Abstract
Aim: Clopidogrel requires CYP2C19 activation to have antiplatelet effects. Pharmacogenetic testing to identify patients with impaired CYP2C19 function can be coupled with clinical decision support (CDS) alerts to guide antiplatelet prescribing. We evaluated the impact of alerts on clopidogrel prescribing.
Materials & methods: We retrospectively analyzed data for 866 patients in which CYP2C19-clopidogrel CDS was deployed at a single healthcare system during 2015–2023.
Results: Analyses included 2,288 alerts. CDS acceptance rates increased from 24% in 2015 to 63% in 2023 (p < 0.05). Adjusted analyses also showed higher acceptance rates when clopidogrel had been ordered for a percutaneous intervention (OR: 28.7, p < 0.001) and when cardiologists responded to alerts (OR: 2.11, p = 0.001).
Conclusion: CDS for CYP2C19-clopidogrel was effective in reducing potential drug-gene interactions. Its influence varied by clinician specialty and medication indications.
Keywords: : clinical, clopidogrel, cytochrome P-450 CYP2C19, decision support systems, electronic health records, genetic testing, patient safety, pharmacogenetics, pharmacy, platelet aggregation inhibitors, precision medicine
Plain language summary
Article highlights.
Clinical decision support systems are utilized to facilitate genotype-guided antiplatelet therapy.
Acute percutaneous coronary intervention indications showed the greatest acceptance for CYP2C19-clopidogrel clinical decision support (CDS).
Despite lack of specific increased dosing recommendations within the Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines clinicians are still utilizing this strategy.
1. Introduction
P2Y12 inhibitors are frequently ordered as part of single or dual antiplatelet therapy to prevent cardiovascular events [1]. Among P2Y12 inhibitors, clopidogrel is most commonly utilized, although medication selections are influenced by patient-specific factors such as the underlying mechanism of a neurovascular or cardiac disease, bleeding risk and compliance [2,3]. Drug-gene interactions are important factors that can influence P2Y12 inhibitor decisions.
Clopidogrel requires activation by the cytochrome P450 enzyme system, primarily CYP2C19, to exert its effect on platelet activation and aggregation. The CYP2C19 gene is highly polymorphic, with over 35 defined alleles with varying functional status [2,4]. Patients who are homozygous for loss of function (LOF) alleles in CYP2C19 do not convert clopidogrel to its active metabolite leading to decreased platelet inhibition. To alert patients and providers about potential drug-gene interactions, the United States Food and Drug Administration (FDA) prescribing information holds a black box warning that recommends use of an alternative platelet inhibitor if this critical pharmacogenetic (PGx) medication consideration exists [5]. Patients with even one CYP2C19 LOF allele have an increased risk of life threatening atherothrombotic events if prescribed clopidogrel [2,6,7].
Given a growing body of evidence that CYP2C19 genotype-guided use of clopidogrel improves patient outcomes, the Clinical Pharmacogenetics Implementation Consortium (CPIC) has issued peer-reviewed, evidence-based guidelines for its use [6,8,9]. Current CPIC guidelines provide recommendations for both neurovascular and cardiovascular indications based on CYP2C19 phenotype, whereas previous iterations were largely focused on patients with acute coronary syndromes (ACS) or undergoing percutaneous coronary intervention (PCI) [2,10]. To facilitate the appropriate use of PGx information to inform medication selections, including CYP2C19 genotyping to inform antiplatelet therapy, healthcare systems have developed clinical decision support (CDS) systems within the electronic health record (EHR), including CDS alerts that provide valuable genetic interpretation and clinical recommendations at the time of medication ordering [11–13]. Questions remain about how to optimize CDS for clinicians at the appropriate time in their decision-making process [14–16]. Herein, we describe the evolution of our CDS system and clinicians' responses to our CYP2C19-clopidogrel PGx CDS. The primary objective of this study is to evaluate the safety of clopidogrel prescribing following PGx interruptive alerts.
2. Materials & methods
2.1. Setting & genotyping methods
Sanford Health is a multisite health system spanning primarily across four states in the upper Midwest, representing the largest rural non-profit healthcare system in the United States. In 2014, Sanford Health implemented a precision medicine program (Sanford Imagenetics) that included an internal medical genetics laboratory and CYP2C19 PGx testing, both stand-alone and as part of a 4-gene panel that was expanded to 8 genes in 2017, 11 genes in 2020 and 14 genes in 2023 [17,18]. Testing was initially performed by quantitative PCR using TaqMan™ probes for alleles classified as tier 1 or tier 2 by the Association of Molecular Pathology (*2, *3, *4, *5, *6, *7, *8 and *17) [4]. Testing was updated in 2020 to use a digital PCR platform, and again updated in December 2023 to use next generation sequencing to test for additional CYP2C19 alleles (*9, *10, *16, *22, *24, *25, *26 and *35) and to facilitate analyses of more genes and variants in the future. In addition to indication-based testing, CYP2C19 genotyping was integrated into an elective genomic testing program offered in primary care settings that included preemptive PGx panel testing and an optional screening for medically actionable disease predispositions [18].
