Randomized trials comparing coronary bypass surgery with medical therapy have shown little difference in subsequent employment [5, 6]. However, convalescence after bypass surgery may be lengthy; this may limit return to work even among successfully treated patients. Coronary angioplasty offers the possibility of substantial relief of angina with a short convalescence, thereby speeding return to normal activities, including employment [7]. We analyzed patterns of employment among patients in the Bypass Angioplasty Revascularization Investigation (BARI), a randomized clinical trial of coronary angioplasty or coronary bypass surgery [8-10].

Methods

Top Methods Results Discussion Author & Article Info References

The Study of Economics and Quality of Life (SEQOL) was conducted in 7 of the 18 BARI clinical centers [11, 12]. Eligible patients had myocardial ischemia severe enough to warrant coronary revascularization and multivessel coronary disease suitable for balloon angioplasty or for bypass surgery.

At study entry, patients enrolled in SEQOL gave detailed information on their employment, including hours worked per week, job type, physical demands, and mental stress [11]. Employment status, including type of work performed, hours worked per week, and time missed from work, was updated every 3 months during follow-up. Patients who stopped work during follow-up gave information on the date that work ended and the main reason for stopping work.

The main outcome measure was total time worked during follow-up, defined as the percentage of full-time equivalent employment of 40 hours a week for 13 weeks per quarter. The time worked in each quarter throughout follow-up was summed over 4 years to provide a summary measure of employment ranging from 0 to 4 full-time equivalent years. Time worked by patients who died during follow-up was included in this analysis. Between-group comparisons were done by using the Wilcoxon rank-sum test and multivariable analysis was done with backward, stepwise linear regression. Confidence intervals were calculated by using the percentile bootstrap method. Analyses were performed by using SAS version 6.12 (SAS, Inc., Cary, North Carolina) and S-PLUS version 3.4, release 1 (MathSoft, Inc., Seattle, Washington).

Results

Top Methods Results Discussion Author & Article Info References

A total of 934 patients were enrolled in SEQOL; of these, 409 were employed at study entry. One hundred ninety-two employed patients were randomly assigned to undergo coronary angioplasty and 217 were assigned to undergo coronary bypass surgery. The baseline clinical and job-related characteristics of the employed angioplasty and surgery patients were well balanced (data not shown).

Over 4 years of follow-up, 157 (82%) patients in the angioplasty group and 177 (82%) patients in the surgery group resumed work. Angioplasty patients who resumed work did so at a median of 4.9 weeks (interquartile range, 2.7 to 10.9 weeks) after randomization compared with 10.9 weeks (interquartile range, 7.7 to 14.4 weeks) for surgery patients (difference, 6 weeks [95% CI, 3.4 to 7.0 weeks]).

After 4 years of follow-up, angioplasty patients had worked an average of 2.23 years compared with 2.13 years for surgery patients (difference, 0.11 years [CI, –0.24 to 0.45 years]). The average time spent working decreased progressively throughout followup in both groups, principally as a result of stopping work entirely rather than reducing the number of hours worked per week. Similar numbers of angioplasty and surgery patients stopped work because of heart disease (49 and 47 patients, respectively), other health conditions (14 and 11 patients), retirement not induced by health problems (23 and 37 patients), or other reasons (28 and 32 patients).

Because the long-term employment of patients who underwent angioplasty did not differ significantly from that of patients who underwent surgery, these two groups were combined to analyze other factors associated with employment. Results of univariable analyses are shown in Table 1, and results of multivariable analysis are shown in Table 2. The strongest factor affecting long-term employment was patient age at study entry. The relation between age and time worked was U-shaped, with patients 60 to 64 years of age at study entry working less during follow-up than both younger and older patients (Table 2). Patients who initially worked full-time worked more during follow-up, as did patients who planned to return to the same job. Patients who had a single, private source of medical insurance at baseline were more likely to continue working (Table 2).

Medical factors were less predictive of long-term employment patterns than were demographic factors and job characteristics. Of the medical characteristics that were significant in the univariable analysis (Table 1), only left ventricular function was significant in the multivariable analysis (Table 2).

Discussion

Top Methods Results Discussion Author & Article Info References

Among patients in a randomized trial, those treated with angioplasty returned to work significantly faster than patients treated with bypass surgery, but this did not translate into greater long-term employment. Long-term employment patterns were associated more with patient age and job characteristics than with measures of illness severity or the method of coronary revascularization.

