The impact of co-morbidity on the cardiovascular risk after single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) remains unclear. We examined the association between a normal versus abnormal SPECT MPI scan on 10-year risk of myocardial infarction, stroke, and all-cause death, overall and according to co-morbidity level. We identified all patients without previous myocardial infarction or cerebrovascular disease, who had an SPECT MPI performed at Aarhus University Hospital Skejby during 1999 to 2011. We categorized the SPECT MPI scan as normal (no defects) or abnormal (reversible and/or fixed defects). Using nationwide medical registries, we obtained information on co-morbidity level (using Charlson co-morbidity index) and outcomes. We used Cox regression to compute hazard ratios with 95% confidence intervals (CIs), adjusting for gender, age, and co-morbidity level. Among 7,040 patients, 4,962 (70%) had normal scans and 2,078 (30%) abnormal scans. Patients with a normal versus abnormal scan had a 10-year risk of 5.7% versus 10.9% for myocardial infarction, 6.0% versus 7.8% for stroke, and 16.5% versus 29.0% for all-cause death. After adjustment, an abnormal scan was associated with increased risk of myocardial infarction (adjusted hazard ratio 1.73, 95% CI 1.37 to 2.18) and all-cause death (1.42, 95% CI 1.23 to 1.65) but not stroke (1.12, 95% CI 0.86 to 1.45). Co-morbidity level did not affect substantially the association between the scan result and the outcomes. In conclusion, independently of co-morbidity level, an abnormal SPECT MPI scan was associated with an increased 10-year risk of myocardial infarction and all-cause death but not stroke.
Single-photon emission computed tomography myocardial perfusion imaging (SPECT MPI) is a noninvasive technique used to assess myocardial perfusion and thereby detect coronary artery disease. The age distribution of the population is changing with a growing proportion of the elderly. Increased life expectancy and sedentary lifestyles increase the prevalence of chronic medical conditions. Currently, 45% of the adult population and 90% of subjects aged >65 years in the United States have at least 1 chronic condition. Because the greatest risk of coronary artery disease exists among the elderly, an increasing number of patients suspected with coronary artery disease also have other coexisting diseases, that is, co-morbidities. As coronary artery disease shares risk factors with many chronic diseases, such as diabetes, chronic pulmonary disease, and obesity, the prevalence of co-morbidity further increases among patients with coronary artery disease. Although the risk of future coronary events after SPECT MPI has been shown previously, no studies have examined the impact of co-morbidity level. We therefore conducted a cohort study to examine the long-term risk of myocardial infarction, stroke, and all-cause death associated with a normal versus abnormal SPECT MPI scan in patients without previous myocardial infarction or cerebrovascular disease and to assess the impact of co-morbidity level.
Methods
The Danish National Health Service provides free, universal, tax-supported health care, guaranteeing all residents unrestricted access to general practitioners and hospitals. Unambiguous individual-level linkage between medical databases was performed using the 10-digit Danish civil personal registration number assigned to each Danish citizen at birth and to residents on immigration.
We conducted this cohort study using the Myocardial Perfusion Imaging Database at Aarhus University Hospital Skejby. Since January 1, 1999, the Myocardial Perfusion Imaging Database has collected information on all SPECT MPI procedures performed at Aarhus University Hospital Skejby. We used this database to identify all adult Danish citizens (aged ≥18 years) without previous myocardial infarction or cerebrovascular disease, who had a technetium-99m sestamibi SPECT MPI performed from January 1, 1999 through last date of registration to the database on April 26, 2011. For each patient, we identified the first SPECT MPI in the study period. The database includes the following variables: gender, age, and scan result. A normal scan was defined as a scan without defects, whereas an abnormal scan was defined by the presence of a reversible and/or fixed defect.
