How can you use patient level data to help rebound from a sluggish product launch?

Our client, a growing specialty pharmaceutical company, has a new product approved for treating two different diseases. After a year on the market, however, the new drug’s uptake is not as fast as expected and sales are underwhelming. The client would like to understand the reasons for the slow launch by examining physician prescribing behavior and refine the business strategy accordingly. Unfortunately, there is a catch: for this particular product, traditional physician prescription data has neither good coverage nor does it have the granularity on physician prescribing behavior required to drive solid business decisions.

Since the product has multiple indications, it is difficult to distinguish sales by indication through traditional physician prescription data.

Our client suspects that the prescribing behavior for the two different indications could provide some clues. They would like to verify their assumptions and determine the marketing strategy accordingly.

KMK leveraged APLD data to separate patients by diseases that the product is indicated to treat, did a deep dive into understanding treatment patterns, and assessed prescribing behavior by physician’s specialty.  The result not only provided a more accurate estimate of the current market share by indication which helped in forecasting, but it also brought valuable business insights to the table for determining the next phase of marketing strategies and tactics, such as developing an enhanced physician education program for generalists and patient segmentation.

Activities

  • Claims data is lacking the necessary clinical information to identify patients accurately.
  • Unlike clinical trials, the dose schedule and frequency of medication actually taken by different patients varies.
  • It is increasingly difficult to identify clinical and financial measurements that are meaningful to payers.
  • There is potential population heterogeneity between database and payer’s population.

Results

  • The product has a much higher market share for one indication (to treat a more severe disease) over the other.
  • Specialists tend to write more prescriptions of the product compared to generalists.
  • It takes a longer time for generalists to prescribe the product than specialists.

Key Learnings

  • The company needs to take extra care in analyzing APLD data.
  • There are missing data issues in APLD data.
  • Better study design can help minimize the bias introduced by missing data.
  • One analysis design won’t fit all business questions.
  • Each individual question should be answered with the appropriate analysis design, then the answers can be integrated in a scientific way to provide synthesis evidence.