It’s commonly known that US healthcare costs 18% of the country’s GDP, or $3 trillion a year. Overall health outcomes are mediocre compared to other developed countries, which generally spend half of that amount per person.
Despite the high amounts we spend on healthcare, it’s never quite clear what we’re paying for, because the billing practices are incomprehensible. It’s especially bad when compared to other things we’re used to paying for.
And it’s not clear why things cost so much. Why does getting a few stitches in an ER cost $5,000? Why does a pill of Tylenol in the hospital cost $55?
In summary, US healthcare is a dysfunctional market. The normal economic laws of supply and demand, and of competition and pricing, are not operating here.
The author presents ten rules of this distorted market:
The major blocs in healthcare are:
Each bloc has opposing interests:
No matter how the incentives are set, participants in healthcare will generally exploit the incentives to their maximum. And if the incentives don’t align with quality and cost-effectiveness, you don’t get quality and cost-effectiveness.
This profit-seeking behavior manifests across all stakeholders:
The maximization of incentives is true of any market, but several factors make healthcare an especially dysfunctional market.
By now, all the major blocs are so deeply entrenched that any zero-sum change is strongly opposed. For example, lowering compensation to doctors might mean cheaper costs for the system, but doctors will staunchly oppose this.
Fragmentation makes it difficult to change anything systemically.
Even if you can’t change the healthcare system yourself, what can you do to reduce your own price of care? The book recommends these actionables:
Question your doctor and hospital about fees before you get treatment.
Wait before getting treatment.
If admitted to the hospital:
Dealing with bills
Look at all the costs of your health insurance.
Lower your costs of drugs.
It’s commonly known that US healthcare costs 18% of the country’s GDP, or $3 trillion a year. Overall health outcomes are mediocre compared to other developed countries, which generally spend half of that amount per person.
Despite the high amounts we spend on healthcare, it’s never quite clear what we’re paying for, because the billing practices are incomprehensible. It’s especially bad when compared to other things we’re used to paying for.
For example, in healthcare, multiple doctors send separate bills with huge amounts, then insurance pays a fraction of the amounts. What if you took a flight and got separate bills from the airline, the pilot, and the flight attendants?
Further, the price for the same procedure costs different amounts depending on where it’s done, who’s doing it, and what insurer you have. What if you paid twice as much for a Toyota Camry in New Jersey as for one in California?
And it’s not clear why things cost so much. Why does getting a few stitches in an ER cost $5,000?
This book is an attempt to answer these questions. Over time, the healthcare market has become dysfunctional. The author presents ten rules of the distorted US healthcare market:
An American Sickness is a fantastic book on US healthcare, but with some weaknesses.
Primarily, the book omits blaming patients for contributing to rising healthcare costs, instead redirecting ire to other participants in healthcare. For example, Rosenthal doesn’t blame patients for wanting the best-in-class care and newest technology, regardless of its cost-effectiveness. She advises patients to be cautious of insurance clauses that “require you to try what the plan considers to be a more ‘cost-effective’ drug.” The scare quotes are unhelpful in getting patients to substitute drugs for generics, which is what she advises in the first place!
She also ignores that 70% of healthcare costs are a direct result of patient behavior, with a majority of cost in cardiovascular disease, diabetes, and obesity being preventable. Immense healthcare savings could be had in keeping to normal weight and adhering to drug regimens. We’ll point out ways that patients contribute to higher healthcare costs with Shortform notes.
Later, the author discusses a panel of solutions for rising healthcare costs, but she doesn’t quantify the effects. For instance, how much would having a single national payer save? How much would allow drug importation save? It’s not clear what the lowest hanging fruits are.
The first and major part of An American Sickness covers the major segments and industries of healthcare. Each chapter contains a history of the industry and how well-meaning policies turned into current perverse incentives.
In the late 1800s, healthcare was unscientific and ineffective. Diseases took a long time to recover from, and people paid for their own healthcare. Health insurance as we know it today didn’t really exist.
The earliest health insurance policies compensated people for income lost while they were sick. Some employers also paid for doctors to be on retainer to care for employees, since long illness absences were a problem.
In the 1920s, Baylor University Medical Center offered a local teachers’ union a catastrophic health plan for $6 per year per person. This included a 21-day stay in the hospital after a deductible of a week. A day in the hospital cost just $5/day, or $105 in today’s dollars. This plan became popular, signing 3 million insured by 1939, and led to the non-profit Blue Cross Plans.
In the 1930s, medical technology improved. Anesthesia, penicillin, and ventilators were discovered. This enabled new effective standards of care. The new technology also increased the cost of care, and insurance had to adjust to cover the higher costs.
At this critical juncture, insurance could have been direct to consumer and private, as with auto and life insurance. But one historical event changed healthcare policy in a dramatic way that would last until today. During World War II, the National War Labor Board froze salaries. Normally, companies used higher salaries to compete for workers, but this was now forbidden. Instead, companies began offering health insurance, which wasn’t frozen. Furthermore, to make this more attractive, the federal government made employer spend on health benefits tax-deductible.
This historical artifact was the origination of employer-sponsored healthcare, the predominant way people are covered today. Population insurance rates exploded from 10% in 1940 to 60% in 1955.
Growth in the insurance industry and public demand for health insurance from employers prompted for-profit insurers like Aetna and Cigna to enter the industry.
The author suggests that for-profit insurers were less bound by the mission statement of non-profit BCBS of “high quality, affordable health care for all.” For-profit insurers could segment the population, focusing on healthier patients and offering lower rates while excluding unhealthy patients. This meant the Blue Cross plans began having to support sicker patients.
In 1994, facing financial difficulty, the Blues’ board allowed member plans to become for-profit. The immediate intention was to raise funds in the stock market to stay afloat. Over time, the plans consolidated and grew, becoming today’s giants like Wellpoint and Anthem BCBS.
