2018 Volume 82 Issue 6 Pages 1623-1631
Background: Although hepatitis C virus (HCV) is a known risk factor for cardiovascular disease, whether antiviral therapy (AVT) can reduce heart failure (HF) hospitalizations is unknown.
Methods and Results: In this population-based cohort study, we used data from the Taiwan National Health Insurance Research Database to evaluate the effect of interferon-based therapy (IBT) on cardiovascular events in patients with chronic HCV infection. Clinical outcomes evaluated included HF hospitalizations; a composite of acute myocardial infarction, ischemic stroke, and peripheral artery disease; all-cause death; and cardiovascular death. Of 83,229 eligible patients with chronic HCV infection, we compared 16,284 patients who received IBT with untreated subjects after propensity score matching. Patients who received IBT were less likely to be hospitalized for HF compared with untreated subjects (incidence density.ID, 0.9 vs. 1.5 events per 103 person-years; hazard ratio.HR, 0.58; 95% confidence interval.CI, 0.42–0.79; P=0.001). Compared with untreated subjects, the treated group had significantly lower risk of composite vascular events (ID, 3.7 vs. 5.0 events per 103 person-years; P<0.001), all-cause death (ID, 5.6 vs. 17.2 events per 103 person-years; P<0.001), and cardiovascular death (ID, 0.2 vs. 0.6 events per 103 person-years; P=0.001).
Conclusions: AVT for chronic HCV infection might offer protection against HF hospitalizations, critical vascular events, and cardiovascular death beyond known beneficial effects.
Hepatitis C virus (HCV) affects approximately 177.5 million people worldwide1 and is a major etiology of severe hepatic complications, including cirrhosis, hepatocellular carcinoma, and liver-associated death. In addition, extrahepatic complications2 also contribute to non-hepatic-related comorbidity,3 of which coronary artery disease (CAD),4 myocarditis, and cardiomyopathy5 have been well documented.
Editorial p 1503
Heart failure (HF) is a disastrous cardiac process and a leading risk factor for hospitalization and cardiovascular death.6 Although CAD and cardiomyopathy are well-known risk factors for HF, HCV has been reported as a risk factor of HF in some studies.5,7 Recent investigations have further shown that patients with chronic HCV infection have high levels of B-type natriuretic peptide (BNP), which is associated with cardiac decompensation.8,9 Therefore, HCV infection may worsen cardiovascular outcomes through cardiac dysfunction beyond the well-known atherosclerotic consequences.
Because of different distributions of genotypes type 1 (50%) and type 2 (45%) in Taiwan,10 the overall sustained viral response (SVR) of interferon-based therapy (IBT) is expected to be 80%,11 which could offer effective eradication. Although several studies have reported the benefits of IBT for both hepatic and extrahepatic complications,12,13 research on the protective effect of IBT with regard to HF hospitalizations is lacking. Therefore, the aim of this study was to investigate whether antiviral therapy (AVT) can reduce these critical events in patients with chronic HCV infection.
This population-based cohort study was retrieved from the Taiwan National Health Insurance Research Database (NHIRD), which has been prospectively recording claims data of all healthcare services in Taiwan since 1995. The national health insurance program in Taiwan is a single-payer, compulsory system, launched in 1995 and covering more than 99% of the population of 23 million people by the end of 2014.14 In the NHIRD, diseases are coded according to the International Classification of Diseases, Revision 9, Clinical Modification (ICD-9-CM), and data are de-identified to protect the anonymity of patients. The protocol of this study was approved by the Research Ethics Committee of Chang Gung Memorial Hospital (Institutional Review Board no. 103-6439B).
