- Email: [email protected]
Age and referral to coronary angiography after an abnormal treadmill thallium test
American Heart Journal Founded in 1925
February 1997 Volume
133, Number 2
CLINICAL INVESTIGATIONS
Acute Ischemic Heart Disease
Age and referral to coronary angiography after an abnormal treadmill thallium test Michael S. Lauer, MD, Fredric J. Pashkow, MD, Claire E. Snader, MS, Sharon A. Harvey, MS, James D. Thomas, MD, and Thomas H. Marwick, MD Cleveland, Ohio
This study investigated the association between age and referral to coronary angiography among ambulatory adults with an abnormal treadmill thallium scan. The subjects studied were 416 consecutive adults who were ->30 years old, under the care of cardiologists, and had an abnormal treadmill thallium scan between 1990 and 1993 at the Cleveland Clinic Foundation. The primary end point was performance of coronary angiography within 90 days of the treadmill test. Coronary angiography was performed in 163 subjects. Coronary angiography was performed in 46% of patients aged 30-49 years, in 53% of those aged 50 to 64 years, in 33% of those aged 65 to 74 years, and in only 18% of those aged ->75 years (X2 test for trend, p < 0.0001). After adjustment for potential confounders, age remained associated with a lower rate of referral to angiography (p < 0.0001). During 2 years of follow-up 34 deaths occurred (14 cardiac), with particularly high mortality rates among those aged >74 years (cumulative rate 31%, 96% confidence interval 16% to 47%). The number of abnormal thallium scan segments was predictive of death (p= 0.02). These data suggest that increasing age is associated with a lower rate of referral to coronary angiography after an abnormal treadmill thallium test. (Am Heart J !997;133:139-46.)
Myocardial revascularization is being increasingly used with improving outcomes in older adults. 1 In the United States Medicare population, the number of coronary angioplasties performed increased from 43,271 in 1987 to 70,869 in 1990, while the number of bypass surgeries performed increased from 79,260 From the Department of Cardiology, Cleveland Clinic Foundation. Received for publication Jan. 26, 1996; accepted May 20, 1996. Reprint requests: Michael S. Lauer, MD, Cardiovascular Imaging Center, Department of Cardiology, Desk F-15, Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195. Copyright © 1997 by Mosby-Year Book, Inc. 0002-8703/97/$5.00 + 0 4/1/76862
to 99,317 during the same years; both procedures were associated with lower short- and long-term mortality rates over this time. 1 Because all myocardial revascularization must be preceded by coronary angiography, which is often itself preceded by noninvasive testing, determining if any association between age and referral to coronary angiography exists is important in the recent era of more revascularization procedures. Few data systematically address this question; one recent study noted an inverse association between age and performance of coronary angiography among patients with acute myocardial infarction admitted to a teaching hospital. 2 This study was designed to determine whether any association exists between age and referral to coronary angiography among adults undergoing noninvasive evaluation of known or suspected coronary disease and whose physician considered them ambulatory enough to be referred to treadmill-thallium testing (as opposed to pharmacologic testing). Analyses were confined to consecutive adults who were ->30 years old, under the care of cardiologists, and who had an abnormal thallium study in our institution between 1990 and 1993. We found t h a t increasing age was associated with a lower rate of referral to coronary angiography. METHODS Study sample. Subjects for this investigation were de-
rived from a group of 416 consecutive adults under the care of cardiologists who were ->30 years old and who had an abnormal symptom-limited treadmill-thallium test at the Cleveland Clinic Foundation between September 1990 and December 1993. All subjects were undergoing their first 139
February 1997
140 Lauer et al.
AmericanHeartJournal
25
II
20
1,,,
oe~
E ..¢ Z
I,,
15
10
h
30 34 38 42 46 50 54 58 62 66 70 74 78 82 86 Age (years) Fig. 1. Distribution of age in the study sample. cardiac stress test at the Cleveland Clinic Foundation. A t h a l l i u m scan was considered a b n o r m a l if either a fixed or t r a n s i e n t perfusion deficit was noted. Subjects with histories of prior cardiac procedures (including a n y angiography), congestive h e a r t failure, or v a l v u l a r or congenital h e a r t disease were excluded from the study. The age distribution of eligible subjects is shown in Fig. 1. During this period, a total of 3299 patients were referred by cardiologists for pharmacologic or exercise stress t h a l l i u m studies; 1341 (41%) h a d n o r m a l exercise results a n d 1542 (47%) h a d normal pharmacologic results. All subjects gave informed consent before undergoing t r e a d m i l l - t h a l l i u m testing. Clinical data. Before each treadmill test, a structured interview and chart review yielded d a t a on symptoms, medications, coronary risk factors, prior cardiac events, and a n u m b e r of cardiac and noncardiac diagnoses. All these data, as well as stress test data, were evaluated prospectively. Resting hypertension was defined as a resting systolic blood pressure ->140 m m Hg, a resting diastolic blood pressure ->90 m m Hg, or t r e a t m e n t with antihypertensive medications. 3 A s s e s s m e n t of diabetes was based on chart review, questioning, and medication use. Prior coronary events consisted p r i m a r i l y of documented myocardial infarction or hospitalization for unstable angina pectoris. Exercise testing. Treadmill testing was carried out according to s t a n d a r d protocols, u s u a l l y the Bruce and modified Bruce protocols. 4, 5 Leaning on h a n d r a i l s during exercise was not allowed. During each stage of exercise, d a t a on symptoms, r h y t h m , h e a r t rate, blood pressure, workload in metabolic equivalents (METs), and ST segm e n t s were collected. Exercise capacity in METs was est i m a t e d from s t a n d a r d published tables. 5 In subjects with
i n t e r p r e t a b l e ST segments (using all leads except a VR to a precision level of 0.5 mm), an ischemic response was considered p r e s e n t if-> 1 m m of horizontal or downsloping ST segment depression occurred 80 msec after the J point, or if-> 1 m m of additional ST segment elevation occurred in leads without pathologic Q waves. Thallium scintigraphy. P a t i e n t s were injected with 2 to 3 pCi of thallium-201 1 m i n u t e before t e r m i n a t i o n of the treadmill test. W i t h i n 10 minutes of stress testing, single positron emission computed tomography (SPECT) imaging was performed over 360 degrees of acquisition with a low-energy, all-purpose collimator with 32 stops, 1.5 to 4 million counts, over 9 to 14 minutes. P a t i e n t s r e t u r n e d 3 to 4 hours after the stress test for redistribution SPECT t h a l l i u m imaging. Images were reconstructed with a H a m m i n g filter and projected in short-axis, vertical, and horizontal long-axis views in a side-by-side display. Ischem i a was i n t e r p r e t e d as the presence of >20% reversibility and scarring was i n t e r p r e t e d as the presence of counts <80% of m a x i m u m (<70% for the posterior wall). Clinical reports of all t h a l l i u m studies were retrieved a n d coded according to a 12-segment system. 6 Each segm e n t was coded as normal, ischemic, or scarred; further coding took into account abnormalities of specific coronary a r t e r y territories (namely, left anterior descending, left circumflex, and right coronary/dominant left posterior descending). The coding of the t h a l l i u m d a t a was blinded with r e g a r d to age a n d other clinical data. Coronary angiography. For subjects who u n d e r w e n t coronary angiography within 90 days of stress testing, coronary angiograms were i n t e r p r e t e d semiqualitatively. Coronary disease was defined as ->50% stenosis in any proximal or middle coronary vessel or major branch. Severe coronary disease was defined as ->50% left m a i n
Volume 133, Number 2 American Heart Journal
Lau8F et al.
