electronic records
zlevxnwt
art1.pdf
ELECTRONIC HEALTH RECORDS SPECIAL FEATURES
AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016 1967
Integrating pharmacogenomics into electronic health records with clinical decision support
J. Kevin Hicks, Pharm.D., Ph.D., DeBartolo Family Personalized Medicine Institute and Department of Population Sciences, H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL.
Henry M. Dunnenberger, Pharm.D., BCPS, Center for Molecular Medicine, NorthShore University HealthSystem, Evanston, IL.
Karl F. Gumpper, B.S.Pharm., M.M.I., BCPS, Department of Pharmacy, Boston Children’s Hospital, Boston, MA.
Cyrine E. Haidar, Pharm.D., BCPS, BCOP, Department of Pharmaceutical Sciences, St. Jude Children’s Research Hospital, Memphis, TN.
James M. Hoffman, Pharm.D., M.S., BCPS, Department of Pharmaceutical Sciences, St. Jude Children’s Research Hospital, Memphis, TN.
Address correspondence to Dr. Hoffman ([email protected]).
Copyright © 2016, American Society of Health-System Pharmacists, Inc. All rights reserved. 1079-2082/16/1201-1967.
DOI 10.2146/ajhp160030
Purpose. Existing pharmacogenomic informatics models, key implemen- tation steps, and emerging resources to facilitate the development of phar- macogenomic clinical decision support (CDS) are described.
Summary. Pharmacogenomics is an important component of precision medicine. Informatics, especially CDS in the electronic health record (EHR), is a critical tool for the integration of pharmacogenomics into rou- tine patient care. Effective integration of pharmacogenomic CDS into the EHR can address implementation challenges, including the increasing vol- ume of pharmacogenomic clinical knowledge, the enduring nature of phar- macogenomic test results, and the complexity of interpreting results. Both passive and active CDS provide point-of-care information to clinicians that can guide the systematic use of pharmacogenomics to proactively opti- mize pharmacotherapy. Key considerations for a successful implementa- tion have been identified; these include clinical workflows, identification of alert triggers, and tools to guide interpretation of results. These consider- ations, along with emerging resources from the Clinical Pharmacogenetics Implementation Consortium and the National Academy of Medicine, are described.
Conclusion. The EHR with CDS is essential to curate pharmacogenomic data and disseminate patient-specific information at the point of care. As part of the successful implementation of pharmacogenomics into clinical settings, all relevant clinical recommendations pertaining to gene–drug pairs must be summarized and presented to clinicians in a manner that is seamlessly integrated into the clinical workflow of the EHR. In some situations, ancillary systems and applications outside the EHR may be in- tegrated to augment the capabilities of the EHR.
Keywords: clinical decision support systems, Clinical Pharmacogenet- ics Implementation Consortium (CPIC), electronic health records, pharma- cogenomics, precision medicine
Am J Health-Syst Pharm. 2016; 73:1967-76
Pharmacogenomics is an important component of precision medicine.1 Although various barriers to implemen- tation remain, there are an increasing number of examples demonstrating the utility of pharmacogenomics-oriented electronic health record (EHR) infor- matics specifically using computerized health records to maintain awareness of pharmacogenomic data to guide drug selection and dosing. In particu- lar, clinical decision support (CDS) has been identified as a critical tool for the
implementation of pharmacogenomics into routine patient care.1-6
The volume and evolving and en- during nature of pharmacogenomic knowledge that must be applied during a patient encounter present challenges with regard to integrat- ing pharmacogenomics into routine care. The Clinical Pharmacogenetics Implementation Consortium (CPIC) has published clinical guidelines per- taining to 13 genes, including thera- peutic recommendations for over 30
SPECIAL FEATURES ELECTRONIC HEALTH RECORDS
1968 AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016
drugs.7-22 Further, the product labeling of over 120 drugs contains genomic information.23,24 Complexity will in- crease as additional guidelines and clinically important pharmacoge- nomic relationships are discovered, including scenarios where multiple genes influence response to or effects of drug therapy.20,21 Moreover, phar- macogenomic data can have clinical utility throughout a patient’s life. Test results in the distant past could influ- ence drug selection and dosing years later and should be used to optimize drug therapy. It may be difficult for cli- nicians to remember both pertinent gene–drug interactions and previous pharmacogenomic test results for a specific patient during the demanding workflow of patient care. Fortunately, the increased use of EHRs across a range of healthcare settings and facili- ties provides solutions to these knowl- edge management challenges.
Even when pharmacogenomic test results are readily available to clini- cians, interpretation of those data can be complex. Clinicians may not be interested in the intricate details of a pharmacogenomic test result; in- stead, they primarily need evidence- based therapeutic recommendations that are consistent with guidelines and health-system policies to opti- mize drug therapy.2 Informatics tools provide a solution, since aspects of the interpretation process can be au- tomated and resources to organize the underlying knowledge required for automation are emerging.18,24
It has been suggested that imple- menting pharmacogenomic CDS presents challenges consistent with other sizable informatics efforts.25 However, EHR vendor support for pharmacogenomics is still emerging and remains limited, which means that substantial institutional effort— more than is typical with some other types of CDS—is required. This is especially a problem for organiza- tions without substantial genomics expertise or clinical informatics sup- port. Challenges include issues relat- ing to storing pharmacogenomic data
(which are valuable over a patient’s lifetime), presenting recommenda- tions to clinicians in a timely man- ner that is seamlessly integrated with clinical workflows, and updating CDS recommendations as the knowledge base changes. This article describes existing pharmacogenomic informat- ics models, identifies key implemen- tation steps, and discusses emerging resources that can facilitate the devel- opment of pharmacogenomic CDS in the EHR.
