The pattern repeats across telehealth platforms and multi-site clinics. A physician performs well.

Patient satisfaction stays high. Revenue per provider exceeds targets.

Then they leave.

Most operators assume compensation.

They increase base salary or restructure RVU splits.

The physician still leaves. Others assume burnout, so they reduce patient volume or add support staff. The physician still leaves.

The real gap surfaces when peptides and compounds enter the treatment mix.

Employed physician models were built for diagnosis and prescription.

The physician sees the patient, orders the test, writes the script.

The system handles the rest. That worked when "the rest" meant a pharmacy filling Metformin or Lisinopril.

Modern therapies broke that model quietly.

When a patient needs Semaglutide, the physician doesn't just prescribe it.

They coordinate timing with the compounding pharmacy.

They adjust dosing based on fulfillment delays. They switch patients between peptides when one goes on backorder.

They manage patient expectations when shipping takes two weeks instead of three days.

None of this shows up in the EMR as clinical work. All of it determines whether the patient stays on therapy.

Employed physicians inside high-performing clinics don't leave because the clinic failed.

They leave because the clinic succeeded at the wrong things.

The infrastructure optimized for volume and reimbursement. It never optimized for supply chain coordination.

Independent practices aren't better at medicine.

They're better at operational control.

The physician who leaves can choose their compounding partner, switch vendors when capacity tightens, and adjust patient flow based on API availability.

They're not more autonomous clinically. They're more autonomous operationally.

The clinics that retain physicians don't fix this with culture or compensation.

They fix it by changing what the physician controls.

That means giving physicians visibility into pharmacy inventory, letting them direct fulfillment timing, and allowing them to adjust patient sequencing based on compound availability.

Most clinics resist this because it feels inefficient. It is inefficient if the goal is maximizing patient volume.

It's necessary if the goal is keeping patients on therapy.

The physicians who stay aren't the ones who care less about outcomes. They're the ones working inside infrastructure that lets them manage outcomes across the entire treatment cycle, not just the consultation.

Retention problems that look like compensation or burnout are usually infrastructure problems.

The system asks physicians to deliver outcomes it wasn't built to support.

Clinics scaling peptides don't collapse from FDA restrictions or state board enforcement. They collapse from pharmacy capacity that worked at 50 patients per month and failed at 500.

The failure happens quietly.

A clinic builds a relationship with a 503A compounding pharmacy. The pharmacy fulfills orders consistently. Lead times stay predictable. The clinic assumes the relationship will scale with volume.

Then the clinic hits capacity. Not clinic capacity. Pharmacy capacity.

The pharmacy starts missing fulfillment windows. Semaglutide ships in 10 days instead of 3. Tirzepatide goes on backorder for two weeks. The pharmacy explains it as a temporary API sourcing issue. The clinic believes it because the pharmacy delivered reliably for months.

The issue isn't temporary. The issue is structural.

Most 503A pharmacies were built to serve a base of prescribers with predictable, distributed volume. A hundred physicians each sending five orders per month creates manageable flow. One telehealth platform sending 500 orders per month creates fulfillment congestion the pharmacy's infrastructure can't absorb.

The pharmacy doesn't refuse the volume. They accept it because they want the relationship. But their sourcing, compounding workflow, and shipping operations weren't designed for velocity. They're designed for variety.

When the pharmacy falls behind, the clinic discovers they never controlled their supply chain.

They had a vendor relationship. Vendor relationships fail under load.

The clinics that survive their own growth don't solve this by finding a bigger pharmacy. They solve it by building redundancy before they need it. That means relationships with multiple 503A pharmacies, direct visibility into API sourcing, and fulfillment models that don't depend on a single partner hitting every window.

This isn't about pharmacy quality. High-quality compounding pharmacies still have capacity limits. Those limits surface fastest with the peptides that drive clinic growth.

Most operators discover the pharmacy problem after patients start churning.

The patients don't leave because the therapy stopped working. They leave because the therapy stopped arriving.

Scaling modern therapies doesn't fail from regulatory risk. It fails from supply chain infrastructure that was never built for the velocity required to make the unit economics work.

The clinics that grow past 500 patients per month aren't the ones with better pharmacy partners.

They're the ones who stopped treating pharmacy relationships as partnerships and started treating them as infrastructure dependencies that need redundancy.

I've spent the past decade building and scaling operations across clinics, telehealth platforms, and compounding pharmacies. I've seen practices succeed at 20 patients per month and break at 200. I've watched regulatory shifts reorganize entire markets overnight. I've built the infrastructure that determines whether modern therapies scale or stall.

This newsletter is about understanding how the system actually functions, where friction hides, and why some approaches compound over time while others don't.

If you're building, running, or participating in modern health in any real way, this layer matters more than most people realize.