Secondary Prevention After Acute Myocardial Infarction

Richard F. Heller, MD, June C. Knapp, SRN, Lisa A. Valenti, and Annette J. Dobson, PhD

The hypothesis that 6 months after acute myocar- dial infarction, adoption of secondary prevention activities would be higher, quality of life better, and blood cholesterol lower in patients randomly allocated to a mail-out intervention program than in those receiving usual care was tested. Patients were aged ~70 years, admitted to hospitals in and around Newcastle, Australia with a suspected heart attack and discharged alive from the hospi- tal. Cluster randomization, based on the patient’s family practitioner, was used to allocate consent- ing patients to an intervention or usual care group. A lowcost mail-out program was designed to help patients reduce dietary fat, obtain regular exercise by walking and (for smokers only) to quit smoking. Supplementary telephone contact was also used. In addition, a letter was sent to the family doctor regarding the benefit of aspirin and p blockers for secondary prevention. Cf eligible patients, 71% participated, and 79% of the 213 iu tervention subjects and 87% of the 237 usual care ones returned a (imonth followup questionnaire. Selfveported fat intake was significantly lower, an “emotional” score obtained from a quality-of life questionnaire was significantly higher in the intervention than in the usual care group, and “physical” and “social” scores for quality of life were slightly higher. Blood cholesterol level and other variables were not different between the groups at 6 months. Simple low-cost programs providing support and advice on liistyle change may be beneficial, particularly in improving pz+ tients’ perceived quality of life.

(Am J Cardiol1993;72:759762)

From the Centre for Clinical Epidemiology and Biostatistics, Faculty of Medicine, The University of Newcastle, Newcastle, New South Wales, Australia. This work was supported by a grant from the National Health and Medical Research Council of Australia. Manuscript received December 3 1, 1992; revised manuscript received and accepted June 1, 1993.

Address for reprints: Richard F. Heller, MD, Centre for Clinical Epidemiology and Biostatistics, David Maddison Clinical Sciences Building, Royal Newcastle Hospital, Newcastle, New South Wales, 2300, Australia.

A t least 45% of subjects in a community-based register of acute myocardial infarction (AMI) have had either previous AMI or other mani-

festations of coronary artery disease.’ The potential for reducing the community burden of AMT through secon- dary prevention is thus considerable. There is evidence that secondary prevention by lowering cholesterol can reduce subsequent rates of both fatal and nonfatal AML2 and that exercise programs after AMI also result in pro- tection against a subsequent event.3,4 However, the efforts directed toward secondary prevention are often minimal, poorly funded in comparison with the care of the acute episode itself and inadequately evaluated. This study examined the role of a simple secondary preven- tion advice package that would be easy to put into prac- tice. The hypothesis tested was that adoption of sec- ondary prevention activities would be higher, quality of life better and blood cholesterol levels lower 6 months after discharge from the hospital following an AMI in patients randomly allocated to a mail-out intervention program than in those receiving usual care.

METHODS Between September 18, 1990 and December 5, 1991,

all subjects aged ~70 years with a suspected heart attack who were registered by the Newcastle collaborating cen- ter of the World Health Organization MONICA Project5 and discharged alive from the hospital (but before their diagnosis, according to the MONICA criteria, could be determined) were considered for inclusion in the study. The study was restricted to patients from the 5 major hospitals near the center of the study area. Patients with renal failure or other special dietary requirements and those considered by their physicians to have “end- stage” heart disease were excluded from the study. Writ- ten consent was obtained in the hospital by nurses from the MONICA study who obtained demograpmc and other data including exercise habits. A previously vali- dated questionnaire (unpublished PhD thesis) was used to assess dietary fat intake, and the results were sum- marized to produce a “fat score.”

Patients were allocated to an intervention or usual care group according to the name of their usual gener- al practitioner. All general practices within the study area were stratified according to the number of doctors within the practice group and randomly allocated to intervention or usual care within those strata. This was to avoid contamination between the 2 sides of the study; although general practitioners of the intervention pa- tients were contacted (see later), they were not asked to be active participants in the study.

