Shoko Ishikawa

Evaluation of the impact of atrial fibrillation on rehospitalization events in heart failure patients in recent years

BACKGROUND Although we have previously reported that the presence of paroxysmal atrial fibrillation (AF) is an independent risk factor for rehospitalization in patients with congestive heart failure (CHF) in a population from 1996 to 2002, the impact of AF configuration as a risk factor in a more recent population remains to be clarified. METHODS AND RESULTS 319 patients with CHF admitted to our institute in 2006-2007 were retrospectively evaluated. The patients were divided into 3 groups in accordance with their basic cardiac rhythm, i.e. sinus rhythm (n=210), chronic AF (n=68), and paroxysmal AF (n=41). During the follow-up period of 19 ± 17 months, there was no significant difference in mortality or rehospitalization events among the 3 groups (p=0.542). In the multivariate analysis, no administration of β-blockers was the only independent risk factor for rehospitalization due to CHF exacerbation. CONCLUSIONS The clinical impact of AF configuration as a risk factor of rehospitalization due to CHF exacerbation was considered to be decreased in recent years.

Author's reply: To PMID 22429679.

I read with great interest the article by Murakami et al. entitled Evaluation of the impact of atrial fibrillation on rehospitalization vents in heart failure patients in recent years” [1]. The authors concluded that the clinical impact of atrial fibrilation as a risk factor for rehospitalization due to congestive heart ailure exacerbation was considered to have decreased in recent ears and mentioned that no administration of -blockers was the nly independent risk factor for rehospitalization in the Cox hazrd multivariate analysis. The study is interesting, but I believe that ome comments regarding the results of the multivariate analysis re warranted. They showed the results of univariate and multivariate analses of the risk factors for rehospitalization in Table 2. They sed the odds ratio as the statistical parameter, but in the Cox roportional hazard model, hazard ratio is usually used. (Odds atio is used in the logistic regression model analysis.) The conrmation of the method of regression model adopted is firstly arranted. Second, they also showed the 95% confidential interval (CI) of he odds ratio of the parameters. For example, odds ratio of age s 0.148 and its 95% CI is −0.062 to 0.0063 (Table 2). However oth odds ratio and hazard ratio are quite unlikely to be negaive values. I wonder 95% CI of what actually is shown. Third, they howed that non-administration of -blocker was the only indeendent risk factor for rehospitalization. However the odds ratio s 0.614, indicating that non-administration of -blocker inhibits he occurrence of the events. Finally, in the multivariate Cox proortion hazard model and multivariate logistic regression analysis, variable per 10 events or more is strongly recommended [2]. n the present study, 17 variables per 99 events, i.e. 1 variable er 5.8 events was assigned. Since it may cause overfitting of staistical analysis, the number of independent variables should be educed. Thus, I hope the authors would reconsider the present results f multivariate analysis and perform the multivariate statistical nalysis again.

Long-term observation of fibrillation cycle length in patients under angiotensin II receptor blocker therapy for chronic atrial fibrillation

The long‐term effect of angiotensin II receptor blockers (ARBs) on atrial fibrillation (AF) is unclear. In this study, we evaluated the change in the fibrillation cycle length (FCL) in patients under long‐term ARB therapy for chronic AF.

Usefulness of ST elevation score by using vector-projected virtual 187-channel ECG for risk stratification in patients with Brugada-type ECG pattern

Although powerful defibrillation devices are available in clinical practice, risk stratification is important in asymptomatic Brugada syndrome. In this study, vector‐projected 187‐channel electrocardiogram (VP‐ECG) was used to calculate the ST‐elevation score in Brugada‐type ECG and test its usefulness in risk stratification.

Preventive Effect of Amiodarone on VT/VF Events in ICD Patients with Structural Heart Diseases

Background: Although amiodarone (AMD) is applied for implantable cardioverterdefibrillator (ICD) patients to reduce VT/VF events, its actual benefit and long‐term efficacy are unclear. In the present study, we retrospectively evaluated the incidence of VT/VF events in ICD patients with and without AMD.