Adverse Drug Events in Children Analyzed
Although practitioners recognize that medication use in infants and children requires extra caution, few data have been available on medication errors in pediatric inpatients. This study was conducted during a six-week period in April and May 1999 at two Boston teaching hospitals, one a freestanding pediatric hospital and the other serving adult and pediatric patients. At both hospitals, physicians handwrote orders.
ADEs were defined as injuries that result from the use of a drug. ADEs associated with medication errors (errors in ordering, transcribing, dispensing, administering, or monitoring) were considered preventable. Medication errors with significant potential for injuring a patient were considered potential ADEs.
Nurses, pharmacists, and physicians identified errors, potential ADEs, and ADEs from patient charts and from voluntary and solicited reports from house physicians, nurses, and pharmacists. Two physicians reviewed the findings, rating ADEs and potential ADEs for preventability and severity of patient injury.
Fifty-five errors per 100 admissions (5.7 per 100 orders) were identified. There were 26 ADEs, or 0.24 per 100 orders; 5 of these were preventable. Potential ADEs totaled 115 (1.1 per 100 orders and 10 per 100 admissions).
By age group, neonates had the highest rate of potential ADEs: 20 per 100 admissions. The neonatal intensive care unit (NICU) at one of the hospitals was included in the study; it had 46 potential ADEs per 100 admissions, compared with 9 for neonates in the other wards studied.
Overall, the greatest percentage of errors were dose related (28%), and 34% of potential ADEs were due to dosing errors. Most of the errors (74%) and potential ADEs (79%) were associated with physician ordering. Many of the ordering errors were corrected before the order reached the pharmacy.
In addition to preventing most potential ADEs, either CPOE or clinical pharmacists on the wards could have prevented four of the five preventable ADEs, said the physician reviewers. For this assessment, the pharmacist's role included full-time participation in work rounds, monitoring medication administration record transcription, communicating with satellite pharmacies, and assisting nurses with dose calculation and administration. One of the study hospitals had ward-based pharmacists, but they were not involved full-time in these activities.
Three fifths of the 115 potential ADEs (68, or 59%) were intercepted. Eighteen (16%) of the potential ADEs were potentially fatal or life-threatening, according to the physician reviewers, while 52 (45%) were serious and 45 (39%) were significant. In comparison, of the five preventable ADEs that occurred, four were classified as serious and one as significant.
The authors compared their findings with a 1992 study in adults. The rates of errors, ADEs, and preventable ADEs were similar in the two studies, but the potential ADE rate was significantly different (1.1% or 115 in the 10,778-patient pediatric study and 0.35% or 35 in the 10,070-patient adult study).
While CPOE has been shown to reduce errors, noted the authors, the same factors that make children vulnerable to errors complicate the development of CPOE systems for pediatrics. For example, pharmacokinetics and appropriate dosages change rapidly in neonates, and a computerized system would need the flexibility to respond to these changes.
Support for the study came from the Risk Management Foundation and the federal Agency for Healthcare Research and Quality (AHRQ). A recent AHRQ summary of research on preventing ADEs and reducing associated hospital costs (www.ahrq.gov/qual/aderia/aderia.htm) notes that ADEs result in more than 770,000 injuries and deaths each year and can cost up to $5.6 million per hospital.
- Kaushal R, Bates DW, Landrigan C et al. Medication errors and adverse drug events in pediatric inpatients. JAMA. 2001; 285:2114-20.
Chain of Errors Leads to Child's Death After Morphine Overdose
A nine-month-old girl died April 11 at Children's National Medical Center in Washington, D.C., after receiving two 10-fold overdoses of morphine. A Washington Post report said a nurse on a medicalsurgical floor administered 5-mg intravenous doses two hours apart to the girl, who was recovering from surgery that day.
The surgeon had handwritten an order for 0.5 mg as needed for postoperative pain but had omitted the zero before the decimal point, and a clerk transcribed the order as 5 mg. The hospital told the Post that the nurse administering the overdoses had more than 10 years' experience and had worked as a contract employee at the hospital for at least 2 years.
After the Post report, the hospital declined to answer questions until root-cause analysis had been completed and a corrective action plan submitted to the Joint Commission on Accreditation of Healthcare Organizations (JCAHO). Therefore, pharmacy staff could not describe for AJHP at press time the hospital's systems for ordering medications and handling controlled substances.
The hospital said in an April 20 statement that "immediate measures regarding the writing, transcription, and administration of medication" had been taken. An overview of the hospital's most recent accreditation visit, in September 2000, on the JCAHO Web site indicates that medication use was rated 1 on a scale of 1 to 5; a rating of 1 means that the organization "consistently meets all major provisions" of the applicable standards.