The intervention began within 1 week of hospital dis- charge when possible and was independent of any re- habilitation services provided as part of routine clinical care. This comprised a letter written to the subject’s gen-

SECONDARY PREVENTION AFTER ACUTE MYOCARDIAL INFARCTION 759

TABLE I Comparison of Participants and Three Main Groups of Nonparticipants

Nonparticipants

Left Refused Refused Hospital

Participants Any Study This Trial Early

Number of patients 450 56 57 58 Men (%) 71 71 77 71 Age groups (%) (years)

25-49 50-59 60-69

Previous myocardial infarction (%)

Currently married (%I* Currently employed

(%)* Current smoker !%)* Current alcohol drinker

(%)*

14 14 12 28 31 27 30 10 55 59 58 62 25 25 23 28

78 27

31 53

71 75 66 31 29 27

37 44 27 57 56 401

*At time of hospital admission for acute event associated with possible recruitment to this trial.

tData on alcohol consumption missing for 3 1% of subjects who left hospital early.

eral practitioner regarding the benefits of aspirin and B- blocking drugs in secondary prevention, and the first of 3 mail-out packages to the subject. The iirst package comprised an introduction and Step 1 (added fat) of the “Facts on Fat” kit (adapted from material developed by the Nutrition Project, Health Promotion Services Branch of the Health Department of Western Australia),6 togeth- er with a walking program and information on the “Quit for Life” program for subjects who smoked on admis- sion to the hospital. Further encouragement to walk was provided in the form of a magnetic reminder sticker. The second package contained Step 2 (takeaways, snack foods and extras) and Step 3 (meat fats) of the “Facts on Fat” kit. In addition, each subject was asked to re- spond to questions regarding whether, and if so, how many times, they had been for a walk or exercised in some way in the past week. Step 4 (dairy food) and Step 5 (“putting it all together”) constituted the third pack- age, together with information regarding local “Walk- ing for Pleasure” groups.

Each step of the “Facts on Fat” kit related to a spe- ciIic source of fat in the diet and began with a quiz to assess the subjects’ knowledge of the sources of fat in their diet. Each subject was given a specific target for fat reduction (based on the fat intake questionnaire com- pleted while in the hospital) and was asked to return a contract agreeing to the target. Further changes identi- fied from the answers to the quiz were dealt with in the letters of encouragement and explanation accompanying each package, and 3 further “contracts” to change aspects of the diet were offered during the intervention period.

Contact with subjects was maintained over the next 4 months with monthly newsletters containing recipes, information on low fat products, and further reinforce- ment of the benefits gained from exercise. A National Heart Foundation booklet, “Planning Fat Controlled Meals,” was included. Two supplementary telephone

TABLE II Comparison of Intervention and Usual Care Groups at Entry to Trial

Intervention Usual Care

Number of participants 213 237 Men (%) 76 68 Previous myocardial infarction (%I* 25 25 Completed high school or higher 26 32

education (%I* Currently married (%)* 77 79 Currently employed full time (%)* 26 28 Ever told to have*:

High blood pressure (%) 53 53 Angina (%) 38 41 High cholesterol (%) 50 55 Diabetes (%) 12 13 Obesity/overweight (%) 36 36

Previous coronary angiography (%I* 28 30 Previous coronary bypass graft (%I* 15 11 Current smoker (%)* 31 31 Current alcohol drinker (%)* 54 51 Never or rarely engaged in leisure- 48 48

time exercise (%)t Mean age (SD)* 59 (8) 58 (8) Mean height (SD)* 171 (9) 170 (9) Mean weight (SD)* 78 (14) 76 (15) Mean fat score (SD)$ 135 (85) 134 (85)

*At time of hospital admission for acute event associated with recruitment to this trial.

iQuestion asked regarding leisure time activity in 7.day period sufficiently long to work up a sweat.

*From interviewer-administered questionnaire 1. Missing values were not included in the denominators used to calculate percentages.

calls were attempted, and patients were encouraged to “phone in” with any questions regarding the program on a toll-free number.

