Although coronary revascularization is common in both Japan and the United States (US), no direct comparison has been performed to demonstrate differences in the clinical characteristics and long-term outcomes of patients in these 2 countries. We analyzed the preprocedural, in-hospital, and long-term data from the Coronary Revascularization Demonstrating Outcome registry (Kyoto, Japan) and the Texas Heart Institute Research Database (Houston, Texas) of 16,100 patients who had undergone elective, initial percutaneous coronary intervention or coronary artery bypass grafting. The Japanese procedures were performed from 2000 to 2002 (n = 8,871, follow-up period 3.5 years, interquartile range 2.6 to 4.3) and the US procedures from 1999 to 2003 (n = 7,229, follow-up period 5.2 years, interquartile range 3.8 to 6.5). The Japanese patients tended to be older (mean age 67.2 vs 62.7 years; p <0.001), to smoke (52.9% vs 46.0%; p <0.001), and to have diabetes (39.2% vs 31.0%; p <0.001) and stroke (16.4% vs 5.0%; p <0.001). The US patients were more obese (body mass index 23.7 vs 29.3 kg/m 2 ; p <0.001), with greater rates of systemic atherosclerotic disease. Both groups had a similar in-hospital mortality rate (Japanese patients 0.9% vs US patients 1.1%; p = 0.19) and crude long-term mortality rate (Japanese patients 27.7/1,000 person-years, US patients 28.2/1,000 person-years; p = 0.35). After adjustment for known predictors, the US group had greater long-term mortality than the Japanese group (hazard ratio 1.71, 95% confidence interval 1.50 to 1.95; p <0.001). This finding was consistent among all high-risk subgroups. In conclusion, the 2 registries showed similar crude outcomes but important differences in patient risk factors such as obesity. In the adjusted analysis, the Japanese patients had better outcomes than did the US patients. Additional study is needed to assess the effect of ethnic and risk factor variations on coronary artery disease.
Physicians in both Japan and the United States (US) perform a high volume of coronary revascularization procedures; however, the ethnicities and lifestyles of their patients differ greatly. In Western countries, several large-scale databases of patients with coronary artery disease (CAD) have been established. Despite the large number of subjects enrolled in these studies, the Asian subgroup has been relatively small. Furthermore, few studies have been conducted in Asian countries, mainly because patients have been hesitant to enroll in clinical trials and sufficient resources are lacking to maintain large databases. Hence, little is known about the clinical presentation and postrevascularization outcomes of Asian patients with CAD or about how these patients compare with their counterparts in the US. To help fill this void, we assessed the clinical characteristics and long-term outcomes of Japanese versus US patients by comparing information from 2 large databases concerning percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) in these 2 countries.
Methods
The Coronary Revascularization Demonstrating Outcome database in Kyoto (CREDO-Kyoto) was a multicenter (n = 29) registry maintained in Kyoto, Japan. The Texas Heart Institute Research Database (THIRDBase) is an ongoing single-center registry maintained at the Texas Heart Institute in Houston, Texas. Details concerning the design of the CREDO-Kyoto and THIRDBase have previously been reported. Both of these comprehensive, longitudinal, clinical registries of patients undergoing coronary revascularization procedures were designed to evaluate associated periprocedural and late events. CREDO-Kyoto enrolled patients from 2000 to 2002, and the THIRDBase has been a continuous, on-going registry since 1993. For the present analysis, we included only those THIRDBase patients enrolled from 1999 to 2003, to match the total number of patients in the CREDO-Kyoto registry. In both countries, all the study patients had undergone initial, elective, isolated revascularization procedures. All PCI patients had undergone placement of a bare metal stent (drug-eluting stents were not available for this use during the study period). The patients were excluded if they had undergone previous PCI or CABG, required valve surgery or peripheral vascular revascularization, or were undergoing primary PCI for an acute myocardial infarction (MI).
The baseline data regarding the patients’ clinical characteristics were obtained prospectively from both registries. The coronary anatomic and procedural characteristics, in-hospital outcomes, and vital status as of December 31, 2006 were assessed for all patients. For the THIRDBase, survivorship was determined from the US Department of Vital Statistics Database. For CREDO-Kyoto, the follow-up data were obtained from hospital charts or by interviewing the patients or referring physicians. Both registries had remarkably high follow-up completion rates (THIRDBase 100%; CREDO-Kyoto 98% at 1 year and 95% at 2 years). The survival analyses included both in-hospital and long-term survival data, and these survival rates were not considered separately.
