Results relate to measurement of four key performance indicators.
Adverse drug events: Results for adverse drug events prevented between Aug. 1, 2009, and Jan. 31, 2010, were reported to the board of trustees. The report describes how real-life human errors were “reprogrammed” using the software technology to prevent overdoses of medications. The mathematics to calculate return on investment was based on three components:
- Reprogramming events.
- Events where the entered amount was at least 2½ times the “soft” limit.
- High-risk drug events likely to cause harm based on inherent risk (e.g., chemotherapy, narcotics, sedatives, insulin and heparin).
During the measurement period, 261 severe potential adverse drug events were prevented by use of the smart pump technology. Using an average cost of adverse drug events from the Institute of Medicine in July 2006 of $8750 per event, a total of $2,283,750 was saved.
Patient controlled analgesia: Patient controlled analgesia, or PCA, errors represent a fourfold higher risk than other reported medication errors due to the nature of the drugs. Internal investigation showed that programming errors with PCA pumps were most often related to the wrong concentration. Changes implemented include:
- Adoption of systemwide PCA physician orders.
- One standard concentration for PCA syringes (fitted into the smart pumps).
- New guidelines for respiratory assessments and frequency of monitoring.
- A two-tiered approach for use of the drug Narcan.
Computer-based learning modules were developed and the electronic nursing documentation system was updated to align with the new requirements. Since implementing the new standards, OhioHealth has experienced zero PCA programming errors reported based on the nationally accepted severity ranking.
End tidal carbon dioxide monitoring:An additional smart pump module noninvasively measures several respiratory parameters to produce a waveform. The technology monitors breathing, detects “no breath,” measures carbon dioxide (exhaled), detects early hypoventilation, and is not affected by concurrent administration of supplemental oxygen. It is implemented at one hospital. An anecdote tells the story.
“An elderly female patient was admitted postoperatively with a PCA pump and end tidal module in place. ‘No breath' and low respiratory rate alarms persisted, so the nurse called the surgeon to discontinue the PCA pump and switch to another method for pain control. Initial readings demonstrated that the patient had high retained CO2, a very slow respiratory rate and ‘no breath' alarms at 30 seconds. Using smart pumps, an alarm would sound that aroused the patient enough to take a breath when she would fall asleep and stop breathing. The family was reassured that the nurses, as well as the smart pump technology, were ‘watching over their mother'.”
Infusion pump programming: OhioHealth reports on nonintercepted medication events according to a nationally accepted severity scale. Events are classified from severity “C” through “I” for all errors that actually reached a patient and were not intercepted. The reported data for four OhioHealth hospitals between January 2007 (the baseline) and January 2010 demonstrates that OhioHealth is steadily decreasing its average number of infusion pump programming errors per month from 12 (the starting point) to seven (the latest reported number).
OhioHealth does not plan to apply for economic stimulus funds for this project; however, other information technology projects are included in our plans for “meaningful use” grant monies.