Pharmacist-Provided Pharmacogenetic Point-of-Care Testing Consultation Service: A Time and Motion Study

Abstract

Background: With recent advances in pharmacogenomics (PGx) comes the potential to customize medication use based on genetic data. Support for PGx has found practical limitations in terms of workflow and turnaround time of a test. However, with the expansion of point-of-care testing (POCT) in pharmacy practice models comes opportunity for PGx testing in the pharmacy setting. Objective: The purpose of this study is to quantify the amount of time spent during each step of a PGx POCT encounter in a community pharmacy setting. Methods: A time and motion study was conducted using a mock community pharmacy space for a simulated PGx-focused encounter to manage antiplatelet therapy following hospital discharge. PGx POCT was conducted using the Spartan RX instrument. Simulated patient encounters were divided into 7 categories. Time spent in each step, as well as total time spent, was tracked. Results: A total of 54 simulated PGx POCT encounters took place with an average time of 9.49 minutes (SD ± 1.38 minutes). Instrument run time adds 60 minutes to the total time required to obtain a result. Duties that could be performed by an appropriately trained pharmacy technician totaled 6.86 minutes. Conclusions: PGx POCT would require 9.49 minutes of pharmacy staff hands-on time for the encounter, which could be reduced to 2.64 minutes of pharmacist time with appropriate pharmacy technician involvement. Time requirements for PGx POCT are similar to that of community pharmacy–based immunizations. Future studies could explore how practice could change if PGx testing were routinely performed in the pharmacy.

Introduction

The ability now seen in modern medicine to harness genetic data to better predict drug response creates great hope for reducing adverse events and improving the chances of finding efficacious therapy for patients.¹ This science of pharmacogenomics (PGx), or the study of how a patient’s genes affect his/her response to drugs, may yield great opportunity for personalizing medicine based on genetic information.¹ Pharmacy organizations have supported the inclusion of PGx into the clinical decision process, with the American Pharmacists Association specifically calling for its inclusion in the Medication Therapy Management (MTM) model.^2-4 MTM is often conducted in the community pharmacy setting, and it involves a direct encounter with patients where pharmacists work to manage medication use in an effort to optimize drug therapy and improve therapeutic outcomes. Given that nearly 90% of patients live within 5 miles of a community pharmacy, the accessibility of the pharmacist plus the clinical application of PGx creates a promising environment for supporting personalized medicine.⁵ Advances in technology have now led to point-of-care testing (POCT) solutions for PGx testing as well.⁶

Multiple studies have demonstrated that incorporating PGx into community pharmacy practice is viable, specifically when supporting antiplatelet therapy optimization.^7,8 However, a common barrier to the integration of PGx into community pharmacy workflow is that current models involve sending a PGx sample off to an off-site laboratory for analysis. This analysis has prompted the need for either a follow-up with the patient after the encounter or a second encounter with the pharmacist.^7,8

Recently, studies have demonstrated the value of using Clinical Laboratory Improvement Amendments (CLIA)–waived POCT as a component of pharmacy-based care models.^9-11 Fundamental to the successful integration of these CLIA-waived POCTs into viable pharmacy practice models was the understanding of how the tests would affect pharmacy workflow and how patients would move through the care model. By describing the workflow of the clinical service, these variables can be characterized. It has been determined for a number of different pharmacy-based care models, which employ CLIA-waived POCT, that pharmacy staff hands-on time throughout the process is typically less than 10 minutes, and the overall time required to deliver the service is driven by the time required to run a test and the amount of patient education required.^12-14 These data were used to manage patient time expectations, anticipate pharmacy staff requirements, and determine pricing for pharmacy-based services.

POCT has been adopted into the clinical workflow within the hospital setting to specifically assess PGx data pertinent to antiplatelet therapy optimization.⁶ However, no studies have assessed the potential viability of a PGx POCT workflow model in the community pharmacy setting. The purpose of this study is to quantify the amount of time spent during each step of a PGx POCT encounter in a community pharmacy setting. By determining workflow logistics and requirements, clinicians will be better able to assess viability of PGx POCT implementation in different practice settings, including the community pharmacy.