Each patient’s history was obtained by interview when the patient arrived at the hospital or clinic, and the details were entered prospectively into the database. The following variables were documented: left ventricular ejection fraction, number of diseased vessels, urgency of the revascularization procedure, presence of hypertension (characterized by blood pressure of >130/90 mm Hg or the current use of antihypertensive medications), severity of angina (Canadian Cardiovascular Society classification), severity of congestive heart failure (New York Heart Association functional status), family history of CAD, previous MI, renal function, need for hemodialysis, presence of diabetes mellitus (characterized by a fasting blood sugar level of >6.87 mmol/L or the use of antidiabetic agents), and the presence of peripheral vascular disease (occlusive or aneurismal vascular disease in the aorta or other peripheral vessels).
The institutional review board of THIRDBase approved the present study. All enrollees provided written informed consent at hospital admission. All analyzed data were stripped of personal identifiers. For CREDO-Kyoto, owing to the retrospective nature of enrollment, the patients had not provided written informed consent; however, when later interviewed for the follow-up evaluation, 73 patients were excluded from the analysis because of their refusal to participate in the study. This process was concordant with the guidelines for epidemiologic studies issued by the Ministry of Health, Labor, and Welfare of Japan.
To compare the 2 registries, we assessed the demographic characteristics and in-hospital mortality, using Pearson’s chi-square test for discrete variables and Student’s t test for continuous variables. We then used the Kaplan-Meier method to draw the survival curve and the log-rank test to identify significant differences in the unadjusted survival rates. We also compared the 2 registries with respect to the survival rates among the subgroups of patients undergoing PCI versus CABG. Logistic regression analysis and Cox proportional hazards models were used to examine the differences in demographic and clinical characteristics with regard to short- and long-term survival, respectively. Multivariate analyses were performed to control for possible confounding factors affecting the association between the registry and outcomes. The variables included in the multivariate models were age ≤65 years, gender, obesity, procedure type (PCI or CABG), history of MI, heart failure, New York Heart Association functional class, peripheral vascular disease, renal function, hemodialysis, hypertension, diabetes, hyperlipidemia, family history of CAD, smoking, the number of diseased vessels, and the geographic location of the registry. Renal function was characterized by a serum creatinine level of ≤179 μmol/L, a serum creatinine level of >179 μmol/L without hemodialysis, and a serum creatinine level of >179 μmol/L with hemodialysis. Cerebrovascular disease was excluded from the multivariate analyses because of differences in the definition of such disease by the 2 registries. To assess the validity of the proportional hazards assumption, we plotted −log [−log(survival)] curves for each category of nominal or ordinal covariates versus the log (analysis time). Because the proportional hazards assumption did not hold for the mode of coronary revascularization, the stratified Cox proportional hazards model using the revascularization procedure selected as the stratification variable was applied for the analysis of the long-term outcomes. Separate models were generated for the subgroups of patients with a high-risk profile (e.g., age ≥65 years, male gender, obesity, hemodialysis, diabetes, and a greater number of diseased vessels). The modeling procedure was performed for the cases for which the necessary data were available (complete case analysis) using Stata, version 10.1, software (StataCorp, College Station, Texas). All tests of significance were 2 tailed, and p values of <0.05 were considered significant.
Results
The series included 16,100 patients—8,871 patients from the CREDO-Kyoto registry (median follow-up period 3.5 years, interquartile range 2.6 to 4.3) and 7,229 patients from the THIRDBase registry (median follow-up period 5.2 years, interquartile range 3.8 to 6.5). The Japanese patients were older and were more likely to be smokers and to have diabetes mellitus and cerebrovascular disease ( Table 1 ). The US patients were more obese, with a greater body mass index. In general, the US patients had a greater prevalence of MI, renal insufficiency, and hypertension. Multivessel disease was seen more frequently in the Japanese patients than in the US patients.