Numerous efforts were implemented to educate clinicians about PGx testing [17,19,20]. Quarterly modules about the genetic basis of precision medicine and implementation of genetic testing that were mandatory for physicians and advanced practice providers were deployed over two years [19,21]. Of the eight educational modules, one was specific to addressing genomic medicine, one was devoted to the application of PGx testing and basics of drug metabolism, including how to utilize CDS; one was devoted to reviewing PGx principles and available CDS tools; and the last was devoted to using genomics to improve management, including application of PGx testing results to medical management [19].
2.2. Clinical decision support system
Figure 1 depicts the components utilized in the CDS system with respect to CYP2C19-clopidogrel. During the early phase of the Sanford Imagenetics, CYP2C19 results were manually entered into the EHR by laboratory directors as genomic tags (phenotypes) rather than discrete laboratory values (genotypes and corresponding phenotypes) in accordance with CPIC diplotype to phenotype mapping tables. Integration of discrete laboratory values into the EHR was quickly implemented to improve efficiencies, avoid manual entry errors and improve medication and patient safety initiatives. Upon laboratory director verification and release of CYP2C19 results, the ordering clinician was notified results were available through automated processes. Interdepartmental collaboration between the cardiology and precision medicine programs led to the creation of a workflow to escalate antiplatelet therapy based on the return of CYP2C19 results. Enhancements within the EHR functionality allowed the PGx pharmacy team to be concurrently notified when CYP2C19 results are verified from the Sanford Medical Genetics Laboratory. Additional features such as identification of actionable drug-gene interactions were later deployed (Figure 2). Expansion of CDS modalities to include passive alerts (require clinicians to click for additional information) and interruptive alerts (information presented to the clinician without additional steps) within procedural areas (cardiac catheterization laboratories) were implemented in 2020 and 2023 respectively [12]. Passive CDS was automatically embedded within the provider result messages, which highlighted drug-gene interactions.
Figure 1.
Components of the clinical decision support system for CYP2C19-clopidogrel. CYP2C19 testing can be ordered via two mechanisms routine order entry or through use of standardized order sets utilized for percutaneous coronary intervention. The DNA sample is processed within the internal Sanford Medical Genetics Laboratory. Once the CYP2C19 result is verified by a laboratory director, a genetic test report is generated within the electronic health record (EHR). CYP2C19 results are simultaneously automatically routed to the ordering clinician, PGx pharmacist and the patient (via the patient portal of the EHR). PGx pharmacists receive an alert to prioritize the review of CYP2C19 when loss of function (LOF) alleles are identified in a patient already prescribed clopidogrel. Clinical decision support alerts, both passive and interruptive, are integrated within the EHR to facilitate genotype-guided antiplatelet therapy.
Figure 2.
Automated pharmacogenetic (PGx) result notification system for the PGx team.
Automated CDS has been the mainstay of proactive clinician guidance for the use of CYP2C19 genotyping in clopidogrel ordering, with alerts based on CPIC and FDA guidelines (Figure 3) [22]. Supplementary Figure S1 depicts one of the earliest alerts utilized for CYP2C19-clopidogrel interaction which comprised of largely the same triggering criteria but have visual enhancements and standardization of content based on feedback. When an order for clopidogrel is attempted for patients who have CYP2C19 LOF alleles, interruptive alerts inform clinicians about the drug-gene interaction and provide single-click response options, with defaults being cancellation of the clopidogrel order and the opportunity to order an alternative P2Y12 inhibitor. In instances where clinicians continue with a clopidogrel order, clinicians are required to explain why they overrode the CDS alert. Within the alert, there are specific options clinicians can endorse to provide rationales for overriding the alert. Clinicians can also provide explanations in a free text field.
Figure 3.
Example of current interruptive alert for CYP2C19-clopidogrel triggering within ambulatory settings.
2.3. Data analyses
The research protocol was approved by the Sanford Health Institutional Review Board (STUDY00003518). All orders entered for clopidogrel in individuals who are carriers of LOF alleles between 2015 and 2023 that triggered a CDS alert were extracted from the EHR. Orders for clopidogrel initiated within the cardiac catheterization laboratory may not have generated a CDS alert depending on the mechanism in which these orders were placed as the orders may not follow standard order entry and pharmacist verification processes. Data validation via a manual chart review was conducted by a pharmacist on the Sanford Imagenetics team.
