MOLECULAR DIAGNOSTICS
Vol. 25 • No. 3 • Page 37
An ever increasing number of clinical institutions seek to improve clinical care by using genetic testing and genomics data to personalize treatment. The confluence of lower cost testing technology, significant clinical studies and published guidelines provide the raw materials to bring precision medicine into the mainstream of clinical practice.
Studies across oncology, cardiology, psychiatry and other specialties show that personalizing medication regimens on the basis of genetic traits, known as pharmacogenetics (PGx), can:
• improve patient health outcomes and experiences by delivering the right drug at the right dose to the right patient at the right time
• reduce costs and risks for healthcare organizations by using a complete picture of the patient to prevent adverse events, protect at-risk populations, and provide the most effective care overall
• improve productivity for clinicians
Integrating Genomics Into Clinical Workflow
One of the biggest challenges associated with incorporating genomics information into treatment decisions is not having the needed information assimilated at the point of care. The sheer volume of available genetic diagnostic tests and the complex flood of molecular data these tests generate make it virtually impossible for physicians to stay current. Clinicians simply do not always know when a genetic test is indicated, or which test should be ordered.
For clinicians to feel empowered to use genetic tests, they must have confidence that they can choose the appropriate test for the clinical problem being addressed. To be practical, genetic test selection and ordering should be integrated into the clinical workflow and applications already in use to help guide decision making.
To help clinicians make sense of complex genetic data, leading healthcare organizations and labs are adopting clinical decision support tools and genomics-based knowledge bases that are continually updated with the latest genetic information. These tools enhance physician productivity and encourage use of genetic testing with limited risk and predictable costs.
Let’s consider a scenario:
To get started with a pharmacogenetics program, healthcare organizations need to establish policies that help guide which genetic test should be ordered on the basis of clinical context. These policies can be derived from consensus care guidelines compiled by scientific and regulatory bodies, as well as organizational judgement of net benefits, costs and risks, such as alert fatigue/desensitization.
When these pre-coded rules are triggered by a physician order, the system can alert the clinician that a genetic test is recommended, and provide one or more suggestions for tests that would be appropriate. The list that is offered should be predetermined by the organization based on their established policy, which may take into account the cost and reimbursement for the test, the preferred provider, and the turnaround time of the laboratory performing the test. For example:
• a physician scheduling a patient for catheterization might be alerted that a genetic test for CYP2C19 should be ordered prior to prescribing Plavix
• a patient with a new diagnosis of leukemia might trigger an alert to test for TPMT to determine how thiopurines should be prescribed
• an Asian patient that is being prescribed Abacavir to treat HIV infection might activate an alert to test for HLA type
• prescribing codeine to a nursing mother could trigger an alert to determine metabolizer genotype to avoid accidental delivery of a morphine overdose to the baby
Based on the genetic test order that is placed, the system can then assemble a complete order set for the test based upon the requirements of the laboratory and data necessary for the test. In all cases, the patient’s current and “considered” medication list must be included so that downstream processing can create alerts that highlight potential gene-drug interactions. The order should also include some indication of the intended use of the result so that the response can be succinct and targeted.
The system places the test order with the appropriate internal or external laboratory and monitors for a response. When the response arrives it should include:
• a report, typically in PDF format, that summarizes the implications of the test for the procedure or condition that triggered the test. For example, a report for a catheterization procedure would include relevant cardiovascular risk factors and medications likely to be prescribed in the course of the procedure (e.g. warfarin or statins)
• a list of alerts for all known medications for which dosing should be personalized on the basis of the patient’s genetic test results. The alert list can be used to notify the physician if any current medications, or medications that are being considered, should be personalized based upon the genetic test results
• discrete genetic test results that may be used for future clinical decision support
As soon as test results are made available in the system, alerts should be reviewed. Upon receiving the alert, the recipient may review the best practice advisory that is associated with the alert, or open the genetic test report to gain a more comprehensive picture of the consequences of the genetic test.
Genetic test results are permanently stored in the patient’s electronic medical record for future reference and use. Within a care organization, analytics applied to aggregated data can inform overall care policies, management of at-risk populations, refinement of clinical decision support patterns, value of care measurements, and negotiation with insurers regarding reimbursement.
Interpreting Genetic Test Results
To gain the benefit of precision medicine, clinicians must be able to understand the implications of genetic tests quickly and completely. Typically, clinicians have limited training in either genetics or pharmacology, and do not have time to regularly synthesize the implications of rapidly advancing science.
PGx introduces a new dimension to the prescribing decision that makes it impossible for a normal physician or pharmacist to consider all the factors for an appropriate course of action within their normal work cadence. As a result, it is important to have clinical decision support tools that enable clinicians to assess both considerations in one place, and iteratively change prescriptions until an optimal solution has been reached for the patient.
Considered in context with the prescribing workflow, genetic testing builds on concepts that clinicians are already familiar with and can be used efficiently in daily practice. But for decision support to be effective, the clinician must see all potential issues with the current medication list together.
Genetic factors add new, complex dimensions to clinical decision-making. In order to leverage genetic intelligence to enhance care, hospitals and labs must integrate new technology into the clinical workflow to guide the use and interpretation of PGx data. With a precision medicine infrastructure in place, organizations can provide patients with the most effective treatments and health management strategies.
Don Rule is CEO, Translational Software, Inc., Bellevue, Wash.