141
Table I. Baseline characteristics according to age Characteristic
Aged 30-49 y r (n = 59)
Aged 50-64 y r (n = 137)
Aged 65-74 y r (n = 169)
Aged >74 y r (n = 51)
p Value
Age (yr) Female (%) Systolic blood pressure (ram Hg) Diastolic blood pressure (ram Hg) Resting heart rate (beats/rain) Prior coronary events (%) Smoker (%) Hypertension (%) B-blocker use (%) Calcium blocker use (%) Lipid lowering therapy (%) Chest pain history (%) Diabetes (%)
43 +_5 6 (10%) 130 + 16 88 -+ 11 77 -+ 12 14 (24%) 20 (34%) 14 (24%) 7 (12%) 10 (17%) 5 (8%) 32 (54%) 9 (15%)
58 -+ 4 18 (13%) 137 +_ 18 88 -+ 10 79 -+ 14 55 (40%) 30 (22%) 66 (48%) 34 (25%) 40 (29%) 11 (8%) 71 (52%) 23 (17%)
69 -+ 3 27 (16%) 144 _+23 86 -+ 11 76 -+ 13 64 (38%) 24 (14%) 86 (51%) 19 (11%) 50 (30%) 22 (13%) 73 (43%) 41 (24%)
78 -+ 3 11 (22%) 141 _+ 20 86 _+ 12 77 + 14 22 (43%) 0 (-) 28 (55%) 6 (12%) 14 (27%) 3 (6%) 21 (41%) 8 (16%)
<0.0001 >0.3 <0.0001 >0.2 >0.3 >0.1 <0.001 0.002 0.006 >0.2 >0.3 >0.7 >0.2
V a l u e s are m e a n ± S D or n u m b e r s (percentages).
a r t e r y stenosis, as three-vessel disease (->70% stenosis in each major coronary a r t e r y system), or as two-vessel disease with a ->70% proximal left anterior descending a r t e r y lesion. Statistical analyses. Subjects were divided into four prespecified age groups, 30 to 49, 50 to 64, 65 to 74, a n d ->75 years. Baseline clinical and exercise continuous variables were compared using the Kruskal-Wallis test, and categoric clinical, exercise, and angiographic variables were compared with the chi-square test or the F i s h e r exact test. Associations between tha]linm results and age as a continuous variable were tested by logistic regression. The p r i m a r y end point of this s t u d y was performance of coronary angiography within 90 days of the treadmill t h a l l i u m test. In u n a d j u s t e d analyses, coronary angiography r a t e s were compared in the four age groups using the Mantel-Haenszel extension test for trend. 7 F o r descriptive purposes, r a t e s of angiography were calculated for predefined subgroups, including those with a n y ischemia revealed by thallium, ischemia noted in the t e r r i t o r y of the left anterior descending artery, and patients with -<2, 3 to 4, and >-5 total a b n o r m a l (whether ischemic or fixed) segments. The inclusion of both ischemic a n d fixed defects allowed for the possibility t h a t physicians m a y refer patients with a n g i n a and extensive fLxed defects to angiography, especially because reinjection was not routinely performed, m a k i n g the differentiation between ischemia a n d scar problematic. The association of age as a continuous variable with referral to coronary angiography was analyzed by multivari: ate logistic regression analyses. Initial models adjusted for potential confounding variables including sex, resting blood pressure, medications, hypercholesterolemia t r e a t e d with lipid-lowering medications, smoking, diabetes, prior coronary events, exercise capacity in METs, h e a r t r a t e a n d blood p r e s s u r e responses to exercise, occurrence of exercise-induced ventricular tachycardia, a n d extent of ischem i a revealed by thallium. Abnormalities revealed by thallium were analyzed four ways: (1) a n y ischemia revealed by thallium, (2) the n u m b e r of ischemic coronary a r t e r y
segments, (3) the presence of ischemia in the left anterior descending a r t e r y distribution, and (4) the total n u m b e r of a b n o r m a l (whether ischemic or fLxed) a b n o r m a l segments. Because models involving these four different t h a l l i u m variables yielde d essentially identical results, only the results of models considering the presence of a n y ischemia revealed by t h a l l i u m are presented. Logistic analyses were also performed to assess for potential interactions between baseline variables and age; no significant interactions were found. In s u p p l e m e n t a r y analyses, only subjects aged ->65 y e a r s were considered; all of these patients can be considered elderly by Medicare criteria, yet we could differentiate between "young" old subjects (aged 65 to 74 years), who h a d s u b s t a n t i a l l y better n a t u r a l histories, and "old" old subjects (aged >74 years). To investigate the potential clinical significance of differentia] rates of referral to coronary angiography as a function of age, 2-year all-cause m o r t a l i t y rates were d e t e r m i n e d in each of the age groups by using the Social Security A d m i n i s t r a t i o n D e a t h M a s t e r File. s D e a t h certificates were reviewed to determine w h e t h e r the cause of d e a t h was cardiac. Cumulative m o r t a l i t y r a t e s were calculated by use of the Kaplan-Meier method with Greenwood 95% confidence intervals; differences in m o r t a l i t y r a t e s were tested with the log r a n k chi-square statistic. The Cox proportional h a z a r d s model 9 was used to e s t i m a t e relative risks for deaths according to age, abnormalities revealed by thallium, and performance of coronary angiography. All analyses were performed with use of the SAS 6.08 statistical package (SAS Institute, Cary, N.C.).
RESULTS Baseline characteristics. O f 416 e l i g i b l e s u b j e c t s , 59 w e r e a g e d 30 to 49 y e a r s , 137 w e r e a g e d 50 to 64 y e a r s , 169 w e r e a g e d 65 to 74 y e a r s , a n d 51 w e r e ->75 y e a r s old. T h e p r i m a r y i n d i c a t i o n s for t r e a d m i l l thallium testing were "rule out coronary disease"
February 1997 American Heart Journal
142 Lauer et al. T a b l e II. E x e r c i s e c h a r a c t e r i s t i c s
Aged 30-49 yr Aged 50-64 y r (n = 59) (n = 137)
Characteristic
Protocol (%) Bruce Modified Bruce CAEP Other Peak METs Peak systolic blood pressure (mm Hg) Peak heart rate (beats/min) Maximum rating of perceived exertion (0-10) Ischemic ST segments* (%) Ventricular tachycardia (%) Anginal chest pain (%) Failure to reach 85% of maximum age-predicted heart rate
Aged >74 yr (n = 51)
1 (2%) 15 (29%) 16 (31%) 19 (38%) 5.0 _+ 1.6 179 _+ 26 126 _+ 19 8.0 ± 2.7 6 (35%) 2 (4%) 6 (12%) 17 (33%)
43 (73%) 10 (17%) 1 (2%) 5 (8%) 10.0 _+ 2.2 183 ± 27 166 _+ 16 9.2 _+ 1.4 11 (26%) 1 (2%) 5 (8%) 8 (14%)
46 (34%) 57 (42%) 7 (5%) 27 (19%) 8.0 + 2.3 185 ± 31 148 ± 21 8.7 _+ 1.9 40 (41%) 2 (2%) 36 (26%) 33 (24%)
17 (4%) 99 (59%) 19 (11%) 34 (26%) 6.5 _+ 2.0 182 _+ 31 135 _+ 20 8.2 _+ 2.4 51 (54%) 3 (2%) 43 (25%) 52 (31%)
1 (2%)
7 (5%)
8 (5%)
(%) Hypotensive response (%)
Aged 65-74 yr (n = 169)
0 (-)
p Value
<0.001
<0.0001 >0.5 <0.0001 0.06 0.02 >0.7 0.007 0.04 >0.2
Values are mean -+ SD or numbers (percentages). CAEP, Chronotropic assessment exercise protocol;MET, metabolic equivalents. *Of subjects with interpretable ST segments.
T a b l e II1. T h a l l i u m s c i n t i g r a p h y c h a r a c t e r i s t i c s
Characteristic
Aged 30-49 y r (n = 59)
Any ischemia (%) Ischemia in the LAD territory (%) Number of ischemic segments Number of scarred segments Number of abnormal segments
35 (59%) 16 (27%) 1.3 ± 1.7 1.6 -+ 1.8 2.9 _+ 1.9
Aged 50-64 y r (n = 137)
69 (50%) ' 37 (27%) 1.0 -+ 1.4 2.4 ± 2.2 3.4 _+ 2.1
Aged 65-74 yr (n = 169)
Aged >74 yr (n = 51)
p Value
76 (45%) 32 (19%) 1.0 ± 1.4 2.6 -~ 2.3 3.5 _+ 2.1
20 (39%) 12 (24%) 0.9 -+ 1.4 2.8 -+ 2.4 3.7 _+ 2.5
>0.1 >0.3 >0.2 0.02 >0.1
Values are mean +- SD or numbers (percentages). LAD, Left anterior descending artery.
