Background and Aims: CR program participation has proven mortality and morbidity benefits in women and men; however, less than 20% eligible patients participate in CR and women, in particular, are reported to be underrepresented in CR programs compared to men. Retrospective surveys have identified a number of factors that may relate to increased CR program participation, such as a strong attending physician endorsement of CR. It is unknown whether prospective manipulation of these factors would result in increased CR participation. We conducted a randomized controlled trial to determine whether more patients admitted for myocardial infarction (MI), unstable angina (UA), coronary angioplasty (PTCA), or coronary artery bypass surgery (CABS), who receive a predischarge personalized written attending cardiologist's (or cardiac surgeon's) endorsement (MD endorsement) encouraging participation in a large tertiary care institution's CR program, in addition to the standard CR referral would actually enroll in the CR program compared to those who received the standard CR referral alone. Patients were given their CR program intake appointment dates prior to hospital discharge.

Methods: Consecutive stable post-MI, post-UA, post-PTCA, or post- CABS in-patients were screened 2 days before anticipated discharge from May 2003 through October 2006. Patients were randomly assigned (stratified by gender and attending MD) to the intervention of a standardized & personalized written attending MD Endorsement accompanying the CR referral versus the CR referral alone (usual care). The primary outcome was CR enrollment as defined by attendance at the CR program within 4 months of index hospital discharge.

Results: Overall, 548 subjects were enrolled and the groups were balanced in terms of gender at baseline with: 87/275 (31.6%) women allocated to the Endorsed group and 85/273 (31.1%) women allocated to the usual care group. There was no difference in the overall (and within gender) proportion of patients who attended the CR program in the Endorsed group compared to standard CR referral group (overall: 58.2% vs 58.2%, P = NS; women: 44.8% vs 49.4%, P = NS; men: 64.4 vs 62.2% P = NS). In addition, there was no significant interaction between the intervention and gender or wait-time to CR intake. However, ignoring intervention, fewer women than men attended the CR program compared to men (47.1% vs 63.3%, P < .0001).

Conclusions: We demonstrated for the first time in a prospective manner that a personalized written MD endorsement encouraging CR attendance in addition to a systematic but standard CR referral has no impact on improving CR program participation compared to a systematic but standard CR referral alone in women or men. In addition, despite a systematic CR referral process inclusive of provision of written documentation of scheduled CR program intake appointment date before hospital discharge, less than two thirds of patients referred actually attended the CR program and substantially fewer women than men actually attended the CR program. The reasons for these discrepancies require evaluation.