One metric of how insurance companies perform is the “medical loss ratio.” This is the percentage of premiums revenue that is spent on medical care.
In 1993, the non-profit Blues spent 95% of premium revenue on medical care, and only 5% on admin, marketing, and dividends.
Today, the standard is closer to 80%. This number is also mandated by Obama’s Affordable Care Act (ACA). But before the ACA, the ratio of some insurers was as low as 64%. As a point of comparison, Medicare is at 98%.
Shortform note: note that the effect of mandating a high medical loss ratio is not straightforward—it can prompt more haphazard spending!
Say an insurer before ACA spends 60% on healthcare, paying the rest for overhead and dividends. ACA passes and now it has to spend 80% of premiums on healthcare. The insurer now has a few (simplified) options:
Keep revenue and healthcare costs the same, and cut admin/dividends until healthcare reaches 80%. (This is closer to what the ACA intended.)
Keep revenue the same, and shift spending that would have gone to administration into paying for healthcare. This would lead to open spending, like departments having to exhaust their budgets at the end of the fiscal year, and less constraint on costs.
Increase premium costs and revenue, and spend the surplus on extra medical care to get to 80%, while keeping administrative costs the same.
The point is that simply mandating that insurance companies spend a percentage of revenue on medical care does not automatically lower healthcare costs.
One would think that insurers have strong incentives to negotiate down prices. After all, if they can lower prices, they can lower premiums, which would recruit more patients.
However, there are strong incentives in play to counteract this:
The incentives are therefore aligned with higher cost care. Because of the 80% medical loss ratio, insurers don’t mind higher payments to vendors. If total healthcare spend grows, so does the 20% of revenues they can use for administration, salaries, and profits. Therefore, insurers benefit when total healthcare costs increase.
Once created to safeguard patients, health insurance eventually became part of the entrenched healthcare ecosystem. Hospitals adapted to their financial incentives, which changed how doctors practice medicine, which changed the types of drugs and devices manufacturers made. We’ll cover all of this in subsequent chapters.
Hospital costs have grown faster than other segments of healthcare, growing 149% from 1997 to 2012, compared to 55% for physician services.
Today, 10-15% of hospital revenue goes to billing administrators and claims processing. In today’s system, it costs a lot to get paid.
Many hospitals began with religious roots (hence the many hospitals with the name Baptist or Presbyterian). They had general social good as their mission.
In the 20th century, health insurance coverage broadened. In the 1960s, Medicare arrived and covered hospital payments. In 1980, 80% of Americans under 65 were covered by insurance.
At this time, reimbursement was generally fee-for-service. Providers charged as much as they could, and insurers generally paid it out. From 1967 to 1983, Medicare payments to hospitals increased from $3 to $37 billion.
With more money rolling in, hospitals hired administrators, who helped steer the organization toward financial performance. Physicians were influenced to focus on more profitable care, told what procedures to perform, given bonuses scaling with revenue they brought in, compared publicly to other doctors on revenue, and even told to attend charm school to make more revenue. Likewise, nurses became “clinical nurse-managers” armed with statistics on billing.
Healthcare costs were getting out of control, so to clamp down on charges, Medicare revised payments to a “diagnosis-related group”—essentially, a fixed bundled amount based on the diagnosis. Medicare would pay a fixed amount for an appendix removal.
Similarly, HMOs had a heyday in the 1990s, where the PCP would be a gatekeeper for followup care and provider access was limited. The author says that patients hated them because they were poorly designed and managed. (Shortform note: however, patients also disliked being unable to get access to any doctor they wanted on demand.)
Hospitals provide the medical service to the patient, then are reimbursed by the payers (insurers) for the service. This practice has led to a complicated set of billing practices and gamesmanship to maximize hospital earnings. Hospitals employ reimbursement consultants to adjust their prices and billing practices to earn more money.
The high sticker price given by hospitals is a negotiating point with payers (insurers). Bigger payers pay a smaller fraction of the list price than smaller payers and the uninsured.
Medicare assigns to every hospital an overall cost-to-charge ratio it considers reasonable. This is meant to constrain the profit percentage that hospitals can make. However, this regulation turned billing into a strategic game. For example, hospitals adjusted their prices to maximize billing. They lowered charges for items that are often not reimbursed (like gauze) and boosted charges for what is reimbursed (OR time, oxygen therapy).
This led to a perplexing practice where a single procedure, like an overnight stay in the hospital, can be billed separately as a wide array of items. This is why a hospital bill can include dozens of charges, including individual Tylenol pills and charges for each separate doctor’s time.
Billing is done aggressively, to the limit of what is acceptable for the service provided. This is called “upcoding.” For example, a simple blood draw can be classified as a level 5 doctor visit. Optional services can be added to increase billing, without necessary patient benefit—using ultrasound to inject steroids in a knee adds $300.
Providers and vendors lobby to get new conditions treated as diseases, since this legitimizes billing for the disease. For instance, if obesity is recognized as a disease, it gets assigned billing codes, which pressures insurers to cover it. (Shortform note: it’s that our conception of health and disease, which should be fairly objective, are molded by the providers and their financial incentives.)
When new treatments appear, there is often a frenzy to promote it and set the billing standards. Because the finances of the new treatment are unpredictable and patient numbers are small, the reimbursement policy may start off being lax; this may let the cat out of the bag and make it difficult to constrain costs later.
Billing can cause changes in how hospitals practice care.
One example: admitting patients for “observation” means admitting them for monitoring before deciding on a treatment. Due to quirks in billing, hospitals get paid more for patients under observation. Essentially, Medicare pays bundled rates for inpatient admissions, but not for outpatient care, which observation fits under. Under observation, hospitals can charge per service as they like.
This has led to situations where patients were admitted under “observation status,” underwent a battery of tests, then were charged far more than if they had merely been admitted as an inpatient.
As hospitals try to maximize how much they get paid, this filters down into how they manage their departments and employees.