(58%) and "follow up coronary disease" (33%). Other baseline characteristics are summarized in Table I. Older subjects were less likely to be smokers and more likely to be taking ~-blockers and have a history of hypertension. Exercise and thallium characteristics. Older subjects were more likely to be assigned to the modified Bruce protocol (p < 0.001), to have a lower exercise capacity (p < 0.0001), and to have a lower peak heart rate (p < 0.0001) (Table II). The most common reason for terminating exercise was fatigue (93% for subjects aged 30 to 49 years, 82% for those aged 50 to 64 years, 80% for those aged 64 to 74 years, and 90% for those aged ->75 years, p = 0.06). No trends were noted for age-related differences in other reasons for exercise termination, including marked ST segment depression, anginal chest discomfort on the treadmill, claudication, blood pressure changes, and arrhythmias. We found a statistically nonsignificant trend toward older subjects being less likely to have ischemia revealed by thallium (Table III). When treated as a
continuous variable, increasing age was associated with a lesser likelihood of any ischemia revealed by thallium (for 5-year increase in age, odds ratio [OR] = 0.88, 95% confidence interval [CI] 0.80 to 0.96, p = 0.005). Age was not associated with the frequency of ischemia revealed by thallium in multiple coronary artery territories or in the territory of the left anterior descending artery, even when treated as a continuous variable. Older subjects had a greater number of scarred segments (that is, fixed defects) (p = 0.02), however, and the total number of abnormal segments (fixed or ischemic) tended to increase in older patients. Age and referral to coronary angiography: univariate analyses. During 90 days of follow-up, 163 (39%) subjects were referred for coronary angiography. Coronary angiography was performed in 46% of subjects aged 30 to 49 years, in 53% of those aged 50 to 64 years, in 33% of those aged 65 to 74 years, and in only 18% of those aged ->75 years (Mantel-Haenszel extension test for trend, ×2 = 17.7, p < 0.0001) (Fig.
Volume 133, Number 2 American Heart Journal
LGUeT et al.
143
o
~ >e-, ft •
e-
.o<
LAD Ischemia (P=0.0006) I'hal Ischemia (P=0.0002) Subjects (P<0.0001)
0.
b~)-f4
Age (years)
>74
Fig. 2. Association of age with rate of referral to coronary angiography according to results of the thallium scan. LAD, Left anterior descending artery; thal ischemia, thallium-documented ischemia.
~ >e-,
"O
=-
Q.
"E'5) • e..
o< O
O.
> 4 segments (P=0.003) 3-4 Segments (P=0.03) segments (P=0.03) b:)-IR.
Age (years)
>74
Fig. 3. Association of age with rate of referral to coronary angiography according to the total number of abnormal (transient or fLxed)segments on the thallium scan. 2). Among subjects with any ischemia revealed by thallium (n = 200), the corresponding proportions were 69%, 72%, 49%, and 25% (Mantel-Haenszel extension test for trend, ×2 = 14.8, p = 0.0002). Among subjects with evidence ofischemia in the left anterior descending artery distribution, referral to coronary angiography occurred in 13 of 16 (81%) subjects aged 30 to 49 years, in 29 of 37 (78%) subjects aged 50 to 64 years, in 16 of 32 (50%) subjects aged 65 to 74 years, and in only 4 of 12 subjects (33%) aged ->75 years (Mantel-Haenszel extension test for trend, X 2 = 11.5, p = 0.0006). Patients were also analyzed according to the total number of abnormal segments (whether ischemic or fLxed defects) imaged by thallium (Fig. 3). In all subgroups, older subjects were less likely to be referred to coronary angiography; this was particularly marked among the 117 patients with ->5 abnormal
segments (Mantel-Haenszel extension test for trend, p = 0.003). Angiographic characteristics. The angiographic results noted among the 163 subjects who underwent angiography are summarized in Table IV. Age was associated with a higher prevalence of any coronary disease (p < 0.001); also, older subjects had higher prevalences of severe disease (p = 0.007), left main arterial disease (p = 0.002 by Fisher exact test), and severe disease of the proximal left anterior descending artery (p = 0.005 by Fisher exact test). Age and referral to coronary angiography multivariate analyses. After adjustment for the extent ofischemia revealed by thallium, clinical characteristics, and exercise findings including functional capacity, increasing age remained associated with a lower rate of referral to coronary angiography (for 5-year inc r e a s e in age, adjusted OR = 0.81, 95% CI 0.73 to X2 = 8 . 8 1 ,
February 1997
144
AmericanHeartJournal
L a u e r et al.
Table IV.
Angiographic characteristics of subjects who underwent cardiac catheterization Characteristic
Aged 30-49 y r (n = 27)
Aged 50-64 yr (n = 72)
Aged 65-74 yr (n = 55)
Aged >74 y r (n = 9)
p Value
Any coronarydisease (%) Severe coronarydisease (%) Left main arterial disease (%) Proximal LAD disease (%)
7 (26%) 5 (19%) 3 (11%) 1 (4%)
55 (76%) 22 (31%) 5 (7%) 19 (26%)
44 (80%) 23 (42%) 11 (20%) 21 (38%)
8 (89%) 7 (78%) 5 (56%) 2 (22%)
<0.001 . 0.007 0.002 0.005
LAD, Left anterior descending artery/
0.90,p < 0.0001). Other predictors for referral to coronary angiography included presence of any ischemia revealed by thallium (adjusted OR = 4.66, 95% CI 2.93 to 7.41,p < 0.0001) and anginal chest pain on the treadmill (adjusted OR = 4.62, 95% CI 2.65 to 8.07, p < 0.0001). Referral to coronary angiography was associated with ischemia revealed by thallium in the subjects aged 50 to 64 years (adjusted OR = 6.61, 95% CI 2.96 to 14.70, p < 0.0001) and in those aged 65 to 74 years (adjusted OR = 3.46, 95% CI 1.83 to 8.55, p = 0.0007) and with anginal chest pain on the treadmill in these groups (aged 50 to 64 years, adjusted OR = 4.96, 95% CI 1.85 to 13.10,p = 0.001; aged 65 to 74 years, adjusted OR = 3.96, 95% CI 1.69 to 7.06,p = 0.0005). Among patients aged >74 years, referral to coronary angiography tended to be associated with anginal chest pain on the treadmill (adjusted OR = 7.26, 95% CI 0.88 to 59.79,p = 0.07), but was not associated with ischemia revealed by thallium (adjusted OR--1.33, 95% CI 0.26 to 6.99, p > 0.7). When analyses were confined to those subjects aged ->65 years, age remained associated with a lower rate of referral to coronary angiography (for each 5-year increase in age, adjusted OR = 0.64, 95% CI 0.44 to 0.95,p = 0.02). In none of these models was sex predictive of referral to angiography. All the covariates considered were tested in other models to look for potential effect modification (interaction) on the association of age with a lower rate of referral to coronary angiography. No significant interactions were noted. Mortality analyses. During approximately 2 years of follow-up, 34 deaths occurred. Cumulative mortality rates were relatively low among subjects aged <65 years, but increased m a r kedl y among those aged 65 to 74 years (2-year mortality rate = 9%, 95% CI 4% to 14%) and those aged >74 years (2-year mortality rate = 31%, 95% CI 16% to 47%) (log r a n k X2 = 33, degrees of freedom = 3,p < 0.0001; relative risk [RR] for 5-year increase in age = 1.77, 95% CI 1.41 to 2.22, p < 0.0001) (Fig. 4). All-cause mortality rates were also associated with the total number of abnormal segments on thallium scan (for each two additional abnormal segments, age-adjusted RR = 1.41, 95% CI
1.06 to 1.88, p = 0.02), but not with referral to coronary angiography (RR = 0.73, 95% CI 0.36 to 1.50, p > 0.3). Of the 34 deaths, 14 were of definite cardiac origin. Increased age was also associated with cardiac death (RR for 5-year increase in age -- 1.41, 95% CI 1.03 to 1.93, p = 0.03). Cardiac death was also associated with the total num ber of abnormal segments on thallium scan (for each two additional abnormal segments, RR = 1.60, 95% CI 1.03 to 2.48, p = 0.04), but it was not associated with referral to coronary angiography (RR = 1.14, 95% CI 0.40 to 3.30,p > 0.8). DISCUSSION Main findings. In this sample of 416 consecutive adults with an abnormal treadmill-thallium scan who were aged ->30 years and under the care of cardiologists, increasing age was associated with a lower rate of referral to coronary angiography. This association existed even after adjusting for sex, standard risk factors, medication use, functional capacify and other exercise parameters, anginal chest pain, and extent of ischemia revealed by thallium. None of the patients had previous coronary angiography and none had a history of congestive h e a rt failure. Even when analyses were confined to those aged ->65 years (enabling comparison between "young" old patients with relatively low mortality rates and "old" oM patients with substantially higher mortality rates), increasing age remained associated with a lower rate of referral to coronary angiography. Analyses were confined to those adults who were considered ambulatory enough by their managing physicians to be referred to treadmill-thallium testing, as opposed to pharmacologic testing; the latter would be more likely performed in subjects with other significant limiting conditions such as osteoarthritis, chronic lung disease, renal failure, and complicated diabetes, which might significantly impact on any decision regarding coronary angiography. Among the patients who were referred to angiography, older patients had substantially higher rates of significant coronary disease, severe disease, and left main arterial disease; all of these findings are
Volume 133, Number 2 AmericanHeartJournal
L a u e r et al.