Methods

Four mock clinical encounter spaces resembling a counseling room in a community pharmacy were used in this time and motion study. Each room contained a Spartan RX point-of-care PGx analyzer system (Spartan Bioscience Inc, Ottawa, Ontario, Canada), a simulated freezer for storing PGx testing supplies, and a consultation table.

Standardized patients presented to the mock pharmacy as posthospital discharge patients that had recently undergone percutaneous coronary intervention following acute coronary syndrome diagnosis. Patients were randomly assigned to present with a prescription for clopidogrel, prasugrel, or ticagrelor. This simulated presentation scenario and workflow model was built to mirror prior community pharmacy–based studies that support patients in similar clinical situations,^7,8 as well as to create a scenario where application of clinical pharmacogenetics implementation consortium (CPIC) guidelines would be appropriate.¹⁵ No analysis of prescription filling time for the prescription took place as those data do not directly relate to PGx POCT service provision. Five pharmacists were provided 2 hours of training specific to PGx and the use of the Spartan RX POCT PGx analyzer, which provides results specific to CYP2C19. The training was similar in content and duration to what was provided in a prior study of PGx implementation in the community pharmacy setting,⁸ but it also included additional training specific to the Spartan RX POCT PGx analyzer. Pharmacists took turns in different clinical encounter spaces throughout the study and conducted the PGx POCT interventions.

Standardized patients completed 1 to 3 iterations of the process each. An observer was assigned to each mock clinical encounter space to track the time of each encounter step using TimeStudy (nuVizz.com). Simulated encounters were composed of 7 steps, and the time spent in each step was timed. The first step represented the time required to greet a patient in the pharmacy waiting area and escort them to the simulated HIPAA (Health Insurance Portability and Accountability Act)–compliant counseling space. During this step, the pharmacist talked with the patient, but it was expected that any discussion of protected health information would be reserved until the patient was in a HIPAA-compliant counseling space, which is when the second step began. Discussion to ensure that the patient was an appropriate candidate for PGx testing based on the clinical scenario, as well as discussion to answer any patient questions prior to sampling, occurred at this step. The third step consisted of the time required for the pharmacist to obtain a PGx sample (set of buccal swabs) from the patient. Once the sample was collected, the fourth step began, which included time required for performing the POCT. The fifth step began when results were obtained from the testing instrument. Time for the pharmacist to review results and apply the results to the patient’s prescription to determine a clinical recommendation was recorded here. Time spent documenting and communicating the results of the recommendation were recorded in step 6. Finally, step 7 represented the time spent counseling the patient regarding the results of the PGx POCT. Encounters did not always occur in a linear manner. Appropriate times were recorded as participants moved in and out of each step.

The practice model that was followed varied in step 6. Twenty-six encounters operated according to a traditional MTM model where the pharmacist would develop a written care plan and fax this recommendation to a prescriber. The remaining encounters operated under a collaborative practice model where a prescriber had previously authorized the pharmacists to provide PGx testing and manage the patient based on expected results.

A fixed instrument run time of 60 minutes was added to the total process time for each subject.¹⁶ The instrument produces test results that identify the presence of *2, *3, and *17 alleles in a sample.¹⁶ The *2 and *3 alleles are the 2 most predominant loss-of-function alleles, and the *17 allele demonstrates gain-of-function. While other alleles have been observed, such other alleles are expressed in less than 0.5% of the individuals in most ethic groups.¹⁷

Owing to the level of professional training required, tasks in steps 1, 2, 3, and 4 were assigned to a trained pharmacy technician, whereas tasks in steps 5, 6, and 7 required the involvement of a pharmacist. Data from the TimeStudy software were downloaded into an Excel spreadsheet and the time for each step of the encounter, total time of steps that could be completed by a pharmacy technician or required a pharmacist, and total encounter time were calculated using descriptive statistics. This project was reviewed by the institutional review board and determined that it did not meet the Federal Definition of research on human subjects and was instead considered quality improvement.

