Among the expected benefits of the conversion to computerized provider order entry (CPOE) from a paper-based system is a reduction in the number of home-medication admission discrepancies that hinder the effectiveness of subsequent therapy. Logic dictates that reducing medication reconciliation errors upon admission will decrease medication errors in general, but we wanted to see what would happen if we eliminated pharmacist review upon admission. At Mercy Health Center, in Oklahoma City, we sought to identify the number of discrepancies and medication errors we were experiencing using a CPOE system during the admission medication reconciliation process, with the end goal of reducing such instances at this crucial juncture in patient care.

Pre- and Post-CPOE Implementation
The governing method of this project was to analyze the impact of a newly implemented CPOE system on admission med rec error discrepancies and error identification in contrast to the previous, paper-based system. To facilitate this process, we defined discrepancies as having one of the following characteristics:

• Medication not listed (omission)
• Incorrect dose, route, or frequency
• Previously discontinued medication listed on patient admission medication list

Pre-CPOE
Prior to implementing CPOE in June of 2010, the process of performing medication reconciliation upon admission was a personnel-centric, paper-based system. In that system, an admitting nurse would collect the initial medication list from the patient for admission reconciliation on the home medication list for admission reconciliation (HMLAR) form. A pharmacy home-medication worksheet was used by the med rec technician (MRT) to verify the initial medication list collected by the nurse upon the patient’s admission to a specific patient care unit. Using the HMLAR form, the MRT then verified the accuracy of the patient-provided medication list by reviewing it with the patient, care giver, appropriate prescriber, or by other means such as contacting the local community pharmacy. The on-duty pharmacist then verified the worksheet for dose, route, frequency, and medication accuracy prior to placement of the final home-medication list into the patient chart. The new medication list was then entered into the eMAR and placed with the patient’s chart for physician verification and signature to change, continue, or discontinue home medications. Any discrepancies in dose, route, omission, or discontinuation of medication were then recorded onto a data collection sheet for analysis. Having this many steps and hand-offs to multiple individuals created numerous opportunities for error.



Post-CPOE
In July of 2010, after having implemented a CPOE system, the process for admission med rec changed significantly. Under this new system, the nurse collects the home medication information upon admission, but then directly enters that information into the patient’s medication profile in the new CPOE system under the heading Prior to Admit (PTA) Medications. The MRT verifies the accuracy of the medication list in the profile and any changes are noted on the medication worksheet and corrected on the PTA medication list. An important distinction is that pharmacist involvement in reviewing the home medication list was initially removed from the new CPOE system.

Admission Med Rec Error Results
Over the three-month study period—June through August of 2010—a total of 18 medication errors were discovered. During the pre-CPOE period in June, four errors were detected including hydrocodone 7.5 mg being ordered as scheduled versus on an as-needed basis and a previously discontinued medication being entered and scheduled to start on the medication profile. The post-CPOE implementation phase saw a significant uptick in medication errors detected as we discovered a total of 12 errors in July and an additional two in August. The more significant errors detected included:

• Levothyroxine 50 mcg—instead of the proper 150 mcg—was entered and given to the patient before the MRT interview and verification of medication dosage
• Lisinopril 40 mg was ordered for an 81 year-old patient who was admitted for a fall
• An incorrect insulin product selection of NovoLIN was entered when NovoLOG should have been selected for patient home sliding scale

Unintended Revelations
It was clear at the end of this study that removing pharmacist verification of home medication information increased the number of medication errors reported. The use of a newly implemented CPOE system demonstrated a relationship between medications reviewed, number of medications correct on admit, and the number of medication discrepancies—the less accurate the initial home medication list, the greater the chance an error would occur. So while our new CPOE system brought much needed standardization, it may also have provided a false sense of security regarding the accuracy of the admission medication list.

It is important to remember that the positive effects of medication reconciliation are not only dependent on proper execution upon admission, but also throughout the continuum of transfers in care and patient discharge. As well, simply because you introduce a computerized system does not mean the information entered therein is correct. For long-term positive effects, pharmacist review of home medication lists at admission is critical and a thorough resolution of all detected errors is necessary to improve the quality and safety of provided services.

Burl Beasley, MPH, MS Pharm, is manager of pharmacy informatics at Community Health Systems in Franklin, Tennessee. Burl received his MPH from University of Oklahoma College of Allied Health and his MS from University of Florida, College of Pharmacy.