We identified 2703 records resulting in the study of 121 relevant articles. Overall, 74 studies (Supplementary Appendix, online only) with a total of 2876 IC patients met inclusion criteria (Fig). Patients represented 82,725 patient-hours of SET.
Most of the SET studies were performed in Western Europe and North
America. Mean patient age ranged from 54 to 76 years, and more men than women participated (Table Frequently reported patient exclusion criteria were IC presenting with pain at rest, critical ischemia, exercise capacity limited by symp- toms of angina, congestive heart failure, chronic obstructive pulmonary disease, arthritis, and poorly controlled hypertension. Exercise programs consisted of treadmill training (n = 49), mixed exercise therapy (n = 8), circuit training (n = 11), strength training (n =
arm ergometry (n = 2), or pole-striding (n = 2). Duration of the exercise programs ranged from 2 weeks to 12 months. Training frequencies varied from one to seven times a week, and 30 to 120 minutes per day (Table II).
Main outcome measurement.
Potential adverse events were explicitly reported in 35 of the 74 studies. Another 32 studies reported no dropouts or dropouts without medical cause. The remaining seven studies were included after positive e-mail conformation. Eight adverse described during >80,000 patient-hours of SET, resulting in an all-cause complication rate of one event in 10,340 patient-hours. Therefore, the risk of complications during SET is considered exceedingly low. Focusing on the two major cardiac complications, an even lower event rate of one event in 41,362 patient-hours was calculated. These results are in line with a prospective registry on complications during exercise for cardiac rehabilitation reporting an almost similar rate of one event per 49,565 patient-hours.26 Moreover, these rates may even be considered relatively high because events that were reported in studies from >30 years ago were also included. Medical treatment has greatly improved the overall prognosis, potentially resulting in even lower current event rates. CPET has been used for >60 years for risk stratification in patients with heart diseases,27 but this technique is no longer recommended as a screening tool for myocardial ischemia in people without known CAD, as suggested by a recent guideline of The National Institute for Health and Clinical Excellence.28 Prediction of a future severe cardiac event during exercise is very difficult, if not impossible.26 In the present review, 28 studies (32% of total patient-hours) included cardiac screening, but because an equal number of adverse events (n ¼ 4) was found in studies with or without cardiac screening, the results may indicate that cardiac screening in IC patients is indeed not useful. However, by the observational and retrospective design of our study, firm conclusions cannot be drawn. Yet, given the very low complication rate, cardiac screening as a tool for a safe execution of SET seems clinically irrelevant.
There is also a downside using CPET as a screening technique. As all screening tests do, CPET also inherently carries a risk of false-positive testing29; for instance, a 15%events of cardiac (n = 6) and noncardiac (n = 2) origin were reported (Table III). One fatal adverse event was reported. The total all-cause event rate was one event per 10,340 patient-hours. Total cardiac and noncardiac event rates were one per 13,788 and one per 41,363 patient- hours, respectively
(Table IV). Sub analysis on cardiac screening. Of 28 studies (26,790 patient-hours of SET [32%]) that included cardiac screening before SET was initiated, 23 used electro- cardiography registration. Four of the eight adverse events were reported in these 28 studies. The remaining four, including a fatal myocardial infarction, were found in the 46 remaining studies that did not included cardiac screening (55,935 patient-hours of SET [68%]). Six studies reported results of 1703 CPETs before SET inclusion. Abnormal CPET findings resulted in 59 patients (3.5%) being excluded from SET participation.