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

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

Hospital Targeting

Switching Analyses

Switching analysis allows us to understand when our drug is chosen which competitive drugs are shunned and when our drug is dropped which competitive drugs benefit. Switching is related to dosing in that it may be hinting at the fact that the physician has not prescribed a stronger dose of the current therapy before moving on to a different therapy (bad if our drug constitutes the current therapy). Two types of switching analyses hold special interest: switching due to physician visits and switching due to hospital visits. The latter is very relevant when studying hospital-retail spillover.

Example - Objectives

The objective here is to understand the impact of hospitalization on the usage of a leading anti-psychotic drug.

  • Do patients consume more or less of our drug after passage through the hospital? What is the impact of hospitalization on the retail consumption of our drug?
  • What is the predominant source of business of our drug: switches, new therapy starts, or add-on’s?
  • In case of switches to our drug, which drugs are shunned? In cases of switches from our drug, which drugs are favored?
  • How do the adoption/rejection dynamics of our drug in retail vary by ICD group and patient profile (age and sex)?
  • Can a “spillover” be defined? How significant is it?

Switch-to’s: Which drugs are being shunned in our favor?

Drugs shunned in favor of our hypnotic drug are:

Drug 1 (43% of switches), Drug 2 (38%), and Drug 3 (9%).

Switch-From’s: Which drugs are being adopted to our detriment?

Drugs favored at the detriment of the drug of interest are:

Drug 2 (46% of switches), Drug 1 (38%), and Drug 3 (9%).

Hypnotic Switching: Switch From vs. Switch To

Are we gaining or losing business?

The bad news is we are losing business (switch from us) at a higher rate than we are gaining business (switch to us). Indeed, 41% of the switches are switch-to vs. a more impressive 59% switch-from. Also, note most of the switches in either direction are simple switches.

Issue #1: Who is the Switcher?

Consider a patient that sees a physician, then another physician, then back to the first physician.

In light of the above, one would certainly not label Dr. John Smith as a switcher since he maintained the therapy initiated by Dr. Jane Doe. This suggests however that if information on Dr. Jane Doe were not available, one would erroneously conclude Dr. John Smith is the switcher. What one needs to do in that case is to look at other patients of Dr. John Smith.

Is the Switcher in the System?

Why bother about hospital information?

In the example above, it would appear the switching was initiated by the treating physician in retail. In actuality, the switching took place in the hospital! This means not having hospital information may lead to erroneous conclusions.

For additional information about this product please send an e-mail to