Institutional members access full text with Ovid®

Share this article on:

Optimizing Risk Stratification in Cardiac Rehabilitation With Inclusion of a Comorbidity Index

Zoghbi, Gilbert J. MD; Sanderson, Bonnie PhD, RN; Breland, Jenny BSN, RN; Adams, Carla BSN, RN; Schumann, Chris MS; Bittner, Vera MD, MSPH

Journal of Cardiopulmonary Rehabilitation: January/February 2004 - Volume 24 - Issue 1 - pp 8-13
Cardiac Rehabilitation: CE

PURPOSE: The risk stratification criteria of the American Association of Cardiovascular and Pulmonary Rehabilitation include guidelines to be used in stratifying cardiac rehabilitation (CR) patients for risk of disease progression (long term) and clinical events (short term). Noncardiac comorbidities are not included as indicators in these criteria. This study was designed to ascertain the prevalence of noncardiac comorbidities among CR patients, and to assess their relation to the current risk stratification algorithm for clinical events.

METHODS: Patients were stratified into high-, intermediate-, and low-risk groups according to the American Association of Cardiovascular and Pulmonary Rehabilitation risk stratification criteria for clinical events (ARSE) at program entry. Within each risk group, age, gender, race, and noncardiac comorbidities were ascertained. Comorbidities were summarized in a comorbidity index (CMI). The relation between clinical events and risk status by ARSE and CMI was evaluated by logistic regression.

RESULTS: Among 490 patients (age, 60 ± 12 years; 35% women; 30% nonwhite) enrolled in CR with ischemic heart disease, the number of comorbidities ranged from 0 to 7 (median, 2; 75th percentile, 3). The patients categorized in the three ARSE groups differed significantly in age and comorbidities. Although ARSE tended to identify patients with a greater comorbidity burden, 38% of the patients with a comorbidity index exceeding the 75th percentile were not classified in the highest ARSE group. Clinical events increased across ARSE and CMI risk strata. Both ARSE and CMI were independent predictors of events in an age-, gender-, and race-adjusted logistic regression analysis (ARSE odds ratio [OR], 1.56; 95% confidence interval [CI], 1.14–2.12; CMI OR, 1.23, 95% CI, 1.03a–1.47). Events were predicted best when both classifications were combined. Exploratory gender-specific analyses suggested that ARSE performed better among men than among women, whereas CMI was a more important predictor among women.

CONCLUSIONS: To appreciate more fully the overall complexity of disease among CR patients, ARSE should be supplemented not only with the inclusion of cardiac risk factors, as suggested in the current guidelines, but also with an assessment of noncardiac comorbidities.

Cardiac rehabilitation (CR) is a multifaceted program that focuses on risk factor modification 1–3 and patient education and aims to slow coronary artery disease progression, reduce coronary heart disease events, enhance quality of life, and reduce mortality. 4–6 Patients enrolled in CR are heterogeneous, with various degrees of illness severity, comorbid conditions, and cardiac disease. 1–3

The American Association of Cardiovascular and Pulmonary Rehabilitation Association (AACVPR) has published risk stratification guidelines to help guide treatment plans and ensure patient safety through appropriate resource allocation during the supervised exercise sessions. 1 These guidelines suggest a dual approach to risk stratification: (1) a stratification for long-term disease progression that takes into account prevalence and control of cardiac risk factors including smoking status, diet composition, dyslipidemia, diabetes mellitus, body weight, hypertension, depression, and physical activity level; and (2) a stratification for occurrence of clinical events in the near term (abbreviated as ARSE in this report) that considers extent of left ventricular dysfunction, presence of dysrhythmias, results of maximal exercise testing, symptoms of ischemia, and absence or presence of clinical depression. Noncardiac comorbidities such as those included in the D’Hoore comorbidity index (CMI) are known to affect prognosis in populations with cardiovascular disease. 7–9 Although one of the noncardiac comorbidities included in this index (diabetes mellitus) is captured in the two-component AACVPR stratification, a detailed assessment of such comorbidities is not currently recommended by the AACVPR guidelines.

The purpose of this study was to determine the prevalence of noncardiac comorbidities in a CR population and to assess their relation to ARSE. It was hypothesized that ARSE would not fully reflect the burden of disease as assessed by the authors’ CMI, and that a combined assessment of ARSE and comorbidities would better predict events during CR and thus serve as the foundation for more appropriate resource allocation in this era of limited resources.

The authors risk-stratified 490 cardiac rehabilitation (CR) patients based on the AACVPR risk stratification criteria for clinical events (ARSE), summarized their noncardiac comorbidities in a comorbidity index (CMI), and related the ARSE and CMI stratification to the occurrence of clinical events during CR. Even though ARSE and CMI stratifications correlated with each other, both were independent predictors of events during cardiac rehabilitation.

From the Division of Cardiovascular Disease (Drs Zoghbi, Sanderson, and Bittner), University Hospital (Dr Sanderson, Ms Breland, Ms Adams, and Mr Schumann), University of Alabama at Birmingham, Alabama.

This study was presented in part at the 74th Scientific Sessions of the American Heart Association, Anaheim, California, November 2001.

The author has no conflict of interest.

Address correspondence to: Gilbert Zoghbi, MD, University of Alabama at Birmingham, Division of Cardiovascular Disease, 306 Lyons-Harrison Research Building, 701 19th Street South, Birmingham, AL 35294-0007 (e-mail:

Editor’s note: Beginning in 2004, JCR will publish an article in each issue for which continuing education credits can be earned. Test questions and instructions are provided at the end of this article.

© 2004 Lippincott Williams & Wilkins, Inc.