Hospitals are increasingly getting rid of unprofitable departments (such as the emergency room, labor and delivery, dialysis, drug treatment, and Medicaid outpatient care) and focusing on profitable offerings (such as orthopedics, cardiac care, stroke center, and cancer care).
They even make money on outsourcing care—hospitals can sell patients to dialysis centers for $40,000-$70,000 per patient.
Physicians are often compensated in proportion to the relative value units (RVUs) they bill, which is based on the complexity of exams and treatment plans. This sets an incentive for physicians to bill as much as they can.
The EMR software that doctors use even give tips on checking the right boxes for upcoding (“you need to check two more boxes for a level 4 visit”).
Hospitals also allow providers to add an outpatient facility fee on top of their billing. Historically, insurers allowed this as an analogue of the inpatient daily rate. However, this has become a strong incentive for doctors to perform procedures in a surgicenter, rather than in their less expensive office.
Medical graduate stipends are paid by federal and state funds. Hospitals get $15 billion to support training, including “indirect payments” to compensate for losses of efficiency in training new doctors. However, the author argues there is little evidence training hospitals are any less efficient than non-teaching hospitals, so these indirect payments are just free money.
The author also argues that hospitals profit from cheap resident labor. The median cost for a resident in 2013 was $135k, including a salary of $50k-80k. Of this total cost, federal support totals $100k, covering most of the cost. However, the value of work done by a resident is $233k, far greater than what the hospital pays the resident.
Hospitals enjoy this cheap labor and lobby to get more residents. The American Hospital Association has lobbied to get foreign medical graduates, especially for less lucrative specialties like nephrology. They also need to respond to regulations that capped the number of hours residents could work, since this creates a relative staffing shortage.
Most hospitals are non-profit. Instead of distributing their profits to shareholders, they spend the operating surplus on expansion, amenities, and executive compensation.
One area hospitals spend a lot of attention on is customer satisfaction. Press Ganey is a survey tool that allows patients to rate their experience at hospitals. They also publish rankings. To try to improve their ratings, hospitals have spent a lot of money on amenities and patient comfort. In the past, hospital rooms with four beds were common; nowadays, single private hospital rooms have become the norm, even though there is little proof this is more cost-effective and many insurers won’t cover them (leaving the patient with the bill).
A focus on customer satisfaction may not lead to better care. If a patient asks for a test, giving it to them increases satisfaction and avoids the trouble of explaining why they don’t need it. Ultimately, the medical system, and we as patients, pay for this.
The author bemoans that hospitals spend so much on new facilities and amenities without increasing the salaries of doctors and medical staff.
Nonprofit hospitals are required to provide “charity care and community benefit” to keep their tax-advantaged status. However, as with much of the healthcare ecosystem, this can be distorted to the hospitals’ advantage.
The ACA requires the IRS to collect each hospital’s quantitative enumeration of charitable activities and their value. Recent research suggests hospitals spend only a fraction of the tax advantage surplus on charity care—a set of 200 hospitals receive $3.3 billion in tax exemptions, but they only spend $1.4 billion on charity care.
This legal status can even open the door for political negotiations. In San Francisco, hospitals planning new construction are required to demonstrate charity care commensurate with size. For California Pacific Medical Center, Sutter Health agreed to create a $20 million Healthcare Innovation Fund, $60 million for affordable housing programs, and an agreement to not raise rates for insurers covering city employees by more than 5% annually.
Physicians take care of patients, but they’re no less concerned with boosting their own pay as the other members of the ecosystem. As we’ll learn, doctors bill in ways to maximize their pay.
Primary care physicians in the United States make 40% more than Germans; orthopedic surgeons make 100% more.
Part of the problem may be medical school debt. In the United States, medical school costs between $120k to $220k (with state schools at the lower range and private schools at the higher range), while it’s free or cheap in many other countries. Medical students graduate with a mean debt of $170k, some of it from undergrad.
The author argues that this debt burden pushes some students into more lucrative specialties, like dermatology and ophthalmology, rather than what they would naturally prefer to practice.
But all this pay may not be enough. One medical student comments that doctors feel a “bizarre martyr complex” where they feel they’re working harder for less money than the rest of America. The author argues this is a symptom of the corporatization of healthcare—doctors are getting less satisfaction from patient care, so they’re looking for dollars to compensate for it.
Before the 1950s, American patients were uninsured, paid doctors out of pocket, and were on an informal sliding scale in proportion to income. Doctors were comfortably middle class, but not wealthy.
In the middle of the 20th century, World War II and the ensuing employer-sponsored insurance, and Medicare Part B (with its essentially unconstrained reimbursements) led to large physician payments and a golden era of compensation. Why did Medicare pay so liberally? Since the medical profession was opposed to Medicare, this was part of a handshake deal to get Medicare passed—the government vowed not to interfere with the practice of medicine, if doctors could avoid opposing Medicare.
The reimbursement structure was based on retrospective “usual and customary” fees. “Usual” was defined as the average price of local providers in the preceding period, and “customary” was defined as the % of the bill that was typically reimbursed (75-90%). The key point is that doctors had every incentive to raise their prices—if they raised their prices today, then they would get reimbursed for more tomorrow, since they had raised the average price of the procedure. During this time, costs exploded, and insurers paid liberally.
This was starting to be a clear problem, so in 1992, Congress and the American Medical Association created the resource-based relative value scale (RBRVS) using relative value units (RVUs). This was an attempt to standardize the amounts paid for medical procedures. The calculation included factors such as:
Furthermore, Medicare set a legal cap for payments to physicians. If a highly valued procedure was approved, other costs had to decrease. If total payments went up, the conversion factor had to decrease.
This system is important and widely used. Private insurers often peg what they’re willing to pay to Medicare’s RBRVS.