1.00 0.90
>
"" ~
2-year survival (95% Cl): 1.oo (-) •
Age:
0.53 0.84)
>74
145
30-49 0.95 (0.91-0.99)D 50-64 (o.8s-o.gs)~ 65-74
=--~0.9'1
0.80
¢n 0.70 0.60 0.50 0.00
i 0.50
i 1.00
F 1.50
i 2.00
Years After Exercise Test
Fig. 4. Kaplan-Meier plot relating age to 2-year cumulative all-cause mortality rates. Note the marked increase in deaths among subjects aged >74 years. consistent with a higher threshold for referral to angiography among older patients. Furthermore, not surprisingly, older patients had substantially higher mortality rates during 2 years of follow-up. Exercise thallium testing is recognized as a feasible, safe, and prognostically useful modality for evaluating older adults with known or suspected coronary disease. 1°12 Coronary angiography and percutaneous interventions can also be safely performed in elderly patients, albeit with somewhat higher complication rates. 13-15 In view of these observations and recent findings t h a t myocardial revascularization outcomes are improving in elderly 1 and very elderly 16 subjects, the inverse association between age and performance of coronary angiography after an abnormal treadmill-thallium scan may be of importance. The lower rate of referral to coronary angiography among older subjects may be a result of lower peak heart rates, with lower associated rates of thalliumdocumentedischemia. The association persisted even after accounting for the thallium results in logistic regression models. However, given the semiqualitative description of the thallium results, a more powerful confounding effect may be operative. Nonetheless, when subjects were stratified according to thallium test results, similar trends were noted. Previous studies. Few data exist regarding any association between age and performance of coronary angiography. One study by Krumholz et al. 16 of gender bias in a teaching hospital found that age was associated with a lower rate of referral to coronary
angiography among survivors of acute myocardial infarction. 2 In a recent report 17 on the use of resources in patients after acute myocardial infarction fee-for-service or health maintenance organization care settings, age was also associated with a lower rate of performing coronary angiography, with an approximate odds ratio for every 10-year increase in age of 0.6. Another report is noted t h a t older patients referred for coronary angiography were more likely to have more severe symptoms, be receiving more aggressive medical therapy, and have more severe angiographic disease t h a n younger subjects. Potential limitations. This is an observational study, with the principal limitation being the possibility of missed confounders. Data on physician age and on patients' responses to their physicians' recommendations were not available. Other important factors for which we did not have information include patient priorities (especially as a trend existed toward less chest pain in older subjects) regarding the benefits and risks of aggTessive revascularization, cognition, affective state, marital status, financial concerns, and familial support. Physician attitudes toward performance of potentially high-risk procedures in elderly patients were not measured and may have played a role. One could argue t h a t these data are unique to patients managed by cardiologists in one tertiary referral center who are referred for evaluation in the exercise laboratory. Older patients may be more likely to be referred to angiography via other routes. In this regard, the data reported by Krumholz et al. 2
146
February 1997 American Heart Journal
L a u e r et al.
are particularly interesting. In that study, patients at a different institution were referred to coronary angiography after sustaining a myocardial infarction and being admitted to the coronary care unit. Consistent with our results, increasing age was associated with a markedly decreased use of coronary angiography. Analogously, in a very recent study, Every et al. 17 found that age was associated with a lower rate of coronary angiography and coronary bypass surgery in patients being treated for acute myocardial infarction in a spectrum of different hospitals. Furthermore; our results parallel those of others in which use of treatments for coronary artery disease have been noted to be underused in elderly patients. For example, Meehan et al.19 studied a Medicare cohort of patients treated for acute myocardial infarction in six different Connecticut hospitals and found that thrombolytic therapy was used in only 43% of eligible patients, B-blockers were used in only 41%, and aspirin was used in only 73%. In view of these observations, and the observations of Krumholz et al. 2 and Every et al., 17 our observations are not unique to one institution or one route of selecting patients for coronary angiography. Nonetheless, w e must emphasize that this population is a relatively selective one of patients who could perform treadmill exercise testing and who had not had prior coronary angiography or congestive heart failure. Whether the inverse association of age with referral to co]:onary angiography represents underuse of coronary angiography in the elderly or appropriate care i s not clear. The poor exercise capacity of the elderly subjects in the current study suggests that these patients are more functionally impaired, which would increase their operative mortality rate. We do not have data on outcomes among patients managed differently (for example, with direct angiography without prior stress testing) or to natural history data from the same age groups. Despite these potential limitations, the association between age and a lower rate of performance of coronary angiography in our cohort was quite marked. Given the increasing use of myocardial revascularization, these findings should strike great interest. Further studies are necessary to confirm these results and to determine whether the lower rate of referral to coronary angiography represents
appropriate care, reflecting the priorities and prognoses of older patients, or whether an opportunity exists to improve clinical outcomes. REFERENCES
1. Peterson ED, Jollis KG, Bebchuk JD, DeLong ER, Muhlbaier LH, Mark DB, et al. Changes in mortality after myocardial revascularization in the elderly: the national Medicare experience. Ann Intern Med 1994; 121:919-27. 2. Krumholz HM, Douglas PS, Lauer MS, PasternakRC. Selectionofpatients for coronary angiography early after myocardial infarction: is there evidence for a gender bias? Ann Intern Med 1992;116:785-90. 3. The fifth report of the Joint National Committee on detection, evaluation, and treatment of high blood pressure (JNC V). Arch Intern Med 1993;153:154-83. 4. Dean AE, Peterson DR, Blackmon JR, Bruce RA. Myocardial ischemia "after maximal exercise in healthy men. Am Heart J 1965;69:11-25. 5. Fletcher GF, Froelicher VF, Hartley LH, Haskell WL, Pollock ML. Exercise standards: a statement for health professionals from the American Heart Association. Circulation 1990;82:1-36. 6. Zaret BL, Wackers FJ. Nuclear cardiology. N Engl J Med 1993;329:77583. 7. Mantel N. Chi-square tests with one degree of freedom: extensions of the Mantel-Haenszel procedure. J Am Stat Assoc 1963;58:690-700. 8. Curb JD, Ford CE, Fressel S, Palmer M, Babcock C, Hawkins CM. Ascertainment of vital status through the National Death Index and the Social Security Administration. Am J Epidemiol 1985;121:754-68. 9. Cox DR. Regression models and life tables (with discussion). J R Stat Soc Br 1972;34:187-220. 10. Ades PA, Grunvald MH. Cardiopulmonary exercise testing before and after conditioning in older coronary patients. Am Heart J 1990; 120:585-9. 11. Hilton TC, Shaw LJ, Chaitman BR, Stocke KS, Goodgold HM, Miller DD. Prognostic significance of exercise thallium-201 testing in patients aged greater than or equal to 70 with known or suspected coronary artery disease. Am J Cardiol 1992;69:45-50. 12. Thompson RF, Crist DM, Osborn LA, Atterbom HA. Treadmill exercise electrocardiography in the elderly with physical impairments. Gerontology i990;36:112-8. 13. Rich MW, Crecelins CA. Incidence, risk factors, and clinical course of acute renal insufficiency after cardiac catheterization in patients 70 years of age or older. Arch Intern Med 1990;150:1237-42. 14. Muller DWM, Shamir KJ, Ellis SG, Topol EJ. Peripheral vascular complications after conventional and complex percutaneous coronary interventional procedures. Am J Cardiol 1992;69:63-8. 15. Clark VL, Dolce J. Unplanned admissions after outpatient cardiac catheterization. Clin Cardiol 1993;16:823-6. 16. Krumholz HM, Forman DE, Kuntz RE, Balm DS, Wei JY. Coronary revascularization after myocardial infarction in the very elderly: outcomes and long-term follow-up. Ann Intern Med 1993;119:108490. 17. Every NR, Fihn SD, Maynard C, Martin JS~ Weaver D for the MITI investigators. Resource utilization in treatment of acute myocardial infarction: staff-model health maintenance organization versus fee-forseryice hospitals, J Am Coll Cardiol 1995;26:401-6. 18. Elder AT, Shaw TRD, Turnbull CM, Starkey IR. Elderly and younger patients selected to undergo coronary angiography. Br Med J 1991; 303:950-3. 19. Meehan TP, Hennen J, Radf°rd MJ, Petrill° MK, Elstein P, Ballard DJ" Process and outcome of care for acute myocardial infarction among Medicare beneficiaries in Connecticut: a quality improvement demonstration project. Ann Intern Med 1995;122:928-36.