Eligibility CriteriaAll patients aged between 20 and 80 years newly diagnosed as HCV carriers (ICD-9-CM codes: 070.41, 070.44, 070.51, 070.54, 070.70, 070.71, V02.62) from January 1, 2003 (the beginning of when IBT was covered by the national health insurance program) were de-identified in the NHIRD and enrolled in this study. Patients who were infected with hepatitis B virus or human immunodeficiency virus, those who died within 6 months after the index date, and those with comorbidities possibly related to studied outcomes were excluded. The comorbidities included (1) major cardiovascular disease (ischemic heart disease (IHD), cerebral vascular accident, peripheral artery disease, HF, deep vein thrombosis, pulmonary embolism, atrial fibrillation, heart transplantation); (2) hepatic-associated illness (liver cirrhosis, hepatic decompensation (esophageal or gastric varices, hepatic encephalopathy, hepatorenal syndrome, portal hypertension and ascites, hepatocellular carcinoma, hepatic surgery, liver transplantation)); (3) non-liver-related malignancies; (4) chronic kidney disease including dialysis; (5) autoimmune disease; and (6) severe psychogenic or mental disorder (Table S1). Finally, we excluded those who accepted incomplete IBT and then the eligible patients were divided into an untreated group and treated group. According to response-guided therapy,11,15 incomplete IBT was defined as IBT course that was less than 16 weeks because of early termination for 1 of 2 reasons: the early virologic response was poor at week 12 according to the national insurance guideline or adverse effects occurred before week 16. Finally, outcome assessment was analyzed after propensity matching the 2 groups based on confounding variables, including sex, age, index year, hospital level, comorbidities, and medications.
Comorbidities and Drug ExposureAll diseases were identified according to the ICD-9-CM codes recorded in the NHIRD, and all comorbidities were confirmed by at least 2 consecutive clinic visits or any hospitalization before the index date. In order to ascertain the accuracy of the diagnoses, we validated the HCV diagnoses using the same criteria as those used in this study by reviewing medical records and the positive predictive rate was 100%. In addition, some diseases were further confirmed by long-term prescription of associated medications. The ICD-9-CM codes used for the comorbidities and clinical outcomes are listed in Table S1. Long-term drug use was defined as any drug prescribed for at least 6 months after the index date. All baseline characteristics contributing to outcomes before propensity score matching are shown in Table 1.
Variable | Treated (n=19,021) |
Incompletely treated (n=1,345) |
Untreated (n=62,863) |
P value |
---|---|---|---|---|
Sex | <0.001 | |||
Male | 10,794 (56.7) | 742 (55.2) | 29,949 (47.6) | |
Female | 8,227 (43.3) | 603 (44.8) | 32,914 (52.4) | |
Age, years | 50.2±11.4 | 52.6±12.2 | 50.4±13.7 | <0.001 |
Age group | <0.001 | |||
20–40 | 3,785 (19.9) | 231 (17.2) | 15,117 (24.0) | |
40–60 | 11,468 (60.3) | 721 (53.6) | 31,316 (49.8) | |
60–80 | 3,768 (19.8) | 393 (29.2) | 16,430 (26.1) | |
Hospital level | <0.001 | |||
Medical center | 6,506 (34.2) | 387 (28.8) | 12,232 (19.5) | |
Regional hospital | 9,532 (50.1) | 652 (48.5) | 21,092 (33.6) | |
District hospital | 1,923 (10.1) | 203 (15.1) | 12,851 (20.4) | |
Clinic | 1,060 (5.6) | 103 (7.7) | 16,688 (26.5) | |
Medical history | ||||
Diabetes mellitus | 1,901 (10.0) | 149 (11.1) | 4,638 (7.4) | <0.001 |
Hypertension | 3,246 (17.1) | 275 (20.4) | 9,313 (14.8) | <0.001 |
Dyslipidemia | 447 (2.4) | 35 (2.6) | 1,690 (2.7) | 0.037 |
COPD | 390 (2.1) | 39 (2.9) | 1,628 (2.6) | <0.001 |
Thyroid disease | 410 (2.2) | 32 (2.4) | 969 (1.5) | <0.001 |
Gouty arthritis | 835 (4.4) | 63 (4.7) | 2,621 (4.2) | 0.293 |
Medication | ||||
Antiplatelet agent | 324 (1.7) | 38 (2.8) | 1,766 (2.8) | <0.001 |
ACEI/ARB | 1,420 (7.5) | 142 (10.6) | 5,342 (8.5) | <0.001 |
β-blocker | 1,507 (7.9) | 182 (13.5) | 5,738 (9.1) | <0.001 |
Calcium-channel blocker | 1,781 (9.4) | 141 (10.5) | 6,519 (10.4) | <0.001 |
Oral hypoglycemia agent | 1,817 (9.6) | 150 (11.2) | 5,790 (9.2) | 0.024 |
NSAID | 7,084 (37.2) | 596 (44.3) | 27,227 (43.3) | <0.001 |
COX-II inhibitor | 593 (3.1) | 67 (5.0) | 3,219 (5.1) | <0.001 |
Diuretic | 112 (0.6) | 31 (2.3) | 1,106 (1.8) | <0.001 |
Steroid | 1,098 (5.8) | 103 (7.7) | 4,678 (7.4) | <0.001 |
Follow-up (years) | 4.2±2.7 | 3.8±2.5 | 5.9±3.0 | <0.001 |
ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; COX-II, cyclooxygenase II; COPD, chronic obstructive pulmonary disease; NSAID, nonsteroidal antiinflammatory drug.