Compared to the “usual and customary” pricing, the intent of RVUs was to limit variation in medical prices and limit the growth of cost. However, this system incentivized procedures with more doctor time per intervention and more training required. If you’re wondering why there’s so much emphasis on specialty care in the United States, this is a big reason why. Expensive procedures like cardiac interventions and surgery are rewarded; less quantifiable practices, like complex neurology diagnoses, were punished.
(Shortform note: notably, RVUs are calibrated to effort, not to quality or outcomes. If a procedure costs more to provide, hospitals and doctors are paid more for it, even if there is a cheaper alternative that works better for the patient.)
According to the author, the median doctor income has risen over the past decade, even as overall real income nationwide has dropped.
Because medical care constantly changes, the Medicare RVU system requires constant updating. Medicare assigned the responsibility of setting RVUs to the AMA. The author considers this as self-defeating as letting the American Petroleum Institute decide what BP and Shell can charge for gas. Doctors get to decide how much they get paid.
Three times a year, the AMA convenes the Relative Value Scale Update Committee (RUC), which operates like a senate serving 26 specialties. Otolaryngology gets 1 representative, general surgery gets 1 representative, and so on.
In these meetings, the different specialties deliberate on how much they should get paid for each procedure they do. As you might expect, there is plenty of politicking involved.
The RUC minutes and votes are public, as shown here.
(Shortform note: Once set, these RVUs have systemic national implications, nudging doctors toward performing the higher compensated procedures.
If this system behaves like a typical economic system, perverse incentives are likely at play here. Imagine how a new medical device might be adopted.
The AMA also earns directly from being in charge of the RVU system. They have copyrighted the billing CPT codes and charge license fees to anyone wishing to associate RVU values with CPT codes. This earns them $70 million a year!)
Like hospitals and insurers, physicians do what they can to increase their pay. Here are a variety of methods that doctors use.
Doctors have created their own businesses to negotiate better terms with insurers.
For example, doctors have created ambulatory surgery centers, which are large enough to allow billing for facility fees. This means that procedures that could have been done in doctors’ offices were moved into the more expensive surgical centers, without any apparent increase in medical outcomes.
The centers themselves don’t participate in insurance networks, so their fees aren’t constrained by insurers’ negotiated rates.
Doctors have also formed LLCs to provide services, then contracted with hospitals. These are especially popular in the specialties of pathology, anesthesiology, radiology, and emergency medicine (PARE). The gain is mutual—by contracting services, hospitals avoid the overhead of malpractice and staffing. It also increases the ability to separate billing and maximize reimbursements. Similarly, these LLCs also evade contracts with insurance, which often means out-of-network payments for patients.
Federal regulation requires hospitals to treat all patients showing up in their emergency rooms. But this regulation doesn’t apply to physicians, who can pick and choose which patients to see.
Thus, hospitals contract with doctors to cover the ER; in exchange, doctors can bill however they like. The contracted doctors might be out of network. So even if you as a patient show up to a covered hospital’s emergency room, the doctors’ bill might not be covered.
A multitude of practices allow doctors to bill for more time in questionable ways. These include:
There have been counters to these practices: Medicare now requires surgeons to be present for key parts of the operation, and physician assistants can bill at 15% of the surgeon’s rate.
But no matter the incentive, doctors will take advantage. For example, in oncology infusion, Medicare paid for the first 90 minutes of the infusion, then a second part for any part of the hour after. This led to a suspicious number of 91-minute infusions.
And to bill for an infusion, doctors technically have to be present in the office for the first 15 minutes of infusion. Not only is this hard to police, but it likely leads to doctors hanging around and checking their watch without improving outcomes.
In the US, oral drugs can be dispensed only by pharmacies and not sold by doctors. However, IV and injectable drugs can be billed by doctors. This leads doctors to buy drugs and bill them with a markup.
Even better, doctors get incentives from pharma (such as free samples, rebates, administrative fees, and grants) to drive adoption of a drug. Doctors then bill for drugs at full price, with a set percentage for markup. Clearly, doctors have an incentive to prescribe more expensive drugs that pay them more, even if they don’t necessarily lead to better patient outcomes.
It works out even better if administering the drug comes with a new procedure to bill for. Here’s a story about Lupron, a hormone treatment.
Prostate cancer patients often had their testicles surgically removed to remove testosterone and limit cancer growth. Instead of castration, Lupron biologically blocked the influence of testosterone. It could be administered at home with subcutaneous injections. However, sales were not a blockbuster.
Lupron manufacturer Takeda then came out with Lupron Depot, a longer-acting formulation that needed to be injected into muscle tissue, and thus required a doctor’s involvement. The strategy was to give doctors a new procedure to bill and allow sales from the drug.
To increase adoption of Lupron Depot in the face of cheaper alternatives, Takeda provided free samples doctors could then bill at full charge. It also offered grants and kickbacks to doctors to prescribe Lupron. Finally, it inflated the wholesale price that Medicare reimburses at (for which they were fined hundreds of millions of dollars.)
Even though Takeda was forced to pay a fine, no criminal penalties were assigned to doctors. Even today, doctors are paid administrative fees to market new drugs.
Doctors push patients to choose higher-priced procedures.
For example, cataract treatments have become very standard and reimbursements have fallen. Enter the femtosecond laser. Medicare wouldn’t reimburse extra for the laser, so ophthalmologists pushed patients to upgrade and pay out of pocket.
This would be fine if patients could make educated decisions about what treatments are most cost-effective for them, but they can’t. Patients have little pricing and quality info to compare across interventions. In effect, they are paying for things unrelated to outcomes, like how much they trust their doctor and what positive emotions they feel.
Doctors also find ways to bill more to provide extra services, such as retainer fees, same-day answer services, or fees for writing prescriptions.
As we’ve seen, surgical procedures earn higher billing. As a result, non-surgical doctors increasingly do procedures, even if they’re not formally trained in it.
For example, nephrologists have largely outsourced dialysis, which is commoditized and doesn’t pay well. But now they perform surgical procedures to create a fistula for vascular access. Traditionally, these have been performed by vascular surgeons with years of training, but nephrologists are taking this over.