We obtained information on co-morbid conditions from inpatient and outpatient clinic hospital diagnoses recorded in the Danish National Patient Registry in the 10 years preceding the SPECT MPI procedure. The Danish National Patient Registry includes data on dates of admission and discharge from all nonpsychiatric hospital admissions in Denmark since 1977 and from emergency room and outpatient clinical visits since 1995. One primary diagnosis and up to several secondary diagnoses are provided for each hospital contact, classified according to the International Classification of Diseases, Eighth Revision , until the end of 1993 and International Classification of Diseases, Tenth Revision , thereafter. We assessed co-morbidity level by means of the widely used Charlson co-morbidity index, which was originally developed to predict 1-year mortality in hospitalized medical patients. This scoring system assigns between 1 and 6 points to 19 conditions (1 point for myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, connective tissue disease, ulcer disease, mild liver disease, and diabetes without end organ damage; 2 points for hemiplegia, moderate to severe renal disease, diabetes with end organ damage, nonmetastatic solid tumor, leukemia, and lymphoma; 3 points for moderate to severe liver disease; and 6 points for metastatic cancer and AIDS). The sum of points serves as a measure of the burden of co-morbidity. We computed the total Charlson score for each patient and then categorized the study population according to co-morbidity burden: score of 0 (normal), score of 1 (moderate), and score ≥2 (severe). Because we excluded all patients with myocardial infarction and cerebrovascular disease, these diseases did not contribute to the scoring.
We used the Danish National Patient Registry to identify all first-time inpatient admissions for myocardial infarction and stroke (ischemic or hemorrhagic stroke) after an SPECT MPI procedure during the study period. First-time admissions were defined as none of these diagnoses back to 1977.
We obtained information on mortality from the Danish Civil Registration System. This registry contains data on date of birth, residence, date of emigration, and exact date of any death for the entire Danish population since 1968, with daily electronic updates.
We followed up all patients from the date of outpatient SPECT MPI procedure or discharge from the admission during which the SPECT MPI procedure was performed (2 patients died during admission and were excluded) until the date of an outcome, death, emigration, 10 years of follow-up, or December 31, 2011, whichever came first. We characterized patients according to gender, age, and co-morbidity.
We used a cumulative incidence method to compute 10-year risk of myocardial infarction, stroke, and all-cause death, overall and according to co-morbidity level, and illustrated graphically the 10-year cumulative incidence function of each outcome. For myocardial infarction and stroke, we used the proportional subhazards model by Fine and Gray. For all-cause mortality, we used the Kaplan-Meier estimator. Death was considered a competing risk in all analyses of nonfatal outcomes. We computed the risk for each outcome within 31 to 365 days and 1 to 10 years of follow-up by co-morbidity levels. We used Cox proportional hazards regression to compute hazard ratios (HRs) with 95% confidence intervals (CIs) for each outcome, overall and stratified on co-morbidity levels, the 2 most prevalent individual Charlson co-morbidities in the study population (diabetes and chronic pulmonary disease), age group, and gender. We adjusted for categories of gender, age (18 to 49, 50 to 59, 60 to 69, and ≥70 years), and co-morbidity level.
We performed 2 sensitivity analyses to examine the influence on the estimators of (1) including only myocardial infarction and stroke diagnoses registered during an acute admission and (2) including all available information on co-morbid conditions in the Danish National Patient Registry (back to its establishment in 1977).
The assumption of proportional hazards was assessed graphically and was found to be appropriate. All statistical analyses were performed using Stata software, version 12.1 (StataCorp LP, College Station, Texas).
The study was approved by the Danish Data Protection Agency (record number: 2011-41-6772).