Results

A total of 54 simulated PGx POCT encounters took place. The average total time of a PGx POCT encounter was 9.49 minutes (SD ± 1.38 minutes). A breakdown of time spent in each step of the analysis is shown in Table 1. Instrument run time adds 60 minutes to the total time required to obtain a result. Of the steps listed in Table 1, arrival and initial encounter, verification questions, sampling, and performance of test required 6.86 minutes and could be accomplished by an appropriately trained pharmacy technician. Steps that would require a pharmacist, notably data interpretation, communication/documenting, and counseling, totaled 2.64 minutes.

Table 1.

PGx POCT Workflow (Data Reported in Minutes).

Workflow Step	Mean	Standard Deviation	Median	Minimum	Maximum	Personnel
1. Arrival/initial encounter	0.68	0.32	0.61	0.19	1.39	Pharmacist or technician
2. Verification questions	1.14	0.76	0.81	0.36	3.69	Pharmacist or technician
3. Sampling	2.27	0.49	2.25	1.22	3.38	Pharmacist or technician
4. Performance of test	2.77	0.88	2.67	1.36	4.65	Pharmacist or technician
5. Data interpretation	0.44	0.32	0.33	0.09	1.76	Pharmacist
6. Communication/documentation	1.01	0.64	1.00	0.00	2.90	Pharmacist
7. Counseling	1.19	0.69	1.04	0.14	2.70	Pharmacist
Total	9.49	1.38	9.55	6.36	12.73

Abbreviations: PGx, pharmacogenomics; POCT, point-of-care testing.

In the encounters where the pharmacist was asked to make decisions and communicate with the patient as if under a collaborative practice agreement, the communication and documentation time was comparable to the time spent when the pharmacist simply faxed the results to another provider (1.04 vs 0.97 minutes). If the collaborative practice agreement allowed for the pharmacist to change the prescription, the pharmacist would be able to counsel as such, and there would be no additional time required for waiting on a response from the prescriber to the pharmacist’s recommendation. Any associated waiting time is not accounted for in the model as that could be highly variable based on local laws, prescribing patters, and interprofessional relationships.

Discussion

This article describes a time and motion study aimed at quantifying the time spent during different steps of a PGx POCT encounter in a community pharmacy. As data support that community pharmacists are interested in providing PGx-based support to their patients,¹⁸ objective data regarding PGx POCT logistics may be helpful to guide workflow model development and service implementation considerations. Especially given that previously published models of PGx support via MTM in the community pharmacy setting do not allow for PGx-guided clinical decisions to be made on the same day, let alone within 70 minutes of the encounter start time, evidence from this analysis may be helpful in identifying innovative workflow models to streamline service delivery. Knowing that workflow and logistical considerations are only one element of the implementation process for new PGx services, the reader should additionally consider related literature that discusses further considerations for clinical implementation of PGx services in community pharmacies.^19,20 For instance, exact models of care provision through a single MTM encounter, multiple MTM encounters, or asynchronous support could be discussed in the context of a specific practice setting for implementation.¹⁹ Issues of prescriber communication and health professional education must also be considered at the time of implementation.^19,20 Additionally, PGx workflow integration and provider training for the specific POCT device should be considered prior to clinical service implementation.²¹ While service implementation in the community pharmacy setting may be limited by patient interest and feasibility, patient interest in such services appears to be relatively high, and minimal time is required for the intervention.²⁰ While several studies have focused efforts on the pharmacogenetic intervention of clopidogrel/CYP2C19, the use of genetic testing to guide drug optimization as described in this analysis could certainly be extrapolated to similar models and other clinical decisions involving PGx data.

The results of this study demonstrate that the active staff (pharmacist and technician) are consistent with the time required for other pharmacy-based services.¹² Additionally, counseling time was consistent with available literature.²⁰ That is important, as the ability to fit a new service into the existing pharmacy workflow is a critical factor in the decision to implement. Comparability with existing services makes the case for implementation more feasible.