Integrating pharmacogenom- ics into the EHR with CDS: Mod- el practices. A growing number of healthcare systems are incorporating pharmacogenomic information into the EHR with CDS. In contrast to other types of CDS, in which vendors supply databases that form the foundation of the CDS program (e.g., various da- tabases for drug–drug interactions),26
pharmacogenomic implementations to date have consisted of customized rules built by the institutions them- selves.4,24,25,27-39 These initial pharma- cogenomic CDS models illustrate the
feasibility of establishing procedures to translate pharmacogenomic data into a predicted phenotype and clini- cal recommendation and to represent these data discretely in the EHR to al- low pharmacogenomic knowledge to be presented as both passive CDS (e.g., patient data reports, clinical consulta- tion notes) and active CDS (e.g., inter- ruptive alerts that notify clinicians to modify drug therapy).24,27 In addition, through its informatics working group CPIC provides model practices through vendor-agnostic implementation re- sources, including descriptions of clin- ical workflows and sample CDS text.18,40
Pharmacogenomic information can be presented passively, such as through interpretive notes or displaying re- sults in the drug order screen. Phar- macogenomic data interpretations entered into the EHR as static notes can provide a summary of test results along with other relevant informa- tion. Such notes are particularly use- ful in selected situations where it is difficult to assign a phenotype—for example, in cases involving duplica- tions of alleles of the gene coding for cytochrome P-450 (CYP) isozyme 2D6 (i.e., CYP2D6 alleles).
Active or interruptive CDS alerts are another well-established approach to presenting pharmacogenetic infor- mation.27,31,36 In the field of pharma- cogenomics, two main types of inter- ruptive alerts should be considered for clinical implementation. Pretest alerts (used when there are no geno- typing results available at the time of prescribing) inform prescribers who are attempting to order a medication affected by pharmacogenomic varia- tion and the associated genotype test is not documented in the EHR. The alert can inform the clinician of po- tential risks associated with prescrib- ing the medicine to a patient with a high-risk phenotype. It can also pro- vide a list of alternative agents to use instead of the high-risk medicine and allow the ordering of the genotype test from the alert window itself.27 Post- test alerts are triggered when a patient with a high-risk (or actionable) geno-
KEY POINTS • Active and passive clinical de-
cision support (CDS) tools in the electronic health record are crucial to maximize the value of pharmacogenomics in routine care.
• Documentation, interpretation, data representation, therapeutic recommendations, and ap- propriate CDS presentation are important considerations when implementing pharmacogenom- ic CDS.
• Resources from the Clinical Pharmacogenetics Implementa- tion Consortium and others are emerging to guide the develop- ment and use of pharmacoge- nomic CDS.
ELECTRONIC HEALTH RECORDS SPECIAL FEATURES
AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016 1969
type is prescribed a therapy that needs to be modified based on the patient’s genetics. This type of alert explains to the clinician the potential problems that might occur if the patient were to receive the medication at standard dosages and can also provide a list of alternative medications. Incorporat- ing these two types of active decision support CDS into the EHR facilitates prescribing the most appropriate drug at optimal starting dosages for indi- vidual patients based on their inher- ited genetic profile, thus limiting the occurrence of adverse effects, opti- mizing pharmacotherapy, and ensur- ing that genotype-guided therapy is used (when available) in every patient.
A systematic process for gathering, evaluating, and translating evidence into clinically useful CDS language and obtaining necessary approvals to implement CDS can be a significant undertaking. Developing a systematic process with standard procedures is necessary. Most CDS model imple- mentations to date have leveraged the existing institutional infrastructure for medication use and CDS deploy- ment. The health system’s pharmacy and therapeutics (P&T) committee has commonly been used for review of CDS recommendations, including the identification of alternative therapies or dosages that complement existing pharmacotherapy and financial poli- cies. P&T committee oversight may be either direct or achieved by forming a subcommittee specifically focused on pharmacogenomics.3,27,31,32 Addi- tionally, health-system informatics integration committees are optimal resources for implementing phar- macogenomic CDS within clinical workflows.
Developing and implementing pharmacogenomic CDS. A success- ful implementation requires institu- tionwide engagement and contribu- tions across multiple disciplines.3 For example, while pharmacists are well positioned to lead the implementa- tion of pharmacogenomic CDS, phar- macists will likely have to facilitate institutional collaboration between
clinical informatics personnel and clinical laboratories to ensure that data entered into the EHR are discrete elements that can be queried by CDS rules. Additionally, engaging the clini- cians who are most likely to prescribe a targeted drug can be helpful in en- suring that the most-desired informa- tion is delivered in a CDS alert. At most practice sites, the wording of CDS alerts is approved by the P&T commit- tee, thus providing oversight of which genetic test results are placed in the EHR and which drugs are the subject of CDS alerts.27,32
Five key concepts. Based on the work of early adopters and CPIC, five key concepts in pharmacogenomics inte- gration into the EHR have emerged: (1) documenting pharmacogenomic results in the EHR in a patient-centric and time-independent manner to facil- itate convenient access over a patient’s life, (2) providing a clinical interpreta- tion of test results (e.g., a predicted phenotype) and clinical recommenda- tions, (3) entering genetic results and interpretations in discrete EHR fields to facilitate CDS and future retrieval as needed, (4) providing drug-specific pharmacotherapy recommendations based on test results and the clinical interpretation, and (5) deploying CDS to guide the application of pharmaco- genomics at the points of prescribing and dispensing. Aside from CPIC, other organizations with experience in this area have published considerations for implementing pharmacogenomics in the EHR.31 Broader technical de- siderata for the implementation of all types of genomic data into the EHR have been published and subsequently updated.41,42
An important initial step in de- veloping pharmacogenomic CDS is determining how test results will be placed in the EHR. Traditionally, ge- nomic results were often reported as scanned documents or PDF files, which made it impossible to drive CDS because results were not represented in a discrete manner. To maximize the potential of pharmacogenomic CDS, test results need to be entered
as discrete EHR data fields. This step may require significant collaboration with the clinical laboratories that re- port genomic results. However, once achieved, discrete data representation facilitates the development of decision support rules and customized alerts.