February 1997 Volume
133, Number 2
CLINICAL INVESTIGATIONS
Acute Ischemic Heart Disease
Age and referral to coronary angiography after an abnormal treadmill thallium test Michael S. Lauer, MD, Fredric J. Pashkow, MD, Claire E. Snader, MS, Sharon A. Harvey, MS, James D. Thomas, MD, and Thomas H. Marwick, MD Cleveland, Ohio
This study investigated the association between age and referral to coronary angiography among ambulatory adults with an abnormal treadmill thallium scan. The subjects studied were 416 consecutive adults who were ->30 years old, under the care of cardiologists, and had an abnormal treadmill thallium scan between 1990 and 1993 at the Cleveland Clinic Foundation. The primary end point was performance of coronary angiography within 90 days of the treadmill test. Coronary angiography was performed in 163 subjects. Coronary angiography was performed in 46% of patients aged 30-49 years, in 53% of those aged 50 to 64 years, in 33% of those aged 65 to 74 years, and in only 18% of those aged ->75 years (X2 test for trend, p < 0.0001). After adjustment for potential confounders, age remained associated with a lower rate of referral to angiography (p < 0.0001). During 2 years of follow-up 34 deaths occurred (14 cardiac), with particularly high mortality rates among those aged >74 years (cumulative rate 31%, 96% confidence interval 16% to 47%). The number of abnormal thallium scan segments was predictive of death (p= 0.02). These data suggest that increasing age is associated with a lower rate of referral to coronary angiography after an abnormal treadmill thallium test. (Am Heart J !997;133:139-46.)
Myocardial revascularization is being increasingly used with improving outcomes in older adults. 1 In the United States Medicare population, the number of coronary angioplasties performed increased from 43,271 in 1987 to 70,869 in 1990, while the number of bypass surgeries performed increased from 79,260 From the Department of Cardiology, Cleveland Clinic Foundation. Received for publication Jan. 26, 1996; accepted May 20, 1996. Reprint requests: Michael S. Lauer, MD, Cardiovascular Imaging Center, Department of Cardiology, Desk F-15, Cleveland Clinic Foundation, 9500 Euclid Ave., Cleveland, OH 44195. Copyright © 1997 by Mosby-Year Book, Inc. 0002-8703/97/$5.00 + 0 4/1/76862
to 99,317 during the same years; both procedures were associated with lower short- and long-term mortality rates over this time. 1 Because all myocardial revascularization must be preceded by coronary angiography, which is often itself preceded by noninvasive testing, determining if any association between age and referral to coronary angiography exists is important in the recent era of more revascularization procedures. Few data systematically address this question; one recent study noted an inverse association between age and performance of coronary angiography among patients with acute myocardial infarction admitted to a teaching hospital. 2 This study was designed to determine whether any association exists between age and referral to coronary angiography among adults undergoing noninvasive evaluation of known or suspected coronary disease and whose physician considered them ambulatory enough to be referred to treadmill-thallium testing (as opposed to pharmacologic testing). Analyses were confined to consecutive adults who were ->30 years old, under the care of cardiologists, and who had an abnormal thallium study in our institution between 1990 and 1993. We found t h a t increasing age was associated with a lower rate of referral to coronary angiography. METHODS Study sample. Subjects for this investigation were de-
rived from a group of 416 consecutive adults under the care of cardiologists who were ->30 years old and who had an abnormal symptom-limited treadmill-thallium test at the Cleveland Clinic Foundation between September 1990 and December 1993. All subjects were undergoing their first 139
February 1997
140 Lauer et al.
AmericanHeartJournal
25
II
20
1,,,
oe~
E ..¢ Z
I,,
15
10
h
30 34 38 42 46 50 54 58 62 66 70 74 78 82 86 Age (years) Fig. 1. Distribution of age in the study sample. cardiac stress test at the Cleveland Clinic Foundation. A t h a l l i u m scan was considered a b n o r m a l if either a fixed or t r a n s i e n t perfusion deficit was noted. Subjects with histories of prior cardiac procedures (including a n y angiography), congestive h e a r t failure, or v a l v u l a r or congenital h e a r t disease were excluded from the study. The age distribution of eligible subjects is shown in Fig. 1. During this period, a total of 3299 patients were referred by cardiologists for pharmacologic or exercise stress t h a l l i u m studies; 1341 (41%) h a d n o r m a l exercise results a n d 1542 (47%) h a d normal pharmacologic results. All subjects gave informed consent before undergoing t r e a d m i l l - t h a l l i u m testing. Clinical data. Before each treadmill test, a structured interview and chart review yielded d a t a on symptoms, medications, coronary risk factors, prior cardiac events, and a n u m b e r of cardiac and noncardiac diagnoses. All these data, as well as stress test data, were evaluated prospectively. Resting hypertension was defined as a resting systolic blood pressure ->140 m m Hg, a resting diastolic blood pressure ->90 m m Hg, or t r e a t m e n t with antihypertensive medications. 3 A s s e s s m e n t of diabetes was based on chart review, questioning, and medication use. Prior coronary events consisted p r i m a r i l y of documented myocardial infarction or hospitalization for unstable angina pectoris. Exercise testing. Treadmill testing was carried out according to s t a n d a r d protocols, u s u a l l y the Bruce and modified Bruce protocols. 4, 5 Leaning on h a n d r a i l s during exercise was not allowed. During each stage of exercise, d a t a on symptoms, r h y t h m , h e a r t rate, blood pressure, workload in metabolic equivalents (METs), and ST segm e n t s were collected. Exercise capacity in METs was est i m a t e d from s t a n d a r d published tables. 5 In subjects with
i n t e r p r e t a b l e ST segments (using all leads except a VR to a precision level of 0.5 mm), an ischemic response was considered p r e s e n t if-> 1 m m of horizontal or downsloping ST segment depression occurred 80 msec after the J point, or if-> 1 m m of additional ST segment elevation occurred in leads without pathologic Q waves. Thallium scintigraphy. P a t i e n t s were injected with 2 to 3 pCi of thallium-201 1 m i n u t e before t e r m i n a t i o n of the treadmill test. W i t h i n 10 minutes of stress testing, single positron emission computed tomography (SPECT) imaging was performed over 360 degrees of acquisition with a low-energy, all-purpose collimator with 32 stops, 1.5 to 4 million counts, over 9 to 14 minutes. P a t i e n t s r e t u r n e d 3 to 4 hours after the stress test for redistribution SPECT t h a l l i u m imaging. Images were reconstructed with a H a m m i n g filter and projected in short-axis, vertical, and horizontal long-axis views in a side-by-side display. Ischem i a was i n t e r p r e t e d as the presence of >20% reversibility and scarring was i n t e r p r e t e d as the presence of counts <80% of m a x i m u m (<70% for the posterior wall). Clinical reports of all t h a l l i u m studies were retrieved a n d coded according to a 12-segment system. 6 Each segm e n t was coded as normal, ischemic, or scarred; further coding took into account abnormalities of specific coronary a r t e r y territories (namely, left anterior descending, left circumflex, and right coronary/dominant left posterior descending). The coding of the t h a l l i u m d a t a was blinded with r e g a r d to age a n d other clinical data. Coronary angiography. For subjects who u n d e r w e n t coronary angiography within 90 days of stress testing, coronary angiograms were i n t e r p r e t e d semiqualitatively. Coronary disease was defined as ->50% stenosis in any proximal or middle coronary vessel or major branch. Severe coronary disease was defined as ->50% left m a i n
Volume 133, Number 2 American Heart Journal
Lau8F et al.