Here’s an abbreviated history of pharmaceutical companies in the US:
An important point: in the US, Medicare is legally prohibited from negotiating prices with pharmaceutical companies. In contrast, single-payer systems like that in the UK have more leverage to choose a subset of drugs to support and squeeze prices down.
Pharma companies use a few strategies to speed up clinical trials to get drugs approved.
The HIV rush for drugs prompted the FDA to allow surrogate measures, or proxies for health (like blood markers) instead of actually proving that the drug cured symptoms over months or years.
Manufacturers promise to carry out follow-up studies, but there is no punishment for not doing so. Nor is there a mechanism to take drugs off the market if they later turn out to be inefficacious.
Orphan drugs treat conditions affecting fewer than 200k patients. To incentivize pharma companies to develop treatments, orphan drugs get a few perks:
For example, the melatonin analogue tasimelteon (Vanda) targeted Non-24, an orphan disease. It was approved despite not showing clinical advantages over melatonin.
The FDA gets $2.5 million per drug application. Thus, the agency has incentives to receive more applications for me-too drugs, even if they improve little beyond what already exists.
The Supreme Court has protected drug advertising under free speech. The United States is only one of two countries allowing drug advertising, along with New Zealand. Other areas like the EU don’t allow commercial speech the same equivalence in free speech, considering drug advertising harmful to the public since it overpromises the benefits of wonder drugs.
Drug advertising began in the 1990s, with “help-seeking ads” where either the drug or disease could be mentioned, but not both. In 1997, ads could contain both the condition and the medicine, but they also had to include all side effects and contraindications
After a drug patent expires, other companies can submit a generic drug for approval. Typically, these far lower the cost of the drug.
However, the process is convoluted.
Generics aren’t a panacea for drug prices either. The economics for generics don’t always work out, causing competitors to pull out and leaving a single manufacturer with a virtual monopoly for a period of time. It’s said that 3 or fewer generic manufacturers are insufficient to provide price competition.
When patents end, pharma companies don’t just call it quits. There are a variety of ways to enjoy continued exclusivity and to fight generics.
By law, generics must be identical in dosage and form to the brand-name drug for a pharmacist to substitute them. States have different laws on the ability of pharmacists to substitute generics for brand-name drugs—some mandate substitution when possible, while others require patient consent.
To stymie the effort of pharmacies to substitute drugs, pharma companies will create new forms of the product to make it harder to substitute a generic with identical dosage and form. This is called “product hopping.” Here’s how it works.
Before their patent expires, the pharma company changes the form, dosing, and route of administration. For example, they might change from a pill that’s swallowed to a chewable pill. These changes require a new application to the FDA.
Then the pharma company destroys all previous forms of the drug, or they raise prices on the previous forms, to push prescriptions to switch to the new form, before generics arrive on the market. Finally, once the generic arrives, it can’t be directly substituted by the pharmacist.
This isn’t a total blocker. The doctor could theoretically prescribe the generic, even if it’s for the older dosage and form. But doctor habit, as well as low marketing from the generics, prevent this from fully happening.
This is said to illustrate market failure, since the brand name having a minor time lead leads to large long term sales differences. (Shortform note: some of the details came not from the book but from this useful FTC amicus brief.)
By law, the same product cannot be on the market as both a prescription and over-the-counter (OTC) product. Further, any company that takes a prescription drug to OTC gets 3 years of market exclusivity (since it requires submitting a new application with safety data).
Once a drug goes OTC, generics manufacturers have 6 months to use up their supply.
The patent holder can then also patent a new formulation or molecule similar to the now OTC drug, and enjoy both OTC and prescription sales. For the drug Flonase, Glaxo employed both strategies.
Typically, drugs that go OTC need a large market size to make up for cheaper prices.
As we’ve discussed, pharma companies sue generic makers, or partner with them to delay the introduction of generics. Here are other tactics that can lead to protection of patented status.
Pharma companies can patent a new drug combination of two off-patent drugs. For example, Duexis is the combination of the two generic drugs, ibuprofen and famotidine.
Pharma companies deliberately use different formulations that are more difficult to prove bioequivalence. For example, sprays, lotions, creams are more difficult to replicate than pills
Pharma companies strike long contracts with hospitals to take their branded drugs. This reduces the market size for new generics.
Pharma companies may even help promote information that suggests the generic substitute is dangerous. In the early 2000s, the anti-nausea drug Zofran faced competition with generic droperidol. Suddenly, the FDA issued a warning that droperidol may cause heart arrhythmias. The generics disappeared from the market, and suddenly only Zofran remained.
What if patients can’t afford their co-pays on the drugs? Pharma companies have a scheme for that too.
The pharma company sets up a nonprofit that helps patients pay co-pays. This is net better for the pharma company anyway—the insurer is paying most of the drug’s cost.
For example, say a drug costs $5000, and the patient’s co-pay is $1250. Pharma covers the co-pay, and it gets $3750 for a sale that wouldn’t otherwise have happened. The drug price remains high, insurance reimburses, and pharma gets marginal sales.
This strategy neutralizes the advantage of cheaper drugs for the consumer—without pricing pressure, consumers pick the drug they like most, or the most expensive one.
The patient’s copay might even be higher with a cheaper competitor that doesn’t have this co-pay assistance program. For the example above, say a competitor has a $1000 drug that costs the patient $200 in co-pay. The patient would obviously rather choose the $0 copay drug, even if it costs the medical system far more.
Medicare forbids co-pay offers directly from pharma firm foundations. However, private insurers aren’t barred from this, and they face competitive pressure to allow it—if one insurer didn’t allow it, patients would quickly leave for a competing insurer. Unfortunately, this sets up a crisis when the patient goes on Medicare and can no longer afford drugs.
Best of all—this co-pay assistance is tax-deductible for pharma companies.