Different levels of interpretation are needed. Depending on the practice site and the clinical laboratory involved, genomic data may be displayed at dif- ferent levels of interpretation in the EHR. Pharmacogenomic test results may be reported by reference labora- tories in terms of a haplotype (i.e., the summary of all genetic variants inher- ited on a single gene); a diplotype (i.e., the summary of genetic variants in- herited from both mother and father); the presence of a genetic variant (e.g., positive or negative); or a predicted phenotype (e.g., poor metabolizer). Figure 1 illustrates the relationship between the different levels of inter- pretations (alleles, diplotypes, pheno- type) a therapeutic recommendation. Some sites have circumvented the need to discretely report the diplotype by adding custom-built phenotype terminologies (e.g., CYP2D6 ultrarap- id metabolizer) into a problem list or “genomics tab” within the EHR.27 The discrete entry of phenotypes into the EHR can serve as a trigger for inter- ruptive CDS alerts. Some institutions heavily favor the therapeutic recom- mendation as the main way to display pharmacogenomic results.37 Given the ever-changing and evolving nature of genomic data, the informatics team needs to build CDS rules in a way that can be updated and modified when new clinically actionable evidence emerges. Ideally, this process should be fully automated to allow for scal- ability and sustainability.
Evaluate options to display re- sults. Once the first two steps of pharmacogenomic CDS implemen- tation (capturing discrete results and determining the level of interpretation to display) are completed, the next step is to decide where to display the results (Figure 2). There are multiple solutions for this step, and a single
SPECIAL FEATURES ELECTRONIC HEALTH RECORDS
1970 AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016
site might use several approaches. Maximizing the clinical value of the genomic result must be a driving prin- ciple behind this decision. Because pharmacogenomic data are useful over a patient’s lifetime, results should be displayed in a time-independent man- ner, thus allowing clinicians an easy way to view and access the result at any time. The data must be displayed in a location commonly accessed in a provider’s routine workflow. For exam- ple, a genetic test result might appear in an order entry or verification screen as well as a laboratory result section. Interpretive reports should also be readily available to ensure that the po- tential clinical impact of the result is known by the clinician.
Clinical recommendations to mod- ify drug therapy are based on the pa- tient’s predicted phenotype. Because there is not a standardized process for assigning a predicted phenotype based on genotype results, there may be rare instances where reference laboratory interpretations differ. For laboratories using whole genome or exome sequencing, rare alleles with little associated phenotypic data may be observed. These examples high- light the importance of developing a systematic process of genotype- to-phenotype assignment to ensure
consistent interpretations and as- sociated CDS across the health sys- tem. Pharmacogenomic translation tables have been used as a guide to translate genotype data to pheno- types and associated CDS rules.24 Links to these tables, consisting of thousands of genotype-to-phenotype translations, can be found in the aforementioned CPIC guidelines.7-22
Because the variability of phar- macogenomic terms can create con- fusion, CPIC conducted a consensus process to standardize terms for clini- cal pharmacogenomic results.43 This process has led to the creation of stan- dardized pharmacogenomic terms for phenotypic and allele functional status that are understood across the pharmacogenomic community.44 Us- ing standard terms will help faciliate sharing of data across diverse EHRs, and institutions that have implement- ed pharmacogenomic testing, as well as laboratories that perfom pharma- cogenomic analyses, are now begin- ning to use these new standardized terms.
Optimal active CDS should be gene–drug specific and consider all aspects of a given clinical scenario. For example, drug-specific alerts must consider all formulations of a drug. For example, oral dapsone should
be avoided in patients who have a glucose-6-phosphate dehydrogenase (G6PD) deficiency associated with he- molytic anemia, but topical dapsone may be given to such patients without the occurrence of hemolytic anemia.45
An ideally designed CDS rule for this gene–drug combination would in- tegrate route of administration and interrupt prescribers only when oral dapsone is ordered in patients with the G6PD deficiency.
Reducing the risk of alert fatigue. A common concern with creating any new CDS alert is that prescribers will be presented with an increased num- ber of interruptive alerts, which could lead to alert fatigue that results in cli- nicians ignoring clinically important messages.46,47 It is important to involve the prescribing clinicians in the CDS design process to ensure that informa- tion displayed is useful in guiding a pa- tient’s care. Given the custom nature of CDS alerts, there may be various opportunities to design highly spe- cific alerts that can limit the potential for alert fatigue. For example, there is a well-established pharmacogenomic relationship between a variant form of HLA-B, the gene coding for a protein in the human leukocyte antigen (HLA) complex that plays a key role in im- mune response, and adverse reactions
Figure 1. Hierarchy of interpretive levels that can be applied in pharmacogenomic clinical decision support. A haplotype summarizes all genetic variants inherited on a single gene (e.g., CYP2D6, the gene coding for cytochrome P-450 isozyme 2D6). A diplotype summarizes the genetic variants inherited from both mother and father (in the examples shown, the first patient has one normal functional copy of the CYP2D6*1 allele and one reduced-function CYP2D6*10 allele; the second patient has deletions of both CYP2D6 alleles). The phenotype is the genetic expression in the patient.