141
Table I. Baseline characteristics according to age Characteristic
Aged 30-49 y r (n = 59)
Aged 50-64 y r (n = 137)
Aged 65-74 y r (n = 169)
Aged >74 y r (n = 51)
p Value
Age (yr) Female (%) Systolic blood pressure (ram Hg) Diastolic blood pressure (ram Hg) Resting heart rate (beats/rain) Prior coronary events (%) Smoker (%) Hypertension (%) B-blocker use (%) Calcium blocker use (%) Lipid lowering therapy (%) Chest pain history (%) Diabetes (%)
43 +_5 6 (10%) 130 + 16 88 -+ 11 77 -+ 12 14 (24%) 20 (34%) 14 (24%) 7 (12%) 10 (17%) 5 (8%) 32 (54%) 9 (15%)
58 -+ 4 18 (13%) 137 +_ 18 88 -+ 10 79 -+ 14 55 (40%) 30 (22%) 66 (48%) 34 (25%) 40 (29%) 11 (8%) 71 (52%) 23 (17%)
69 -+ 3 27 (16%) 144 _+23 86 -+ 11 76 -+ 13 64 (38%) 24 (14%) 86 (51%) 19 (11%) 50 (30%) 22 (13%) 73 (43%) 41 (24%)
78 -+ 3 11 (22%) 141 _+ 20 86 _+ 12 77 + 14 22 (43%) 0 (-) 28 (55%) 6 (12%) 14 (27%) 3 (6%) 21 (41%) 8 (16%)
<0.0001 >0.3 <0.0001 >0.2 >0.3 >0.1 <0.001 0.002 0.006 >0.2 >0.3 >0.7 >0.2
V a l u e s are m e a n ± S D or n u m b e r s (percentages).
a r t e r y stenosis, as three-vessel disease (->70% stenosis in each major coronary a r t e r y system), or as two-vessel disease with a ->70% proximal left anterior descending a r t e r y lesion. Statistical analyses. Subjects were divided into four prespecified age groups, 30 to 49, 50 to 64, 65 to 74, a n d ->75 years. Baseline clinical and exercise continuous variables were compared using the Kruskal-Wallis test, and categoric clinical, exercise, and angiographic variables were compared with the chi-square test or the F i s h e r exact test. Associations between tha]linm results and age as a continuous variable were tested by logistic regression. The p r i m a r y end point of this s t u d y was performance of coronary angiography within 90 days of the treadmill t h a l l i u m test. In u n a d j u s t e d analyses, coronary angiography r a t e s were compared in the four age groups using the Mantel-Haenszel extension test for trend. 7 F o r descriptive purposes, r a t e s of angiography were calculated for predefined subgroups, including those with a n y ischemia revealed by thallium, ischemia noted in the t e r r i t o r y of the left anterior descending artery, and patients with -<2, 3 to 4, and >-5 total a b n o r m a l (whether ischemic or fixed) segments. The inclusion of both ischemic a n d fixed defects allowed for the possibility t h a t physicians m a y refer patients with a n g i n a and extensive fLxed defects to angiography, especially because reinjection was not routinely performed, m a k i n g the differentiation between ischemia a n d scar problematic. The association of age as a continuous variable with referral to coronary angiography was analyzed by multivari: ate logistic regression analyses. Initial models adjusted for potential confounding variables including sex, resting blood pressure, medications, hypercholesterolemia t r e a t e d with lipid-lowering medications, smoking, diabetes, prior coronary events, exercise capacity in METs, h e a r t r a t e a n d blood p r e s s u r e responses to exercise, occurrence of exercise-induced ventricular tachycardia, a n d extent of ischem i a revealed by thallium. Abnormalities revealed by thallium were analyzed four ways: (1) a n y ischemia revealed by thallium, (2) the n u m b e r of ischemic coronary a r t e r y
segments, (3) the presence of ischemia in the left anterior descending a r t e r y distribution, and (4) the total n u m b e r of a b n o r m a l (whether ischemic or fLxed) a b n o r m a l segments. Because models involving these four different t h a l l i u m variables yielde d essentially identical results, only the results of models considering the presence of a n y ischemia revealed by t h a l l i u m are presented. Logistic analyses were also performed to assess for potential interactions between baseline variables and age; no significant interactions were found. In s u p p l e m e n t a r y analyses, only subjects aged ->65 y e a r s were considered; all of these patients can be considered elderly by Medicare criteria, yet we could differentiate between "young" old subjects (aged 65 to 74 years), who h a d s u b s t a n t i a l l y better n a t u r a l histories, and "old" old subjects (aged >74 years). To investigate the potential clinical significance of differentia] rates of referral to coronary angiography as a function of age, 2-year all-cause m o r t a l i t y rates were d e t e r m i n e d in each of the age groups by using the Social Security A d m i n i s t r a t i o n D e a t h M a s t e r File. s D e a t h certificates were reviewed to determine w h e t h e r the cause of d e a t h was cardiac. Cumulative m o r t a l i t y r a t e s were calculated by use of the Kaplan-Meier method with Greenwood 95% confidence intervals; differences in m o r t a l i t y r a t e s were tested with the log r a n k chi-square statistic. The Cox proportional h a z a r d s model 9 was used to e s t i m a t e relative risks for deaths according to age, abnormalities revealed by thallium, and performance of coronary angiography. All analyses were performed with use of the SAS 6.08 statistical package (SAS Institute, Cary, N.C.).
RESULTS Baseline characteristics. O f 416 e l i g i b l e s u b j e c t s , 59 w e r e a g e d 30 to 49 y e a r s , 137 w e r e a g e d 50 to 64 y e a r s , 169 w e r e a g e d 65 to 74 y e a r s , a n d 51 w e r e ->75 y e a r s old. T h e p r i m a r y i n d i c a t i o n s for t r e a d m i l l thallium testing were "rule out coronary disease"
February 1997 American Heart Journal
142 Lauer et al. T a b l e II. E x e r c i s e c h a r a c t e r i s t i c s
Aged 30-49 yr Aged 50-64 y r (n = 59) (n = 137)
Characteristic
Protocol (%) Bruce Modified Bruce CAEP Other Peak METs Peak systolic blood pressure (mm Hg) Peak heart rate (beats/min) Maximum rating of perceived exertion (0-10) Ischemic ST segments* (%) Ventricular tachycardia (%) Anginal chest pain (%) Failure to reach 85% of maximum age-predicted heart rate
Aged >74 yr (n = 51)
1 (2%) 15 (29%) 16 (31%) 19 (38%) 5.0 _+ 1.6 179 _+ 26 126 _+ 19 8.0 ± 2.7 6 (35%) 2 (4%) 6 (12%) 17 (33%)
43 (73%) 10 (17%) 1 (2%) 5 (8%) 10.0 _+ 2.2 183 ± 27 166 _+ 16 9.2 _+ 1.4 11 (26%) 1 (2%) 5 (8%) 8 (14%)
46 (34%) 57 (42%) 7 (5%) 27 (19%) 8.0 + 2.3 185 ± 31 148 ± 21 8.7 _+ 1.9 40 (41%) 2 (2%) 36 (26%) 33 (24%)
17 (4%) 99 (59%) 19 (11%) 34 (26%) 6.5 _+ 2.0 182 _+ 31 135 _+ 20 8.2 _+ 2.4 51 (54%) 3 (2%) 43 (25%) 52 (31%)
1 (2%)
7 (5%)
8 (5%)
(%) Hypotensive response (%)
Aged 65-74 yr (n = 169)
0 (-)
p Value
<0.001
<0.0001 >0.5 <0.0001 0.06 0.02 >0.7 0.007 0.04 >0.2
Values are mean -+ SD or numbers (percentages). CAEP, Chronotropic assessment exercise protocol;MET, metabolic equivalents. *Of subjects with interpretable ST segments.
T a b l e II1. T h a l l i u m s c i n t i g r a p h y c h a r a c t e r i s t i c s
Characteristic
Aged 30-49 y r (n = 59)
Any ischemia (%) Ischemia in the LAD territory (%) Number of ischemic segments Number of scarred segments Number of abnormal segments
35 (59%) 16 (27%) 1.3 ± 1.7 1.6 -+ 1.8 2.9 _+ 1.9
Aged 50-64 y r (n = 137)
69 (50%) ' 37 (27%) 1.0 -+ 1.4 2.4 ± 2.2 3.4 _+ 2.1
Aged 65-74 yr (n = 169)
Aged >74 yr (n = 51)
p Value
76 (45%) 32 (19%) 1.0 ± 1.4 2.6 -~ 2.3 3.5 _+ 2.1
20 (39%) 12 (24%) 0.9 -+ 1.4 2.8 -+ 2.4 3.7 _+ 2.5
>0.1 >0.3 >0.2 0.02 >0.1
Values are mean +- SD or numbers (percentages). LAD, Left anterior descending artery.
