The total annual cost of continued digoxin therapy was estimated to be $688 million. The corresponding costs for the alternative, where digoxin was discontinued, were estimated to be $1,094 million. Overall, the net savings resulting from continuation of digoxin were estimated to be $406 million/year, comprising $247 million for patients with and $159 million for those without ACE inhibitor therapy. On the basis of the range of net cost of digoxin withdrawal in one-way sensitivity analysis, the largest source of uncertainty came from the epidemiologic assumptions about the prevalence of congestive heart failure in the United States (range $324 million), the proportion of patients with congestive heart failure meeting the criteria previously outlined (range $365 million) and the baseline incidence of hospital admissions for patients meeting the criteria (range $257 million).
Significant uncertainty in the net cost of digoxin withdrawal was also attributable to the assumptions regarding the initial effectiveness of digoxin (described in terms of the relative risk of treatment failures with digoxin during the first 12 weeks). For patients receiving angiotensin-converting enzyme inhibitors, by varying the assumed relative riskfrom 0.06 to 0.47, the estimated net cost of digoxin withdrawal varied from $474 million down to $307 million. For patients not receiving angiotensin-converting enzyme inhibitors, by varying the assumed relative risk from 0.21 to 0.94 the estimated net cost of digoxin withdrawal varied from $545 million to $194 million. Additional uncertainty results from the extrapolation of the trials' data from the first 12 weeks to the rest of the year. If digoxin continued to be as effective during weeks 13 to 52 as during the first 12 weeks, the net savings from digoxin withdrawal increased to $534 million. Even if digoxin had no effect during weeks 13 to 52, a small net savings ($21 million) is expected to accrue.
The multiway sensitivity analysis showed that choosing different values for the assumptions for each of 10,000 model iterations could lead to a wide range of cost estimates, with a 90% range of certainty that extends from $106-822 million. More than 99% of the model iterations yielded net savings with continuation of digoxin therapy.