Haplotype
Diplotype
Phenotype
Therapeutic recommendation
Examples: *1, *2, *3, *17
Examples: CYP2D6 (*1/*10) or CYP2D6 (*5/*5)
Example: CYP2D6 poor metabolizer
Example: Do not use codeine. Please select an alternative therapeutic agent.
ELECTRONIC HEALTH RECORDS SPECIAL FEATURES
AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016 1971
to carbamazepine; this relationship is noted in the Food and Drug Adminis- tration (FDA)–approved labeling as a boxed warning.48 Individuals with the variant allele HLA-B*15:02 who receive carbamazepine are at an increased risk for serious adverse events, includ- ing Stevens–Johnson syndrome and toxic epidermal necrolysis.11 The vari- ant allele is most common in patients of Asian ancestry (and this higher frequency is noted in the FDA boxed warning). When designing a CDS alert for this pharmacogenomic relation- ship, data on Asian ancestry could be leveraged to improve alert specificity; for example, a CDS alert could prompt HLA-focused genetic testing when carbamazepine is ordered but would be presented only if the patient were of Asian descent.49 This approach, however, could pose a problem be- cause race and ancestry data may not always be reported in the EHR.50
Strategic design decisions. Another consideration in developing pharma- cogenomic CDS is to determine if all aspects of the CDS initiative should be built by the health system or if consul- tants and additional vendors should be engaged. As published models have illustrated, select hospitals have the informatics infrastructure neces- sary to develop and maintain pharma- cogenomic CDS. However, as the use of pharmacogenomics expands, this option might become harder to sus- tain. New pharmacogenomic knowl- edge and evidence are emerging con- stantly, which may prompt changes in custom CDS alerts; with time, the on- going maintenance may become over- whelming. Ancillary systems integrat- ed with the health-system EHR could help minimize the time required to create, store, and maintain CDS rules. In addition, software designed specif- ically for pharmacogenomics may of- fer additional features and increased flexibility, including patient-specific portals for secure review of pharma- cogenomic test results and genetics- specific CDS rule engines. Also, with the execution of complex CDS rules outside the primary EHR, pharma- F
ig u re
2 .
G en
er al
o ve
rv ie
w o
f st
ep s
n ee
d ed
t o
s u cc
es sf
u lly
in co
rp o
ra te
p h ar
m ac
o g
en o
m ic
t es
t re
su lts
in to
t h e
el ec
tr o
n ic
h ea
lth r
ec o
rd (E
H R
) f o
r ac
tiv e
cl in
ic al
d ec
is io
n
su p
p o
rt .
O rd
e r
a m
u lti
g e
n e
p h
a rm
a co
g e
n o
m ic
te st
S to
re d
is cr
e te
g
e n
e tic
r e
su lts
Y e
s
N o
A ct
iv e
C lin
ic a
l D e
ci si
o n
S u
p p
o rt
T o
o l
In te
rf a
ce w
ith la
b o
ra to
ry in
fo rm
a tio
n s
ys te
m
E xt
ra ct
d is
cr e
te d
a ta
u si
n g
s ta
n d
a rd
p ro
ce d
u re
s
A re
th e r
e su
lts re
tu rn
e d in
a d
is cr
e te
e le
ct ro
n ic
fo rm
a t?
Y e
s
N o
M e
rg e
g e
n e
tic re
su lts
w ith
c lin
ic a
l in
te rp
re ta
tio n
s
P a
rs e
o u
t a
n d
s e
le ct
o n
ly c
lin ic
a l
in fo
rm a
tiv e
r e
su lts
A re
a ll
g e n e tic
r e su
lts cl
in ic
a lly
in fo
rm a tiv
e ?
M ig
ra te
g e
n e
tic re
su lts
in to
t h
e E
H R
S to
re a
n d
d is
p la
y re
su lts
in
a p
a tie
n t
ce n
tr ic
a
n d
t im
e -i n
d e
p e
n d
e n
t fa
sh io
n
A ct
iv e
C lin
ic a
l D e
ci si
o n
S u
p p
o rt
T o
o l
SPECIAL FEATURES ELECTRONIC HEALTH RECORDS
1972 AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016
cogenomic CDS does not become detrimental to the performance of the health system’s primary EHR. Tak- ing this approach cannot eliminate all the challenges of pharmacogenomic CDS. For example, these systems do not eliminate the complexities that result from the different ways clini- cal laboratories provide results and organize data in the EHR. Further, if these ancillary systems are not fully integrated with the EHR, the proc- ess of patient care is slowed because clinicians have to access another re- source to retrieve a patient’s pharma- cogenomic information. Additional processes may be needed to make certain clinicians reliably access phar- macogenomic information. Finally, as with any clinical software, the benefits must be weighed against challenges such as additional costs, security is- sues, and the need to develop inter- faces with the EHR.
Resources to support develop- ment of pharmacogenomic CDS. As previously described, individu- al health systems develop custom, institution-specific pharmacogenom- ic CDS because database vendors that support other types of CDS in widely used commercial EHRs do not cur- rently provide pharmacogenomic CDS content. While standard tools for developing rule-based CDS within commercial EHRs have been success- fully used, substantial site effort is re- quired at each health system to orga- nize relevant medication and genomic knowledge, define the clinical work- flow, and prepare alert text. Vendor- agnostic resources that should help reduce the burden on individual sites are emerging from CPIC and the Dis- playing and Integrating Genetic Infor- mation Through the EHR (DIGITizE) initiative sponsored by the National Academy of Medicine (NAM, formerly the Institute of Medicine).
