When discussing hospital pharmacy practice, it is important to remember that it has operational and clinical components, both of which may benefit from analytics.
Operational issues generally relate to the purchase, storage, and distribution of medications in the various locations within the hospital.
Clinical issues relate to the appropriate use of those medications to treat various diseases, managing cost-effectiveness of therapy, dealing with potential issues related to drug use (such as allergies and drug interactions), patient education about their medications, and, occasionally, research.
The line between them can become blurry at times, because even clinical practice has an operational component. For example, diversion analytics have features of both fields.
It is also important to realize that pharmacy cost is primarily driven by the cost of drugs the pharmacy manages and not by headcount. So, from a cost perspective, analytics that help manage the cost of drugs are more important than analytics aimed at optimizing headcount. Managing this cost involves both clinical and operational components.
From an outcomes perspective, cost is ultimately secondary to maintaining medication availability. When a patient needs a medication, it must be available when there are no suitable alternative.
The pharmacy therefore performs a balancing act to ensure just the right amount of inventory is on the shelf. This guarantees enough medication is available to allow medical care to be rendered in a timely fashion and minimizes waste from expiring medication and needless consumption of operational dollars
Modern hospital pharmacies tend to use decentralized inventory stores in automated dispensing cabinets (ADC) in patient care areas, in addition to the stores within the pharmacy itself to make the most commonly used medications immediately available. This improvement in medication availability comes at the cost of complicating the process of managing the balancing act previously described. Variations in patient populations being managed can further complicate this assessment.
In addition to this, there are hundreds of medications in short supply that the pharmacy constantly struggles to manage handle. To manage these shortages, the pharmacy needs visibility into all the locations in which such a medication may be stored and its use to plan mitigation of a new shortage.
Operational Analytics
Operational analytics can therefore answer the following questions:
- Is the pharmacy sufficiently staffed?
- Are there scheduling changes that would better align staffing with the flow of work?
- Is the pharmacy inventory being managed properly?
- What is the inventory turnover in the pharmacy ?
- How often does the pharmacy run out of medications on its shelves?
- What is the cost of medications that expired and had to be discarded?
- Could ordering more frequently reduce the standing inventory?
- Are each of the decentralized inventories being managed properly?
- Are each of the ADCs stocked with the right drugs given current patient need?
- Are those medications stored in the correct quantities to ensure availability to the patient care staff?
- Does the pharmacy refill activity conflict with the needs of caregivers to withdraw medications for administration to patients?
- How often does the pharmacy make “emergency” trips to refill a needed medication?
- How does the pharmacy manage drug shortages?
- How many shortages is the pharmacy currently managing?
- Could supplies of a medication not being used be relocated to better handle the shortage?
- For a substitute drug, is there enough in stock to replace the drug in short supply?
The complexity of having to deal with decentralized inventories makes older traditional mechanisms of operational management unsustainable. Therefore, pharmacies need analytical tools to acquire and organize the transactional information from both the pharmacy and decentralized inventory systems and to compute what changes may need to be made in purchasing and distribution of medications to meet the needs of the patient care environment, to minimize loss to expiration, and to manage medications in short supply. The pharmacy also needs analytical tools to look at the ebb and flow of work within the pharmacy, to see how the various personnel are spending their time, and whether they are scheduled to meet the demands of the work.
Clinical Analytics
Clinical pharmacists spend their time attempting to ensure that the medications being administered to patients are appropriate from both a safety/efficacy and cost perspective. For example,
- Is a patient being administered medications intravenously when they would receive the same clinical effect with oral drugs?
- Is the antibiotic ordered for a patient effective against the microbe with which they are infected?
- Is the combination of medications a patient is taking likely to cause an adverse reaction?
- Are there more cost-effective medications that might be used?
- Are there patients whose doses need to be customized based on their situation? If so, perform dosing calculation services.
Analytic systems that monitor laboratory results, vital signs, new medication orders, and other key information can provide the answers to these and similar questions.
Diversion Analytics
Controlled substances are medications that have clinical benefit but can cause substance abuse disorder. Recent evidence has shown that caregivers with substance abuse disorder can divert these medications for personal use by falsifying medication administration records, making it difficult to detect. Identifying potential diverters requires complex analytical modeling of normal behavior within a caregiver population to identify those whose usage stands out as abnormal and, therefore, possibly indicative of diversion. Further investigation is always required, but the analytics can help identify persons of interest. In the absence of these analytics, investigators struggle with where to begin.
In summary, this discussion highlights how hospital pharmacy has become an increasingly complex profession that increasingly requires analytics to address the complexity.
The ideas expressed herein are my own, and not necessarily those of my employer.