Index date was defined as the date when HCV infection was first diagnosed in the untreated group, and the date when IBT was first prescribed in the treated groups. The observation period was from January 1, 2003 to December 31, 2013 and clinical outcomes were HF hospitalization, critical vascular event and CV death. HF hospitalization and composite vascular event were defined according to the principle diagnosis of admission. In order to evaluate the diagnostic accuracy of HF hospitalization from HCV patients, a validation study was conducted in 150 randomly selected HCV-infected patients admitted for HF (principle diagnosis coded as 428) during 2005–2013. Through review of the medical records, including echocardiogram, laboratory data and symptom presentations, 147 hospitalized subjects were confirmed as definite HF according to the discharge diagnosis, clinical symptoms/signs, BNP level and echocardiographic evidence. The positive predictive rate was 98%. The composite vascular event was defined as an admission for any of acute myocardial infarction (AMI), ischemic stroke, and peripheral artery disease and the individual diagnosis had been previously validated.12 In addition, cardiovascular death was defined as death caused by cerebral vascular accident (ICD-9 codes 430–437), IHD (ICD-9 codes 410–414), HF (ICD-9 code 428), cardiomyopathy (ICD-9 codes 425), or cardiac arrhythmia (ICD-9 426–427). To decrease the influence of interferon-related cardiac complications, we excluded those subjects with outcome-associated events during the 6 months prior to therapeutic period.
Statistical AnalysisGiven that the current study used retrospectively collected nonrandomized data, there was an inherent risk of selection bias, which may have confounded the results. To minimize this bias, we performed propensity score matching16 before comparing the treated and untreated cohorts. Covariates included in the propensity score are shown in Table 1 (except for follow-up years) plus index year of enrollment. We set the tolerance level as 0.2-fold of the SD of the propensity score (caliper=0.2) to warrant the quality of matching using the greedy nearest neighbor matching algorithm.
Baseline characteristics were compared among groups before matching (treated vs. incompletely treated vs. untreated) using the chi-square test for categorical variables or one-way analysis of variance for continuous variables. We compared characteristics between groups after matching (treated vs. matched untreated) using the chi-square test for categorical variables or Student’s t-test for continuous variables. Incidence of time to event outcomes (HF hospitalizations, composite vascular events, all-cause death, and cardiovascular death) was estimated using incidence density (ID) as the number of events per 1,000 person-years after the index date. Incidence of time to event outcomes between the treated and matched untreated groups was compared using a Cox proportional hazard model adjusted for propensity score. Data analyses and propensity score matching were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA).