Alerts could fire multiple times for clopidogrel orders initiated during a single clinical encounter if a specific clopidogrel order was viewed multiple times (e.g., order entry, modifications, or order sign) or due to multiple orders for clopidogrel. Given this, alerts could fire for clinicians whom are not the final prescriber of clopidogrel (e.g., registered nurse prepping orders). Only the last response to an alert for a given encounter or the final medication order associated with that encounter was included in statistical analyses, except where noted. In addition, antiplatelet agent orders were limited to maintenance orders, defined as orders with daily dosages of less than 300 mg. Loading doses were omitted to focus on sustained normative responses to CDS alerts. Medication orders responses were categorized as either having accepted or overridden alert recommendations based on an alternative antiplatelet order or increased dose of clopidogrel (greater than 75 mg) vs. an order of clopidogrel at a normal maintenance dose, respectively. Analyses of reasons for overriding CDS alerts combined responses to structured response options and open-ended responses. Data in free text fields were reviewed by two Sanford Imagenetics pharmacists and categorized into one of seven distinct classifications (contraindications to alternative therapy, patient declines alternative therapy, potential side effects with alternatives, financial concerns such as cost, indication other than ACS, provider discretion and an “other” category that included blank responses, responses of “not applicable” and responses from five or fewer respondents such as “dose increase”, “patient had spinal” and “patient is not on Plavix”).
Repeated measures analyses that examined whether alert recommendations were accepted or overridden used a generalized mixed-effects model to account for within-patient and within-provider effects, given that CDS alerts that could trigger for multiple encounters for individual patients and individual clinicians. Covariates were identified and included based on a criterion of theoretical and situational justification as well as preliminarily showing a difference between accepted vs. overridden recommendations in bivariate analyses at p < 0.05. Such possible covariates included: year of the alert, clinician specialty, indication, whether the patient had a prior clopidogrel order and history and timing of PCI. For the purposes of this analysis, acute PCI was defined as intervention within the last 30 days, whereas other PCIs outside of the 30-day window were classified as history of PCI. Chi-squared tests and Fisher's exact tests were used for nominal variables and Wilcoxon ranked sum test was used for linear variables. A reverse-stepwise methodology was used to decide on the final model fit. Specifically, all variables that were statistically significant at p < 0.05 in bivariate analyses were included in initial statistical models. They were then removed one-at-a-time while considering coefficient significance and magnitude and comparisons in model deviance when individual covariates were removed, and models were re-run. Variables were retained in the model if they were significant at a Bonferroni-corrected p-value of 0.05/n in adjusted analyses, where n was the final number of included variables. We also ran stratified analyses to examine whether specific cutoffs in dates might affect model estimates, but omit findings because no differences were observed. Data were analyzed using R version 4.2.
3. Results
3.1. Ordering response following CDS alerts
From 2015 to 2023, 2,288 distinct clinical encounters triggered drug-gene alerts for clopidogrel orders for 866 patients. Descriptive statistics of encounters, clinicians and patients can be found in Table 1. The majority of alerts (95%) were triggered for patients who were intermediate metabolizers, 5% of alerts that triggered for patients who were poor metabolizers. There were 613 distinct clinicians that comprised all these responses.
Table 1.
Characteristics of encounters, clinicians and patients at the time of a CDS alert.
