Every formulary change is a bet. When a P&T committee moves a drug from Tier 2 to Tier 3, or replaces a brand with a biosimilar, or adds a prior authorization requirement, they are betting that the financial benefit of the change will outweigh the disruption cost. That disruption cost is not just administrative overhead. It includes members who switch drugs (some to clinically equivalent alternatives, some to clinically worse options), members who file appeals, members who contact customer service, and members who leave the plan entirely at open enrollment.

Understanding these behavioral dynamics is what we call patient stickiness modeling, and it is one of the least developed analytical capabilities in the formulary management space.

The Behavioral Spectrum

When a formulary change affects a member's current medication, that member enters a decision process. Research and claims data analysis consistently show five behavioral categories:

  1. Compliant switchers (40-50%). These members fill the alternative drug without complaint. They may not even notice the change, particularly if their pharmacist handles the substitution at the point of dispensing.
  2. Reluctant switchers (15-25%). These members switch but contact customer service first, or discuss the change with their physician. They add operational cost but ultimately comply.
  3. Appeal filers (5-15%). These members request an exception to remain on their current medication. The appeal rate correlates strongly with the clinical significance of the affected drug. Specialty medications see higher appeal rates than generics.
  4. Non-adherent (5-10%). Some members simply stop taking their medication rather than switching or appealing. This is the most clinically dangerous outcome and generates downstream costs in ER visits and hospitalizations.
  5. Plan leavers (3-8%). Members who value their current medication enough to switch health plans at open enrollment. This is the most financially significant behavioral category for the plan, because losing a healthy member over a formulary change trades a small pharmacy savings for the loss of that member's entire premium.

Variables That Predict Behavior

Not all formulary changes produce the same behavioral response. The variables that most strongly predict member behavior include:

Building a Stickiness Model

A practical stickiness model uses historical claims data and enrollment data to predict the behavioral distribution for a proposed formulary change. The core data requirements are:

The most valuable insight from stickiness modeling is not the point estimate of member behavior. It is the sensitivity analysis. How does the behavioral distribution change if we communicate the change 90 days in advance instead of 30? What if we offer a grandfathering period for members on the drug for more than two years? These policy levers can dramatically shift the cost-benefit calculus of a formulary change.

Financial Impact Integration

The stickiness model becomes most powerful when integrated with the financial impact model. A formulary change that saves $2M in pharmacy cost but causes 500 members to leave the plan (at an average lifetime value of $15,000 per member) is a net negative. Without the stickiness model, the P&T committee sees only the $2M savings. With it, they see the full picture.

This is the kind of analytical capability that separates sophisticated formulary operations from manual ones. The data exists in every health plan's claims and enrollment systems. The modeling techniques are well-established. The gap is in building systems that connect these data sources and present the integrated analysis to decision-makers at the point of decision. That is where technology needs to go.