Cardiac rehabilitation may not provide a quality of life benefit in coronary artery disease patients

Patients attending Phase II CR were significantly younger compared to patients who did not attend CR and thus subsequent analyses were age-adjusted. In relation to socioeconomic status, 24% (n = 37) of CR patients were identified as belonging to low, which was comparable to 16% (n = 11) in the CR non-participants (Table 1). The CR patients were more likely to have undergone angiography following an admission with an acute coronary syndrome (ACS) (Table 2), with 41% (n = 62) having experienced a recent myocardial infarction compared with 17% (n = 12) who did not have CR (p < 0.001). Despite this, the extent of angiographic disease was similar between groups as was the initial therapy although there was a slight trend towards medical therapy for those not participating in CR (Table 2).

Scores on the HRQoL indices indicated the CR patients had a better quality of life in terms of both general health status and angina-specific morbidity at baseline (Table 2). Scores on the physical limitation domain of the SAQ were significantly higher in the CR patients compared to CR non-participants. The angina frequency domain was also higher (indicating less angina) in the CR patients, however this was not statistically significant. Consistent with this, scores on the SF-36 PCS tended to be higher in the CR patients. In addition, elevated psychosocial distress as identified by the SF-36 MCS, was lower in the CR patients (Table 2).

Predictors of CR participation were assessed in the overall group of patients presenting for angiography who were potentially eligible for referral. Following univariate analysis, backward elimination retained the following predictors: age, recent myocardial infarction and a prior history of musculoskeletal disease. In the final model, only recent myocardial infarction (OR = 2.8, 95% CI 1.5-5.3, p = 0.001) remained as a significant predictor of CR participation.

At 6 months following angiography, at which time the CR patients had completed their rehabilitation programme, there were no differences between groups in relation to cardiac events (Table 3). Furthermore there were no differences between groups in changes in medical therapy or angina limitations as assessed via the CSCC system (Table 3).

Figures 2 and 3 summarise the HRQoL indices over the 6-month period for the 2 study groups. The HRQoL indices were measured at 1 and 6 months although at the 1-month assessment time point, only 70% (n = 46) of the CR participants had commenced the program and none had completed the 6-week program. By 6 months, all the CR participants had completed the CR program. The SAQ-derived physical limitation, angina frequency, and quality of life scales all showed significant improvement over the 6-month study period. Both study groups also demonstrated statistically significant improvements in the SF-36 PCS and MCS over the 6-month period. However, no differences were observed in the rate of improvement from baseline over 6 months between the two groups in SAQ and SF-36 scores (Figures 2 and 3). Furthermore, cross-sectional analysis on the 1 and 6 month endpoints revealed no significant differences for both the SAQ and SF-36 scales between the two groups.

As discussed above and reported in previous studies [30], the patients who did not participate in CR were the most disabled and thus may well have been those that would have derived the most benefit. Five patient-related factors may have influenced the findings in this study and should be closely considered in further studies. Firstly, CR efficacy in relation to impact on HRQoL has been shown to be age-dependent with previous studies reporting that younger (i.e. <41 years) and older (>65 years) patients benefitting the most from CR with no benefits observed in those 41–65 years [21, 29]. In this study, the average age of the CR participants was 60 years, which may have influenced the findings. Secondly, almost half of the CR patients were identified as experiencing elevated psychosocial distress and this has a major impact on HRQoL. Thus if the CR program also focused on the treatment of psychosocial distress, an incremental improvement in HRQoL may have also been achieved. Thirdly, the CR study population predominantly underwent this therapy following a recent myocardial infarct. Muller-Nordhorn et al [21] showed that only patients undergoing CR following revascularisation procedures showed improvement in their SF-36 scores whereas recent myocardial infarct patients did not. Fourthly, patient participation in CR was 31%, which is comparable to rates reported by other studies, which have shown 14-43% of potential cardiac patients participate in CR programs [31‐33]. Although all the study patients were eligible for CR, why the non-participants did not avail themselves to this opportunity is unclear. The focus by the health workers on the acute presentation [34, 35] may have under-emphasized the importance in chronic conditions and thus education of both health care workers and patients is required. Logistic factors may have also influenced patient participation, such as transport availability/cost and session times. Finally, although all the CR patients completed the program, the commitment they made to the therapy was not assessed.

CR programs are heterogeneous in content although professional societies are developing guidelines with core components to standardize the therapy [36, 37]. Hence the findings from this study may not be applicable to other CR programs. In particular, tailored programs for different CAD presentations or populations may be useful to maximize benefits for patients, as suggested by Piepoli et al [38]. For example, tailored gender-specific programs have been shown to provide increased quality of life improvement for women [39]. Other aspects that require further considerations include CR program delivery and duration. The hospital-based CR program in this study may be better delivered as a telephone or on-line program, which may assist in patient participation and provide long-term access to all. This is important since recent studies have shown that a continued intervention following CR is effective in improving cardiovascular outcomes [40, 41]. Also the duration of the CR program in this study was only 6 weeks whereas Piepoli [38] highlights that such short-term approaches are unlikely to yield long-term benefits or impact on quality of life.

The current study was a longitudinal case–control design and the limitations of this design may have influenced the findings. Firstly, the non-randomised allocation to CR therapy may have been influenced by a selection bias although the two groups were similar in their baseline HRQoL indices. Secondly, the study period was limited to 6 months and longer follow-up periods may have influenced the results. Of note, a meta-analysis of CR therapies showed improvement in all-cause mortality only at 3-year follow-up and not earlier [42]. Thirdly, important information such as referral and drop out rates, reasons for non-inclusion in CR, such as ambulation, were not available in this study. The lack of standardized data collection and thus insufficient program information causes difficulty when assessing biases [43]. Lastly, although the study was powered to detect a difference of 6 points on the SF-36 PCS, the total sample size was relatively small; larger studies are required and should investigate smaller but clinically relevant differences in HRQoL. It should also be noted that HRQoL assessment is best undertaken by patient interview where possible as opposed to mail-out surveys to increase response rate.