Evidence Based Review Article

The Journal of Informed Pharmacotherapy 2004;15:200.

Serum Digoxin Concentrations and Outcomes in Patients with Heart Failure: Is There Any Association?

Reviewer: Robert McCollom
Reviewer抯 e-mail address: robert_mccollom@rhss.bc.ca

Reviewer's profession/specialty:
Pharmacist, ICU/Critical Care

Original Citation

Rathore SS, Curtis JP, Wang Y, Bristow MR, Krumholz HM.  Association of serum digoxin concentration and outcomes in patients with heart failure. JAMA 2003;289:871-878.  PubMed Citation

Overall Study Question

This study was a post-hoc analysis of data from the 1997 Digitalis Investigation Group (DIG) trial. (1)  The DIG trial was a 3-year randomized, double-blind, placebo-controlled trial of the effect of digoxin on mortality and hospitalization rates in adult patients with heart failure, a left ventricular ejection fraction of 45% or less and a normal sinus rhythm.  The DIG investigators reported that digoxin afforded no overall mortality benefit and only a modest reduction in hospitalizations among patients with heart failure and depressed left ventricular systolic function.  The main objective of the current study was to assess the association of variations in serum digoxin concentration (SDC) with mortality and hospitalizations in a subgroup of male patients (mean age 63 years) enrolled in the DIG study. 

Of the 6800 male and female patients originally enrolled in the DIG trial, 3,782 men were included in the current analysis.  Of these patients, 1171 were randomly assigned to digoxin and 2611 were randomly assigned to placebo.  Those patients allocated to the digoxin group who had valid SDCs assessed at 1 month following randomization were identified and subsequently divided into 3 subgroups based on SDC thresholds (0.5-0.8 ng/mL, 0.9-1.1 ng/mL and > 1.2 ng/mL) determined from previous studies.

What was the declared relationship between authors and sponsors for this study?

This was an unfunded analysis of data obtained from a public use copy of the DIG study and was conducted independently from the original DIG investigators.

Are the Results of the Study Valid?

1. Was assignment of patients randomized?

Yes.  This study was a post-hoc analysis of a cohort of males who participated in the original randomized, double-blind placebo-controlled DIG trial.  Patients were originally randomly assigned to digoxin or placebo based on a published algorithm based on age, sex, weight and renal function. Adjustments to the initial dose were allowed depending on prior digoxin dosages and concomitant pharmacotherapy.

2. Were all patients who entered the trial properly accounted for and attributed at its conclusion?

Yes.  Among the men assigned to placebo, 28 were excluded from analysis because they died in the month following randomization.  This was done to replicate the same process undertaken for the digoxin-assigned patients.

3. Were patients, their clinicians, and study personnel 'blind' to treatment?

Yes. Patients and data investigators were blinded with respect to treatment, and cause of death was determined without knowledge of assigned treatment group.

4. Were the groups similar at the start of the trial?

The baseline characteristics of the digoxin and placebo groups in the main DIG trial appeared to be similar.  However, there were significant differences in baseline characteristics when comparing SDC subgroups.  Patients with a SDC of 0.5 to 0.8 ng/mL tended to have a lower incidence of NYHA class III/IV heart failure as compared to those with higher SDC ranges.  This subgroup was also younger, had a lower occurrence of four or more heart failure signs or symptoms, had better renal function (higher GFR and lower serum creatinine) and had used less nitrates and diuretics than the higher SDC groups.  This would suggest that these patients were less sick and this may have contributed to the higher survival in this patient group.

5. Aside from the experimental intervention, were the groups treated equally?

Yes.  There appeared to be no differences in the additional treatment received by the study patients. 

What were the Results?

1. How large was the treatment effect? 

Of the 1171 men with SDCs assessed at 1 month, 572 (49%) had an SDC between 0.5 and 0.8 ng/mL, 322 (27%) had an SDC of 0.9 to 1.1% and 277 (24%) had an SDC of 1.2 ng/mL or higher.  There was no overall difference in the primary endpoint (all-cause mortality) among men assigned to placebo or those digoxin recipients who had SDCs assessed (36.2% vs. 36.6%, 95% confidence interval [CI] �9 to 3.8, p = 0.80). 

