Successful whole bowel irrigation in self-poisoning with potassium capsules

To the Editor:

Overdose of slow-release (SR) potassium chloride is uncommon, but life-threatening (1–3). Potassium chloride capsules are radio-opaque (1,4), but effective traditional gastrointestinal (GI) decontamination is difficult to achieve (1). The Position Paper on whole bowel irrigation (WBI) (5) recommended that WBI be considered for toxic ingestions of SR drugs. We present a patient who was admitted on three occasions because of self-poisoning with potassium capsules and was successfully treated twice with WBI.

A 28-year-old woman was admitted to the emergency department because of self-poisoning with potassium capsules. Her medical history included pacemaker in situ, borderline personality, and two previous episodes of potassium poisoning.

Episode 1. Two years previously, the patient was admitted 3 h after ingesting 100 capsules of SR potassium chloride (Kalium retard^®, 750 mg, 10 mmol K⁺). Serum potassium level on admission was 7.2 mmol/L. Shortly after refusing treatment, the patient had a cardiac arrest. Cardiopulmonary resuscitation and administration of IV adrenaline resulted in ventricular fibrillation. She was defibrillated, intubated, and transferred to the ICU. The serum potassium level was 9.2 mmol/L 2 h after admission; the ECG demonstrated pacemaker capture. Blood pressure was 90/35 mmHg. No GI decontamination was performed. Despite IV treatment with calcium, sodium bicarbonate, and dextrose-insulin and oral sodium polystyrene sulfonate resin, the serum potassium level remained elevated (9.2 → 6.9 → 7.5 → 9.5 mmol/L) during the following 8 h. Four hours of hemodialysis normalized the serum potassium level. No further treatment was necessary and the patient was transferred to the psychiatric department 2 days later.

Episode 2. The second episode occurred 4.5 months later. The patient presented 1 h after ingesting 100 capsules of SR potassium chloride (Kalitabs^®, 750 mg). Serum potassium level was 5.1 mmol/L. She was transferred to the ICU, where she again refused treatment. ECG displayed pacemaker rhythm and blood pressure was 140/60 mmHg. Despite her refusal, the patient was anesthetized and intubated, and gastric lavage was performed with limited results. Serum potassium level was 6.9 mmol/L 2 h after admission. Treatment with IV dextrose-insulin and sodium bicarbonate and oral sodium polystyrene sulfonate resin was initiated. Chest X-ray, performed 3 h after admission to check the position of a central venous line, showed numerous capsules in the stomach and WBI was initiated. Over the next 12 h, 10 L of polyethylene glycol electrolyte solution was administered by nasogastric tube. This resulted in diarrhea containing large numbers of capsule fragments and some whole capsules. During this period, the serum potassium level never exceeded 6.9 mmol/L (6.2 → 6.8 → 5.7 → 5.3 → 4.4 mmol/L). Twenty-four hours later, the patient was extubated and transferred to the psychiatric clinic.

Episode 3. On the third overdose, she presented 2.5 h after ingesting 70 capsules SR potassium chloride (Kalitabs^®, 750 mg). The serum potassium level was 7.1 mmol/L and serum creatinine 69 μmol/L (0.78 mg/dL). ECG revealed sinus rhythm of 70 beats/minute with prominent peaked T-waves. In the ICU, she vomited once. Four capsules were found in the vomit. She again refused to cooperate but was anesthetized and intubated. Sodium bicarbonate and dextrose-insulin were administered IV. Abdominal X-ray revealed multiple capsules (Fig. 1) and WBI was initiated. Over the next 6 h, 4 L of a polyethylene glycol electrolyte solution was administered by nasogastric tube causing diarrhea that contained numerous capsule fragments and several whole capsules. The serum potassium levels during that period were 6.5 → 3.9 → 4.9 → 5.5 → 4.4 mmol/L. Potassium level remained normal without further treatment. A second abdominal X-ray approximately 11 h after the first showed no capsules in the GI tract (Fig. 2A and B). Eight hours later, the patient was transferred from the ICU.

Fig. 1.  Abdominal X-ray from the patient's third overdose episode, taken approximately 3.5 h after ingestion and displaying numerous radio-opaque capsules in the stomach (arrow).

Fig. 2.  (A, B) Abdominal X-rays from the patient's third overdose episode, taken approximately 11 h after the first X-ray. No capsules were visible in the gastrointestinal tract.

In the first episode, abdominal X-ray and WBI were not performed. The 8-h period of sustained hyperkalemia, despite vigorous potassium-lowering treatment before hemodialysis, was due to continuous absorption. In the second episode, the presence of capsules on chest X-ray prompted the ICU staff to consult the Poisons Center, which recommended WBI. This treatment, combined with traditional potassium-lowering measures, rendered hemodialysis unnecessary. In the third episode, an emergency abdominal X-ray revealed multiple capsules. WBI was immediately started and reduced the need for prolonged potassium-lowering measures and shortened the ICU stay.

WBI has been used for many years for gut cleansing prior to radiographic investigations (6). The intervention was introduced as a GI decontamination procedure after poisoning in the late 1980s (7). Volunteer studies indicate that WBI variably reduces drug absorption (8–10). Several reports of WBI in poisoned patients have been published, but only one involving SR potassium (1). WBI should not be used routinely in poisoned patients, but the clinical courses of these three episodes suggest that it should be considered in massive self-poisoning with SR potassium chloride.