Meta-analysis: Digoxin contributes to lower hospitalization, poses neutral effect on death

Issue: October 2015

LONDON — Digoxin was associated with lower hospitalization and had a neutral effect on death in randomized controlled trials, and was only linked to a higher death rate in observational studies, according to a systematic review and meta-analysis presented at the European Society of Cardiology Congress.

When the researchers performed a meta-regression, they found that the differences in death rates could be attributed to baseline differences in those taking digoxin vs. those taking placebo or no treatment, Dipak Kotecha, MBChB, PhD, MRCP, FESC, FHEA, said during a presentation. Kotecha and colleagues simultaneously published their findings in the BMJ.

Dipak Kotecha

“Confounding is the main reason why observational data continue to show increases in mortality associated with digoxin use,” Kotecha said. “This suggests that digoxin should continue to be considered as a treatment option in patients with both [HF] and atrial fibrillation until further randomized data become available.”

Kotecha, from the University of Birmingham Centre for Cardiovascular Sciences, United Kingdom, and colleagues analyzed the effect of digoxin, a drug prescribed to treat HF and AF, vs. placebo or no treatment in studies published between 1960 and 2014. The primary outcome was all-cause mortality and the main secondary outcome was all-cause hospital admission.

They performed a systematic review of 52 studies, covering 621,845 patients, and a meta-analysis of 75 study analyses, including 4,006,210 patient-years of follow-up.

Baseline differences

In the systematic review, Kotecha and colleagues determined that, compared with those assigned placebo or taking no treatment, those assigned or taking digoxin were 2.4 years older (95% CI, 1.3-3.6); had lower left ventricular ejection fraction (33% vs. 42%); were more likely to have diabetes (OR = 1.38; 95% CI, 1.21-1.57); were more likely to use diuretics, a marker of HF severity (OR = 3.22; 95% CI, 2.21-4.68).

In the meta-analysis, they determined that the pooled RRs for death were as follows:

for unadjusted observational studies: RR = 1.76; 95% CI, 1.57-1.97;
for adjusted observational studies: RR = 1.61; 95% CI, 1.31-1.97;
for propensity-matched studies: RR = 1.18; 95% CI, 1.09-1.26; and
for randomized controlled trials: RR = 0.99; 95% CI, 0.93-1.05.

Baseline differences between treatment groups had a major effect on the mortality findings, Kotecha said, including the use of diuretics (P = .004).

The researchers also found that a neutral association with mortality was more likely to be reported in studies of higher quality and lower risk for bias (P < .001).

Regardless of study type, digoxin was associated with reduced all-cause hospital admission (RR = 0.92; 95% CI, 0.89-0.95), according to the researchers. All studies with this endpoint evaluated digoxin in patients with HF, not AF, Kotecha said.

“We have good data from randomized trials supported by observational data that digoxin does not affect mortality in [HF], but reduces hospital admissions to a small extent,” Kotecha said during the presentation. “The data would suggest that there is no difference in those patients with concomitant [AF], similar to our findings in the beta-blocker trials (Kotecha D, et al. Lancet. 2014;doi:10.1016/S0140-6736(14)61373-8). There are still very poor amounts of data for patients with [AF] without [HF].”

Underuse of effective treatment

In a related editorial in the BMJ, Graham D. Cole, MB, BChir, MA, and Darrel P. Francis, MA, MB, BChir, FRCP, wrote that, “When randomized data point in one direction, we should resist the temptation to obscure the message by presenting observational data with an opposite association.”

Cole and Francis, from International Centre for Circulatory Health, National Heart and Lung Institute, Imperial College London, wrote, “The penalty patients have paid for clinicians allowing observational data (pointing to increased mortality) to be stacked up against data from randomized controlled trials (pointing to reduced admissions and no increase in mortality) might have been underuse of an effective treatment. The 28% reduction in admissions for [HF] in the DIG trial could have provided worthwhile relief from the high burden of [HF] worldwide.” – by Erik Swain

References:

Cole GD, Francis DP. BMJ. 2015;doi:10.1136/bmj.h4662.

Kotecha D, et al. Clinical Trial Update I – Cardiovascular Diseases: Prevention, Outcomes, Quality. Presented at: European Society of Cardiology Congress; Aug. 29-Sept. 2, 2015; London.

Ziff OJ, et al. BMJ. 2015;doi:10.1136/bmj.h4451.

Disclosures: Kotecha reports receiving honoraria from Menarini, receiving professional development support from Daiichi Sankyo and serving as the head of the steering committee for the Beta-blockers in Heart Failure Collaborative Group and for the upcoming RATE-AF trial of digoxin vs. beta-blockers in patients with AF. Cole and Francis report no relevant financial disclosures.