At completion of the 6-month follow-up period, the same intervention package (without contracting and tele- phone calls) was offered to all subjects in the usual care group*

The study outcomes were assessed at 6 months by a mailed questionnaire to each participant and by a request to visit a laboratory for a blood cholesterol mea- surement. The subject could visit any laboratory, and an open letter asked the laboratory to send the results and an invoice for payment to the study investigators. The questionnaire included an assessment of fat intake using a previously validated, self-administered questionnaire,7 details of drugs being taken, exercise performed, inter- ventions and investigations performed, and employment history since hospital discharge, as well as current quali- ty of life. Reminders were mailed to subjects not retum- ing a questionnaire or to those from whom a blood cho- lesterol result was not received.

The quality-of-life questionnaire was designed espe- cially for coronary artery disease. It was developed by Oldridge et al* who based it on a previously validated measure of quality of life with congestive heart failure.9 Based on the answers to the questionnaire, we per- formed a factor analysis that identified 3 separate fac- tors: “emotional” (including 15 questions), “physical” (9 questions) and “social” (7 questions). This was dis- cussed in more detail previously.lO

StatIstical methods For descriptive purposes, partic- ipants and nonparticipants were compared regarding

760 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 72 OCTOBER 1,1993

CABLE III Differences Between Intervention and Usual Care Groups at Six-Month Follow-Up

Intervention Usual Care Difference

(95% confidence interval)

Returned 6-month questionnaire

Returned to work (of those originally employed)

Since hospital discharge Cardiac catheter Coronary angioplasty Coronary artery bypass graft z 1 hospital readmission

Currently receiving aspirin Currently receiving p blockers Exercise (3 times weekly) Mean fat score (SD)* Current smoker Mean blood cholesterol (SDYf

mmol/L mg/dl

Mean HDL cholesterol (SD)t mmol/L mgldl

168/213 (78.9%) 2071237 (87.3%)

40/61 (66%) 50166 (76%)

60 (36%) 11 (7%) 29 (17%) 47 (28%)

113 (67%) 61 (36%)

112 (67%) 13.1 (6.5)

18 (11%)

5.73 (1.01) 221.6 (39)

1.14 (0.41) 44.1 (16)

73 (36%) 16 (8%) 35 (17%) 60 (29%)

147 (71%) 81 (39%)

124 (61%) 15.3 (7.1)

26 (13%)

5.61 (0.98) 216.9 (38)

1.10 (0.32) 42.5 (12)

-8.4% t-15.3--1.5)

-10.2% c-26.0-5.6)

0.4% (-9.4-10.2) -1.2% t-6.6-4.2)

0.5% C-7.3-8.3) -0.8% (-10.0-8.4) -3.7% C-13.1-5.7) -2.8% C-12.7-7.1)

6.7% t-3.1-16.5) -2.2 C-3.6--0.8)

- 1.9% (8.4-4.6)

0.12 (0.10-0.34) 4.6 C-4-13)

0.04 (-0.05-0.13) 1.5 (- 1.9-5.0)

*From self-administered questionnaire 2.7 tin 145 intervention and 177 usual care subjects. Unless otherwise stated, percentages are of subjects who returned questionnaire.

TABLE IV Mean Scores on Quality-of-Life Factors at Six-Month Follow-Up

Intervention; Usual Care; Difference Factor No. (mean f SD) No. (mean * SD) 195% confidence interval)

Emotional (15 questions)

Physical (9 questions)

Social (7 questions)

168 (5.40 i- 1.12) 207 (5.15 f 1.24) 0.25 (0.01-0.49)

168 (5.40 2 1.21) 206 (5.22 k 1.29) 0.18 C-0.07-0.43)

168 (5.86 A 1.10) 205 (5.75 rt 1.14) 0.11 C-0.12-0.34)

several demographic variables including gender, age and history of AMI, using &i-square tests.

At baseline and 6-month follow-up, the intervention and usual care groups were compared by calculating dif- ferences and 95% confidence intervals for differences in proportions and means using the normal distribution. The mean scores of the 3 factors derived from the re- sponses to the quality-of-life questionnaire (emotional, physical and social) were compared between interven- tion and usual care groups, and corresponding 95% con- fidence intervals were obtained.