| Variable | Missing Data | Japan (n = 8,871) | United States (n = 7,229) | p Value | PCI | CABG | ||||
|---|---|---|---|---|---|---|---|---|---|---|
| Japanese Patients (n = 6,510) | US Patients (n = 4,076) | p Value | Japanese Patients (n = 2,361) | US Patients (n = 3,153) | p Value | |||||
| Age (years) | 0 | 67.2 ± 10.0 | 62.7 ± 11.1 | <0.001 | 67.2 ± 10.2 | 62.4 ± 11.6 | <0.001 | 67.1 ± 9.4 | 63.1 ± 10.6 | <0.001 |
| Women | 0 | 29.1% | 29.9% | 0.28 | 29.7% | 32.4% | 0.003 | 27.6% | 26.7% | 0.46 |
| Body mass index (kg/m 2 ) | 439 | 23.7 ± 3.2 | 29.3 ± 5.8 | <0.001 | 23.7 ± 3.3 | 29.6 ± 6.1 | <0.001 | 23.5 ± 3.2 | 28.9 ± 5.4 | <0.001 |
| Coronary artery bypass grafting | 0 | 26.6% | 43.6% | <0.001 | ||||||
| History of myocardial infarction | 80 | 26.3% | 34.3% | <0.001 | 23.3% | 33.6% | <0.001 | 34.5% | 35.1% | 0.62 |
| History of heart failure | 99 | 15.4% | 14.0% | 0.013 | 12.1% | 11.4% | 0.30 | 24.7% | 17.4% | <0.001 |
| New York Heart Association functional class IV | 105 | 1.4% | 7.4% | <0.001 | 1.5% | 6.3% | <0.001 | 1.3% | 8.7% | <0.001 |
| Peripheral vascular disease | 76 | 11.3% | 14.7% | <0.001 | 8.1% | 12.7% | <0.001 | 20.1% | 17.3% | 0.008 |
| Cerebrovascular disease | 75 | 16.4% | 5.0% | <0.001 | 14.1% | 4.1% | <0.001 | 22.7% | 6.1% | <0.001 |
| Valve disease | 85 | 7.6% | 6.1% | <0.001 | 7.7% | 6.0% | 0.001 | 7.3% | 6.2% | 0.12 |
| Renal insufficiency | 195 | 6.4% | 11.4% | <0.001 | 5.8% | 10.4% | <0.001 | 7.9% | 12.6% | <0.001 |
| Hemodialysis | 0 | 4.0% | 1.1% | <0.001 | 3.7% | 1.2% | <0.001 | 4.7% | 0.9% | <0.001 |
| Hypertension | 73 | 69.5% | 74.3% | <0.001 | 69.0% | 72.8% | <0.001 | 70.7% | 76.1% | <0.001 |
| Diabetes mellitus | 76 | 39.2% | 31.0% | <0.001 | 36.5% | 27.6% | <0.001 | 46.6% | 35.5% | <0.001 |
| Hyperlipidemia | 87 | 51.3% | 61.5% | <0.001 | 50.2% | 62.1% | <0.001 | 54.4% | 60.8% | <0.001 |
| Family history of coronary artery disease | 537 | 15.7% | 33.9% | <0.001 | 15.4% | 34.5% | <0.001 | 16.6% | 33.1% | <0.001 |
| Smoking | 238 | 52.9% | 46.0% | <0.001 | 52.5% | 44.7% | <0.001 | 54.1% | 47.7% | <0.001 |
The foregoing trends were also seen when the PCI and CABG patients were analyzed separately ( Table 1 ), except that the Japanese CABG patients had a greater prevalence of peripheral vascular disease. In Japan, the proportion of PCI versus CABG procedures did not differ among patients with single-vessel disease, but PCI was performed more frequently in patients with multivessel disease ( Figure 1 ). To treating 3-vessel disease, CABG was preferred in the US (82.2%), but PCI and CABG were performed with a similar frequency in Japan (48.5% and 51.5%, respectively).
Both registries had similar in-hospital outcomes, regardless of the revascularization procedure used. Overall, the in-hospital mortality rate was 0.9% for the Japanese patients and 1.1% for the US patients (p = 0.19), indicating that the procedural complication rates were similar between the 2 groups. When the in-hospital mortality rate was analyzed according to the revascularization procedure performed, it was similar for both PCI (0.45% in Japan vs 0.27% in the United Stages; p = 0.15) and CABG (2.2% in Japan vs 2.3% in the US; p = 0.99). Table 2 lists the multivariate predictors of in-hospital outcome. When the results were adjusted for these confounding variables, the US patients had a greater risk of in-hospital death compared to the Japanese patients (odds ratio 1.60, 95% confidence interval 1.02 to 2.51; p = 0.039).