| Characteristic | Overall |
Response to alert | p-valuea | |
|---|---|---|---|---|
| N = 2288 | Overridden (N = 924) | Accepted (N = 1364) | ||
| Year, n (%) | <0.001*** | |||
| 2015 | 29 (1.3) | 22 (76) | 7 (24) | |
| 2016 | 22 (1.0) | 11 (50) | 11 (50) | |
| 2017 | 120 (5.2) | 64 (53) | 56 (47) | |
| 2018 | 225 (9.8) | 95 (42) | 130 (58) | |
| 2019 | 340 (15) | 122 (36) | 218 (64) | |
| 2020 | 334 (15) | 151 (45) | 183 (55) | |
| 2021 | 380 (17) | 145 (38) | 235 (62) | |
| 2022 | 406 (18) | 153 (38) | 253 (62) | |
| 2023 | 432 (19) | 161 (37) | 271 (63) | |
| Metabolizer status, n (%) | 0.004** | |||
| Intermediate | 2,169 (95) | 861 (40) | 1,308 (60) | |
| Poor | 119 (5.2) | 63 (53) | 56 (47) | |
| Age at firing, Median (IQR) | 73 (65–80) | 72 (64–80) | 74 (66–80) | 0.016* |
| Alert type, n (%) | 0.30 | |||
| General | 1818 (79) | 744 (41) | 1074 (59) | |
| Outpatient | 470 (21) | 180 (38) | 290 (62) | |
| Specialty of responding clinician or setting, n (%) | <0.001*** | |||
| Cardiology | 1375 (60) | 447 (33) | 928 (67) | <0.001*** |
| Inpatient | 273 (12) | 112 (41) | 161 (59) | |
| Neurology | 56 (2.4) | 38 (68) | 18 (32) | |
| Other | 213 (9.3) | 90 (42) | 123 (58) | |
| Primary care | 324 (14) | 212 (65) | 112 (35) | |
| Vascular surgery | 47 (2.1) | 25 (53) | 22 (47) | |
| Clinician type, n (%) | <0.001*** | |||
| Cardiology fellow | 140 (6.1) | 24 (17) | 116 (83) | |
| Certified nurse practitioner | 523 (23) | 206 (39) | 317 (61) | |
| Licensed practical nurse | 119 (5.2) | 46 (39) | 73 (61) | |
| Physician | 620 (27) | 310 (50) | 310 (50) | |
| Registered nurse | 586 (26) | 197 (34) | 389 (66) | |
| Other | 293 (13) | 140 (48) | 153 (52) | |
| Unknown | 7 | 1 | 6 | |
| Patient Charlson score, (IQR) | 6.00 (4.00–8.00) | 6.00 (4.00–8.00) | 6.00 (4.00–8.00) | 0.57 |
| Medication Indication, n (%)b | ||||
| Stroke | 166 (7.3) | 87 (52) | 79 (48) | 0.001** |
| Acute coronary syndrome | 261 (11) | 98 (38) | 163 (62) | 0.32 |
| Coronary artery disease | 518 (23) | 195 (38) | 323 (62) | 0.15 |
| Carotid stenosis | 52 (2.3) | 32 (62) | 20 (38) | 0.002** |
| Peripheral vascular disease | 136 (5.9) | 47 (35) | 89 (65) | 0.15 |
| Acute percutaneous intervention (PCI) | 547 (24) | 75 (14) | 472 (86) | <0.001*** |
| Had a prior clopidogrel order, n (%) | 2,105 (92) | 870 (41) | 1,235 (59) | 0.002** |
Percentages in columns reporting characteristics stratified by the response of the final clinician to view a particular alert report the likelihood that CDS was accepted.
*p < 0.05; **p < 0.01; ***p < 0.001.
May complement other indications.
Of these encounters, 1,293 (56.5%) alert recommendations were accepted by the first clinician to view the alert (Supplementary Table S1), with 1,364 (59.6%) alert recommendations ultimately being accepted based on the last clinician to view the alert. Of these 1,364 accepted recommendations, 423 (31%) were associated with orders for an alternative antiplatelet agent and 941 (69%) were associated with orders for clopidogrel at an increased dose (>75 mg per day). The cardiology specialty had 2.5 greater odds of utilizing an increased dose strategy as compared with all other specialties (OR: 2.48 [2.08, 2.97], p < 0.05). All subsequent covariate identification and modeling were evaluated based on the final alert response to CDS.
Potential covariates for multivariable analyses were identified in bivariate analyses that compared accepted and overridden recommendations (Table 1). There was a trend that CDS acceptance rate increased yearly (p < 0.05), increasing from 24% in 2015 to 63% in 2023. Patient age was also higher for alert recommendations that were accepted rather than overridden (median = 74 vs. 72, respectively, p < 0.05). Orders placed by cardiology were more likely to be an alternative antiplatelet or increased dose of clopidogrel (67%) as compared with all other specialties (48%, p < 0.001). Analyses of indications showed that alert recommendations were less likely to be accepted if alerts triggered after a stroke than if alerts triggered for other indications (48% vs. 61%, respectively, p < 0.05). Alert recommendations were also less likely to be accepted for patients with a history of clopidogrel use than for patients who were clopidogrel naive (59 vs. 70%, respectively, p = 0.002). Finally, alert recommendations were six-times more likely be accepted for patients with an acute PCI than for other indications (p < 0.001).