RESULTS During the study period, there were 635 patients who

were discharged from the hospital and fulfilled the study entry criteria. Patients who agreed to participate in the study were recruited until the planned sample size of 450 was achieved. Fourteen subjects were too ill to be approached and were not considered further. There were 56 patients registered by the MONICA Project who re- fused to participate: 3 refused to participate in the MONICA study, and the other 53 provided data for that project, but refused to participate in any associated study. A further 57 subjects agreed to consider being involved in other studies, but refused to participate in the

1

secondary prevention study for various reasons. In addi- tion, 58 patients left the hospital too early to be inter- viewed by the study nurses and thus were not invited to participate in the secondary prevention study.

A comparison of the study participants and 3 main groups of nonparticipants is shown in Table I. The only statistically signilicant difference was that in the group who left the hospital too early to be interviewed (usual- ly because the coronary event was not severe), more subjects were aged ~50 years.

Of 450 patients randomly allocated, 213 were in the intervention and 237 in the usual care group. The num- bers were not equal, because the allocation process was according to the general practitioner (see Methods). Table II shows that the 2 groups were similar in baseline char- acteristics. There were no signilicant differences in demo- graphic characteristics, medical history or patterns of con- sumption of tobacco, alcohol and fat. Fii percent (n = 107) of the intervention group had a clinical discharge diagnosis coded as AMI compared with 55% (n = 128) of the usual care group (p = 0.55). Of the intervention and usual care groups, 96 and 97%, respectively, had a MONICA diagnosis of ‘ ‘definite” or “possible” AMI.

Six-month follow-up questionnaires were received from 168 of 213 intervention subjects (79%) and from

SECONDARY PREVENTION AFTER ACUTE MYOCARDIAL INFARCTION 761

207 of 237 usual care subjects (87%), although 6 inter- vention and 3 usual care subjects had died during the 6 months and were not sent follow-up questionnaires. The difference in follow-up response rates was statistically significant (-8.4%; 95% confidence interval -15.3 to -1.5; p = 0.0171). Follow-up cholesterol measurements were received from 87 and 86% of subjects who returned a questionnaire in the intervention and usual care groups, respectively. Table III shows some charac- teristics of the 2 study groups at follow-up. The mean fat intake score was lower in the intervention group (p = 0.002), but there were no differences between the 2 groups regarding other outcome variables. For both groups, angiography and coronary bypass grafts had in- creased, and cigarette smoking decreased substantially.

Table IV shows the mean scores for the 3 factors from the quality-of-life questionnaire; each was higher in the intervention than in the usual care group, and for the emotional factor, the difference was statistically sig- nilicant (p = 0.04).

The 6-month questionnaire also asked about use of preventive services provided in the course of usual clini- cal care. There was a very low use of dietary, exercise or antismoking services by both groups; approximately 10% of patients had received rehabilitation services, most of which were offered after coronary artery bypass grafting.

DISCUSSION Only 4.50 of 635 potential AM1 survivors (71%)

agreed to participate in the study. Although we did not identify major differences between respondents and non- respondents (Table I), this program, offered outside the usual care program in the context of a research study, had limited appeal to subjects recovering from AMI. Although the response rate may limit the external valid- ity of the study, it did not affect the internal validity, because groups were comparable at baseline after ran- domization (Table II). Adherence to the intervention pro- gram was good, but few subjects used the extra help available in the community, such as the “Quit for Life” program to stop smoking, and “Walking for Pleasure” groups.

The response rate to 6-month follow-up was lower in the intervention group and suggests that at least some subjects were put off by the intervention process. The extent of the bias that this produced in the outcome assessment is unknown, but it is possible that subjects who did not adopt secondary prevention advice were the ones who did not respond and that those who did re- spond may have been eager to please the investigators.

The overall change in risk factors was no different between the groups, although the stated dietary fat in- take was lower in the intervention than in the usual care group. It is probable that the amount of dietary change was insufficient to produce a detectable change in blood cholesterol level. This study was designed to have suffi- cient power to detect a 5% difference in cholesterol lev- els, because we thought that this would have clinical importance. However, a description of 16 trials of Step

1 diets (similar in magnitude to our dietary intervention) produced changes in blood cholesterol concentrations ranging from -2.1 to +l% over 4 to 10 years,” and the present study would not have detected a change of that size.

