Here is a number I cannot get out of my head. More than $500B is lost around the world every year, not to drugs that do not work, but to patients who stop taking ones that do. It is one of the largest and least discussed sources of waste in healthcare, and almost none of it shows up as a line item anyone owns.
Fewer than 1 in 3 patients is still on a novel therapy at 12 months. Some never fill the prescription at all: fewer than half of novel scripts written are ever filled. The rest start with real intent, and then over the months that follow, they drift away, one small decision at a time.
I think of it as a leaky bucket. Our industry spends a decade and a fortune proving a therapy works: the trials, the approval, the launch. Then we pour all of that hard-won value into a bucket that begins leaking the moment the patient leaves the exam room. We work hard to fill it faster, and rarely ask why it will not hold water.
A story that repeats
Picture a patient who starts a new injectable for a chronic condition. Week one goes well. At week three she has a reaction that frightens her, nothing dangerous, but no one told her it might happen. She has a question on a Sunday night and no obvious place to ask it. She skips the next dose to see whether she feels better. By the time her refill is due she has half-decided she is done, and the missed refill, weeks later, is the first moment any system notices.
Nothing in that story is unusual, and no one in it failed at their job. The Hub enrolled her. The rep did the visit. The app sent its reminders. Each part worked, and the patient still slipped through the space between them. That space, between the handoffs, is where most of the leak lives.
Why we ended up here
It is not for lack of effort or care. It is that almost every system we built points at the start. Enrollment, activation, first fill: these are clean, countable moments with dates attached, and teams have gotten genuinely good at moving them. They are also where budgets and incentives concentrate.
Staying on therapy is different. It happens over months, at home, in a hundred small private decisions no dashboard was built to see. So we measure and optimize what we can count and hope the rest follows. It usually does not, and the hoping is expensive.
Where the value actually leaks
When you sit with real patients, the drop-off is rarely dramatic. A side effect at week three. A copay that doubled at renewal. A prior authorization that lapsed through no fault of the patient. A slow erosion of belief that the therapy is worth the hassle. Each one is survivable with a timely answer, and each one is invisible to a program watching only refill claims.
By the time a missed refill appears in a report, the decision was usually made weeks earlier. Refill data is an accurate record of when a patient left. It almost never tells you why, and it never tells you in time to help. The leak is not one moment you can staff around. It is the whole stretch after the start.
What it takes to close it
Closing the bucket is not a caring problem or an effort problem. It is a design and data problem, and those are solvable. It takes three things working together: a unified view of the patient's experience so the pieces stop talking past each other, the ability to predict who is drifting from the behavioral signals that come before the missed dose, and the means to intervene at the moment that changes the decision.
That is the work we do at BrightInsight, and it is why we lead with proof rather than promises. In a rigorous real-world analysis by a global biopharma partner, 13% more patients stayed on therapy at 12 months than in traditional patient support programs, a 7 percentage-point lift that widens further at 24 months. That kind of gain compounds across a population, quarter after quarter, where it is worth millions in reclaimed value in the first year alone.
The leaky bucket has been treated as a cost of doing business for as long as I have worked in this field. It does not have to be. The therapies are ready. The least we owe patients, and the brands that spent years developing them, is a system that helps people stay.
Who should own it
One reason the leak persists is that no single function owns it. Acquisition has a team and a budget. Persistence lives in the seams between commercial, medical and patient services, which is a polite way of saying it is no one's job. Value that is no one's job does not get protected.
The brands that pull ahead give persistence an owner, a number and a method, the way every other industry treats retention. That is a decision a leadership team can make in an afternoon, and it changes what the whole program optimizes for.
If you run a patient program, the leak is costing more than any single line on your dashboard shows. That is the problem we exist to solve, and we are glad to walk through what closing it looks like for your brand.
1. Shaw, Gina. “Half of Novel Specialty Prescriptions Go Unfilled, IQVIA Data Show.” Specialty Pharmacy Continuum, April 10, 2026.
2. HealthPrize Technologies and Capgemini. Pharmaceutical Companies Lose $637 Billion in Revenue Annually Due to Medication Nonadherence. Norwalk, Conn.: PRNewswire; November 16, 2016. (Reported here conservatively as more than $500B.)
3. IQVIA Institute for Human Data Science. Understanding the Use of Medicines in the U.S. 2025: Evolving Standards of Care, Patient Access, and Spending. April 2025.
Persistence result: real-world analysis by a global biopharma partner (12-month observation, three independent cohorts, six indications, 6,000+ patients analyzed).