Of 83,229 patients with chronic HCV infection enrolled in this study, 19,021 (mean age, 50.2±11.4 years) received IBT >16 weeks, 1,345 (mean age, 52.6±12.2 years) received incomplete treatment, and 62,863 (mean age, 50.4±13.7 years) did not receive treatment (Figure 1). Untreated patients were mainly followed at district hospitals and clinics, whereas treated patients were followed at higher-level hospitals (medical centers and regional hospitals). Patients who received IBT but with interruptions, had higher rates of comorbidities (diabetes mellitus, hypertension, chronic obstructive pulmonary disease) and were more likely to be prescribed medications (angiotensin-converting enzyme inhibitors/angiotensin-receptor blockers, β-blockers, oral hypoglycemic agents) compared with the other 2 groups (Table 1). Distribution of characteristics after propensity score matching was well-balanced and homogeneous between the treated and untreated cohorts (Table 2).
Flowchart of study design and subject selection. AMI, acute myocardial infarction; CVA, cerebral vascular accident; HCV, hepatitis C virus; HF, heart failure; IBT, interferon-based therapy; PAD, peripheral artery disease.
Variable | Treated (n=16,284) |
Untreated (n=16,284) |
P value |
---|---|---|---|
Sex | 1.000 | ||
Male | 9,056 (55.6) | 9,056 (55.6) | |
Female | 7,228 (44.4) | 7,228 (44.4) | |
Age, years | 49.8±11.7 | 49.7±12.3 | 0.717 |
Age group | 1.000 | ||
20–40 | 3,499 (21.5) | 3,499 (21.5) | |
40–60 | 9,630 (59.1) | 9,630 (59.1) | |
60–80 | 3,155 (19.4) | 3,155 (19.4) | |
Hospital level | 0.460 | ||
Medical center | 5,150 (31.6) | 5,265 (32.3) | |
Regional hospital | 8,245 (50.6) | 8,116 (49.8) | |
District hospital | 1,841 (11.3) | 1,831 (11.2) | |
Clinic | 1,048 (6.4) | 1,072 (6.6) | |
Medical history | |||
Diabetes mellitus | 1,315 (8.1) | 1,262 (7.7) | 0.277 |
Hypertension | 2,434 (14.9) | 2,354 (14.5) | 0.211 |
Dyslipidemia | 354 (2.2) | 383 (2.4) | 0.280 |
COPD | 331 (2.0) | 284 (1.7) | 0.056 |
Thyroid disease | 315 (1.9) | 307 (1.9) | 0.746 |
Gouty arthritis | 650 (4.0) | 625 (3.8) | 0.475 |
Medication | |||
Antiplatelet agent | 288 (1.8) | 268 (1.6) | 0.392 |
ACEI/ARB | 1,167 (7.2) | 1,127 (6.9) | 0.386 |
β-blocker | 1,223 (7.5) | 1,188 (7.3) | 0.459 |
Calcium-channel blocker | 1,424 (8.7) | 1,390 (8.5) | 0.502 |
Oral hypoglycemia agent | 1,361 (8.4) | 1,324 (8.1) | 0.456 |
NSAID | 6,207 (38.1) | 6,080 (37.3) | 0.147 |
COX-II inhibitor | 537 (3.3) | 518 (3.2) | 0.552 |
Diuretic | 103 (0.6) | 87 (0.5) | 0.244 |
Steroid | 935 (5.7) | 1,010 (6.2) | 0.079 |
Follow-up (years) | 4.4±2.8 | 4.3±2.7 | 0.003 |
Abbreviations as in Table 1.
Incidence of HF hospitalizations was significantly lower in the treated group than in the untreated group (ID, 0.9 vs. 1.5 events per 103 person-years; hazard ratio (HR), 0.58; 95% confidence interval (CI), 0.42–0.79; P=0.001) (Figure 2A; Table 3).
Incidence densities of heart failure hospitalizations (A) and composite vascular events (B) between treated and untreated groups after 1:2 matching.
