Experience is the Best Teacher

David M. Gaba, M.D.; Steven K. Howard, M.D.

In a previous edition of the Anesthesia Patient Safety Foundation Newsletter we requested that you report interesting safety-related events. The goal of this column is to advance safety by collecting event reports from anesthesia providers, analyzing the reports, and informing the anesthesia community about these occurrences and the lessons to be learned from them. The reports we have received to date have each been instructive.

“What harm can there be in giving a little potassium?”

A staff anesthesiologist at a university medical center was concurrently supervising two cases, each with an anesthesia resident. One case involved a patient undergoing resection of a cerebral AVM which had been proceeding uneventfully for 12 hours. The resident told the attending that he wanted to administer some potassium (K+) because the last serum K+ was 2.1 mEq/1. The staff commented that the hypokalemia was due, in part, to the hyperventilation which was in routine use for this neurosurgical case. The resident acknowledged this effect but argued that the diuresis produced by the routine administration of mannitol had increased urinary K+ losses, and besides, ‘…what harm can there be in giving a little potassium?’ The staff anesthesiologist left the decision up to the resident and went to the next room to check on his other case.

About an hour later the case was taken over by the on-call resident after a hand-over from the original resident who informed his colleague that he had given 40 mEq of potassium about an hour before. Shortly after the hand-over, the attending returned to check on the case. He found the resident reaching up under the drapes. The resident said that the arterial trace had dampened and that he was checking to see if there was a kink. The staff looked at the monitor, found that the ECG, arterial, pulmonary artery, and central venous pressure tracings were all flat, and that the end-tidal C02 had dropped markedly. The surgeons were informed that the patient’s heart had arrested and CPR was initiated.

Because the staff anesthesiologist suspected that the arrest was secondary to iatrogenic hyperkalemia, he insisted that all bags of IV solution be taken down and replaced with freshly opened bags of saline. After approximately six minutes of CPR the cardiac rhythm became sinus tachycardia with good arterial perfusion. A serum K+ drawn after the resuscitation showed a K+ of 9.2 mEq/l! The surgeons completed the procedure expeditiously and the patient was taken to the ICU in barbiturate coma. The patient subsequently did well with no deficit linked to the intraoperative arrest.

Afterwards, the staff and the residents were able to reconstruct the chain of events. The first resident elected to administer 40 mEq of KCl diluted in 100 ml of saline using a buretrol “minidrip” infusion set (60 drops = I ml) piggybacked into an infusion port on a blood pump peripheral IV set (See Figure). The buretrol was hung at a higher level than the peripheral IV which was running at a ‘keep open’ rate for the entire case. After the 100 ml of potassium-containing solution had dripped in, the infusion set was removed before the on-call resident took over the case. When the on-call resident opened the IVs to give some volume, a bolus of K+ (residing in the blood pump IV set and the remaining lactated Ringer’s solution) was inadvertently administered.

The report states that the hospital has since formulated rules that K+ will not be administered intravenously unless its flow rate is controlled by a mechanical infusion device (pump or drip controller).

Analysts’ Comments

The main culprit in this case was the administration of a drug into a large volume IV component with a slow rate of IV flow. In addition, the buretrol was hung at a higher level than the main IV which resulted in a ‘reservoir’ of drug that was rapidly infused when the IV was opened. Note that the hospital’s new policy, while sensible for other reasons, would not necessarily prevent a recurrence of this event, since the potassium containing solution could still be administered into the “reservoir” by a mechanical device. A better policy change might be to require all piggybacked drugs to be infused using ports which are as close to the patient as possible and with adequate flow of the carrier solution to ensure the drug reaches the patient in a timely fashion.

There are other interesting issues raised by this case. When the cardiac arrest unexpectedly occurred, the resident’s initial reaction was to assume that the arterial wave form had dampened due to a mechanical artifact without verifying that the patient was OK. The persistent attribution of abnormalities to artifacts in the face of abundant evidence to the contrary is a type of error the psychologists call a ‘fixation error”. Whenever seeing an abnormality, the “burden of proof’ is on the anesthetist to prove that the patient is OK before assuming it’s an artifact. In this case, feeling a pulse, listening for heart sounds, or looking at the ECG, filling pressures, and capnogram would each have demonstrated that the ‘damped” arterial tracing was a real and lethal finding.

It is not clear in this case whether a more thorough hand-over between the residents would have helped avoid this complication, although it is possible that the on-call resident might have realized that administering the KCl in the proximal portion of the IV set could result in a reservoir effect.

Regardless, complete transfer of information about a case is imperative for safe patient care whenever a change of personnel takes place.

A final word about potassium. Unlike many drugs, for which the inadvertent administration of a few n-d would be inconvenient, the administration of potassium can be immediately lethal. Extraordinary care should be used however it is administered.

Perhaps this case answers the question: “What harm can there be in giving a little potassium?’

Dr. Gaba, Associate Professor of Anesthesia at Stanford University Medical Center and a Staff Anesthesoiologist at the Palo Alto Veterans Affairs Medical Center, is Secretary of the APSF. Dr. Howard is a Clinical Instructor of Anesthesia at Stanford University Medical Center and a Staff Anesthesiologist at the Palo Alto Veterans Affairs Medical Center.

Figure 1. IV set as described in the case report. The level of the buretrol is higher than the peripheral IV solution allowing for retrograde flow and a reservoir of potassium containing solution in the IV set. Potassium level in the remaining LR was measured at 9 mEq/l.