Drug commercialization depends on reaching the right physicians at the right moment. Identifying those physicians and understanding the patients they treat has long depended on real-world data, or RWD. Billing claims, electronic health records, or EHRs, and prescription data offer some useful insights, but their partial, rearview perspectives are surpassed by the precision, granularity and timeliness of de-identified laboratory data.
Understanding the differences among RWD sources is key to commercial success. Claims data traditionally has been a primary input for commercial planning. Yet the source has significant limitations, in part because it often identifies patients and physicians after a treatment decision has been made. EHRs offer richer clinical context but typically aggregate and summarize test results rather than preserve the raw values that support precise analyses.
In contrast, lab data provides test results that are a direct measure of biological processes. Biopharma companies can look at tests ordered or test results to identify diagnostic gaps, understand disease progression, and uncover testing and treatment trends. Importantly, commercial teams receive those insights early enough to act on them.
“Lab data gives commercial teams a window into the patient journey before the physician writes a prescription and, in some cases, before they even reach a diagnosis. That visibility changes how teams reach physicians at critical moments in the diagnostic journey and drive adoption of therapies,” Parag More, Executive Director of Healthcare Analytics Solutions for Lifesciences at Quest Diagnostics, said.
Improving patient identification
Lab data identifies physicians treating patients who are eligible for a given therapy today, as well as those caring for people who may benefit from second- or third-line therapies. Because lab results are available within days of testing, commercial teams can engage those physicians during the short window between diagnosis and treatment when prescription decisions are made.
Acute myeloid leukemia, or AML, illustrates the value of this timing. AML is deemed a medical emergency, and immediate treatment has been the standard of care for decades.1,2 Treatment decisions rest on test results, such as FLT3 mutation status, that determine a patient’s eligibility for targeted therapies. Access to test results empowers teams to identify the treating physician and deliver education on appropriate therapies at the precise moment a drug is being selected.
The value of lab data extends beyond the timely but retrospective identification of diagnosed patients. Increasingly, teams are using the resource prospectively, looking at historical trends of biomarkers over time to uncover people who may be misdiagnosed or suspected of having a disease. The approach supports a more nuanced understanding of the market.
There is a clear need for improvements to the diagnostic pathway. In rare diseases, the average time between symptom onset and confirmed diagnosis is 4.7 years, and 73% of patients are misdiagnosed at least once.3 In many disease areas, the addressable market is larger than the diagnosed population.
Integrating claims and lab data can shorten the journey to a rare-disease diagnosis. Claims records may reveal patterns of specialist referrals that point toward a clinical suspicion, while lab data can expose gaps in the patient’s diagnostic journey, such as confirmatory testing that was never ordered or results that are misaligned with the recorded diagnosis.
Used in combination, the two data sources empower commercial teams to pinpoint physicians whose patients may need further diagnostic workups to either confirm or exclude a condition. That opens a door for biopharma companies, first to support physicians with guidance on the right diagnostic approach and ultimately to introduce them to available treatment options.
Fueling predictive AI models
Artificial intelligence and machine learning, or AI/ML, models based on lab data are supporting the shift toward proactive identification of patients who may benefit from a treatment. Because lab data provides actual test results, the source is a preferred input for training predictive models that researchers have found can shorten the diagnostic journey.
A machine-learning algorithm trained on complete blood count data identified 95 undiagnosed patients at high risk of myelodysplastic syndrome, leading to confirmed diagnoses of the blood cancer.4 Another team used machine learning and lab data to improve detection of common variable immunodeficiency disease.5
The algorithms work prospectively, identifying undiagnosed patients whose health records match those of the known population and pointing commercial teams toward their treating physicians. Commercial teams can then educate physicians on the patient journey and recommend specific tests to confirm a diagnosis, expanding the treatable patient population.
Adding predictive analytics to lab data’s established ability to identify existing patients is enhancing the value of test results to commercial teams. Needing evidence-based ways to identify the right patients and providers, biopharma companies risk misunderstanding the market and overlooking opportunities if they exclude lab data from their strategies.
Lab data is becoming the launchpad for precision commercialization, faster therapy adoption and better health outcomes. Contact Quest Diagnostics to learn how its unique repository of real-world lab data is empowering companies to drive therapy adoption.
References
- Burchenal, J. H., Murphy, M. L. & Tan, C. T. Treatment of acute leukemia. Pediatrics 18, 643–660 (1956).
- Burnett, A. K. et al. A comparison of low-dose cytarabine and hydroxyurea with or without all-trans retinoic acid for acute myeloid leukemia and high-risk myelodysplastic syndrome in patients not considered fit for intensive treatment. Cancer 109, 1114–1124 (2007).
- Faye, F. et al. Time to diagnosis and determinants of diagnostic delays of people living with a rare disease: results of a Rare Barometer retrospective patient survey. Eur J Hum Genet 32, 1116–1126 (2024).
- Lason, A. et al. A multi-center study of a machine learning algorithm for identifying undiagnosed patients with myelodysplastic syndrome based on complete blood count data. Blood 146, 5629–5629 (2025)
- Johnson, R. et al. Electronic health record signatures identify undiagnosed patients with common variable immunodeficiency disease. Sci Transl Med 16, eade4510 (2024).