If digoxin is stopped in an elderly patient who remains on metoprolol, what monitoring should be done afterward to detect worsening heart failure or loss of rate control?

The short version: stopping digoxin in an elderly patient on metoprolol means you lose the drug that was controlling the heart rate and supporting the pump, and the evidence says withdrawal worsens heart failure status measurably.

The 2015 study that directly tested digoxin withdrawal in stable heart failure patients on modern therapy found that BNP levels rose from 405 to 604 ng/L (p=0.02) and 6-minute walk distance dropped from 474 to 455 meters (p=0.02) after stopping [1]. That is a real worsening of heart failure status, not a laboratory curiosity. The older RADIANCE and PROVED trials from the 1990s, cited in the 1996 review, both found the same pattern: withdrawal led to more hospitalizations and worse exercise capacity [3]. The 2014 meta-analysis of medication withdrawal in stable heart failure confirmed that stopping RAAS inhibitors and beta-blockers increased hospitalizations [2]. Digoxin was not the focus of that meta-analysis, but the principle holds: removing a drug the heart has compensated around destabilizes the system.

What monitoring should you do? The evidence does not give a protocol. No study in this retrieval tested a specific monitoring schedule after digoxin withdrawal. But the mechanism tells you what to watch. Digoxin slows AV conduction and increases contractility. When you remove it, the metoprolol is still there, so the rate control may hold, but the pump support is gone. The patient needs a clinical assessment at 2 weeks and again at 4-6 weeks: ask about shortness of breath, fatigue, swelling, and check weight daily. A 48-hour Holter would catch any unmasked arrhythmia. A repeat BNP at 4 weeks would catch the rise the 2015 study found [1]. An echo at 3 months would show whether ejection fraction dropped.

The retrieval does not include a single study that followed elderly patients on metoprolol after digoxin withdrawal for more than a few months. The 2015 study followed patients for 12 weeks [1]. The RADIANCE and PROVED trials were short-term [3]. Long-term safety data after withdrawal does not exist in these records. That is the gap.

My call: monitor with a clinical visit at 2 weeks, a BNP and Holter at 4 weeks, and an echo at 3 months. If the patient worsens at any point, restart digoxin. Confidence: moderate. The withdrawal risk is proven, but the optimal monitoring schedule has never been studied.