While beginning the reverse stepwise methodology, year, CYP2C19 metabolizer status, patient age, whether the order was made by cardiology, clinician type, prior history of clopidogrel use, medication indication of stroke or carotid stenosis and whether the patient had an acute PCI were included. After completing the stepwise process, the final model was fit with the year of the alert (OR: 1.23 per year [1.07, 1.40; Table 2]), whether the order was made by cardiology (2.11, [1.34, 3.34], p = 0.001) and whether the patient had an acute PCI (OR: 28.7 [15.3, 53.8], p < 0.001). Clinician and patient ID had intraclass correlation coefficients (ICCs) of 0.25 and 0.79 respectively.
Table 2.
Summary of multivariable statistical models predicting acceptance of CDS alerts for clopidogrel.
| Characteristic | OR (95% CI) | p-value |
|---|---|---|
| Year of alert (per year) | 1.23 (1.07 to 1.40) | 0.003 |
| Cardiology | 2.11 (1.34 to 3.34) | 0.001 |
| Acute PCI | 28.7 (15.3 to 53.8) | <0.001 |
CI: Confidence interval; OR: Odds ratio.
3.2. Override reasons
Given that override reasons were consistent over time, analyses of reasons that clinicians overrode CDS alerts were limited to the patients first encounter. Furthermore, given that multiple override reasons could be listed for one CDS alert instance, override reasons were limited to those provided when a clopidogrel order was signed. Out of 866 distinct patients, 251 override responses (Table 3) were available for 264 orders which CDS alerts were overridden. Clinician discretion in overriding the CDS alert was listed as the most common override reason (n = 64) accounting for 25%. Next most frequent was patient refusal (n = 52) and side effects (n = 50) accounting for 21% and 20%, respectively. Contraindications to alternative therapy (n = 23) and a non-ACS indication (n = 17) accounting for 9% and 7%, respectively. Financial concerns, including cost, was the least common with only 9 alerts (4%). All remaining responses were attributed to “other” and accounted for 36 (14%) alerts.
Table 3.
Reasons clinicians endorsed for overriding CDS alerts.
| Override categories | N (%) |
|---|---|
| Provider discretion to continue clopidogrel | 64 (25%) |
| Patient declined alternative therapy | 52 (21%) |
| Potential side effects with alternative therapy | 50 (20%) |
| Contraindications to alternative therapy | 23 (9%) |
| Indication other than acute coronary syndromes | 17 (7%) |
| Financial concerns | 9 (4%) |
| Othera | 36 (14%) |
Includes blank responses, responses of “not applicable” and responses from five or fewer respondents such as “dose increase”, “patient had spinal” and “patient is not on Plavix”.
4. Discussion
This work provides additional data on acceptance of CDS alert recommendations for genotype-guided clopidogrel prescribing as well as insights into reasons for overriding CDS recommendations. We found that clinicians ordered increased doses of clopidogrel in response to 41% of CDS alerts and ordered alternative antiplatelet agents in response to 18% of CDS alerts, for an overall acceptance rate of recommendations of nearly 60%. Acceptance rates improved over time, and were higher in cardiology settings and especially high when patients had an acute PCI. In instances where clinicians continued with clopidogrel at standard doses despite CDS alerts, clinicians most often cited patient preferences and their own discretion as their rationales. Taken together, response suggest that automated CDS alerts for drug-gene interactions about clopidogrel were successful in improving the likelihood that medication orders were consistent with current recommendations for the use of genotype-guided antiplatelet therapy.
Findings are encouraging given the potential of genotype-guided antiplatelet therapy to balance safety and efficacy concerns. Acceptance rates previously described in the literature range from 15 to 70% [22–25]. Our data, showing higher acceptance rates than many of those studies, may reflect the ongoing efforts to inform and support the use of PGx testing. Sanford Health systematically prioritized outreach to clinicians about PGx testing, including mandatory genetics education for physicians and advanced practice providers. These efforts appear successful, as acceptance of recommendations improved over time in our analyses; and prior work showed that clinicians reported increased confidence ordering PGx tests and increased perceived utility of PGx testing after completion of educational modules [19,21]. Sanford Health also increased the availability of PGx specialists to provide assistance to clinicians across the health system, beginning in 2014, and improved PGx workflows to more proactively address existing drug-gene interactions [17,19]. Moreover, the strength of evidence about genotype-guided prescribing in the literature and guideline recommendations has been growing [2,10]. These developments likely explain improvement in clinicians' attitudes over time, as well as the strong adherence to genotype-guided prescribing by cardiology we observed with additional buy in from non-cardiac specialists [21].
Our study also highlights how medication indications influence responses to CDS alerts. We found that recommendations for CDS alerts that occurred after patients received a PCI within the previous 30 days had an 86% acceptance rate compared with 51% when an acute PCI was not present. While evidence about the clinical impact of genotype-guided antiplatelet therapy for some indications is mixed, evidence of benefits following PCI are particularly strong [6,7,26,27]. It is likely that clinicians responding to CDS alerts in our study are more aware of the nuances of existing recommendations for genotype-guided antiplatelet therapy and how the evidence base may be less robust for non-PCI indications [2].
