How Big Data Can Help Decrease Negative Drug Interactions

It’s a classic, if extreme case: a patient died. This particular patient was undergoing treatment for, among other things, rheumatoid arthritis, under the care of a physician. He contracted an infection, for which the physician prescribed amoxicillin; he was also given leflunomide (another drug used to treat RA). When his RA continued to get worse, the patient doubled his weekly dose of (previously prescribed) methotrexate without telling his doctor. This increased dose negatively interacted with the new medications, and he was dead within days.

Adverse drug reactions (ADRs) caused either by individual drugs or drug combinations are one of the top 10 causes of death in the U.S. today, and over 20% of those reported ADRs are due to interactions of co-administered drugs.

(Related: 3 Keys to Getting Provider Data Management Right)

If you’re a group health benefits advisor, it’s important to stay current on this topic — understanding how to effectively use the available data can help promote healthier outcomes for your members, reduce potentially harmful or dangerous acute care episodes, and cut overall health related costs.

Even if you have only limited involvement in health insurance sales or no involvement at all, you should know something about this topic because, in an age of pharmaceutical cornucopia, where individuals may be on multiple medications to treat a variety of morbidities, it is important to make informed and empowered decisions regarding one’s health. Increasing collaboration, transparency, and sharing knowledge with health care practitioners will only serve to benefit one’s overall well-being. We can all get sick, and can be subject to the possibility of a negative drug interaction.

So, while the above patient’s story is certainly a worst-case scenario, ADRs are by no means uncommon. While not always deadly, those negative drug interactions lead to numerous emergency room visits and hospital stays, and can affect not only a person’s health for years to come, but also their professional care networks – from pharmacists, to physicians, to other members of the health care system.

The Challenge

Pharmacists and providers typically exist in separate spheres. Each only has access to their own, limited information about a patient’s medication, and shares little of what they know. Only the payer has access to the full picture of the patient’s care from the data generated by that care. The challenge we’re facing is how to get pharmacies, providers, and payers to be able to work together collaboratively to better manage care for patients with comorbidities and multiple medications, to ultimately avoid ADRs.

Multiple barriers stand in the way of sharing this necessary information, however. To begin with, while payers have access to the raw data, most can only do so much to review, synthesize, analyze, and, ultimately, share it. A payer’s database is the singular location where all of the patient’s claims are aggregated – in contrast to hospitals, physicians, or independent pharmacies that may only have access to their respective slice of that information. To further complicate this, it’s not at all uncommon for a patient to receive prescriptions from more than one source (PCPs, specialists, urgent care, etc.) in more than one health system, and filled in more than one pharmacy. The increasing presence of polypharmacy further narrows the chances of one pharmacy having access to a complete patient medication history. Even some patients themselves might not have access to or awareness of this information. Without easy and comprehensive access to payer data, how can any single pharmacy have the knowledge necessary to intervene at the point of sale to prevent ADRs?

Another hurdle in the process is how to disseminate the data, once aggregated and understood – the success of any program begins with payer data, but it does not end there. To share important data and insights with providers, payers must create, implement, and staff special programs. This is important, yet difficult work: it requires an automated system that identifies the negative drug interaction, the prescribing parties, and who to contact and how to contact them, while also evaluating what recommendations to make instead of the potentially harmful drug combination, and measuring the impact of any changes. Without proof of a positive return on investment, most of these programs will never see implementation.

Toward A Holistic Solution

It’s true that some PBMs have programs aimed at the reduction of ADRs, and it’s true that they achieve positive outcomes – they are certainly better than nothing. However, they cannot measure up to a holistic, payer-centric model. PBM programs have no access to medical records, so they stop short of being able to track and confirm the actual impact of a potential interaction. These programs are, overall, less sensitive than the payer-based systems in the market today. And their impact necessarily stops at the pharmacy level – while they can notify a pharmacist before a prescription is filled, they don’t have the ability to prevent the prescription by contacting the physician and stopping the problem at its source.

Some payers are beginning to take steps toward a more comprehensive approach, implementing a new, data-driven program. The program begins by comparing payer data on pharmacy claims against a database of clinically recognized negative drug interactions. This helps it to identify any instances where the payer has paid for two prescriptions within a short timeframe that have known potentially negative interactions. Second, it analyzes medical claims data from the period after the interaction, looking for indications of the known side effects – any diagnoses or procedures that would be expected given the particular interaction, for example. From this, the program creates a dashboard compiling all this information, along with contact information for the providing physicians, for the payer’s care coordinators to notify them. Lastly, the program analyzes the ongoing records to evaluate the impact, both medically as well as financially, of this outreach.

Starting with payer data, then, this new method closes the gaps to piece together a more clear, comprehensive, and holistic picture of a patient’s particular care journey. By aggregating data from multiple sources, the new method encourages collaboration and communication between payers, pharmacies, and providers, all focused on the singular goal of lowering health care costs while providing superior health care, but now traveling on that same road together.

To achieve success with a managed care system, payers need to unlock the insights in their data, to help make better decisions and drive targeted care interventions, improving both health outcomes and cost. This will ultimately help sustain the goal of providing more benefits to patients’ health (and wallets) over other, fee-for-service based systems. To successfully do so requires an ever-evolving interplay of medical knowledge, large-scale data analytics, and business approaches.

— Connect with ThinkAdvisor Life/Health on Facebook, LinkedIn and Twitter.

Dane Pearson and Brian Moy are directors in the health care practice of AArete, a global consultancy specializing in data-informed performance improvement.