Established in 2009, CPIC provides clinical practice guidelines that enable the translation of genetic laboratory test results into actionable prescrib- ing decisions for specific drugs.51 CPIC guidelines closely follow NAM’s “Stan-
dards for Developing Trustworthy Clinical Practice guidelines.”23 CPIC members have expertise in various aspects of pharmacogenomics, and many are involved in the implemen- tation of clinical pharmacogenomics. CPIC recognized that successful adop- tion of pharmacogenomics in routine clinical care requires a curated and machine-readable database of phar- macogenomic knowledge suitable for use in an EHR with CDS. In 2013, CPIC formed the CPIC Informatics Working Group to support the adoption of CPIC guidelines into the clinical electronic environment. Starting with guidelines focused on HLA-B genotype and risks of abacavir use, CPIC has implement- ed a primary strategy of systematically incorporating a set of implementation resources applicable to any EHR into all CPIC guidelines.18,41
By developing comprehensive tables that help clinicians translate genotype information to phenotype predictions and associated clinical rec- ommendations using human-readable and structured text, CPIC provides foundational resources for all EHR implementation steps. The transla- tion tables relate pharmacogene “star allele” nomenclature, allele function, and phenotype, which allows results from multiple formats to be translated into a uniform and manageable data set. The CPIC guidelines include dip- lotypes of known functional signifi- cance as a supplementary table, and a complete table with all possible dip- lotypes (in one case, over 700) is post- ed to PharmGKB, a federally funded pharmacogenomics knowledge base managed by Stanford University.12 For each diplotype, the functional effect on the protein, the representation of gene and drug in standard vocabular- ies, and clinical workflows incorpo- rating test results are presented. Sug- gested text for preprescription and postprescription consultations regard- ing each diplotype is presented; this text may be used within CDS alerts. The standardized-nomenclature tables help with potential data transfer both inside an EHR and outside the EHR to
additional resources. The implementa- tion workflows suggest best practices and facilitate a common understand- ing of the implementation process across the multidisciplinary team.
NAM’s DIGITizE effort has pro- vided a forum for EHR vendors, clini- cal laboratories, and academic centers with experience implementing clini- cal genomics to collaborate on how genomic data can be consistently represented and integrated into the EHR.52 The group has focused exten- sively on pharmacogenomics and, in 2015, published an implementation guide pertaining to both the HLA- B*5701 allele and risks of abacavir use and to the impact of certain variants of the TPMT gene on thiopurine dos- ing. The guide outlines a framework, with specific details, for implement- ing CDS rules for these gene–drug pairs into the EHR. The guide reviews using standard clinical vocabular- ies, including the Logical Observa- tion Identifiers Names and Codes (LOINC) system (Regenstrief Institute, Inc., Indianapolis, IN) and the Sys- tematized Nomenclature of Medicine (SNOMED) Clinical Terms product (SNOMED CT, International Health Terminology Standards Development Organisation, Copenhagen, Denmark) to represent pharmacogenomic re- sults and provide active CDS.
Future directions. Although preci- sion medicine is often thought of as a genomics-based model, the con- cept of precision medicine goes well beyond an individual’s genomic pro- file to include family history, envi- ronmental exposures, lifestyle, and many other factors. To fully embrace the power of precision medicine, ge- nomic knowledge needs to be inte- grated with other knowledge types to optimize pharmacotherapy. Pharma- cists already seek to combine all their patient-specific knowledge (e.g., vari- ous laboratory values, data on renal and liver function), and CDS provides a framework to integrate and present all relevant patient-specific factors to optimize pharmacotherapy. Combin- ing all drug interaction knowledge
ELECTRONIC HEALTH RECORDS SPECIAL FEATURES
AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016 1973
(e.g., gene–drug and drug–drug inter- action data) may be an initial step to integrate different knowledge types. Pharmacogenomic knowledge could be integrated with existing commer- cial knowledge bases already widely used to provide drug–drug interaction alerts in EHRs. Ideally, CDS should provide clinicians with a single clini- cal recommendation that considers all drug interactions and patient fac- tors to optimize drug therapy.
EHR genomic infrastructure de- velopment efforts have mostly fo- cused on building decision support tools that take into consideration single-gene data. As our knowledge of genomics improves, it is clear that
multiple genes can influence the me- tabolism of certain drugs and that electronic infrastructure development must move toward more complex de- cision support algorithms capable of providing a summative clinical rec- ommendation that takes into consid- eration multiple genetic testing results and other clinical factors. In instances where two genes influence prescribing practices, as with the genes CYP2C9 and VKORC1 and warfarin prescrib- ing, decision support triggers often consist of a summative phenotype (in this case, warfarin sensitive) that describes the impact of both genes. However, it is unlikely that a single summative phenotype can adequately
represent all combinatorial gene ef- fects in the future. The scenario of one drug being affected by multiple genes is likely to become more common as next-generation sequencing becomes the norm for genotyping patients, and decision support rules will need to evolve in order to accommodate this new type of data.53
St. Jude Children’s Research Hospi- tal has implemented CDS alerts using a “two gene–one drug” model using amitriptyline and the phenotypes for CYP2D6 and CYP2C19. Rules were cre- ated to build a CDS alert that takes into account all of the possible CYP2D6 and CYP2C19 phenotype combina- tions, and clinical recommendations
Figure 3. Screenshot showing a pharmacogenomic clinical decision support alert that takes into account nongenetic risk factors. In this case, the recommendation displayed to clinicians is specific to the patient’s age, route of medica- tion administration, and cytochrome P-450 isozyme 2C19 (CYP2C19) phenotype. Used with permission of Cerner Corporation.