Pharmacy benefit managers are hired by insurers and employers to negotiate drug purchases with pharmacies. Per antitrust law, insurers cannot compare prices they pay for medicine, but PBMs can.
The largest PBMs are Express Scripts, CVS Caremark, and OptumRx.
PBMs pocket a percentage of discounts they negotiate. Thus they put drugs that net them the best deals, not necessarily the drugs that are most in demand.
Furthermore, they may treat individual drugs as strategic pieces in a chess game. They can drop certain drugs from coverage to pressure pharma companies for better deals, with patients as collateral damage as they scramble to find alternatives.
Medicare Part D was started in 2006 as a way to make drugs more affordable for patients. Before this, Medicare seniors had to pay out of pocket.
Part D had a unique co-pay structure:
The unintended effect of Part D was that seniors now had guaranteed drug coverage and ample budgets to spend on medicines. Pharma companies quickly pushed up prices for essential medicines, sometimes to multiples what they used to cost.
As economics would predict, every incentive and opportunity is exploited by market participants.
Factor VIII is a treatment for the bleeding disorder hemophilia. Here’s a story about how this old treatment has become progressively more expensive over time.
The medical device industry is dominated by a few major players: Medtronic, St. Jude Medical, Boston Scientific, Stryker, and Zimmer Biomet. Like other segments of healthcare, they have consolidated over time.
Many device makers began in hardware or consumer electronics. Medtronic started in 1949 as a medical equipment repair shop
In 1969, surgeon Denton Cooley implanted the first artificial heart in a patient for 3 days without FDA approval.
In 1976, the FDA defined three classes of devices that need different levels of approval.
The scrutiny in class 2 is so much lower that most devices are submitted under this designation, including devices you might consider to be potentially life-threatening such as joint replacements and surgery clips. Class 2 applications outnumber class 3 by 60 times.
The bar for class 2 is relatively low. You need only claim that your device is “substantially equivalent” to a product already sold and used for the same purpose.
Class 2 requires far less testing than class 3. Only 10% of FDA class 2 applications contain clinical data—many devices aren’t even tested in animals before placing into humans, and most offer no clinical trials at all. Class 2 devices aren’t required to be proven “safe and effective” like drugs are.
As a result, the FDA has little issue approving the devices. FDA cleared 85% of devices under the “substantially equivalent” criterion. The FDA spends 1200 hours reviewing class 3 devices, compared to only 20 hours for 510(k) requests.
Some manufacturers don’t even seek FDA approval for new models, since they don’t consider the new version “substantially different” from older models. However, patients don’t always know the device is experimental and don’t consent to it.
Despite the fact that the devices are “substantially equivalent,” if a specific device is recalled, there is no recall mechanism to recall equivalent devices. Furthermore, complications are hard to detect—generally, reporting of complications is voluntary, unless the FDA mandates it.
The easy clearance of “substantially equivalent” devices means they’re largely interchangeable and thus don’t have a clear product or performance advantage. As a result, marketing and sales agents need to intervene to push products.
This results in a variety of questionable relationships between industry and doctors.
Surprisingly, device reps can serve as unpaid assistants in the operating room. Theoretically, they’re on hand only when a device is used. They don’t scrub into surgery, but they point to things that should be used in specific places and help troubleshoot. These device reps make money through commission only.
The natural bias will be to favor devices from companies whose reps are more likable and helpful. Since they like the rep and know she’s paid on commission, they’ll skew to higher-priced items. As a device rep reports, doctors “feel somewhat obligated to use the most expensive device because they obviously called you in for it.” In response, hospitals have demanded device reps be registered and get permission to enter premises, like lobbyists.
Says a device rep, “it’s nearly impossible to break into the market - it’s all about these relationships.”
Devices are often largely interchangeable, so there is little differentiation between them.
To address this, manufacturers use proprietary screws, tools, and “operating systems” to generate complexity. Because the requirements are often changing, this pushes doctors to ask for more device reps on hand to help figure out how to use the devices.
Manufacturers use pricing pressure to push adoption of the new version of the device. For example, they price the older version at the same price as the new one, so it stops being cost-effective.
What can hospitals do to lower device costs? They can limit the range of devices used, so they gain bargaining power over the manufacturer. They can also close down product marketing, requiring proposals to get quotes from manufacturers to compete with each other.
As payers tightened up their spending, providers looked for other ways to increase billing. One opportunity came in testing and ancillary services, where they restructured the business models to better profit. The doctor, hospital, and staff all benefit from increased testing.
Tests done in hospitals are more expensive than those at third-party labs (eg Quest). However, as a patient, you often don’t get the choice of where to send your tests, and the options aren’t often clear.
Testing in general in the United States has inflated pricing. An MRI costs $160 in Japan. It costs $3500 in a US hospital.
The incentives to increase billing for testing trickle down to doctors, who get rewarded through bonuses for billing. Hospitals are complicit. One doctor ordered EEGs for kids to detect undiagnosed seizures, until most patients turned out not to have seizures at all.
The more tests a doctor orders, the more she can bill. This has changed medical practice to heighten the testing done:
The game to increase billing is to unbundle them and charge the maximum per individual test.
For example, a hospital overnight stay is billed at an all-inclusive rate. However, pre-operative tests are done on an outpatient basis, so they can be billed separately from the bundled rate.
Another example is to unbundle a test into as many separate components as possible. The standard CHEM-7 test can be split into seven separate tests, increasing billing significantly.
It’s too much to suggest that doctors and hospitals only order tests to increase billing. Of course, testing has a legitimate use. However, factors such as the incentives described above can push testing beyond their cost-effective point.
Another factor includes patient demands. Patients increasingly want more data now, and they may personally ask for excessive testing. Doctors may sometimes find it easier to placate patients by ordering tests than to instruct them about why the test yields little more information.
Doctors themselves worry about missing something. In this litigious medical environment, doctors practice defensive medicine to lower their risk of liability in a malpractice lawsuit. This means more testing to avoid a catastrophic mistake.