Notably, other strategies Sanford Health implemented to support PGx testing, including CYP2C19 genotyping to inform antiplatelet therapy, may have influenced the results of our analyses. Clinicians in our study who ordered PGx testing for patients who were already taking clopidogrel may have been vigilant about CYP2C19 results or may have been informed about CYP2C19-clopiogrel interactions by a pharmacist with PGx experience [17]. If so, they may have altered patients' antiplatelet therapies before clopidogrel was re-ordered, thereby avoiding CDS alerts. Our prior data support this conjecture: 75% of patients initiated on clopidogrel were adjusted to an appropriate alternative therapy within 10 days of testing when PGx testing was completed within 7 days of antiplatelet initiation [9].
In addition, Sanford Health offers panel PGx testing that allows CYP2C19 genotypes to be characterized before a need for antiplatelet therapy arises. Our prior analyses showed that preemptive CYP2C19 genotyping had a strong influence on the initial antiplatelet agent choice of patients with ACS or PCI such that patients with intermediate or poor metabolizer phenotypes were typically initiated on prasugrel or ticagrelor rather than clopidogrel [9].
Our work also provides early evidence into the rationale in which providers override clopidogrel CDS alerts. Providers were most likely to continue with clopidogrel orders due to patient and clinician preference followed by side effects or contraindications to alternative therapy. Notably, current CDS alerts for clopidogrel were not dose-centric, and if clinicians elected to increase the dose of clopidogrel in response to CYP2C19 LOF allele carrier status, the alert would still trigger. Cardiology was 2.5-times more likely to utilize an increased dose strategy as compared with all other prescribers.
As the precision medicine program at Sanford Health continues to evolve, we aim to continually refine CDS alerts based on provider feedback and EHR capabilities. Given the success with cardiology champions, we plan to partner with additional specialties to yield more inclusion of drug-gene interactions within routine clinical care. Future state may include the use of genetically tailored ordersets to further guide medication prescribing.
Our work has limitations. CPIC guidelines were updated to include additional indications for genotype-guided antiplatelet therapy during the study time frame. Only the final antiplatelet agent order following an encounter where an alert triggered could be ascertained rather than the immediate clinician responses. Clinical decisions and considerations outside of discrete orders are not capturable within our dataset and therefore cannot be examined, thereby limiting a complete understanding of the providers' decision-making process and the influence of CDS alerts. Ordering providers are notified with PGx results in the inbasket, in which CDS highlights actionable drug-gene interactions. Secondarily, the PGx pharmacy team was notified of results with specific actionable alerts for drug-gene interactions with significant risk for adverse events. Additionally, current clopidogrel CDS do not account for dose, therefore, prescribers will encounter the alert with increased dosage clopidogrel. Additional limitations include generalizability to other healthcare institutions where genomics educational efforts and infrastructure may be less robust [18,19]. Cardiology was early adopters of incorporating PGx testing into routine clinical care as well as placing a high value on precision medicine initiatives [19,21].
5. Conclusion
This study highlights an important role of point-of-ordering CDS for CYP2C19-clopidogrel interactions. Partnership with provider champions has aided in successful integration and acceptance of PGx CDS alerts. While cardiology is more likely to order PGx congruent antiplatelet therapy, integrated CDS has demonstrated an impact on PGx congruent antiplatelet therapy across all specialties.
Supplementary Material
Acknowledgments
The authors would like to thank C Larson for his work developing the alerts, M Kara for her work abstracting data from medical records, and the Sanford Medical Genetics Laboratory for analyzing and reporting PGx results.
Funding Statement
KD Christensen was supported by a Robert H Ebert Career Development Award from the Harvard Pilgrim Health Care Institute.
Supplemental material
Supplemental data for this article can be accessed at https://doi.org/10.1080/14622416.2024.2394014
Author contributions
Conceptualization: A Massmann, J Van Heukelom, M Weaver, A Schultz, D Figueroa, A Stys, TP Stys, KD Christensen. Data curation: A Massmann, J Van Heukelom, M Weaver, KD Christensen. Formal analysis: M Weaver, KD Christensen. Writing – original draft: A Massmann, J Van Heukelom, M Weaver, A Schultz, D Figueroa, KD Christensen. Writing – review and editing: A Massmann, J Van Heukelom, M Weaver, A Schultz, D Figueroa, A Stys, TP Stys, KD Christensen.