SPECIAL FEATURES ELECTRONIC HEALTH RECORDS
1974 AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016
were incorporated into each possible combinatorial alert. It is expected that understanding of how different com- binations of variants present in drug transporter, drug target, and metabo- lizer genes influence pharmacothera- py will grow.
As an example of the integration of nongenetic factors into pharma- cogenomic CDS rules, posttest alerts regarding CYP2C19 phenotype sta- tus and voriconazole use recently developed at St. Jude take into ac- count a patient’s age and the route of administration of this antifungal agent. When voriconazole is ordered for a patient who is a CYP2C19 poor metabolizer, the rules engine also searches for the patient’s age and the route of administration. The clinician is then presented with a dosage rec- ommendation that takes into account genetics, age, and route of adminis- tration (Figure 3). Because voricon- azole therapeutic drug monitoring is used frequently in patients at St. Jude, the rules were also customized to be presented to clinicians only with a new voriconazole order prior to ther- apeutic drug monitoring. Also, the alert is not triggered if voriconazole has been ordered in the past 30 days or if serum voriconazole concentrations were obtained for therapeutic drug monitoring purposes in the month prior to ordering of the drug because dosing decisions should be based on the serum concentrations. These ad- ditional limits to the alert were an at- tempt to limit unnecessary alerts as part of St. Jude’s ongoing efforts to un- derstand the risks of alert fatigue and make corresponding CDS system re- finements.54,55 The implementation of pharmacogenomic data into the EHR with CDS must also be considered in the broader context of an organiza- tion’s overall approach to clinical ge- nomics and the EHR, and the general core requirements for genomics in the EHR have been suggested.41,42
Given the large size of genomic data sets, it is unlikely that these data will be stored in the EHR, and ancil- lary systems that would be integrated
with the EHR have been suggested as a solution.56-58 A commonly drawn analogy is the use of a picture ar- chiving and communication system (PACS) in radiology; with PACS tech- nology, the raw data files are stored outside the EHR and accessible only to experts, but summary informa- tion to guide patient care is readily available in the EHR. These ancillary systems may summarize and store pharmacogenomic data at the phe- notype, allele, or allelic subvariant levels. Pharmacogenomics may serve as an initial-use case for health sys- tems working to integrate genomic data into routine clinical care and for solving problems likely to be encoun- tered in broader clinical genomic in- formatics implementation programs. Ideally, solutions developed for phar- macogenomic CDS should be capa- ble of incorporating different types of genomic data and expandable to larger clinical genomic implementa- tion projects, especially considering that an organization’s approach to storing and presenting genomic data may change over time.
Conclusion. The EHR with CDS is essential to curate pharmacogenomic data and disseminate patient-specific information at the point of care. As part of the successful implementation of pharmacogenomics into clinical settings, all relevant clinical recom- mendations pertaining to gene–drug pairs must be summarized and pre- sented to clinicians in a manner that is seamlessly integrated into the clini- cal workflow of the EHR. In some situations, ancillary systems and ap- plications outside the EHR may be in- tegrated to augment the capabilities of the EHR.
Acknowledgments Kelly E. Caudle, Pharm.D., Ph.D., and Donald K. Baker, Pharm.D., M.B.A., are acknowledged for their guidance and feedback during the development of this article.
Disclosures Dr. Haidar and Dr. Hoffman are funded by National Institutes of Health grant R24GM115264, awarded to CPIC, and by
ALSAC; they have declared no other potential conflicts of interest. The other authors have declared no potential conflicts of interest.
References 1. Johnson JA, Weitzel KW. Advancing
pharmacogenomics as a component of precision medicine: how, where and who? Clin Pharmacol Ther. 2016; 99:154-6.
2. Relling MV, Evans WE. Pharmacoge- nomics in the clinic. Nature. 2015; 526:343-50.
3. Owusu-Obeng A, Weitzel KW, Hatton RC et al. Emerging roles for pharma- cists in clinical implementation of pharmacogenomics. Pharmacotherapy. 2014; 34:1102-12.
4. Dunnenberger HM, Crews KR, Hoffman JM et al. Preemptive clinical pharma- cogenetics implementation: current programs in five US medical centers. Annu Rev Pharmacol Toxicol. 2015; 55:89-106.
5. Johnson SG. Leading clinical phar- macogenomics implementation: advancing pharmacy practice. Am J Health-Syst Pharm. 2015; 72:1324-8.
6. Wilke RA, Xu H, Denny JC et al. The emerging role of electronic medical records in pharmacogenomics. Clin Pharmacol Ther. 2011; 89:379-86.
7. Gammal RS, Court MH, Haidar CE et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for UGT1A1 and atazanavir prescribing. Clin Pharmacol Ther. 2016; 99:363-9.
8. Saito Y, Stamp LK, Caudle KE et al. Clinical Pharmacogenetics Imple- mentation Consortium (CPIC) guide- lines for human leukocyte antigen B (HLA-B) genotype and allopurinol dosing: 2015 update. Clin Pharmacol Ther. 2015; 99:36-7.
9. Hicks JK, Bishop JR, Sangkuhl K et al. Clinical Pharmacogenet- ics Implementation Consortium (CPIC) guideline for CYP2D6 and CYP2C19 genotypes and dosing of selective serotonin reuptake inhibitors. Clin Pharmacol Ther. 2015; 98:127-34.
10. Birdwell KA, Decker B, Barbarino JM et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for CYP3A5 genotype and tacrolimus dosing. Clin Pharmacol Ther. 2015; 98:19-24.
11. Leckband SG, Kelsoe JR, Dunnenberger HM et al. Clinical Pharmacogenetics Implementation Consortium guide- lines for HLA-B genotype and car- bamazepine dosing. Clin Pharmacol Ther. 2013; 94:324-8.
ELECTRONIC HEALTH RECORDS SPECIAL FEATURES
AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016 1975
12. Ramsey LB, Johnson SG, Caudle KE et al. The Clinical Pharmacogenetics Implementation Consortium guide- line for SLCO1B1 and simvastatin- induced myopathy: 2014 update. Clin Pharmacol Ther. 2014; 96:423-8.
13. Relling MV, McDonagh EM, Chang T et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for rasburicase therapy in the context of G6PD deficiency genotype. Clin Pharmacol Ther. 2014; 96:169-74.
14. Clancy JP, Johnson SG, Yee SW et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines for ivacaftor therapy in the context of CFTR genotype. Clin Pharmacol Ther. 2014; 95:592-7.
15. Crews KR, Gaedigk A, Dunnenberger HM et al. Clinical Pharmacogenetics Implementation Consortium guide- lines for cytochrome P450 2D6 geno- type and codeine therapy: 2014 update. Clin Pharmacol Ther. 2014; 95:376-82.
16. Muir AJ, Gong L, Johnson SG et al. Clinical Pharmacogenetics Imple- mentation Consortium (CPIC) guide- lines for IFNL3 (IL28B) genotype and PEG interferon-a-based regimens. Clin Pharmacol Ther. 2014; 95:141-6.
17. Relling MV, Gardner EE, Sandborn WJ et al. Clinical Pharmacogenetics Implementation Consortium guide- lines for thiopurine methyltransfer- ase genotype and thiopurine dosing: 2013 update. Clin Pharmacol Ther. 2013; 93:324-5.
18. Martin MA, Hoffman JM, Freimuth RR et al. Clinical Pharmacogenet- ics Implementation Consortium guidelines for HLA-B genotype and abacavir dosing: 2014 update. Clin Pharmacol Ther. 2014; 95:499-500.
19. Scott SA, Sangkuhl K, Stein CM et al. Clinical Pharmacogenetics Imple- mentation Consortium guidelines for CYP2C19 genotype and clopidogrel therapy: 2013 update. Clin Pharmacol Ther. 2013; 94:317-23.
20. Johnson JA, Gong L, Whirl-Carrillo M et al. Clinical Pharmacogenetics Im- plementation Consortium Guidelines for CYP2C9 and VKORC1 genotypes and warfarin dosing. Clin Pharmacol Ther. 2011; 90:625-9.
21. Hicks JK, Swen JJ, Thorn CF et al. Clinical Pharmacogenetics Imple- mentation Consortium guideline for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants. Clin Pharmacol Ther. 2013; 93:402-8.
22. Swen JJ, Nijenhuis M, de Boer A et al. Pharmacogenetics: from bench to byte—an update of guidelines. Clin Pharmacol Ther. 2011; 89:662-73.
23. Caudle KE, Klein TE, Hoffman JM et al. Incorporation of pharmacoge- nomics into routine clinical prac- tice: the Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline development process. Curr Drug Metab. 2014; 15:209-17.
24. Hicks JK, Crews KR, Hoffman JM et al. A clinician-driven automated system for integration of pharmacogenetic interpretations into an electronic medical record. Clin Pharmacol Ther. 2012; 92:563-6.
25. Herr TM, Bielinski SJ, Bottinger E et al. Practical considerations in genomic decision support: the eMERGE experi- ence. J Pathol Inform. 2015; 6:50.
26. Troiano D, Jones MA, Smith AH et al. ASHP guidelines on the design of database-driven clinical deci- sion support: strategic directions for drug database and electronic health records vendors. Am J Health-Syst Pharm. 2015; 72:1499-505.
27. Bell GC, Crews KR, Wilkinson MR et al. Development and use of active clinical decision support for preemp- tive pharmacogenomics. J Am Med Inform Assoc. 2014; 21:e93-9.
28. Pulley JM, Denny JC, Peterson JF et al. Operational implementation of prospective genotyping for person- alized medicine: the design of the Vanderbilt PREDICT project. Clin Pharmacol Ther. 2012; 92:87-95.
29. Lazaridis KN, McAllister TM, Babovic-Vuksanovic D et al. Imple- menting individualized medicine into the medical practice. Am J Med Genet C Semin Med Genet. 2014; 166C:15-23.
30. Weitzel KW, Elsey AR, Langaee TY et al. Clinical pharmacogenetics imple- mentation: approaches, successes, and challenges. Am J Med Genet C Semin Med Genet. 2014; 166C:56-67.
31. Peterson JF, Bowton E, Field JR et al. Electronic health record design and implementation for pharmacoge- nomics: a local perspective. Genet Med. 2013; 15:833-41.
32. Hoffman JM, Haidar CE, Wilkinson MR et al. PG4KDS: a model for the clinical implementation of pre- emptive pharmacogenetics. Am J Med Genet C Semin Med Genet. 2014; 166C:45-55.
33. Shuldiner AR, Relling MV, Peterson JF et al. The Pharmacogenomics Research Network Translational Phar- macogenetics Program: overcoming challenges of real-world implemen- tation. Clin Pharmacol Ther. 2013; 94:207-10.