Use of secondary prevention agents (aspirin and p- blocking drugs) was not different between groups. A re- cent study by our group concentrating on encouraging general practitioners to increase use of those agents in patients with AM1 was also unsuccessful (unpublished Master’s thesis).

This study was designed as a low-level intervention that could easily be implemented, and the level of inten- sity may have been too low to reduce the major risk fac- tors. The alternative explanation that a large amount of effort is already put into secondary prevention as part of usual care, leaving little scope for improvement by our program, is plausible for cigarette smoking, but unlikely for the other outcomes. Although the hospitals in the study area all use National Heart Foundation advice booklets and audio- or videotapes during hospital- ization, there were no comprehensive post-hospital reha- bilitation programs during the time of the study.

The intervention program appears to have improved quality of life, although the questionnaire used provides only a subjective assessment. If this is a true effect, it may be important, and could lead to simple post-hospi- tal discharge programs to provide support and encour- agement to patients. Another study, although not ran- domized, has shown a benefit of a rehabilitation program on psychosocial function. l2 We plan to examine this aspect further in a study aimed at improving quality of life among patients surviving AMI.

1. Dobson AJ, Alexander HM, Al Roomi K, Gibberd RW, Heller RF, Malcolm JA, Steele PL. Methodological issues in interpreting trends in MONICA event rates. Rev Epidemiol Sante Publique 1990;38:397-402, 2. Rossouw JE, Lewis B, Rifkind BM. The value of lowering cholesterol after myccadial infarction. N Engl J Med 1990;323: 1112-l 119. 3. Oldridge NB, Guyatt GH, Fischer ME, Rinn AA. Cardiac rehabilitation after myocadial infarction: combined experience of randomised clinical trials. JAMA 1988;260:945-950. 4. Lau J, Antman EM, Jimenez-Silva J, Kupelnick B, Mosteller F, Chalmers TC. Cumulative m&-analysis of therapeutic trials for myccadial infarction. N Engl J Med 1992;321:2118-254. 5. WHO MONICA Project principal Investigators. The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. .I Clin Epidemiol 1988;41: 105-l 14. 6. Miller M, Swanson M, Coli T, Clark K, Maggiore P. Facts on fat: a commu- nity nutrition education campaign. J Food Nutr 1987;44:6165. 7. Dobson AJ, Blijlevens R, Alexander HM, Croce N, Heller RF, Higginbotham N, Pike G, Plotioff R, Russell A, Walker R. Short fat questionnaire: a self-admin- istered measure of fat intake. Aust J Public Health 1993:17:144-149. 8. Oldridge N, Guyatt G, Jones N, Cmwe J, Singer J: Feeny D, McKelvie R, Runions J, Streiner D, Torrance G. Effects on quality of life with comprehensive rehabilitation after acute myocardial infarction. &I J Cardiol 1991;67:1084-1089. 9. Guyatt GH, Nogradi S, Halcrow S, Singer J, Sullivan MJ, Fallens EL. Devel- opment and testing of a new measure of health status for clinical trials in heart fail- ure. J Gen Intern Med 1989:4:101-107. 10. Lii L L-Y, Valenti LA; Knapp J, Dobson A, Plot&off R, Higginbotham N, Heller RF. A self-administered quality-of-life questionnaire after acute myocardia~ infarction: a factor analytic appr&h. J Clin E$idemiol; in press. 11. Ramsay LE, Yea WW, Jackson PR. Dietary reduction of serum cholesterol concen~tion: time to think again. BA4J 1991;303:953-957. 12. Dracup N, Moser DK, Marsden C, Taylor SE, Guzy PN. Effects of a multi- dimensional cardiopulmonary rehabilitation program on psychosocial function. Am J Cardiol 1991;68:31-34.

762 THE AMERICAN JOURNAL OF CARDIOLOGY VOLUME 72 OCTOBER 1.1993