Key result
Inpatients undergoing CRT implantation had more comorbidities than outpatients, but propensity-matched analysis showed no significant difference in heart failure hospitalization or composite outcomes at 2 years.
Why the study?
Does inpatient versus elective outpatient status prior to cardiac resynchronization therapy (CRT) implantation impact long-term clinical outcomes?
Does inpatient versus elective outpatient status prior to cardiac resynchronization therapy (CRT) implantation impact long-term clinical outcomes?
Answers are grounded in this paper and cite their sources.
While inpatients undergoing CRT implantation have more advanced symptoms and comorbidities, propensity-matched analysis shows similar clinical outcomes to outpatients, reinforcing the clinical intuition to stabilize patients prior to elective device implantation.
This editorial refers to ‘Inpatient vs. elective outpatient cardiac resynchronization therapy device implantation and long-term clinical outcome’ by O.A. Ajijola et al., on page 1745 Ajijola et al.1 from Massachusetts General Hospital (MGH) in Boston address the question of whether in-hospital or outpatient status prior to cardiac resynchronization (CRT) implantation has an impact on clinical outcomes. The results of this analysis will not surprise most clinicians. Not unexpectedly, inpatients undergoing CRT implantation have a higher risk of heart failure hospitalization compared with elective outpatient implantation. The authors also confirm that patients with prior admissions for heart failure have a greater risk for rehospitalization. Additionally, the authors have found that a prolonged in-hospital course is associated with a worse clinical outcome. In the recently published European Cardiac Resynchronisation Therapy Survey, based on 2438 implantations from 141 centres in 13 countries, we found that 82% of patients were electively admitted for implantation with a median duration of hospitalization of only 3 days and 57% of patients had been admitted to hospital for heart failure decompensation during the previous year. A CRT-defibrillation (CRT-D) device was implanted in 73% of patients and a CRT-pacing (CRT-P) device was implanted in 27% of patients; 26% of patients had previously had a device implanted, either a permanent pacemaker or an implantable cardioverter-defibrillator (ICD).2 In the paper by Ajijola et al., there are several issues that could lead to important selection bias. This is a retrospective analysis of 196 outpatients and 105 inpatients. Clearly, rigorous adjudication of heart failure hospitalization in this retrospective analysis must have been difficult, especially in an open, non-randomized analysis. MGH is a well-known tertiary referral centre and only patients who receive their entire follow-up at MGH over a 4-year period were included. Only successful CRT implantations were captured. Patients who had been previously implanted with a device were also excluded. One of the main finding in the study is that modest differences between clinical outcomes could be explained by the extent of co-morbidity. This is confirmed by the development of the propensity score which, using appropriate co-variates, demonstrated no statistical significant differences in the incidence of heart failure hospitalization or the composite endpoint that included heart failure hospitalization, death, left ventricular assist device implantation, or cardiac transplantation over a 2-year period between the inpatient and outpatient groups. An important piece of missing data concerns the type of the device that was implanted. It has recently been demonstrated by Romeyer-Bouchard et al. that CRT-D implantation may be associated with a higher risk of device-related complications compared with CRT-P, which could possibly affect the results if CRT-D devices were more often implanted in the sicker inpatient population.3 It would also have been relevant to compare the medical treatments, especially with regard to beta-blocker treatment. In the 301 patient cohort, it was found that inpatients had more advanced symptoms and substantially more co-morbidity, especially diabetes mellitus, atrial fibrillation, and renal dysfunction. One would expect that a patient requiring a hospitalization due to decompensation of heart failure or a cardiovascular event would be sicker than patients scheduled for elective CRT implantation with less extensive comorbidity. This represents a marker of the underlying severity of heart failure and, to a large degree, is secondary to the dominating pathophysiology. Certainly, a degree of cardio-renal syndrome would be operative in such patients. It is also well known that previous admissions for heart failure and a prolonged in-hospital course are associated with a poor prognosis. As the authors emphasize, much of this will appear intuitive to experienced clinicians. In addition to the previous limitations, it is important to acknowledge that there were indeed important temporal changes in practice during the data collection period between 2003 and 2007. Many centres have developed dedicated multi-disciplinary device clinics with uniform device optimization protocols and well-organized follow-up. There have been improvements in operators’ skills, and in operating techniques, and more recently refinements in the guideline recommendations on device therapy in heart failure.4 The data provided by Ajijola et al. may not really help us to define the optimal timing for CRT implantation following an episode of decompensation of heart failure. The clinical status of a patient will usually determine whether or not they require hospitalization. A patient's response to therapy will determine the time required to reach ambulatory status. Table 1 compares the baseline characteristics of these two groups and demonstrates some of the important differences both with regard to the cardiovascular status and to the extent of comorbidty as demonstrated by the difference between the two groups in the proportion of patients with NYHA IV class heart failure (21 vs. 2). A significant difference between the groups with regard to renal insufficiency is likely an important factor explaining the results. Operators will obviously prefer to implant a device in a stable outpatient. Patients who respond well and can be stabilized on medical therapy should have the device implanted on an elective basis. We would expect that these patients would have a better clinical outcome compared with a patient who is too ill to be discharged from hospital. This would be true of any invasive or surgical procedure. I suppose the take-home message is that we should make every effort to stabilize patients prior to device implantation and perform the procedure electively. But I think we knew that. Conflict of interest: none declared.
No takes yet. Share an insight, caveat, or question.
Kenneth Dickstein (2010) conducted an editorial in Heart failure requiring cardiac resynchronization therapy (n=301). Inpatient status prior to CRT implantation vs. Elective outpatient status was evaluated on Heart failure hospitalization or a composite of heart failure hospitalization, death, left ventricular assist device implantation, or cardiac transplantation (p=NS). Inpatients undergoing CRT implantation had more comorbidities than outpatients, but propensity-matched analysis showed no significant difference in heart failure hospitalization or composite outcomes at 2 years.
Synapse has enriched 10 related papers. Consider them for comparative context: