If you buy a car and it turns out to be a dud – one that spends more time at the dealership than in your driveway – every state in the country has a law that lets you hand the keys back and get your money returned. We call them lemon laws. They’ve been on the books for forty years. The logic is simple: you paid for something to work, it didn’t, so you don’t have to pay for it.
Now consider what happens in pharmacy.
A plan approves a $180,000-a-year biologic for a patient with rheumatoid arthritis. The Prior Authorization is rigorous – labs, chart notes, the whole bit. The patient starts therapy and religiously refills their prescription. The plan pays. Twelve months…Two years…Still paying…Three years…Still paying.
Did the medication work? Did it achieve some semblance of a clinical endpoint? Did it demonstrate even a shred of clinical efficacy? Nobody asked. Nobody has to ask. The “check the box” PA was satisfied at month one, and nothing in the system requires anyone to confirm the drug is actually doing what it was supposed to do.
To be fair, it’s not that the drug is “never” rechecked. There is usually a recertification at some interval. But for the most part what recertification usually asks: “Is the medication still appropriate for this patient?” Of course it is. It was appropriate enough to prescribe in the first place, and nothing has changed about the diagnosis. The prescriber confirms what was already confirmed, the box gets checked again, and the refills continue. Nowhere in that process does anyone ask the only question that matters: is the drug actually working?
Overly complex guidelines that are forced to be distilled into overly simple questions sets, that are asked over and over. That’s not formulary management.
At least with a lemon car you know if it doesn’t start in the morning. In our current health care system there is no rigorous process in place that assures patients using biologics – drugs with subtle clinical endpoints, that achieve modest and spotty clinical efficacy – are getting some sense of clinical value.
The Quiet Math of “Just Refill It”
Specialty drugs (man how I hate that moniker, but that’s a blog for another day) now account for over half of total pharmacy spend despite being prescribed for a smallish fraction of patients. The fastest-growing line item on most plan P&Ls isn’t a mystery – it’s the autopilot refill of expensive medications, year after year, with no clinical accountability tied to whether the patient is actually getting better.
The dirty secret is that the clinical response rates achieved in practice are significantly less robust than what the clinical trials (and associated marketing) might imply.
Many new therapies are approved for use compared to placebo in controlled trials – not compared to usual care. Take the case of the new CFTR Modulators for cystic fibrosis, approved in clinical trials that evaluated these agents against placebo using the rationale that these drugs have a unique mechanism of action and that a placebo-controlled trial is the only way to isolate safety concerns. In these studies, baseline ppFEV1 was ~60%, and patient’s using the CFTR’s achieved a 3-4% improvement in ppFEV1 (typically considered NOT to be an improvement that is perceptible to patients). Now some might argue that the real benefit of these CFTR agents is lung function preservation – the problem is that the trials lasted ~24 weeks (not enough time to assess function preservation). The trials also demonstrated “impressive” reductions in pulmonary exacerbations and hospitalizations – but, not to beat a dead horse, against PLACEBO. So now payers are obligated to approve these drugs (at ~$30K per month) not having any idea of the relative value over simple mucolytics (the current standard of care) – if in fact there are any!
The potential for low value in newly approved drugs is not limited to newfangled genetic mechanisms of actions. Now payors have to contend with low value from drugs that have been resurrected as new dosage forms. Recently approved extended-release amantadine hits the market at ~$1600 per month (it better come with wings and a beer). Also approved in a placebo-controlled trial using troublesome dyskinesia (apparently there are neurologists out there that think there is pleasant dyskinesia). In clinical trials ER amantadine showed an improvement of 2.9 hours without dyskinesia OVER PLACEBO. The best comparative information available are pharmacokinetic studies that show better AUC with ER…no actual data of benefit over prompt amantadine (~$30 per month).
Where’s the lemon law that says if ER amantadine improves my dyskinesia no better that prompt amantadine, I get my $1600 back?
Why Legacy PA Can’t Solve This
Legacy PA – and even AI-enabled “faster PA” – is a gate at the front door. It asks, “Did the prescriber check the right boxes to start therapy?” It does not ask, “Is the patient actually improving on this medication?” And it certainly doesn’t ask that question at month six, or year two, or year four.
AI is making the front-door problem worse, not better. Most AI PA tools are optimized to approve faster, not approve smarter. They speed up the rubber stamp. More specialty meds get approved, faster, with less friction – and once they’re in the door, the same autopilot kicks in.
We are getting more efficient at saying yes to drugs we never check on again.
What a Lemon Law for Pharmacy Actually Looks Like
You don’t need new legislation to fix this. You need to go back to the original intent of PA – assure a triage of therapy that gets the patient to the most efficacious agent at the lowest cost.
Value-based drug contracting – real value-based contracting, not the PR version – means the plan only pays for a medication if the patient is demonstrably improving against the clinical metrics that matter for that disease state. DAS28 for rheumatoid arthritis. HbA1c for diabetes. PASI for psoriasis. eGFR for kidney disease. The metrics already exist. Prescribers already know what good looks like and monitor (or should be monitoring) these endpoints. The data is already being generated.
What’s missing is the platform to collect those metrics on a continuous basis and tie recertification of an expensive medication to whether it is, in fact, working.
When you build that platform, three things happen:
The patient who is responding stays on therapy without interruption. Good outcome, good story, no change.
The patient who is not responding either gets moved – quickly – to a medication that has a better chance of working for them or to their previous therapy that was just as effective and lower cost. Better outcome for the patient. They are no longer trapped on an expensive drug that isn’t doing its job just because the original PA went through.
And payors stop wasting premium dollars on lemons. Not because someone decided to be stingy, but based on clinical reality of efficacy. The drug had its chance. It didn’t deliver. The plan returns the keys.
Value Base Pricing – Pharmacy’s Lemon Law
The auto industry figured out forty years ago that if a product doesn’t work, the buyer shouldn’t be the one stuck with the bill. Pharmacy benefit management is woefully overdue for the same type of accountability – the accountability that value-based pricing brings.
The clinical metrics to drive value-based pricing exist. The contracting frameworks exists. What’s been missing is a platform that ties them together – a platform that authorizes and reauthorizes therapy using readily available progress notes and lab reports that are already being collected by legacy prior authorization, but are being used ineffectively.
Pay for performance. Stop paying for lemons.
If you’re a plan or PBM looking to put clinical teeth into your prior authorization, we’d welcome the conversation.
PS. If your current PA vendor tells you they already do this, ask them to show you their recertification denial rate on patients who failed to hit clinical response. The silence will be informative.
Jim Notaro, RPh, PhD is a pharmacist and seasoned managed care professional with extensive experience in pharmacy benefits management. He has worked across multiple plans overseeing pharmacy benefits. In 2000 he founded CSS Health, a provider of medication therapy management and quality metric improvement software and services, eventually selling it to Clarest Health.