View source for Galanter WL, Polikaitis A, DiDomenico RJ. A trial of automated safety alerts for inpatient digoxin use with computerized physician order entry. J Am Med Inform Assoc. 2004 Jul-Aug; 11(4): 270-277


A Trial of Automated Safety Alerts for Inpatient Digoxin Use with Computerized Physician Order Entry

Reducing medication errors for hospitalized patients would lead to fewer adverse drug events and would therefore lower the cost of care, improve outcomes and increase the overall quality of care delivered. Digoxin is a drug used to treat heart failure and arrhythmias, and can reach toxic levels in patients when existing electrolyte levels are abnormal. This study aimed to evaluate the effectiveness of clinical decision support (CDS) in improving the safe use of digoxin. Over a period of six months, CDS alerts were delivered at the time of digoxin ordering and when a patient’s electrolyte levels were newly reported as low. The control data was taken from a chart review of the six months prior to CDS implementation. This study measured the speed and magnitude of clinician response.

The effectiveness of two alerts were evaluated. Synchronous alerts were informational alerts given to the clinician at the time of digoxin ordering. Alerts were invoked when digoxin was ordered for a patient without current digoxin or electrolyte levels available, or with low potassium or magnesium levels reported. Asynchronous alerts were invoked upon laboratory results indicating that a patient has newly abnormally low electrolyte levels. Asynchronous alerts were delivered via printouts and e-mail.

The results showed a great improvement in the study group’s response to synchronous alerts regarding unknown electrolyte and digoxin levels. Prior to CDS implementation, only 9% of clinicians had ordered potassium levels, and only 12% had ordered magnesium levels within the first hour after the digoxin order. With the alerts invoked, the respective numbers increased to 57% and 40%. At 24 hours, the control group orders were 49% for potassium and 44% for magnesium compared to the study group who showed additional increases to 81% and 66%.

When low levels of potassium or magnesium was reported at time of digoxin order through synchronous alerts, the alerts were not statistically improved. 

When laboratory reports showed new abnormally low electrolyte levels, clinicians in the study group were quite responsive to the asynchronous alerts. They ordered magnesium supplements at a rate of 49% compared to a control rate of 5%. Study group clinicians ordered potassium supplements at a rate of 35% versus 6%. 

Comments: This study was quite limited, as it seems alerts would be even more effective if it were easy to add on additional orders at the time of digoxin order, rather than just make a suggestion. Additionally, the CPOE should ensure clinicians are more aware of the clinical situation, including relevant electrolyte levels during digoxin ordering. If current levels have not been reported, then make it easy to order them along with the digoxin. This is less annoying than an alert after the fact.

The study did not measure the effectiveness of the increased orders on patient outcome. As it is the patient outcomes that clinicians want to improve, this needs to be studied further.