Outcomes | Treated (n=16,284) |
Untreated (n=16,284) |
P value‡ |
---|---|---|---|
Heart failure hospitalization | |||
Events, n (%) | 63 (0.39) | 106 (0.65) | |
Incidence density§ | 0.9 (0.7–1.1) | 1.5 (1.2–1.8) | |
HR (95% CI) | 0.58 (0.42–0.79) | Ref. | 0.001 |
Composite vascular events | |||
Events, n (%) | 259 (1.59) | 346 (2.12) | |
Incidence density§ | 3.7 (3.3–4.1) | 5.0 (4.5–5.5) | |
HR (95% CI) | 0.73 (0.62–0.85) | Ref. | <0.001 |
All-cause death | |||
Events, n (%) | 397 (2.44) | 1,203 (7.39) | |
Incidence density§ | 5.6 (5.1–6.1) | 17.2 (16.2–18.2) | |
HR (95% CI) | 0.32 (0.29–0.36) | Ref. | <0.001 |
CV death | |||
Events, n (%) | 15 (0.09) | 39 (0.23) | |
Incidence density§ | 0.2 (0.1–0.3) | 0.6 (0.4–0.8) | |
HR (95% CI) | 0.37 (0.20–0.67) | Ref. | 0.001 |
#Any one of acute myocardial infarction, ischemic stroke or peripheral arterial occlusion disease; §incidence density: no. of events per 1,000 person-years; ‡adjusted for propensity score. CI, confidence interval; CV, cardiovascular; HR, hazard ratio.
In Figure 2B, Incidence of composite vascular events was lower in the treated group (ID, 3.7 events per 103 person-years; 95% CI, 3.3–4.1) than in the untreated group (ID, 5.0 events per 103 person-years; 95% CI, 4.5–5.5), showing a significant difference (HR, 0.73; 95% CI, 0.62–0.85; P<0.001). Incidence of AMI (ID, 0.5 vs. 0.9 events per 103 person-years; HR, 0.59; 95% CI, 0.39–0.90; P=0.014) and ischemic stroke (ID, 3.0 vs. 3.9 events per 103 person-years; HR, 0.77; 95% CI, 0.65–0.93; P=0.005) were significantly lower in the treated group than in the untreated group (Table 4). In contrast, the incidence of peripheral artery disease was comparable between groups (ID, 0.3 vs. 0.4 events per 103 person-years; HR, 0.74; 95% CI, 0.41–1.34; P=0.322) (Table 4).
Outcomes | Treated (n=16,284) |
Untreated (n=16,284) |
P value‡ |
---|---|---|---|
Acute myocardial infarction | |||
Events, n (%) | 36 (0.22) | 60 (0.37) | |
Incidence density§ | 0.5 (0.3–0.7) | 0.9 (0.7–1.1) | |
HR (95% CI) | 0.59 (0.39–0.90) | Ref. | 0.014 |
Ischemic stroke | |||
Events, n (%) | 214 (1.31) | 269 (1.65) | |
Incidence density§ | 3.0 (2.6–3.4) | 3.9 (3.4–4.4) | |
HR (95% CI) | 0.77 (0.65–0.93) | Ref. | 0.005 |
Peripheral arterial disease | |||
Events, n (%) | 19 (0.12) | 25 (0.15) | |
Incidence density§ | 0.3 (0.2–0.4) | 0.4 (0.3–0.5) | |
HR (95% CI) | 0.74 (0.41–1.34) | Ref. | 0.322 |
§Incidence density: no. of events per 1,000 person-years; ‡adjusted for propensity score. Abbreviations as in Table 3.
The incidence of both all-cause death and cardiovascular death was approximately 3-fold higher in the untreated group than in the treated group (all-cause death: ID, 5.6 vs. 17.2 events per 103 person-years; HR, 0.32, 95% CI, 0.29–0.36; cardiovascular death: ID, 0.2 vs. 0.6 events per 103 person-years; HR, 0.37; 95% CI, 0.20–0.67), revealing favorable outcomes in the treated group (Figure 3A,B). The causes of death in both cohorts are listed in Table S2.
Incidence densities of all-cause mortality (A) and cardiovascular deaths (B) between treated and untreated groups after 1:2 matching.
