Jean-Patrick Tsang, PhD & MBA (INSEAD)
Tel: (847)920-1000

Igor Rudychev, PhD
Tel: (847) 679-8278

Patient Level Data

What is Patient Level Data?

Patient-level data captures encounters of the individual patient with the healthcare system. This includes prescriptions, diagnoses, procedures, physician visits, hospitalizations, lab tests, etc.  To ensure anonymity of the patient, the latter is assigned a "de-identified" number. Although this number prevents one from identifying the patient (that's the goal), the same ID is used each time that patient in question interacts with the healthcare system. This is extremely powerful since this means patients can be tracked longitudinally (over time)!

To date, the gold standard has been physician-level data up. Patient-level data takes physician-level data one step further by specifying which patients (via de-identified ID's) get which therapy from which physician.

Applications of Patient Level Data

There are numerous applications of longitudinal patient-level data for the pharmaceutical industry.  In addition to the areas where physician-level data is traditionally used, patient-level data is extremely important for pharmaceutical sales and marketing, market research, brand management, sales operations, outcomes research, R&D, and clinical trials.

Classical applications of patient-level data include:

  • Compliance and Persistence
  • Sources of Business (New Therapy Starts, Switching, Add-On, Concomitancy)
  • Treatment Algorithms
  • Length of Therapy
  • Lifetime Value of Patient and Physician
  • Patient and Physician Segmentation
  • Etc.

Advanced Applications of patient-level data include:

  • Physician Influencers and Spheres of Influence
  • Identification of Key Opinion Leaders (KOL's)
  • Hospital-Retail Spillover
  • Molecular Targeting
  • Source of Business Promotion Response
  • DTC Campaign Effectiveness and DTP/DTC Equilibrium
  • Sales Forecasting (Based on Patients, not Rx's)
  • Predictive Modeling (Based on Markov Models)
  • Sales Force Sizing, Alignment, and Compensation
  • Sample Optimization (Based on Patient-Base of Physician)
  • Etc.