Financial disclosure
KD Christensen was supported by a Robert H Ebert Career Development Award from the Harvard Pilgrim Health Care Institute. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
Competing interests disclosure
The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript.
Writing disclosure
No writing assistance was utilized in the production of this manuscript.
Ethical conduct of research
The authors state that they have obtained appropriate institutional review board approval.
Data availability statement
The data that support the findings of this study are available from the corresponding author, [AM], upon reasonable request.
References
Papers of special note have been highlighted as: • of interest; •• of considerable interest
- 1.Writing Committee Members Lawton JS, Tamis-Holland JE, et al. 2021 ACC/AHA/SCAI Guideline for Coronary Artery Revascularization: a report of the American College of Cardiology/American Heart Association Joint Committee on clinical practice guidelines. J Am Coll Cardiol. 2022;79(2):e21–e129. doi: 10.1016/j.jacc.2021.09.006 [DOI] [PubMed] [Google Scholar]
- 2.Lee CR, Luzum JA, Sangkuhl K, et al. Clinical Pharmacogenetics Implementation Consortium Guideline for CYP2C19 genotype and clopidogrel therapy: 2022 update. Clin Pharmacol Ther. 2022;112(5):959–967. doi: 10.1002/cpt.2526 [DOI] [PMC free article] [PubMed] [Google Scholar]; •• In the presence of CYP2C19 genotypes, recommendations are available for clopidogrel prescribing based on clinical indication. Robust data exists for the clinical impact of these variations for acute coronary syndrome and following percutaneous coronary intervention with emerging evidence for other indications.
- 3.Bhatia K, Ladd LM, Carr KH, et al. Contemporary antiplatelet and anticoagulant therapies for secondary stroke prevention: a narrative review of current literature and guidelines. Curr Neurol Neurosci Rep. 2023;23(5):235–262. doi: 10.1007/s11910-023-01266-2 [DOI] [PubMed] [Google Scholar]
- 4.Pratt VM, Del Tredici AL, Hachad H, et al. Recommendations for clinical CYP2C19 genotyping allele selection: a report of the association for molecular pathology. J Mol Diagn JMD. 2018;20(3):269–276. doi: 10.1016/j.jmoldx.2018.01.011 [DOI] [PubMed] [Google Scholar]; • The Association for Molecular Pathology curates lists of alleles to be clinically relevant in the development of pharmacogenetic assays.
- 5.PLAVIX-clopidogrel tablet . Package Insert. Sanofi-aventis U.S. LLC. September 30, 2022. Accessed May 29, 2024. https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=de8b0b67-eb25-4684-83b5-7ad785314227
- 6.Mega JL, Close SL, Wiviott SD, et al. Cytochrome P-450 polymorphisms and response to clopidogrel. N Engl J Med. 2009;360(4):354–362. doi: 10.1056/NEJMoa0809171 [DOI] [PubMed] [Google Scholar]
- 7.Mega JL, Simon T, Collet JP, et al. Reduced-function CYP2C19 genotype and risk of adverse clinical outcomes among patients treated with clopidogrel predominantly for PCI: a meta-analysis. JAMA. 2010;304(16):1821–1830. doi: 10.1001/jama.2010.1543 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Relling MV, Evans WE. Pharmacogenomics in the clinic. Nature. 2015;526(7573):343–350. doi: 10.1038/nature15817 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Massmann A, Christensen KD, Van Heukelom J, et al. Clinical impact of preemptive pharmacogenomic testing on antiplatelet therapy in a real-world setting. Eur J Hum Genet. 2024;32:895–902. doi: 10.1038/s41431-024-01567-1 [DOI] [PMC free article] [PubMed] [Google Scholar]; • Pre-emptive genotyping showed high PGx congruent antiplatelet prescribing. No differences were observed in major adverse cardiovascular events (MACE) or bleeding events within 1 year between preemptive or early reactive PGx testing cohorts.
- 10.Scott SA, Sangkuhl K, Stein CM, et al. Clinical Pharmacogenetics Implementation Consortium guidelines for CYP2C19 genotype and clopidogrel therapy: 2013 update. Clin Pharmacol Ther. 2013;94(3):317–323. doi: 10.1038/clpt.2013.105 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Aquilante CL, Trinkley KE, Lee YM, et al. Implementation of clopidogrel pharmacogenetic clinical decision support for a preemptive return of results program. Am J Health-Syst Pharm AJHP Off J Am Soc Health-Syst Pharm. 2024;81(12):555–562. doi: 10.1093/ajhp/zxae008 [DOI] [PMC free article] [PubMed] [Google Scholar]; • A multidisciplinary approach was imperative to successful implementation of CYP2C19-clopidogrel testing and development of clinical decision support.