When stratified by serum digoxin concentration, men with lower SDCs of 0.5 to 0.8 ng/mL had a 6.3% lower all-cause mortality rate (p = 0.005; number needed to treat [NNT] = 15) and 5.9% lower rate of hospitalizations (61.9% vs. 67.8%; 95% CI -1.5 to �.2; NNT = 17) as compared to the placebo group.  Compared to this subgroup, there was no associated reduction in mortality in the subgroup of men with SDCs of 0.9 to 1.1 ng/mL (absolute difference 2.6%, p = 0.36).  In contrast, men with SDCs > 1.2 ng/mL had an 11.8% increased all-cause mortality rate compared with men in the placebo group (95% CI, 5.7 to 18.0, p < 0.001).

2. How precise was the estimate of the treatment effect?

The 95% CI around the absolute mortality benefit of 6.3% (SDC 0.5 to 0.8 ng/mL subgroup vs. placebo) was �.5 to �1. The 95% CI around the non-significant reduction in mortality between this subgroup and the SDC 0.9 to 1.1 ng/mL subgroup was �0 to 8.3, p = 0.36).  Finally, the 95% CI around the absolute all-cause mortality benefit of 11.8% (SDC > 1.2 ng/mL vs. placebo) was 5.7 to 18.0.

Will the Results Help Me in Caring for My Patients?

1. Can the results be applied to my patient care?

Because this was a post-hoc analysis of non-randomized treatment subgroups, the possibility of confounding influences cannot be excluded and it should be considered hypothesis generating only. Additionally, since it only evaluated men with heart failure in sinus rhythm and an LVF EF <45%, generalizability to the general population is limited.

2. Were all clinically important outcomes considered?

Yes. The primary outcome was all-cause mortality at follow-up (mean 37 months). Secondary endpoints were death due to cardiovascular causes, death due to worsening heart failure and hospitalization for worsening heart failure.

3. Are the likely treatment benefits worth the potential harms and costs?

Digoxin is an effective and inexpensive agent for symptom relief, despite its inability to reduce overall mortality rates in heart failure. This analysis suggests that the effectiveness and safety of digoxin in men with heart failure may be optimized by monitoring and maintaining SDC between 0.5 and 0.8 ng/mL. This may require an increase in the number of serum digoxin determinations, with an associated increase in monitoring costs. It is yet to be determined if the benefits outweigh the additional costs.


Since the publication of the DIG trial, concerns have been raised regarding the relative efficacy and safety of a SDC greater than 1.0 ng/mL. Recent data have suggested that the maintenance of a low (< 0.09 ng/mL) SDC is as effective in heart failure as higher levels and is associated with a lower incidence of toxicity. What the present analysis does is provide further support for the suggestion that a SDC that has traditionally been considered as 搉ormal�, may in fact be harmful and that the target therapeutic range for digoxin should be lowered to 0.5-0.8 ng/mL.

How the results of this analysis are translated into everyday clinical practice is difficult to determine. The question of where digoxin fits in the in the treatment of heart failure is probably more important than the issue of the appropriate SDC. Because of the well-established mortality benefits of ACE Inhibitors, beta blockers, and spironolactone, therapy with these agents should be optimized first.  Clinicians must then decide, before any concerns about appropriate SDCs, whether the benefits of using digoxin outweigh its risks. For those who do not believe in the benefits of digoxin, these results will probably have little impact on their current practice. Those who believe that digoxin has a role in alleviating symptoms of heart failure, prolonging exercise tolerance, improving quality of life and reducing hospitalizations, it may now be clinically prudent to attempt to achieve a SDC of 0.5-0.8 ng/mL. Whether it is practical or possible to maintain SDCs in this narrow range is debatable.  It would certainly necessitate more routine use of digoxin level determinations. Whether the benefits of maintaining SDCs within the lower, narrower therapeutic range offset the added cost and workload remains to be determined.  A randomized, controlled trial (including the routine use of beta-blockers) is needed to confirm these findings.

Reviewer Competing Interests

None declared.


  1. Anon. The effect of digoxin on mortality and morbidity in patients with heart failure. The Digitalis Investigation Group.  N Engl J Med 1997;336:525-33.

Copyright � 2004 by the Journal of Informed Pharmacotherapy. All rights reserved.