We further analyzed the beneficial effects of IBT among subgroups with regard to HF hospitalizations, composite vascular events, and all-cause death (Figure 4, Table S3). The results showed that the beneficial effects on HF hospitalizations were comparable among subgroups (P for interaction >0.05). The benefit of IBT on the risk of composite vascular events was more obvious in patients with a history of gout than in those without (P for interaction=0.037). In terms of all-cause death, male and younger patients were associated with greater beneficial effects (P for interaction=0.002 and <0.001, respectively). In addition, the mean age of male patients was younger than that of female patients (47.8±12.0 years vs. 52.3±10.8 years).
Incidence densities and hazard ratios of heart failure hospitalizations between the treated and untreated groups of patients with hepatitis C virus infection regarding age, sex, hypertension, diabetes, dyslipidemia, chronic obstructive pulmonary disease and gouty arthritis.
This was the first large-scale nationwide cohort study to support that effective viral eradication maybe offer comprehensive cardiovascular protection, including HF and cardiovascular death.
HF Beyond Systemic AtherosclerosisA meta-analysis17 reported that HCV can be thought as a systemic atherosclerotic risk factor that contributes to early-age myocardial infarction.18 Although CAD is a major etiology of HF,6 ischemic cardiomyopathy cannot entirely explain the underlying mechanism of worsening pulmonary hypertension and cardiac dysfunction in patients with chronic HCV infection. Viral myocarditis and cardiomyopathy have been reported as highly prevalent,5 and may be associated with direct HCV replication and indirect immunologic consequences in the myocardium.19,20 Younossi et al reported that patients with HCV infection had a higher incidence of HF (adjusted odds ratio, 2.49; 95% CI, 1.04–5.96),7 and that they were younger than patients without HCV infection. Other studies have reported that diastolic dysfunction,21 cardiac mass increasing as hypertrophy,22 and cardiac fibrosis23 can lead to early myocardial remodeling, as demonstrated on cardiac magnetic resonance imaging24 and myocardial scans.25
Antonelli et al reported that cryoglobulinemic patients had significantly higher levels of NTproBNP and interleukin 6 compared with controls.26 Despite the association between higher BNP level and left ventricular systolic impairment,8 recent studies have further reported a strong link between myocardial injury and severity of liver damage, with increased inflammatory activity recorded as higher levels of tumor necrosis factor-α,27 C-reactive protein,27 cardiac enzymes28 and circulating advanced oxidation protein products29 among patients with chronic HCV infection.
Even if we did not know the types of HF in our study, volume overload with increased pulmonary pressure was associated with advanced cirrhosis and impaired renal function, as indicated by high use of diuretics in untreated patients. Although the mechanisms of cirrhotic cardiomyopathy and hepatocardiac syndrome30 are yet to be clarified, both manifestations are associated with portocardiac shunt-induced pulmonary hypertension.31 Furthermore, studies using echocardiography and stress ventriculography have revealed that cardiac dysfunction can be reversed after liver transplantation in patients with HCV infection.32 Thus, untreated patients with HCV infection would demonstrate worse pulmonary hypertension without defined ischemic cardiomyopathy because of advanced cirrhotic deterioration.
Effect of IBT on Cardiac Deterioration and DeathBecause of adverse effects, or drug intolerance, the proportion of subjects receiving complete IBT was relatively low (22.8%); however, the ratio was similar as in a previous study (19% of HCV-selected patients).33 In addition, the high sustained viral response (SVR) in Taiwan was 64–79% for genotype 1 and 74–95% for genotype 2,11 especially for IL-28B C-allele patients. Many Taiwanese studies have supported that AVT can reduce rates of ischemic stroke, acute coronary syndrome, and end-stage renal disease beyond the known hepatic benefits,12 as seen in the significant reduction in composite vascular events in this study. We provide further correlation between AVT and HF hospitalizations in patients with chronic HCV infection from real-world practice. The beneficial effects on both vascular events and HF hospitalization might have contributed to the significant reduction in cardiovascular deaths in this study.
