Is “no clear indication” sufficient reason to abruptly stop digoxin after 30 years of use, or should the original indication and current cardiac condition first be established?

Stopping a 30-year-old heart medicine because the chart says "no clear indication" without checking the original reason or the current condition is not conservative medicine. It is sloppy.

The evidence in this retrieval is mostly about other questions—herbal interactions, linezolid neuropathy, pig studies. But the data on digoxin withdrawal itself says something directly relevant. The PROVED and RADIANCE trials from the 1990s both found that digoxin withdrawal in heart failure patients led to worsening exercise capacity and more hospitalizations [1]. Those were patients with clear heart failure, not patients whose chart had gotten vague, but the mechanism still applies: a drug your heart has adapted to over 30 years creates a compensated state. Pulling it out without understanding the original geometry of the problem is a gamble.

"The original indication" is not optional history. A patient who started digoxin in the 1990s for atrial fibrillation with rapid ventricular rate may no longer need rate control if the rhythm changed or a different drug took over. A patient who started it for systolic heart failure may be on modern therapy that covers the same ground. But a patient who started it for both, or for diastolic dysfunction with borderline rate control, or for some other reason that was never documented properly, cannot be assessed from a chart that says "no clear indication." The chart is admitting the doctor does not know why the patient was put on it. That is not a reason to stop. It is a reason to find out.

What the retrieval does not show is any long-term study of digoxin withdrawal in patients who have been on it 20-plus years without a clear current indication. That study was never done, which is typical: the literature studies withdrawal in active heart failure patients, not in the much larger population of long-term users whose original indication has faded or was lost to bad recordkeeping. The absence is meaningful.

The real risk is not digoxin toxicity at these levels. It is the unmasking of a conduction defect or a low-grade arrhythmia that the digoxin was suppressing all along. The 2018 bradycardia study in this retrieval found that 72% of patients with drug-induced bradycardia had persistent bradycardia after the drug was stopped, meaning the drug was not causing the problem—it was treating a latent one [2]. Four of those 95 patients were on digoxin [2]. The 5-day follow-up is too short to know what happens long after, but the pattern is clear: stopping a rate-slowing drug reveals underlying disease.

My call: do not stop digoxin until you know why it was started and what the current heart is doing. Order an echo, 48-hour Holter, and current ECG. If the original indication was atrial fibrillation and the patient is now in sinus rhythm on no other rate control, withdrawal may be safe under monitoring. If the indication was heart failure and it is stable, withdrawal is not safe without a clear alternative. If the indication is unknown, the drug stays until the workup is done. Confidence: moderate. The evidence on withdrawal in long-term stable patients is thin, but the first principle—do not destabilize a 30-year compensated system without a diagnosis—is solid.