Several points of descriptive information also were collected. The average days’ supply trended down from a mode of 10 days to 7 days. This reduction in days’ supply likely was seen because staff became more aware of the customary amount required to bridge the veteran until the CMOP supply arrived by mail. Postintervention data showed a 70% utilization of partial-fill reasoning codes. The reasons for partial filling of prescriptions are summarized in Table 2.
Discussion
Following the implementation of the updated partial-fill guidance and provision of education to the FVAMC veterans and staff, a noteworthy cost savings was observed with respect to both material and workload expenditures. This large reduction in expenditure likely was not related to a reduction in the total prescription volume, as the number of total prescriptions filled by the CMOP and at FVAMC were similar in both the preliminary and postintervention periods. When the results of this 3-month QI project were extrapolated, the annual projected cost avoidance was $91,949.12.
Of note, there was no established process for adjudicating appeals to the partial-fill guidance. Any extenuating circumstances that fell outside the guidance were addressed by the outpatient pharmacy supervisor. There was no formal documentation for these disputed cases. Since there was no prespecified supervisory override code, the most appropriate partial-fill code was entered into the comments field for these scenarios. As such, there is no way to distinguish precisely how many of these partial fills were escalated to a supervisory level for a decision.
The positive fiscal impact noted from the implementation of this project should not be viewed as the only utility for such guidance. Though not directly measured within the confines of this project, a reduction in pharmacy staff time spent on partial-fill prescriptions will likely result in shorter pharmacy wait times, line lengths, and streamlining of pharmacy workflow. When the pharmacy staff is free to work on pressing issues rather than on continually educating veterans on the partial-fill or refill process, many will benefit. Veteran satisfaction was not directly measured during this project but could be an interesting topic to review as a future study.
Each VA facility is unique, with its own challenges for implementation of a project such as this. Nevertheless, the incorporation of a formal guidance and education process, perhaps adapted to the indi vidual facility’s needs may be considered for overall pharmacy operations QI.
Limitations
During the preliminary data collection period, FVAMC and its catchment area were impacted by a natural disaster, Hurricane Matthew. Based on a review of the text entered into the comments field for all partial fills, about 4% of the partial fills completed in the preliminary phase can be attributed to the hurricane. The effects of this hurricane may have potentially increased the number of partial fills completed in the preliminary phase compared with that in the postintervention phase, due to the number of veterans who were temporarily or permanently displaced from their homes. This increase in partial fills and associated expenses preintervention likely caused a slightly higher cost savings to be reflected in the postimplementation phase than what would have traditionally been observed without extenuating factors.
Several other limitations must be considered for this QI project. The implementation phase, during which all education and training was completed, was only 1 month. A longer implementation period and more opportunities to educate veterans and staff might have created a greater impact on the results. Additionally, because there were no data collected on New Patient Orientation attendance for this project, it is unclear exactly how many veterans received refill-process education through this outlet.
Though all staff members were trained on the appropriate process, it was discovered during interim analysis that several pharmacists were not following the partial-fill guidance, potentially negatively impacting the results. It is likely that staff would have benefited from continual reeducation of the process throughout the entirety of the project, as the restriction of partial filling was a novel concept to many. In addition to continual reeducation of current employees, any new hires would likely need this information as part of initial training.
Cost variance in the type of partial fills completed between the preliminary and postintervention phases also may have negatively impacted the results. The postintervention phase contained 2 high-cost classes of drugs (antivirals and immunoglobulins) that received multiple partial fills but were not partialed in the preliminary phase, which increased the raw drug cost in the postintervention phase.
