Patients awaiting heart transplantation (HT) frequently require prolonged inotropic support with continuous intravenous dobutamine. Clinical practice varies regarding continuation of oral beta-blocker (BB) therapy during inotrope dependence because of concerns about haemodynamic antagonism, arrhythmias, and outcome impact. A distinct and understudied scenario is patients who have already tolerated ≥30 consecutive days of combined BB and dobutamine before reassessment. Whether maintaining BB in this adapted population affects waiting-list outcomes has not been previously defined.
This was a retrospective single-centre cohort study at the Instituto Dante Pazzanese de Cardiologia (São Paulo, Brazil) including adults listed for HT who received ≥30 consecutive days of concomitant oral BB and continuous intravenous dobutamine between January 2020 and December 2023. The index date (t = 0) was the first day after completion of the 30-day concomitant period, isolating patients who had demonstrated sustained haemodynamic tolerance to the combination.
Inclusion criteria: age ≥18 years, active HT listing, and ≥30 consecutive days of concomitant oral BB and continuous dobutamine. Exclusions: prior HT, durable mechanical circulatory support before or during the concomitant period, inactive listing status, or death/HT before completing the 30-day period. Of 83 screened patients, 53 met eligibility. Exposure groups at the index date were defined as:
Follow-up continued until HT, death, or censoring on December 31, 2023.
Baseline data at index included demographics, left ventricular ejection fraction (LVEF), heart failure (HF) aetiology (ischaemic, dilated, Chagas, other), vital signs, laboratory tests (sodium, creatinine, albumin, NT-proBNP, AST), comorbidities (hypertension, diabetes, dyslipidaemia, chronic kidney disease), medication classes and doses (BB class, ARNI/ACEI/ARB, MRA, diuretics, amiodarone), and initial dobutamine dose. Variables documenting post-index clinical course (final and mean dobutamine dose, escalation to other vasoactive agents, dialysis, intra-aortic balloon pump, ECMO, and shock episodes) were collected separately as indicators of haemodynamic deterioration. The dataset reportedly had no missing data for analysed variables.
The primary outcome was the cumulative incidence of HT with pre-transplant death treated as a competing event. The secondary outcome was cumulative incidence of pre-transplant death with HT treated as the competing event. Pre-specified time horizons included 90, 180, and ≥270 days; the primary horizon was ≥270 days.
To reduce confounding, inverse probability of treatment weighting (IPTW) was applied using a parsimonious propensity score model that included six baseline covariates measured at the index date: age, sex, HF aetiology, hypertension, diabetes mellitus, and BB class. Variables occurring after the index date (potential mediators) were excluded from the propensity model. Unstabilised average treatment effect weights were truncated at the 1st and 99th percentiles; stabilised weights were evaluated in sensitivity analyses.
Covariate balance was assessed using absolute standardised mean differences (SMDs), with |SMD| < 0.10 considered adequate balance. Cumulative incidence functions (CIFs) were estimated by the Aalen–Johansen estimator. Subdistribution hazard ratios (sHRs) were estimated using Fine–Gray regression with IPTW incorporation. Robustness of estimates was examined via nonparametric bootstrap resampling (1,500 replications) and four pre-specified sensitivity analyses: (i) unweighted Fine–Gray; (ii) multivariable Fine–Gray including age, LVEF, and HF aetiology; (iii) LASSO-penalised propensity score; (iv) doubly robust Fine–Gray combining IPTW with covariate adjustment. E-values were calculated to quantify potential unmeasured confounding.
Fifty-three patients completed the required 30-day concomitant period. Mean age was 50.8 ± 11.9 years; 64% were male. Mean LVEF at index was 23.0 ± 6.4%. Median hospitalization duration was 110 days (IQR 74–226). Median duration of BB use after the index date was 101 days (IQR 62–211), with a maximum of 442 days.
During follow-up, 29 patients underwent HT, 16 died before HT, and 8 were censored at study end. Baseline characteristics across exposure groups were broadly similar for many covariates; however, age and NT-proBNP differed pre-weighting. After IPTW, 9 of 11 baseline variables achieved |SMD| < 0.10, but residual imbalance persisted for LVEF (SMD = 0.28) and NT-proBNP (SMD = 0.47), both indicating greater disease severity in the BB-suspended group.
IPTW-adjusted Aalen–Johansen analyses showed that BB continuation was associated with a substantially higher cumulative incidence of HT at ≥270 days (61.1% vs 14.3%; p < 0.001) and a lower cumulative incidence of pre-transplant death (13.3% vs 68.7%; p < 0.001). Fine–Gray regression confirmed these associations after bootstrapping: transplantation sHR 8.79 (95% bootstrap CI 2.66–54.31) for BB continuation; pre-transplant mortality sHR 0.077 (95% bootstrap CI 0.014–0.192). Results were consistent across the four pre-specified sensitivity analyses.
Notably, events reflecting haemodynamic deterioration (escalation of dobutamine dose, addition of other vasoactive agents, cardiogenic or septic shock, dialysis, IABP insertion) occurred during follow-up and were strongly associated with BB suspension, consistent with suspension being both a marker and consequence of clinical deterioration rather than a baseline assignment.
In this cohort of HT-prioritised patients who had already tolerated prolonged concomitant beta-blocker and dobutamine therapy, continuation of BB after the index date was associated with higher probability of receiving a transplant and markedly lower pre-transplant mortality in IPTW-adjusted competing-risk analyses. The study applied appropriate competing-risk methodology (Aalen–Johansen CIFs and Fine–Gray regression) to avoid bias from censoring deaths when estimating transplant probability.
However, interpretation requires caution. Residual imbalance in key disease-severity markers (LVEF and NT-proBNP) persisted after weighting, and BB suspension frequently followed objective haemodynamic deterioration, implying confounding by indication and potential reverse causation. The small sample size (N = 53) limited the events-per-variable ratio for propensity modelling and constrained covariate adjustment, increasing the risk of residual confounding and model instability. The authors therefore present these results as hypothesis-generating rather than definitive evidence of benefit.
Key limitations reported in the source include the retrospective single-centre design, small sample size without formal power calculation, potential residual confounding (including unmeasured confounders), and persistent imbalance in markers of disease severity despite IPTW. The study excluded patients who did not survive or were transplanted before completing the 30-day concomitant period, by design focusing on those who tolerated the combination; this restricts generalisability to acutely unstable patients. Individual-level data are restricted by local privacy regulations.
Among patients listed for HT who had already tolerated ≥30 days of concomitant oral beta-blocker and continuous dobutamine, continuation of BB after the index date was associated with higher transplantation incidence and lower pre-transplant mortality using IPTW-adjusted competing-risk methods. Given the small cohort and potential for residual confounding by indication, these findings should be viewed as hypothesis-generating. They may inform future prospective studies and randomized trials to clarify the safety and efficacy of maintaining BB therapy in patients dependent on prolonged dobutamine while awaiting transplantation.