In June of 2017, Walgreens and LabCorp (a clinical laboratory offering a number of specialty tests, eg, genomic testing via polymerase chain reaction, HIV genotyping, and phenotyping, etc)²² developed a partnership to offer LabCorp patient services and specimen collection within the pharmacy area. The initial 7 “LabCorp at Walgreens” sites will be opened in Denver, North Carolina, and Deerfield, Illinois.²³ Walgreens, also one of the early adopters in pharmacist immunization services,²⁴ has a history of setting the precedent when it comes to large-scale implementation of new or progressive pharmacy services. As was the case with immunizations, the increased uptake of such laboratory testing in community pharmacies beyond those who have obtained CLIA waivers may open the door for moderate complexity tests administered by pharmacists in a variety of community-based practice settings. In the case of this test, the 60-minute run time of the test could express additional logistical difficulties, as it may require the encounter to be broken into 2 components while the patient returns to the pharmacy at a later time or waits in the pharmacy for the duration of the instrument run time. However, while the duration of the instrument run time may be different for different tests, other POCT offered in community pharmacies also require instrument run time, so the issue of waiting for POCT test completion is not unique to this specific POCT. Certainly nondispensing services like POCT and immunizations have the opportunity to disrupt the standard dispensing workflow, but these services have continued to grow to meet patient demand and the needs of the health care system.

It should be noted that PGx data have greater application than to selection of antiplatelet medications. Specific guidelines are available for many medications through CPIC, several of which are influenced by CYP2C19.¹⁵ This specific analysis assumed that the only medications requiring potential adjustment had to do with antiplatelet needs. However, it is possible that further clinical decisions may be appropriate for other medications, with examples including sertraline or voriconazole.¹⁵ While additional clinical decisions for medications that have other uses may require additional time, there may also be additional clinical value through the application of PGx data to other medications. Such additional time should not be taken to mean a doubling or tripling of the overall intervention time, as many of the steps, such as verification questions and performance of test would not need to be repeated when applying data to other drug-gene pairs.

Training needs in this study were consistent with prior work for the specific pharmacogenetic management of CYP2C19/clopidogrel.⁸ However, if additional drug-gene pairs are to be included in a more robust PGx analysis, additional training may be necessary. Formal training opportunities have been discussed elsewhere.¹⁹ However, clinicians may find it helpful to become familiar with resources like CPIC, PharmGKB, and the Genetics/Genomics Competency Center (G2C2) as a way to deepen their understanding of PGx in clinical practice.^15,25,26

Limitations include that the study was conducted in a simulated clinical encounter space and not a community pharmacy during normal business hours, so distractions that may exist in an operational health care environment may have been altered. This approach was suggested to minimize workflow disruption to an operational community pharmacy and mitigate any risk related to accidental disclosure of protected health information but does have some minimal environmental differences as compared with an operational community pharmacy. The study did not consider the accuracy of the pharmacists’ interpretation or recommendation. While accuracy is certainly a critical concern in providing pharmacogenetic testing, accuracy is unlikely to affect the time to interpret the result or provide the recommendation. Additionally, categories of time that could be accomplished by a pharmacy technician were not modeled via an appropriately trained pharmacy technician and instead were modeled by an appropriately trained pharmacist. However, it could be argued that some of the steps, such as PGx sampling, would be unlikely to be substantially different given that sampling requires a specific duration of swabbing. Finally, the pharmacist training on the PGx process occurred on the same day as the study observations. As such, the pharmacists had limited repetitions prior to the study, which may have lengthened the observations.

Conclusion

PGx POCT would require 9.49 minutes of pharmacy staff hands-on time for the simulated encounter, plus 60 minutes of instrument run time. Involving appropriately trained pharmacy technicians could reduce the pharmacist hands-on time to 2.64 minutes, creating an efficient practice model that could more easily fit into existing community pharmacy workflow. Time requirements for PGx POCT are similar to that of community pharmacy–based immunizations. Future studies could explore how practice could change if PGx testing were routinely performed in the pharmacy.