Results
Patient characteristics are described in Table 1 . We identified 7,040 patients without previous myocardial infarction or cerebrovascular disease and with a first-time SPECT MPI in the study period. Among the patients, 4,962 (70%) had a normal scan and 2,078 (30%) an abnormal scan ( Table 1 ). A total of 25 patients (0.4%) emigrated. Age and gender are described in Table 1 .
| Characteristic | Normal Scan (n = 4,962) | Abnormal Scan (n = 2,078) | Total (n = 7,040) |
|---|---|---|---|
| Gender | |||
| Female | 3,115 (63%) | 844 (41%) | 3,959 (56%) |
| Male | 1,847 (37%) | 1,234 (59%) | 3,081 (44%) |
| Age (years) | |||
| 18−49 | 962 (19%) | 264 (13%) | 1,226 (17%) |
| 50−59 | 1,461 (30%) | 512 (25%) | 1,973 (28%) |
| 60−69 | 1,481 (30%) | 670 (32%) | 2,151 (31%) |
| ≥70 | 1,058 (21%) | 632 (30%) | 1,690 (24%) |
| Median age (years) | 60 | 64 | 61 |
| Co-morbidity level ∗ | |||
| Normal | 3,216 (65%) | 1,096 (53%) | 4,312 (61%) |
| Moderate | 980 (20%) | 487 (23%) | 1,467 (21%) |
| Severe | 766 (15%) | 495 (24%) | 1,261 (18%) |
| Frequent individual co-morbidities | |||
| Diabetes | 480 (10%) | 310 (15%) | 790 (11%) |
| Chronic pulmonary disease | 552 (11%) | 291 (14%) | 843 (12%) |
∗ Levels of co-morbidity were based on Charlson co-morbidity index scores as follows: 0 (normal), 1 (moderate), and ≥2 (severe).
Patients with normal scans were more likely to have low co-morbidity compared with patients with abnormal scans and less likely to have moderate and severe co-morbidity. Also, they were less likely to have diabetes and chronic pulmonary disease ( Table 1 ).
We identified 314 first-time hospitalizations for myocardial infarction, 261 first-time hospitalizations for stroke, and 756 deaths within a maximum of 10 years of follow-up. The median follow-up time was 5.0 years (interquartile range 2.5 to 7.9 years) for myocardial infarction, 5.1 years (interquartile range 2.5 to 8.0 years) for stroke, and 5.2 years (interquartile range 2.6 to 8.2 years) for all-cause death. The overall 10-year risk of myocardial infarction was 5.7% (95% CI 4.7 to 6.8) in patients with a normal scan and 10.9% (95% CI 9.3 to 12.7) in patients with an abnormal scan ( Table 2 ). The cumulative incidence of myocardial infarction throughout the follow-up period is presented in Figure 1 . The corresponding adjusted HR during this 10-year period comparing an abnormal scan with a normal scan was 1.73 (95% CI 1.37 to 2.18; Table 2 ). When stratified on age group and gender, we found that the association between scan result and myocardial infarction was strongest in the youngest age group and among women (not shown).
| Outcome | Scan Result | No. of Events | Risk % (95% CI) | Hazard Ratio (95% CI) | |
|---|---|---|---|---|---|
| Unadjusted | Adjusted ∗ | ||||
| Myocardial infarction | Normal | 153 | 5.7 (4.7 to 6.8) | 1 (Reference) | 1 (Reference) |
| Abnormal | 161 | 10.9 (9.3 to 12.7) | 2.30 (1.84 to 2.87) | 1.73 (1.37 to 2.18) | |
| Stroke | Normal | 159 | 6.0 (4.9 to 7.1) | 1 (Reference) | 1 (Reference) |
| Abnormal | 102 | 7.8 (6.3 to 9.5) | 1.35 (1.05 to 1.73) | 1.12 (0.86 to 1.45) | |
| All-cause death | Normal | 387 | 16.5 (14.7 to 18.5) | 1 (Reference) | 1 (Reference) |
| Abnormal | 369 | 29.0 (26.3 to 31.9) | 1.96 (1.70 to 2.26) | 1.42 (1.23 to 1.65) | |
For stroke, the 10-year risk was 6.0% (95% CI 4.9 to 7.1) in patients with a normal scan and 7.8% (95% CI 6.3 to 9.5) in those with an abnormal scan ( Table 2 ). The cumulative incidence of stroke during the follow-up period is presented in Figure 2 . No substantial association was observed between an abnormal scan and stroke (adjusted HR 1.12, 95% CI 0.86 to 1.45; Table 2 ).