(58%) and "follow up coronary disease" (33%). Other baseline characteristics are summarized in Table I. Older subjects were less likely to be smokers and more likely to be taking ~-blockers and have a history of hypertension. Exercise and thallium characteristics. Older subjects were more likely to be assigned to the modified Bruce protocol (p < 0.001), to have a lower exercise capacity (p < 0.0001), and to have a lower peak heart rate (p < 0.0001) (Table II). The most common reason for terminating exercise was fatigue (93% for subjects aged 30 to 49 years, 82% for those aged 50 to 64 years, 80% for those aged 64 to 74 years, and 90% for those aged ->75 years, p = 0.06). No trends were noted for age-related differences in other reasons for exercise termination, including marked ST segment depression, anginal chest discomfort on the treadmill, claudication, blood pressure changes, and arrhythmias. We found a statistically nonsignificant trend toward older subjects being less likely to have ischemia revealed by thallium (Table III). When treated as a
continuous variable, increasing age was associated with a lesser likelihood of any ischemia revealed by thallium (for 5-year increase in age, odds ratio [OR] = 0.88, 95% confidence interval [CI] 0.80 to 0.96, p = 0.005). Age was not associated with the frequency of ischemia revealed by thallium in multiple coronary artery territories or in the territory of the left anterior descending artery, even when treated as a continuous variable. Older subjects had a greater number of scarred segments (that is, fixed defects) (p = 0.02), however, and the total number of abnormal segments (fixed or ischemic) tended to increase in older patients. Age and referral to coronary angiography: univariate analyses. During 90 days of follow-up, 163 (39%) subjects were referred for coronary angiography. Coronary angiography was performed in 46% of subjects aged 30 to 49 years, in 53% of those aged 50 to 64 years, in 33% of those aged 65 to 74 years, and in only 18% of those aged ->75 years (Mantel-Haenszel extension test for trend, ×2 = 17.7, p < 0.0001) (Fig.
Volume 133, Number 2 American Heart Journal
LGUeT et al.
143
o
~ >e-, ft •
e-
.o<
LAD Ischemia (P=0.0006) I'hal Ischemia (P=0.0002) Subjects (P<0.0001)
0.
b~)-f4
Age (years)
>74
Fig. 2. Association of age with rate of referral to coronary angiography according to results of the thallium scan. LAD, Left anterior descending artery; thal ischemia, thallium-documented ischemia.
~ >e-,
"O
=-
Q.
"E'5) • e..
o< O
O.
> 4 segments (P=0.003) 3-4 Segments (P=0.03) segments (P=0.03) b:)-IR.
Age (years)
>74
Fig. 3. Association of age with rate of referral to coronary angiography according to the total number of abnormal (transient or fLxed)segments on the thallium scan. 2). Among subjects with any ischemia revealed by thallium (n = 200), the corresponding proportions were 69%, 72%, 49%, and 25% (Mantel-Haenszel extension test for trend, ×2 = 14.8, p = 0.0002). Among subjects with evidence ofischemia in the left anterior descending artery distribution, referral to coronary angiography occurred in 13 of 16 (81%) subjects aged 30 to 49 years, in 29 of 37 (78%) subjects aged 50 to 64 years, in 16 of 32 (50%) subjects aged 65 to 74 years, and in only 4 of 12 subjects (33%) aged ->75 years (Mantel-Haenszel extension test for trend, X 2 = 11.5, p = 0.0006). Patients were also analyzed according to the total number of abnormal segments (whether ischemic or fLxed defects) imaged by thallium (Fig. 3). In all subgroups, older subjects were less likely to be referred to coronary angiography; this was particularly marked among the 117 patients with ->5 abnormal
segments (Mantel-Haenszel extension test for trend, p = 0.003). Angiographic characteristics. The angiographic results noted among the 163 subjects who underwent angiography are summarized in Table IV. Age was associated with a higher prevalence of any coronary disease (p < 0.001); also, older subjects had higher prevalences of severe disease (p = 0.007), left main arterial disease (p = 0.002 by Fisher exact test), and severe disease of the proximal left anterior descending artery (p = 0.005 by Fisher exact test). Age and referral to coronary angiography multivariate analyses. After adjustment for the extent ofischemia revealed by thallium, clinical characteristics, and exercise findings including functional capacity, increasing age remained associated with a lower rate of referral to coronary angiography (for 5-year inc r e a s e in age, adjusted OR = 0.81, 95% CI 0.73 to X2 = 8 . 8 1 ,
February 1997
144
AmericanHeartJournal
L a u e r et al.
Table IV.
Angiographic characteristics of subjects who underwent cardiac catheterization Characteristic
Aged 30-49 y r (n = 27)
Aged 50-64 yr (n = 72)
Aged 65-74 yr (n = 55)
Aged >74 y r (n = 9)
p Value
Any coronarydisease (%) Severe coronarydisease (%) Left main arterial disease (%) Proximal LAD disease (%)
7 (26%) 5 (19%) 3 (11%) 1 (4%)
55 (76%) 22 (31%) 5 (7%) 19 (26%)
44 (80%) 23 (42%) 11 (20%) 21 (38%)
8 (89%) 7 (78%) 5 (56%) 2 (22%)
<0.001 . 0.007 0.002 0.005
LAD, Left anterior descending artery/
0.90,p < 0.0001). Other predictors for referral to coronary angiography included presence of any ischemia revealed by thallium (adjusted OR = 4.66, 95% CI 2.93 to 7.41,p < 0.0001) and anginal chest pain on the treadmill (adjusted OR = 4.62, 95% CI 2.65 to 8.07, p < 0.0001). Referral to coronary angiography was associated with ischemia revealed by thallium in the subjects aged 50 to 64 years (adjusted OR = 6.61, 95% CI 2.96 to 14.70, p < 0.0001) and in those aged 65 to 74 years (adjusted OR = 3.46, 95% CI 1.83 to 8.55, p = 0.0007) and with anginal chest pain on the treadmill in these groups (aged 50 to 64 years, adjusted OR = 4.96, 95% CI 1.85 to 13.10,p = 0.001; aged 65 to 74 years, adjusted OR = 3.96, 95% CI 1.69 to 7.06,p = 0.0005). Among patients aged >74 years, referral to coronary angiography tended to be associated with anginal chest pain on the treadmill (adjusted OR = 7.26, 95% CI 0.88 to 59.79,p = 0.07), but was not associated with ischemia revealed by thallium (adjusted OR--1.33, 95% CI 0.26 to 6.99, p > 0.7). When analyses were confined to those subjects aged ->65 years, age remained associated with a lower rate of referral to coronary angiography (for each 5-year increase in age, adjusted OR = 0.64, 95% CI 0.44 to 0.95,p = 0.02). In none of these models was sex predictive of referral to angiography. All the covariates considered were tested in other models to look for potential effect modification (interaction) on the association of age with a lower rate of referral to coronary angiography. No significant interactions were noted. Mortality analyses. During approximately 2 years of follow-up, 34 deaths occurred. Cumulative mortality rates were relatively low among subjects aged <65 years, but increased m a r kedl y among those aged 65 to 74 years (2-year mortality rate = 9%, 95% CI 4% to 14%) and those aged >74 years (2-year mortality rate = 31%, 95% CI 16% to 47%) (log r a n k X2 = 33, degrees of freedom = 3,p < 0.0001; relative risk [RR] for 5-year increase in age = 1.77, 95% CI 1.41 to 2.22, p < 0.0001) (Fig. 4). All-cause mortality rates were also associated with the total number of abnormal segments on thallium scan (for each two additional abnormal segments, age-adjusted RR = 1.41, 95% CI
1.06 to 1.88, p = 0.02), but not with referral to coronary angiography (RR = 0.73, 95% CI 0.36 to 1.50, p > 0.3). Of the 34 deaths, 14 were of definite cardiac origin. Increased age was also associated with cardiac death (RR for 5-year increase in age -- 1.41, 95% CI 1.03 to 1.93, p = 0.03). Cardiac death was also associated with the total num ber of abnormal segments on thallium scan (for each two additional abnormal segments, RR = 1.60, 95% CI 1.03 to 2.48, p = 0.04), but it was not associated with referral to coronary angiography (RR = 1.14, 95% CI 0.40 to 3.30,p > 0.8). DISCUSSION Main findings. In this sample of 416 consecutive adults with an abnormal treadmill-thallium scan who were aged ->30 years and under the care of cardiologists, increasing age was associated with a lower rate of referral to coronary angiography. This association existed even after adjusting for sex, standard risk factors, medication use, functional capacify and other exercise parameters, anginal chest pain, and extent of ischemia revealed by thallium. None of the patients had previous coronary angiography and none had a history of congestive h e a rt failure. Even when analyses were confined to those aged ->65 years (enabling comparison between "young" old patients with relatively low mortality rates and "old" oM patients with substantially higher mortality rates), increasing age remained associated with a lower rate of referral to coronary angiography. Analyses were confined to those adults who were considered ambulatory enough by their managing physicians to be referred to treadmill-thallium testing, as opposed to pharmacologic testing; the latter would be more likely performed in subjects with other significant limiting conditions such as osteoarthritis, chronic lung disease, renal failure, and complicated diabetes, which might significantly impact on any decision regarding coronary angiography. Among the patients who were referred to angiography, older patients had substantially higher rates of significant coronary disease, severe disease, and left main arterial disease; all of these findings are
Volume 133, Number 2 AmericanHeartJournal
L a u e r et al.
