Stark Law Explained
31 year-old anti-self referral rules that raise questions for digital health and value-based care
Waiting Room is a blog that explains incentive misalignments in healthcare and health tech. Today, we’re covering the Stark Law.
The 1989 Ethics in Patient Referrals Act seemed simple enough. Congressman Pete Stark found that physicians who owned labs had a higher rate of referring patients to tests than those that did not. Surprise, surprise. So Stark set out to eliminate the financial incentives of unnecessary tests and procedures by prohibiting physician self-referral. The most recognized example of Stark is against physicians sending patients to an entity with which they have a compensation arrangement or ownership interest.
Let’s illustrate this. You’re a primary care doctor in central Florida (the retirement haven where seniors have high rates of STDs and are apparently not shy about having sex… anyways…). Because you treat seniors, a number of your patients experience knee pain associated with arthritis. Dr. Jones is an orthopedist - a specialist in skeletal disorders - at a different practice and you often refer seniors to her.
Stark violation: You receive a $1K kickback from Dr. Jones for every patient sent to her practice.
Stark compliant: You refer seniors to Dr. Jones because she’s in network for most insurance companies and you think she provides great care.
The first scenario is obviously egregious. The second scenario is worth highlighting because physicians build relationships with each other and prioritize referring to specialists that are responsive, have availability, provide high quality care, etc. Dr. Jones may generate increased revenue because of your referrals but there’s nothing illegal about being a good doctor. Patients always have the right to choose their specialist but are strongly swayed by physician recommendations, even if it means picking more expensive care.
Stark prohibits many other practices (with big penalties):
Miscoding healthcare claims to obtain greater reimbursement - Adventist Health System fined $115M
Billing insurance (particularly Medicare and Medicaid) for services that did not occur - Tuomey Healthcare System fined $237M
Referring patients for medically unnecessary procedures - St. Joseph Medical Center fined $22M
Paying physician salaries and bonuses far above fair market value - The Christ Hospital fined $108M (this may be an example of using the Lord’s name in vain)
Goes without saying but don’t be that guy ⬇️
Now an increasing number of organizations want to modify and even repeal Stark in the face of newly adopted technology and payment models. Some reasons/notable areas of impact:
Telehealth companies like Doctor on Demand and Hims/Hers employ physicians (directly or indirectly) to conduct appoints. Those physicians have to be compensated at fair market value for identifiable services - and the contracts cannot take into account the volume or value of referrals. In legalese, the physician’s employment must be “commercially reasonable” even if no referrals are ever made.
But what about startups that give physicians equity? If physicians know that the value of that equity is tied to revenue generated from more appointments, lab tests, prescriptions, durable medical equipment (DME) authorizations... the incentives to increase unnecessary healthcare spend may violate Stark. Telehealth companies often outsource appointments to third-party clinicians as a way to protect against this (see: Wheel).
I’m not poo-pooing telemedicine or it’s incredible expansion of healthcare access. Many brands have nailed consumer marketing and attract thousands of customers that need care. We have yet to see CMS or commercial insurers come down on any telehealth companies for defying Stark but I think quite a few tow a fine line.
Value-Based Care (VBC)
Payers like Medicare or a commercial insurer traditionally reimburse physicians per service provided. A new type of payment scheme, value-based care, ties reimbursement to outcomes rather than volume. Chronic kidney care is a common VBC model: instead of going directly to a nephrologist or jumping into an expensive kidney transplant program, a patient navigates between a primary care physician, nutritionist, and a social worker who coordinate care together.
The department of Health and Human Services has released comments in the past on Stark impeding VBC; insurers can’t easily incentivize / reward physicians who reduce costs, coordinate care, achieve quality standards, and use remote monitoring tools for earlier intervention. Physicians can apply for fraud and abuse waivers but new CMS proposals may waive Stark entirely for VBC arrangements.
Group Purchasing Organizations (GPOs)
Healthcare has a systemic supply chain issue. We saw that fragility in full form when hospitals battled each other for ventilators to treat Covid-19 patients.
GPOs reduce the supply chain problem by consolidating purchasing power: steered by a committee of clinicians, GPOs negotiate discounts and make purchases on behalf of member hospitals and healthcare facilities. One of the largest GPOs, Vizient, represents almost $100B in annual purchasing volume. GPOs buy ERP software, property insurance, employee benefits, IT, staffing services, pharmaceuticals, equipment, etc.
There are GPOs owned by the same parent company that owns some of the GPO’s member hospitals. Stark does limit GPOs from selling units to members at more than 3% of the purchase price. But you guessed it. There are quite a few opportunities for self-dealing and biased vendor selection…
I’d love to hear ideas for making telehealth, VBCs, and GPOs “Stark-proof.” I assume operating in good faith is the best way to comply and avoid any semblance of kickbacks. But is Stark even the right anti-fraud mechanism? Ideally, the government and the public would be able to see a map of healthcare ownership. New disclosure / transparency requirements could have a positive freezing effect. United/Optum, CVS/Aetna, large health systems, and PE firms own so many subsidiaries, holding companies, and parent companies that it’s possible self-dealing happens more than we know. CMS can only catch so much.
If you liked this post from Waiting Room, subscribe for more.