Prevalence of Potential Cardiac Resynchronization Therapy Candidates and Actual Use of Cardiac Resynchronization Therapy in Patients Hospitalized for Heart Failure

Cardiac resynchronization therapy (CRT) is increasingly used as an additional therapy for patients with chronic heart failure (CHF) in New York Heart Association (NYHA) class III or IV on optimal medical therapy who have a left ventricular ejection fraction (LVEF) ≤35% and an increased QRS duration (≥120 ms).1,2 A recent meta-analysis also indicates that patients with milder heart failure could potentially benefit from this therapy.3 Despite the evidence regarding the effects of CRT, recent surveys suggest that this therapy is probably underused, with a large variation between hospitals regarding implantation rates in patients hospitalized with CHF.4 Also, although implantation rates for implantable cardioverter defibrillator and CRT-D have increased significantly in Europe, there remain major (and unexplained) differences between countries.5 Finally, data on the prevalence of CRT candidates in hospitalized CHF patients are limited. The aims of the present study are to evaluate, in patients admitted with NYHA class III-IV heart failure, (i) the prevalence of eligible CRT candidates, (ii) the actual use of CRT, and (iii) potential clinical reasons for not using CRT in eligible CRT candidates. This study was based on data from a prospective heart failure registry, evaluating all patients admitted with NYHA class III–IV heart failure in two hospitals in Belgium in 2008. A total of 368 consecutive patients who were admitted to cardiology wards in two hospitals: AZ Maria Middelares Gent, Belgium (305 patients), and AZ VUB Brussels, Belgium (63 patients), were evaluated. Patients already treated with CRT on admission (n = 5) and patients who died during hospitalization (n = 26) were excluded. Additional 15 patients were also excluded because of incomplete clinical data or missing echocardiographic data. The remaining 322 patients formed the study population. In all patients, clinical data and heart failure medication at discharge were recorded. Echocardiography was used to measure LVEF (Simpson's method) and left ventricular end-diastolic diameter. QRS duration was measured from the surface electrocardiogram and creatinine levels were used as a measure of kidney function. Several co-morbidities were also documented including the presence of ischaemic/non-ischaemic aetiology, a previous history of admission for CHF, a previous history of peripheral arterial disease, and a previous history of stroke or dementia. Potential CRT candidates were defined as patients with an LVEF 20 ms. During the study period, potential treatment options (including CRT) were discussed every week during the heart failure staff meetings in the presence of the treating heart failure specialist, an electrophysiologist, and a heart failure nurse. The decision to implant a CRT-P or CRT-D device was at the discretion of the treating physician. Patients who actually received CRT during hospitalization or during the 6 months after discharge were documented. Statistical analysis was performed using SPSS v 17.0 for Windows (SPSS Inc., 2008, Chicago, IL, USA). Data are presented as mean ± standard deviation of the total number (%). Differences between groups were calculated with the unpaired Student t-test or Fisher's exact test as indicated. A P-value of 35% and QRS > 120 ms. Of the 79 potential CRT candidates, only 18 patients (23%) actually received a CRT device during hospitalization or during the 6 months following discharge. Only 6% of the total population and 14% of all patients with LVEF ≤ 35% were treated with a new CRT device. Of note, only 49% of all patients with a QRS duration ≥120 ms had QRS duration of more than 150 ms. Differences in characteristics between potential CRT candidates who did and did not receive CRT are shown in Table 1. Figure 2 shows the presence of different co-morbidities in CRT candidates who did and did not receive CRT. All co-morbidities listed were more frequently present in patients who did not receive CRT. In particular, the prevalence of advanced age and the history of stroke or dementia were significantly higher in CRT candidates who did not receive CRT. The results of this study show that 25% of all patients admitted with NYHA class III–IV CHF to cardiology wards are potential candidates for CRT. However, in daily practice only one in four of these potential CRT candidates actually receives this therapy. Thus, only 6% of the total population of CHF patients and only 14% of the patients with an LVEF ≤ 35% received CRT. This last figure is in line with a recent large American survey showing that on average only 4.8% of the patients admitted with CHF and LVEF ≤ 35% left the hospital with a new CRT device.4 Our results also indicate that typical left bundle branch block is present in ~70% and a QRS > 150 ms in ~50% of the potential CRT candidates. Clearly, patient characteristics and co-morbidities play an important role in patient selection and they are often different from patients enrolled in randomized controlled trials.6,7 Our results are in line with a recent large European CRT survey,6 indicating that patients who did not receive a device were older, had a narrower QRS, and more frequently had an ischaemic cardiomyopathy or first admission for CHF. Also, all of the co-morbidities that we evaluated were more frequently present in patients that did not receive CRT. Nevertheless, chronic obstructive pulmonary disease and creatinine >2 mg/dL, two important risk factors for CHF mortality,8,9 were present in 17% and 12% of the patients with a new CRT device. In addition, atrial fibrillation, which is currently only a class IIa indication for the reduction of morbidity in CRT candidates with pacemaker dependency,2 was present in 11% of the CRT recipients. The strength of this study lies in its detailed description of parameters, influencing the use and non-use of CRT in patients hospitalized with CHF. However, the small sample size is a major limitation. In clinical practice, advanced age, patient characteristics, and co-morbidities are all associated with the non-use of CRT. Conflict of interest: none declared.