Successful coronary angioplasty avoids the need for general anesthesia, cardiopulmonary bypass, and a large chest incision, allowing patients to return to normal activity sooner than patients undergoing coronary bypass surgery. Timing of return to work after medical illness and surgical procedures is governed by the intrinsic rate of physical recovery, physician advice, employer policies, and social convention. We did not assess the latter three factors in this study and could not evaluate their relative importance to resumption of employment. Our results suggest that the speed with which a patient can resume normal activities, including employment, is sensitive to the results of medical interventions. Therapies, such as angioplasty, that allow a more rapid convalescence have medical, social, and economic benefits.

Surgery patients in BARI have been shown to have fewer anginal symptoms, better objective measures of exercise tolerance, and reduced need for repeated revascularization procedures [9, 10], but these results did not improve long-term employment. Other randomized trials of angioplasty compared with bypass surgery [13-15] or with medical therapy [15-17] have also shown little effect of clinically successful coronary revascularization on employment. These results are consistent with observations in less selected populations [2, 3, 18] and indicate that nonmedical factors are the most important determinants of long-term employment. The seriousness of coronary revascularization may prompt many patients to reconsider their goals in life, leading them to reduce work and redirect themselves toward their families or other priorities, even after a successful clinical result.

Older age was strongly correlated with a lower rate of long-term employment, with one intriguing exception. Patients 65 years of age or older who were working at study entry worked significantly more than patients 60 to 64 years of age (Table 2). One potential explanation is that patients 65 years of age and older who were still working had already passed up one opportunity to retire and thus would be more likely to continue working even after coronary revascularization, whereas patients aged 60 to 64 at study entry had not yet faced a retirement decision. This interpretation tends to be corroborated by the work experience of the entire cohort of 934 patients, which included those who had retired before randomization and those who were still working at study entry. The amount of time worked in the entire cohort decreased progressively with age, from 2.19 years for patients younger than 55 years of age to 1.54 years for patients 55 to 59 years, 0.81 years for patients 60 to 64 years of age, and 0.24 years for patients 65 years of age or older. Patients who worked past the traditional age of retirement seem to be self-selected for continued employment, even in the face of major surgery.

Another interesting observation was that patients who had a single source of private health insurance, probably obtained through their job, tended to work an average of 0.53 years more during follow-up (Table 2). Patients with preexisting conditions, such as heart disease, may be reluctant to leave their jobs unless they can obtain health insurance through another source. Unfortunately, before patients are eligible for Medicare at 65 years of age, few sources of private health insurance are available for those who lose their employment-based coverage. Our findings suggest that ability to maintain health insurance coverage may be an important consideration for patients with heart disease who are deciding whether to retire.

Several caveats should be considered in the interpretation of our findings. First, patients were selected for BARI on the basis of stringent inclusion and exclusion criteria; therefore, they may not be representative of all employed patients with coronary artery disease. Our findings comparing coronary angioplasty and coronary bypass surgery, however, are based on a randomized comparison and therefore have high validity. In addition, these results are consistent with those of earlier nonrandomized comparisons performed in more representative populations of patients with coronary disease [7]. In addition, we did not examine whether angioplasty or surgery might offer specific advantages for patients with particular types of jobs.

In conclusion, coronary angioplasty allows faster recovery and resumption of employment than does bypass surgery. Long-term employment patterns in patients with coronary artery disease are more strongly associated with demographic and job characteristics than with severity of illness or type of treatment.

Dr. Brooks: University of Pittsburgh, Graduate School of Public Health, 127 Parran Hall/130 DeSoto Street, Pittsburgh, PA 15261.

Dr. Rogers: University of Alabama, 334 Lyons-Harrison Building, 701 South 19th Street, Birmingham, AL 35294-0007.

Dr. Pitt: University of Michigan, 3910 Taubman Building, 1500 East Medical Center Drive, Ann Arbor, MI 48109.

Drs. Reeder and Smith: Mayo Clinic, 200 First Street, SW-West 16A, Rochester, MN 55905.

Dr. Ryan: Boston University Medical Center, Division of Cardiology, 88 East Newton Street, Boston, MA 02118.

Dr. Whitlow: The Cleveland Clinic Foundation, Department of Cardiology, F25, 9500 Euclid Avenue, Cleveland, OH 44195.

Dr. Wiens: St. Louis University Medical Center, BARI Laboratory 14FDH, 3635 Vista Avenue at Grand Boulevard, Box 15250, St. Louis, MO 63110-0250.

Dr. Mark: Division of Cardiology, Duke University Medical Center, Box 3485, Durham, NC 27710.