Despite the rare cardiotoxicity of interferon,34 the effect of IBT on HCV-associated dilated cardiomyopathy has been reported.35 Viral eradication may resolve myocardial defects, but HCV relapse would worse perfusion again.23 Standard IBT has been shown to decrease the C-reactive protein level,36 and improve cardiovascular biomarkers with a non-hepatic origin.37 In addition, serum levels of creatine kinase-MB and cardiac troponin T have been shown to decrease along with serum HCV titers during IBT.38 Therefore, successful viral eradication would decrease myocardial injury and avoid further fibrotic remodeling.
Lee et al reported that anti-HCV seropositive patients with undetectable serum HCV RNA level showed a lower rate of cardiovascular death than those with detectable HCV RNA level (HR, 1.16 vs. 1.53; P=0.026).39 In the current study, treated patients showed a more significant reduction in cardiovascular death than untreated patients (HR, 0.37; P<0.001). Nahon et al13 demonstrated that patients who achieved SVR had lower rates of vascular events, extrahepatic death, and major adverse cardiac events, including HF and cardiovascular death in a 5-year observational study. Although some limitations, including nonuniform therapeutic agents and small sample size, restricted its clinical implication, their findings are compatible with our investigation in terms of preventing HF hospitalizations and critical cardiovascular complications.
Study LimitationsThere are several limitations to this study. First, details of infection duration, genotype, virologic response, laboratory data (liver biochemistry function, platelet count), and personal health habits were unavailable. However, we surveyed essential data, including date of diagnosis, therapeutic age, duration of therapy, and combined agents. Second, information on severity of underlying disease was limited, which may have affected the outcomes of the analysis. Therefore, we used information on outpatient medications within 1 year, comorbidities, and hospital level, and excluded major illnesses associated with cardiovascular events. In addition, we used propensity score matching to balance multiple variables in both cohorts and to decrease severity bias. Third, HF is a syndrome related to clinical symptoms that are ill-defined according to ICD-9-CM codes. We also did not know the types and severity of HF. However, we strengthened the accuracy of the diagnosis according to discharge ICD-9-CM codes, and the diagnostic methods were rechecked with 98% sensitivity.
Beyond the known extrahepatic benefits of AVT for chronic HCV infection, our results further showed that IBT might offer significant protection against HF hospitalizations, critical vascular events, and cardiovascular death. In the coming era of direct-acting antiviral drugs, large prospective studies including multivariate analysis and sensitivity testing will allow for greater understanding of this issue.
This study was supported by grants from the Chang Gung Medical Research Fund of Chang Gung Memorial Hospital, Chiayi, Taiwan (CMRPGME0011). We thank Alfred Hsing-Fen Lin and Zoe Ya-Jhu Syu for the assistance with statistics.
Conflict of Interest Statement: No conflict of interests to declare.
Ethical Approval: The Institutional Review Board of Chang Gung Memorial Hospital: (103-6439B).
Grant Support: This study was supported by grants from the Chang Gung Medical Research Fund of Chang Gung Memorial Hospital, Chiayi, Taiwan (CMRPGME0011).
Writing Assistance: For ensure that the language is clear and free of errors, this manuscript has been corrected by professional editors at Taiwan Editage, a division of Cactus Communications. Funding source is from primary investigator.
Author Contributions: All authors reviewed the manuscript and gave final approval for submission. Conception and design of the study: M.-S.L., C.-M.C., M.-L.C.; acquisition of data: M.-Y.C., S.-T.C., P.-H.C., W.-Y.L., T.-J.H.; analysis and interpretation of data: M.-S.L., W.-Y.L., T.-J.H., T.-H.C.; drafting of the manuscript: M.-S.L., C.-M.C.; critical revision of the manuscript for important intellectual content: C.-M.C., M.-L.C., Y.-S.L.
Supplementary File 1
Table S1. ICD-9-CM codes used for diagnoses
Table S2. Causes of death in the cohort after propensity score matching
Table S3. (A) Subgroup analysis: composite vascular events (acute myocardial infarction, ischemic stroke, peripheral arterial disease), (B) Subgroup analysis: all-cause death
Please find supplementary file(s);
http://dx.doi.org/10.1253/circj.CJ-17-1118