Clinical Informatics Outcomes Research Group. Inpatient verbal orders and the impact of computerized provider order entry
In 2000, the Institute of Medicine reported that approximately $2 billion in health care spending is the result of preventable adverse drug events. The medication errors leading to these events can occur at any phase in the medication use cycle. The medication use cycle consists of the prescribing, transcribing, preparing, dispensing, administration and documentation phases. Studies have shown the largest number of errors in adult patients occur in the prescribing phase. Although not as well studied, the available evidence suggests prescribing errors account for approximately ¾ of all medication errors in children. The prescribing phase is carried out in one of three ways; the physician writes the order in the chart, enters the order directly into a electronic system or gives the order verbally to member of the care team. The verbal order method of prescribing is generally seen as the most vulnerable and is discouraged by a number of organizations. Cincinnati Children’s Hospital Medical Center (CCHMC) looked specifically at this method of prescribing to determine whether the implementation of computerized provider order entry (CPOE) would decrease the rate of verbal order entry.
Methods: CCHMC collect baseline data from December 2001 to August 2002. During this period, 100 randomly selected discharge charts were audited weekly for verbal orders and unsigned verbal orders. Staff auditing the charts identified verbal orders by the notation “verbal order” or “vo” entered next to the order. “The rate of verbal orders was defined as the number of verbal orders divided by the total number of orders sampled.” Upon implementation of the CPOE system all data concerning verbal orders was collected electronically. The overall data collection period was from December 2001 to January 2004.
Results: The baseline data revealed that 23% of all orders were placed verbally and 43% of the verbal orders were unsigned. Where as in the 21 months after CPOE implementation, the verbal order entry rate decreased to 10% and only 9% were unsigned. Of the 10 % of verbal orders place, 38% were medication orders. The majority of verbal order entry occurred during the week and outside the hours of 5 am to 8 am. Psychiatry units were responsible for 74% of the verbal orders placed.
Conclusion: Kaplan et al. attributed the decrease in verbal order entry to the ability to place orders from any place within the institution, increased nursing enforcement of policies that discouraged verbal order entry and computer order entry was more efficient than verbal order entry. The decrease in unsigned orders was a combination of process improvement and CPOE system functionality. The CPOE system notified physicians at logon that there were unsigned orders needing signature. Additionally, the Health Information Management Department monitored compliance with the policy. If a clinician was out of compliance with the policy, an email was sent with a list of unsigned orders and a link to the policy. Thus, they concluded, that CPOE combined with process improvement can have a significant and positive impact on the rate of verbal order entry and rate of unsigned verbal orders.