34. Van Driest SL, Shi Y, Bowton EA et al. Clinically actionable genotypes
among 10,000 patients with preemp- tive pharmacogenomic testing. Clin Pharmacol Ther. 2014; 95:423-31.
35. Bielinski SJ, Olson JE, Pathak J et al. Preemptive genotyping for personal- ized medicine: design of the right drug, right dose, right time—using genomic data to individualize treatment proto- col. Mayo Clin Proc. 2014; 89:25-33.
36. Goldspiel BR, Flegel WA, DiPatrizio G et al. Integrating pharmacogenetic information and clinical decision sup- port into the electronic health record. J Am Med Inform Assoc. 2014; 21:522-8.
37. O’Donnell PH, Bush A, Spitz J et al. The 1200 patients project: creating a new medical model system for clinical im- plementation of pharmacogenomics. Clin Pharmacol Ther. 2012; 92:446-9.
38. Crews KR, Cross SJ, McCormick JN et al. Development and implementation of a pharmacist-managed clinical pharmacogenetics service. Am J Health-Syst Pharm. 2011; 68:143-50.
39. Rasmussen-Torvik LJ, Stallings SC, Gordon AS et al. Design and an- ticipated outcomes of the eMERGE- PGx project: a multicenter pilot for preemptive pharmacogenomics in electronic health record systems. Clin Pharmacol Ther. 2014; 96:482-9.
40. Hoffman JM, Dunnenberger HM, Hicks JK et al. Developing knowl- edge resources to support precision medicine: principles from the Clinical Pharmacogenetics Implementation Consortium (CPIC). J Am Med Inform Assoc. 2016; 23:796-801.
41. Masys DR, Jarvik GP, Abernethy NF et al. Technical desiderata for the integration of genomic data into electronic health records. J Biomed Inform. 2012; 45:419-22.
42. Welch BM, Eilbeck K, del Fiol G et al. Technical desiderata for the integra- tion of genomic data with clinical decision support. J Biomed Inform. 2014; 51:3-7.
43. Devine EB, Lee CJ, Overby CL et al. Usability evaluation of pharmacoge- nomics clinical decision support aids and clinical knowledge resources in a computerized provider order entry system: a mixed methods approach. Int J Med Inform. 2014; 83:473-83.
44. Caudle KE, Dunnenberger HM, Freimuth RR et al. Standardizing terms for clinical pharmacogenetic test results: consensus terms from the Clinical Pharmacogenetics Imple- mentation Consortium (CPIC). Genet Med. Epub ahead of print. 2016 Jul 21.
45. Pamba A, Richardson ND, Carter N et al. Clinical spectrum and sever- ity of hemolytic anemia in glucose 6-phosphate dehydrogenase–deficient
SPECIAL FEATURES ELECTRONIC HEALTH RECORDS
1976 AM J HEALTH-SYST PHARM | VOLUME 73 | NUMBER 23 | DECEMBER 1, 2016
children receiving dapsone. Blood. 2012; 120:4123-33.
46. Carspecken CW, Sharek PJ, Longhurst C, Pageler NM. A clinical case of elec- tronic health record drug alert fatigue: consequences for patient outcome. Pediatrics. 2013; 131:e1970-3.
47. Van der Sijs H, Aarts J, Vulto A, Berg M. Overriding of drug safety alerts in computerized physician order entry. J Am Med Inform Assoc. 2006; 13:138-47.
48. Sherry ST, Ward MH, Kholodov M et al. dbSNP: the NCBI database of genetic variation. Nucleic Acids Res. 2001; 29:308-11.
49. Hoffman MA, Williams MS. Electronic medical records and personalized medicine. Hum Genet. 2011; 130:33-9.
50. Liao KP, Cai T, Gainer V et al. Electronic medical records for discovery research in rheumatoid arthritis. Arthritis Care Res. 2010; 62:1120-7.
51. Relling MV, Klein TE. CPIC: Clinical Pharmacogenetics Implementation Consortium of the Pharmacogenom- ics Research Network. Clin Pharmacol Ther. 2011; 89:464-7.
52. Health and Medicine Division, Na- tional Academy of Sciences. DIGITizE: Displaying and Integrating Genetic In- formation Through the EHR. www.na- tionalacademies.org/hmd/Activities/ Research/GenomicBasedResearch/ Innovation-Collaboratives/EHR.aspx (accessed 2016 Jan 12).
53. Yang W, Wu G, Broeckel U et al. Comparison of genome sequencing and clinical genotyping for pharma- cogenes. Clin Pharmacol Ther. 2016; 100:380-8.
54. Hoffman JM, Baker DK, Howard SC et al. Safe and successful implementa- tion of CPOE for chemotherapy at a children’s cancer center. J Natl Compr Canc Netw. 2011; 9:S36-50.
55. McDaniel RB, Burlison JD, Baker DK et al. Alert dwell time: introduction of a measure to evaluate interruptive clini- cal decision support alerts. J Am Med Inform Assoc. 2016; 23(e1):e138-41.
56. Starren J, Williams MS, Bottinger EP. Crossing the omic chasm: a time for omic ancillary systems. JAMA. 2013; 309:1237-8.
57. Chute CG, Kohane IS. Genomic medicine, health information technol- ogy, and patient care. JAMA. 2013; 309:1467-8.
58. Shirts BH, Salama JS, Aronson SJ et al. CSER and eMERGE: current and potential state of the display of genetic information in the electronic health record. J Am Med Inform Assoc. 2015; 22:1231-42.
Copyright of American Journal of Health-System Pharmacy is the property of American Society of Health System Pharmacists and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