The sensitivity analyses including only nonfatal outcome diagnoses registered during an acute admission showed a slightly increased adjusted HR for myocardial infarction (1.89, 95% CI 1.46 to 2.43) and stroke (1.24, 95% CI 0.93 to 1.66; not shown).
Ten-year mortality risk in patients with a normal scan was 16.5% (95% CI 14.7 to 18.5) and 29.0% (95% CI 26.3 to 31.9) in patients with an abnormal scan ( Table 2 ). The cumulative mortality risk throughout the follow-up period is presented in Figure 3 . The adjusted HR for all-cause death comparing an abnormal scan with a normal scan was 1.42 (95% CI 1.23 to 1.65; Table 2 ).
Ten-year risk estimates for the outcomes according to co-morbidity level are presented in Table 3 . An abnormal scan was associated with an increased risk of the outcomes in virtually all co-morbidity levels, both within 10 years ( Table 3 ) as well as within 31 to 365 days and 1 to 10 years (not shown).
| Co-morbidity Level | Scan Result | No. of Events | Risk % (95% CI) | Hazard Ratio (95% CI) | |
|---|---|---|---|---|---|
| Unadjusted | Adjusted ∗ | ||||
| Normal (0 points) | |||||
| Myocardial infarction | Normal | 69 | 4.3 (3.2 to 5.6) | 1 (Reference) | 1 (Reference) |
| Abnormal | 67 | 8.8 (6.8 to 11.1) | 2.46 (1.75 to 3.44) | 1.91 (1.35 to 2.70) | |
| Stroke | Normal | 86 | 4.9 (3.8 to 6.3) | 1 (Reference) | 1 (Reference) |
| Abnormal | 49 | 7.4 (5.4 to 9.8) | 1.40 (0.98 to 1.99) | 1.13 (0.79 to 1.63) | |
| All-cause death | Normal | 134 | 9.8 (8.0 to 12.1) | 1 (Reference) | 1 (Reference) |
| Abnormal | 115 | 18.3 (15.3 to 21.8) | 1.99 (1.55 to 2.55) | 1.40 (1.08 to 1.81) | |
| Moderate (1 point) | |||||
| Myocardial infarction | Normal | 41 | 6.2 (4.3 to 8.6) | 1 (Reference) | 1 (Reference) |
| Abnormal | 48 | 13.8 (10.2 to 17.9) | 2.20 (1.45 to 3.35) | 1.88 (1.22 to 2.90) | |
| Stroke | Normal | 28 | 5.4 (3.4 to 8.0) | 1 (Reference) | 1 (Reference) |
| Abnormal | 27 | 8.0 (5.3 to 11.6) | 1.77 (1.04 to 3.00) | 1.64 (0.95 to 2.83) | |
| All-cause death | Normal | 118 | 22.7 (18.6 to 27.6) | 1 (Reference) | 1 (Reference) |
| Abnormal | 104 | 35.6 (29.8 to 42.3) | 1.57 (1.20 to 2.04) | 1.31 (1.00 to 1.72) | |
| Severe (≥2 points) | |||||
| Myocardial infarction | Normal | 43 | 11.0 (7.6 to 15.1) | 1 (Reference) | 1 (Reference) |
| Abnormal | 46 | 12.9 (9.4 to 16.8) | 1.62 (1.07 to 2.46) | 1.41 (0.91 to 2.19) | |
| Stroke | Normal | 45 | 9.8 (6.4 to 14.0) | 1 (Reference) | 1 (Reference) |
| Abnormal | 26 | 8.4 (5.4 to 12.2) | 0.86 (0.53 to 1.40) | 0.81 (0.49 to 1.33) | |
| All-cause death | Normal | 135 | 34.5 (28.7 to 41.1) | 1 (Reference) | 1 (Reference) |
| Abnormal | 150 | 47.3 (41.1 to 54.0) | 1.66 (1.32 to 2.10) | 1.48 (1.16 to 1.89) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