1.00 0.90
>
"" ~
2-year survival (95% Cl): 1.oo (-) •
Age:
0.53 0.84)
>74
145
30-49 0.95 (0.91-0.99)D 50-64 (o.8s-o.gs)~ 65-74
=--~0.9'1
0.80
¢n 0.70 0.60 0.50 0.00
i 0.50
i 1.00
F 1.50
i 2.00
Years After Exercise Test
Fig. 4. Kaplan-Meier plot relating age to 2-year cumulative all-cause mortality rates. Note the marked increase in deaths among subjects aged >74 years. consistent with a higher threshold for referral to angiography among older patients. Furthermore, not surprisingly, older patients had substantially higher mortality rates during 2 years of follow-up. Exercise thallium testing is recognized as a feasible, safe, and prognostically useful modality for evaluating older adults with known or suspected coronary disease. 1°12 Coronary angiography and percutaneous interventions can also be safely performed in elderly patients, albeit with somewhat higher complication rates. 13-15 In view of these observations and recent findings t h a t myocardial revascularization outcomes are improving in elderly 1 and very elderly 16 subjects, the inverse association between age and performance of coronary angiography after an abnormal treadmill-thallium scan may be of importance. The lower rate of referral to coronary angiography among older subjects may be a result of lower peak heart rates, with lower associated rates of thalliumdocumentedischemia. The association persisted even after accounting for the thallium results in logistic regression models. However, given the semiqualitative description of the thallium results, a more powerful confounding effect may be operative. Nonetheless, when subjects were stratified according to thallium test results, similar trends were noted. Previous studies. Few data exist regarding any association between age and performance of coronary angiography. One study by Krumholz et al. 16 of gender bias in a teaching hospital found that age was associated with a lower rate of referral to coronary
angiography among survivors of acute myocardial infarction. 2 In a recent report 17 on the use of resources in patients after acute myocardial infarction fee-for-service or health maintenance organization care settings, age was also associated with a lower rate of performing coronary angiography, with an approximate odds ratio for every 10-year increase in age of 0.6. Another report is noted t h a t older patients referred for coronary angiography were more likely to have more severe symptoms, be receiving more aggressive medical therapy, and have more severe angiographic disease t h a n younger subjects. Potential limitations. This is an observational study, with the principal limitation being the possibility of missed confounders. Data on physician age and on patients' responses to their physicians' recommendations were not available. Other important factors for which we did not have information include patient priorities (especially as a trend existed toward less chest pain in older subjects) regarding the benefits and risks of aggTessive revascularization, cognition, affective state, marital status, financial concerns, and familial support. Physician attitudes toward performance of potentially high-risk procedures in elderly patients were not measured and may have played a role. One could argue t h a t these data are unique to patients managed by cardiologists in one tertiary referral center who are referred for evaluation in the exercise laboratory. Older patients may be more likely to be referred to angiography via other routes. In this regard, the data reported by Krumholz et al. 2
146
February 1997 American Heart Journal
L a u e r et al.
are particularly interesting. In that study, patients at a different institution were referred to coronary angiography after sustaining a myocardial infarction and being admitted to the coronary care unit. Consistent with our results, increasing age was associated with a markedly decreased use of coronary angiography. Analogously, in a very recent study, Every et al. 17 found that age was associated with a lower rate of coronary angiography and coronary bypass surgery in patients being treated for acute myocardial infarction in a spectrum of different hospitals. Furthermore; our results parallel those of others in which use of treatments for coronary artery disease have been noted to be underused in elderly patients. For example, Meehan et al.19 studied a Medicare cohort of patients treated for acute myocardial infarction in six different Connecticut hospitals and found that thrombolytic therapy was used in only 43% of eligible patients, B-blockers were used in only 41%, and aspirin was used in only 73%. In view of these observations, and the observations of Krumholz et al. 2 and Every et al., 17 our observations are not unique to one institution or one route of selecting patients for coronary angiography. Nonetheless, w e must emphasize that this population is a relatively selective one of patients who could perform treadmill exercise testing and who had not had prior coronary angiography or congestive heart failure. Whether the inverse association of age with referral to co]:onary angiography represents underuse of coronary angiography in the elderly or appropriate care i s not clear. The poor exercise capacity of the elderly subjects in the current study suggests that these patients are more functionally impaired, which would increase their operative mortality rate. We do not have data on outcomes among patients managed differently (for example, with direct angiography without prior stress testing) or to natural history data from the same age groups. Despite these potential limitations, the association between age and a lower rate of performance of coronary angiography in our cohort was quite marked. Given the increasing use of myocardial revascularization, these findings should strike great interest. Further studies are necessary to confirm these results and to determine whether the lower rate of referral to coronary angiography represents
appropriate care, reflecting the priorities and prognoses of older patients, or whether an opportunity exists to improve clinical outcomes. REFERENCES
1. Peterson ED, Jollis KG, Bebchuk JD, DeLong ER, Muhlbaier LH, Mark DB, et al. Changes in mortality after myocardial revascularization in the elderly: the national Medicare experience. Ann Intern Med 1994; 121:919-27. 2. Krumholz HM, Douglas PS, Lauer MS, PasternakRC. Selectionofpatients for coronary angiography early after myocardial infarction: is there evidence for a gender bias? Ann Intern Med 1992;116:785-90. 3. The fifth report of the Joint National Committee on detection, evaluation, and treatment of high blood pressure (JNC V). Arch Intern Med 1993;153:154-83. 4. Dean AE, Peterson DR, Blackmon JR, Bruce RA. Myocardial ischemia "after maximal exercise in healthy men. Am Heart J 1965;69:11-25. 5. Fletcher GF, Froelicher VF, Hartley LH, Haskell WL, Pollock ML. Exercise standards: a statement for health professionals from the American Heart Association. Circulation 1990;82:1-36. 6. Zaret BL, Wackers FJ. Nuclear cardiology. N Engl J Med 1993;329:77583. 7. Mantel N. Chi-square tests with one degree of freedom: extensions of the Mantel-Haenszel procedure. J Am Stat Assoc 1963;58:690-700. 8. Curb JD, Ford CE, Fressel S, Palmer M, Babcock C, Hawkins CM. Ascertainment of vital status through the National Death Index and the Social Security Administration. Am J Epidemiol 1985;121:754-68. 9. Cox DR. Regression models and life tables (with discussion). J R Stat Soc Br 1972;34:187-220. 10. Ades PA, Grunvald MH. Cardiopulmonary exercise testing before and after conditioning in older coronary patients. Am Heart J 1990; 120:585-9. 11. Hilton TC, Shaw LJ, Chaitman BR, Stocke KS, Goodgold HM, Miller DD. Prognostic significance of exercise thallium-201 testing in patients aged greater than or equal to 70 with known or suspected coronary artery disease. Am J Cardiol 1992;69:45-50. 12. Thompson RF, Crist DM, Osborn LA, Atterbom HA. Treadmill exercise electrocardiography in the elderly with physical impairments. Gerontology i990;36:112-8. 13. Rich MW, Crecelins CA. Incidence, risk factors, and clinical course of acute renal insufficiency after cardiac catheterization in patients 70 years of age or older. Arch Intern Med 1990;150:1237-42. 14. Muller DWM, Shamir KJ, Ellis SG, Topol EJ. Peripheral vascular complications after conventional and complex percutaneous coronary interventional procedures. Am J Cardiol 1992;69:63-8. 15. Clark VL, Dolce J. Unplanned admissions after outpatient cardiac catheterization. Clin Cardiol 1993;16:823-6. 16. Krumholz HM, Forman DE, Kuntz RE, Balm DS, Wei JY. Coronary revascularization after myocardial infarction in the very elderly: outcomes and long-term follow-up. Ann Intern Med 1993;119:108490. 17. Every NR, Fihn SD, Maynard C, Martin JS~ Weaver D for the MITI investigators. Resource utilization in treatment of acute myocardial infarction: staff-model health maintenance organization versus fee-forseryice hospitals, J Am Coll Cardiol 1995;26:401-6. 18. Elder AT, Shaw TRD, Turnbull CM, Starkey IR. Elderly and younger patients selected to undergo coronary angiography. Br Med J 1991; 303:950-3. 19. Meehan TP, Hennen J, Radf°rd MJ, Petrill° MK, Elstein P, Ballard DJ" Process and outcome of care for acute myocardial infarction among Medicare beneficiaries in Connecticut: a quality improvement demonstration project. Ann Intern Med 1995;122:928-36.