| Variable | Odds Ratio | 95% Confidence Interval |
|---|---|---|
| Age ≥65 years | 2.31 | 1.53–3.51 |
| Women | 1.24 | 0.83–1.85 |
| Body mass index ≥25 kg/m 2 | 0.78 | 0.51–1.18 |
| Coronary artery bypass grafting | 3.90 | 2.53–6.03 |
| Previous myocardial infarction | 1.48 | 1.04–2.13 |
| History of heart failure | 2.14 | 1.43–3.21 |
| New York Heart Association functional class IV | 1.13 | 0.61–2.10 |
| Peripheral vascular disease | 1.75 | 1.19–2.57 |
| Renal function | ||
| Serum creatinine ≤179 μmol/L | 1.00 | — |
| Serum creatinine >179 μmol/L | 2.97 | 1.85–4.77 |
| Hemodialysis | 5.46 | 3.10–9.63 |
| Hypertension | 1.16 | 0.75–1.80 |
| Diabetes mellitus | 1.50 | 1.04–2.16 |
| Hyperlipidemia | 0.65 | 0.45–0.93 |
| Family history of coronary artery disease | 1.00 | 0.64–1.56 |
| Smoking | 1.56 | 1.07–2.29 |
| No. of diseased vessels | ||
| 1 Vessel | 1.00 | — |
| 2 Vessels | 1.47 | 0.86–2.51 |
| 3 Vessels | 1.82 | 1.05–3.15 |
| United States patients | 1.31 | 0.82–2.08 |
The crude long-term mortality rate was similar in both Japan and the US (27.7 vs 28.2/1,000 person-years, respectively). Figure 2 shows the Kaplan-Meier survival curves for each registry. The crude hazard ratio for the American patients was 1.02 (95% confidence interval 0.93 to 1.13; p = 0.65). This trend did not alter when survival was analyzed according to revascularization procedure ( Figure 3 ).
Several significant predictors of mortality were identified ( Table 3 ). The univariate model showed that the strongest predictors of mortality (p <0.001) in the Japanese group were age >65 years, body mass index >25 kg/m 2 , a history of heart failure, the need for hemodialysis, and the presence of left circumflex artery disease. In the US group, the strongest predictors of mortality were age >65 years and variables associated with an atherosclerotic process, such as a history of MI, a history of heart failure, peripheral vascular disease, renal function, hemodialysis, hypertension, diabetes mellitus, hyperlipidemia, and a family history of CAD.
| Variable | Japan | United States | ||
|---|---|---|---|---|
| HR | 95% CI | HR | 95% CI | |
| Age ≥65 years | 2.74 | 2.29–3.27 | 3.01 | 2.65–3.43 |
| Women | 1.04 | 0.90–1.20 | 1.43 | 1.26–1.63 |
| Hypertension | 1.09 | 0.94–1.27 | 1.77 | 1.51–2.08 |
| Diabetes mellitus | 1.39 | 1.22–1.59 | 1.78 | 1.57–2.02 |
| Hyperlipidemia | 0.59 | 0.51–0.67 | 0.61 | 0.54–0.69 |
| Family history of coronary artery disease | 0.66 | 0.53–0.82 | 0.65 | 0.56–0.74 |
| Smoking | 0.91 | 0.80–1.05 | 1.08 | 0.96–1.22 |
| Body mass index ≥25 kg/m 2 | 0.46 | 0.38–0.55 | 0.56 | 0.49–0.64 |
| Previous myocardial infarction | 1.45 | 1.26–1.67 | 1.36 | 1.21–1.54 |
| History of heart failure | 3.13 | 2.71–3.62 | 3.56 | 3.12–4.07 |
| New York Heart Association class IV | 3.17 | 2.25–4.46 | 2.96 | 2.50–3.50 |
| Peripheral vascular disease | 2.20 | 1.86–2.60 | 2.60 | 2.27–2.98 |
| Cerebrovascular disease | 1.78 | 1.52–2.07 | 2.42 | 1.98–2.97 |
| Renal function | ||||
| Serum creatinine level ≤179 μmol/L | 1.00 | — | 1.00 | — |
| Serum creatinine level >179 μmol/L | 6.98 | 5.54–8.78 | 2.68 | 2.30–3.13 |
| Hemodialysis | 6.12 | 5.07–7.40 | 8.83 | 6.68–11.7 |
| No. of diseased vessels: | ||||
| 1 Vessel | 1.00 | — | 1.00 | — |
| 2 Vessels | 1.48 | 1.23–1.78 | 1.22 | 1.04–1.43 |
| 3 Vessels | 2.09 | 1.74–2.52 | 1.59 | 1.31–1.93 |
| Left anterior descending artery disease | 1.26 | 1.05–1.51 | 0.99 | 0.87–1.14 |
| Left circumflex artery disease | 1.43 | 1.23–1.65 | 1.30 | 1.14–1.49 |
| Right coronary artery disease | 1.59 | 1.37–1.85 | 1.21 | 1.06–1.37 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