- 12.Van Heukelom J, Weaver M, Max C, et al. Early evaluation of a novel alert within cardiac procedural areas to facilitate genotype-guided antiplatelet therapy. JACCP J Am Coll Clin Pharm. 2024;7(4):jac5.1956. doi: 10.1002/jac5.1956 [DOI] [Google Scholar]; • Genotype-guided antiplatelet prescribing significantly increased after cardiac catheterization laboratory alerts for CYP2C19-clopidogrel were implemented.
- 13.Green ED, Gunter C, Biesecker LG, et al. Strategic vision for improving human health at The Forefront of Genomics. Nature. 2020;586(7831):683–692. doi: 10.1038/s41586-020-2817-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wake DT, Smith DM, Kazi S, Dunnenberger HM. Pharmacogenomic clinical decision support: a review, how-to guide, and future vision. Clin Pharmacol Ther. 2022;112(1):44–57. doi: 10.1002/cpt.2387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Smith DM, Wake DT, Dunnenberger HM. Pharmacogenomic clinical decision support: a scoping review. Clin Pharmacol Ther. 2023;113(4):803–815. doi: 10.1002/cpt.2711 [DOI] [PubMed] [Google Scholar]
- 16.Carter JPL, Critchlow J, Jackson S, et al. Pharmacogenomic alerts: developing guidance for use by healthcare professionals. Br J Clin Pharmacol. 2022;88(7):3201–3210. doi: 10.1111/bcp.15234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Petry N, Baye J, Aifaoui A, et al. Implementation of wide-scale pharmacogenetic testing in primary care. Pharmacogenomics. 2019;20(12):903–913. doi: 10.2217/pgs-2019-0043 [DOI] [PubMed] [Google Scholar]
- 18.Christensen KD, Bell M, Zawatsky CLB, et al. Precision population medicine in primary care: the Sanford Chip experience. Front Genet. 2021;12:626845. doi: 10.3389/fgene.2021.626845 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Hajek C, Hutchinson AM, Galbraith LN, et al. Improved provider preparedness through an 8-part genetics and genomic education program. Genet Med Off J Am Coll Med Genet. 2022;24(1):214–224. doi: 10.1016/j.gim.2021.08.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Blout Zawatsky CL, Leonhard JR, Bell M, et al. Workforce considerations when building a precision medicine program. J Pers Med. 2022;12(11):1929. doi: 10.3390/jpm12111929 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Preys CL, Blout Zawatsky CL, Massmann A, et al. Attitudes about pharmacogenomic testing vary by healthcare specialty. Pharmacogenomics. 2023;24(10):539–549. doi: 10.2217/pgs-2023-0039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Petry NJ, Curtis B, Feldhege E, et al. Impact of automated best practice advisories on provider response to CYP2C19 genotyping results for patients on clopidogrel. J Pharm Pract. 2023;36(3):487–493. doi: 10.1177/08971900211049589 [DOI] [PubMed] [Google Scholar]
- 23.Lemke LK, Cicali EJ, Williams R, Nguyen KA, Cavallari LH, Wiisanen K. Clinician response to pharmacogenetic clinical decision support alerts. Clin Pharmacol Ther. 2023;114(6):1350–1357. doi: 10.1002/cpt.3051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Herr TM, Peterson JF, Rasmussen LV, Caraballo PJ, Peissig PL, Starren JB. Pharmacogenomic clinical decision support design and multi-site process outcomes analysis in the eMERGE Network. J Am Med Inform Assoc JAMIA. 2019;26(2):143–148. doi: 10.1093/jamia/ocy156 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Weitzel KW, Elsey AR, Langaee TY, et al. Clinical pharmacogenetics implementation: approaches, successes, and challenges. Am J Med Genet C Semin Med Genet. 2014;166C(1):56–67. doi: 10.1002/ajmg.c.31390 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Cavallari Larisa H, Lee Craig R, Beitelshees Amber L, et al. Multisite investigation of outcomes with implementation of CYP2C19 genotype-guided antiplatelet therapy after percutaneous coronary intervention. JACC Cardiovasc Interv. 2018;11(2):181–191. doi: 10.1016/j.jcin.2017.07.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Pereira NL, Rihal C, Lennon R, et al. Effect of CYP2C19 genotype on ischemic outcomes during oral P2Y12 inhibitor therapy. JACC Cardiovasc Interv. 2021;14(7):739–750. doi: 10.1016/j.jcin.2021.01.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author, [AM], upon reasonable request.