Phoenix Children’s uses big data to eliminate pediatric medication errors

Using data analytics to comb through eight years of patient records, Phoenix Children’s Hospital has eliminated all overdosing medication errors among pediatric patients for the last six years.

By analyzing more than 750,000 patient records and identifying specific dosage thresholds, the hospital developed an automated alert system that informed clinicians when a dose is too high, according to an op-ed by David Higginson, chief information officer at the Phoenix Children’s Hospital in Healthcare IT News, a HIMSS publication.

Dosing errors among pediatric patients can have severe consequences for a patient population in which age and weight vary considerably. Computerized physician order entry has been shown to reduce medical errors, while also creating new problems involving dosing.

RELATED: Lean IT strategy pays off in Phoenix Children’s modernization initiative

The automated system uses the dosage thresholds to issue “soft stop” and “hard stop” alerts when an order is entered, forcing the physician to confirm an order with the pharmacist, Higginson wrote. The hospital has not had an overdose at the prescribing stage since the system was implemented in 2011, and the success of the initiative has prompted other units of the hospital to adopt automated alerts for isolation orders and pregnancy screenings for specific patient populations.

Automated systems have been shown to reduce medical errors by more than half, and other children’s hospitals have sought the use of algorithms and big data to automatically detect potential overdoses. Recent research has found incorporating genetic information into EHRs can also reduce adverse reactions to medications.