefits from a largely African-American sample, its design prevents us from generalising or learning from that sample. The effectiveness of telephone-based follow-up is another encouraging element of the study.
Information about the effectiveness of smaller, less resource-intensive efforts to improve transitions is critical; other options such as Accountable Care Organisations, integrated care programmes or postacute bundling experiments,5 require more extensive systems change. Thus, further research would be welcome.