Pathology is particularly used by dermatologists, gastroenterologists, and urologists.
Patients have little choice over where biopsies are sent for analysis, and doctors aren’t aware of how much they cost. Sending to a particular pathology center might cost 5 to 10 times more than a commercial center.
In the 1990s, the Stark Law limited referrals for testing where the provider had a financial interest. This was meant to reduce self-dealing.
In 2002, Congress allowed exceptions for certain types of care rendered in physicians’ offices, like X-rays and physical therapy. In turn, many specialists hired pathologists to process samples in-house.
Over time, pathologists formed their own LLCs and contracted their services with hospitals. As is the theme in healthcare, many firms consolidated, leading to major international diagnostics companies like Miraca.
Commercial pathology companies want the business of hospitals, so they court hospitals through sponsoring fellowships and paying the salaries of doctors in training.
Some medical practices outsource pathology to national labs, but they bill directly as though the tests were done in-house.
Alarming pathology diagnoses are re-analyzed in-house, leading to double billing.
Ambulances are run by cities, hospitals, or private companies (the largest of which are Rural/Metro and American Medical Response).
For such a critical service, billing can be lax. Said a city budget analyst, “we could keep raising rates for ambulances because the private insurance companies always paid.” Ambulance rides earn the city of New York $200 million each year.
Many ambulance providers refuse to contract with insurers, leaving the bills with patients. Of course, patients who need ambulances in emergencies are in little position to ask about billing as they get picked up by the ambulance.
As we’ve seen multiple times already, ambulances unbundle their billing to increase reimbursements. There may be separate charges for an oxygen tank, ice pack, and gauze.
As is typical in healthcare, relationships make a difference in who gets the business. Ambulance crews get assignments through a central government dispatcher, and then they get leeway on where to take patients. Hospitals entice ambulance crews to go to their hospitals with comfortable lounges and food. Some hospitals also run their own ambulances to compete for patients.
Physical therapy was a $26.6 billion industry in 2014. The industry has grown rapidly in the past few decades because Medicare gradually lifted the cap on therapy. Starting at $100 per year in 1979, the cap rose to $1500 in 1999, $1860 in 2010, and finally lifting the cap entirely in 2018.
The billing practices are loose. Therapists can deploy extenders like doctors do, and patients can see physical therapists without physician referral.
A physical therapy consult and follow-up has become standard in every hospital discharge, even though there is little proof of efficacy in outcomes in many diseases.
Medicare Advantage pays a fixed monthly fee for elderly care services to take care of members. The fee increases with the burden of illness.
So health care plans contract with companies to conduct home visits at $300 per head, and they can earn $2,000-4,000 more per person by uncovering new illnesses. Of course, the incentive goes toward diagnosing new conditions.
Billing codes have gotten so complex that both insurers and providers have outsourced claims to third-party contractors. They help handle billing, coding, and collections, and they often get paid a percentage of the billing they obtain.
Disease codes were first used for epidemiological purposes. The WHO created the ICD to formalize the classification of conditions.
In 1979, the US used ICD codes for Medicare/Medicaid claims, creating their own version of ICD, the ICD-CM.
Getting a code for a condition is a big deal, because it becomes classified as a formal disease, and it obligates insurers to pay for it. Obesity got its code in 2013.
Three types of billing codes are now prevalent: CPT, HCPCS, and ICD.
Because medicine advances quickly, codes often change. This has spawned an industry dedicated to billing codes, with contractors arising both to help providers bill for more and to help insurers pay less.
The key to maximizing billing is to understand that different billing codes have different prices.
Perversely, this can cause overtreatment to fit the higher-paying, more complicated code.
To put some limitations on prices, insurance companies have negotiated rates with hospitals for billing codes. Unfortunately, hospitals have been shown to charge the uninsured 2.5x more than those covered by insurance.
Billing is essentially a tug of war between providers and insurers.
Providers will bill for whatever they can. Not only does this mean billing for the most expensive treatments allowable, sometimes third-party coders will inappropriately code for treatments that never happened. One patient with a Jewish name found a charge for a circumcision for his newborn son, even though they didn’t even have a circumcision performed.
In turn, insurers will deny claims as overreaching. They also outsource precertification to contractors, firms that check for coverage and get money by denying care.
With the increasing complexity of billing, single doctors have found it too difficult to run all the checks for a patient’s treatment, including whether the doctor participates in the patient’s insurance plan, whether a procedure is covered, and whether the procedure needs authorization from the insurer.
This is one reason independent doctor’s offices are rare. Instead, doctors have banded together into independent physician associations to better handle overhead costs.
To help empower patients to make more informed care decisions, patient advocates can now be hired by employers to estimate local prices and give advice on where to get care. If an advocate costs, say, $50 per employee, they can help save money on the average patient.
Patient advocates also make money by reselling price data to hospitals and insurers who want to know how much their competitors are getting reimbursed. However, advocates don’t share their price lists with customers, since this would remove part of the need to hire them.
According to An American Sickness, medicine’s history began with more moral ambitions and less profit motive.
A massive change happened when the Cystic Fibrosis Foundation (CFF) sold its rights to drug royalties for $3.3 billion. In 2000, the CFF invested in Aurora Biosciences, which was then acquired by Vertex Pharmaceuticals. The company released a new drug Kalydeco, which was FDA approved in 2012 and cost $300,000 per year. Two years later, the CFF cashed in its rights to drug royalties and received over $3 billion.
This provoked non-profit organizations to embrace a new business model: “venture philanthropy.” They now invest in pharma and device companies expecting to earn a financial profit. For example, the Juvenile Diabetes Research Fund invested $17 million in Medtronic for a glucose sensor, invested another $4.3 million in BD, and formed a new investment fund with PureTech Ventures.
As another shift, disease-centered nonprofits and charitable foundations are now often funded directly by pharma companies, leading to some confounding of interests. Pharma participates in foundation events to recruit patients and talk about new treatments.
The close relationship between nonprofits and healthcare vendors now complicates the incentives of nonprofits.
It appears as though fund-raising is now the primary metric for foundations, rather than patient wellbeing. Said a critic of JDRF, “if the March of Dimes were operating like today’s foundations, we’d have iron lungs in 5 colors controllable by your phone, but we wouldn’t have a polio vaccine.”
The American Medical Association is the professional organization representing doctors. It nominally supports the well-being of doctors and influences medical guidelines. It publishes the well-known medical journal JAMA.
However, like the disease-focused non-profits we just discussed, it engages in a variety of profitable practices that are somewhat questionable:
Nowadays, only 25% of doctors join the AMA, partly because individual specialties now have their own professional organizations, each with their own lobbying efforts. Each specialty is focused on protecting its own doctors, even at the expense of patient cost and convenience.
Examples of what they lobby for:
Examples of what doctors in general lobby for:
The American Board of Internal Medicine mandates activities from its member doctors, including more testing and recertification programs and attending conferences. This is required for bonus payments from Medicare and admitting privileges from hospitals.
The ABIM can then charge high prices for their own certification programs.
Healthcare systems in local regions have consolidated to provide negotiating power against employers and insurers.
Say you’re the only major medical provider in town—you might have literally the only maternity ward in the region. The local patient population needs access to you. Therefore, employers have to buy insurance that provides you in-network. Therefore, insurers have to meet your demands, especially around pricing, to sign you.
The pricing effect is real—studies show that hospital mergers in concentrated markets cause prices to increase by over 20%. Low-competition areas show symptoms of higher premiums, higher medical prices, and possibly suboptimal care and overtreatment
Counter-intuitively, the prices may not be lowered with competition. The large players might set a high price, which emboldens smaller players to raise the prices as well. This is just one way that the healthcare market is dysfunctional, compared to the idealized economics free market.
Large health systems span a huge range of services and consist of hospitals, ambulatory care clinics, nursing facilities, and large physician networks.
Being part of a health system opens up opportunities for more aggressive billing. For example, say a health system purchases a primary care office. Now a visit to a primary care doctor can tack on facility fees to the charges; doctors can also bill visits and tests as hospital charges, which earn higher rates. All this is allowed despite the fact that the doctor visit hasn’t changed or gotten more effective.
The patient can suffer under these changes. Some insurers require a higher deductible for hospital fees, and so the higher costs get pushed onto patients.
With their size, hospitals also demand other terms favorable to them, like forcing certain procedures (e.g. drug infusions) to be done in hospitals.
Size begets size. Large healthcare systems can pressure smaller players into being acquired in order to access the system’s services. It can also pressure insurers into dropping smaller players.
Sutter Health is an exemplar of aggressive consolidation, buying hospitals and restructuring around more profitable services. It downsizes less profitable hospitals to the bare minimum to meet Medicare’s 25-bed critical access designation, which raises reimbursement rates.
Obama’s 2009 HITECH Act gave $19 billion in incentives to providers to adopt EMRs. While this advanced the adoption of technology, it had a critical omission in not requiring intercompatibility between EMRs.
Like many software companies, EMR vendors such as Epic desire low intercompatibility to increase switching costs. Once you get used to a system, if you can’t easily export your records, you tend to stick with the system you’re currently using.
Because EMRs aren’t intercompatible, they can’t talk to each other. Patient records at one hospital are difficult to transport to other hospitals.
Non-intercompatibility is good for providers too. Large health systems can charge smaller players fees to get into its medical system. It can also impede importing of outside test data, which raises friction for doctors and thus encourages the use of testing in the hospital (with its outsized prices). The EMRs are complicit in this too—they can change the default workups to use in-hospital testing.
(Shortform note: If you’ve noticed we’ve skipped Chapter 10, that’s because it’s about healthcare as a business, and we’ve integrated its points into previous chapters.)
One of Obama’s hallmark achievements was passing the Affordable Care Act. While it made good progress in some areas, it ultimately fell very far from its original sweeping vision. Here’s a discussion of its achievements and its pitfalls.
Here’s how the Affordable Care Act succeeded:
Here’s how the Affordable Care Act failed:
It’s not clear to patients that their lives have improved with the Affordable Care Act. Healthcare spending doesn’t seem to have been strongly controlled, and premiums have risen for many patients. This has led many to call for the repeal of the ACA.
Now that you understand the healthcare system more, reflect on problems you’ve personally encountered.
What are some grievances or complaints you’ve had about the healthcare you’ve received?
Now that you understand more about how healthcare operates as a business, what do you see as some of the causes of the problems?
Now that you understand why and how American healthcare is so dysfunctional, what can you do about it?
Part II of An American Sickness discusses steps you can take to improve your personal healthcare and possibly make a dent in the American healthcare system. Each following chapter contains 1) practical tips on how to make medicine personally better for you, and 2) legal reforms that, if passed, would lead to more systemic change.
This chapter describes healthcare systems in other developed countries, ranging in the degree of government intervention. Generally, they show better quality for lower cost compared to the US healthcare system.
What it is: National governments set fees for health services and negotiate prices with vendors.
Where: In Germany, Japan, Belgium
Benefits:
What it is: A single authority, usually the central government, pays most of the money to providers.
Where: In Canada, Australia, Great Britain, Taiwan
Benefits:
What it is:
Where: Singapore
Benefits:
According to An American Sickness, many Americans don’t fill prescriptions because of cost. Why? It’s difficult to do price comparisons, prices can change from month to month, and, depending on your insurance, it’s unclear what things will cost.
(Shortform note: this is the last chapter of the book and is included here because it’s short.)
Despite the promises of digital technology to improve healthcare, much has not panned out. Many health startups tend to add layers to cost, rather than making it simpler.
Here are the book’s recommendations on digital tools for healthcare: