top of page

What's a Placebo Worth?

August 9, 2018. The Case of Epidural Steroid Injections (ESI)

The Inspiration for this Post.

At the cranky age of 73 I was referred to a neurosurgeon because my chronic low back pain had abruptly worsened. He spent a couple minutes perusing the MRI while I described my complaint. Then, without committing to a diagnosis, he said "You need to get an epidural," meaning an Epidural Steroid Injection (ESI).

Exercising due diligence, I googled PubMed. There I uncovered a profile of patients who are NOT helped by ESI. According to the stats, ESI is worthless when the pain is chronic, linked to stenosis, not aggravated by physical activity, and not radicular (doesn't run down the leg). Advanced age is another negative.

That profile described me perfectly. Clearly, I was the wrong candidate for ESI. So I sought out a second opinion and ended up seeing two more pain doctors at an orthopedic center. But their response was identical to the first: "Let me schedule you for an epidural." I insisted on a referral to physical therapy instead.

As practitioners of "evidence based medicine" my doctors were surely aware I was not a good candidate for ESI. But the opportunity to bill insurance upwards of $500 for a 25-minute procedure was just too good to pass up. They cherry-picked that opportunity from my MRI scan. Like most old men, my MRI revealed multiple signs of degenerative disease, any one of which could be used as a justification to bill Medicare for an ESI. That's the way "evidence based medicine" works in the real world.

The remainder of this post explains why my experience is not an isolated incident. In a nutshell...

Despite the fact that steroid injections provide short-term pain relief only for a precisely selected subgroup of patients, over the past 25 years the procedure has been indiscriminately expanded to virtually to everyone with a complaint of back pain. In other words, "pain management" profiteers are exposing the wrong patients to a risky procedure that's no more beneficial than a shot of saline.

ESI Illustrates the General Problem of Big Bills for Small Effects.

With no superior health benefit to show for it, America's healthcare industry consumes almost twice as much GDP as other rich countries. An oft-cited reason is the medical profession's penchant for "aggressive diagnosis" resulting in over-utilization. This is a euphemism for the prescribing of expensive tests and treatments whose benefit to patients is highly dubious.

These patients/consumers could get a lot more value for their money if the healthcare market operated more like the market for sneakers, pizza and tattoos. That's because shoppers for those items usually have access to good information about the their price and quality. And shoppers' ability to choose what to buy, or not to buy at all, motivates suppliers to deliver real value. Armed with decent information, even the dullest consumer is unlikely to get conned into a purchase that costs more than it's worth.

That's not the way it works in America's health care system. In the market for medical services, consumer choice is replaced by supplier discretion. Patients lack the expertise to evaluate treatment options, so the suppliers (physicians) choose for them. Whatever the doctors prescribe, insurance coverage reduces the cost to the patient and thereby disables an important check against excessive pricing. Even worse, Medicare and private insurers accept the "fee-for-service" model favored by the AMA, which means doctors are reimbursed for the procedures they perform rather than patient outcomes. So even if a procedure's therapeutic value is nil, the physician gets paid the same "reasonable and customary" fee, about $100 per hour.

Thanks to these perverse incentives, medical specialists are peddling treatments whose full price (insurance + out-of-pocket) is far greater than their therapeutic value. Some popular treatments are only marginally effective*; and others are actually no better than a placebo. If a professionally approved intervention performs no better than a placebo, it should be named for what it is: FAKE TREATMENT. (And if the fake treatment puts patients at risk, MALPRACTICE would be more accurate).

Billing for Fake Treatments: The Sorry Tale of Epidural Steroid Injections.

ESI never earned FDA approval because "the effectiveness and safety of epidural administration of corticosteroids have not been established." In 1999, a major Cochrane review concluded that the pooled results of randomized control studies failed to demonstrate the effectiveness of ESI. The fact that some of those studies were biased by the authors' financial stake in the procedure casts further doubt on its effectiveness (see graph).

Bias in reported benefit of ESI

Since 1999 an avalanche of new evidence** has decisively exposed ESI for what it is: a fake treatment. The data show that injecting the spinal area with steroids is no better than a placebo for relieving all varieties of back pain except one: Lumbar Radicular Pain. But even here the pain relief is only temporary and "of small magnitude." So today, treatment guidelines issued by professional societies*** limit approval of ESI to lumbar radicular pain and nothing else.

Yet non-radicular pain accounts for a majority of ESIs actually performed. Despite the mounting evidence against the expansive use of ESI, pain doctors followed the money, aided and abetted by the willingness of Medicare and insurers to reimburse them. In 1994, just 0.6% of Medicare recipients were treated with epidural injections, but by 2006 the rate had septupled to 4.3%, where it more or less remained through 2014. Of the 2,256,000 ESIs funded by Medicare in 2014, only 38% could be justified by a radicular pain diagnoses; the other 62% amounted to fake treatments.

So at $500 a pop, the bogus ESIs cost Medicare about $700,000,000 in 2014. And this does not include the loot extracted from folks under 65, ancillary costs like MRIs, and the harms inflicted by the procedure's side effects. The cumulative cost of this fake treatment since 1994 is surely measured in billions.

Is a Placebo-Treatment Better than Nothing?

Now that medical societies recognize that ESI is no better than a placebo for non-radicular pain, their new guidelines recommend remedies such as massage, acupuncture, spinal manipulation, tai chi, yoga, exercise, mindfulness, and spinal manipulation. The irony is, some of these recommendations might be no better than a placebo. The evidence is still too weak to tell. Doctors admit that they sometimes resort to placebo-treatments when their substantive tool kit is empty.

So, if it's okay for a physician to prescribe yoga or exercise, why not a steroid injection? To borrow from James Carville, it's the side effects stupid!

The potential side effects of ESI range from headache to paralysis and death; but the side effects of exercise are not just less adverse, they are hugely positive. (In case you haven't heard, sitting, inactivity, and obesity are the new smoking). So, unlike exercise, ESI treatment exposes the patient to risk without reward, in blatant disregard of "first, do no harm."


In my personal search for relief from low back pain, three specialists chose to ignore the guidelines of their own profession. By failing to mention the benign alternatives, they tried to lock me into a risky procedure -- a high pressure sales tactic worthy of an accomplished grifter.

At fault is fee-for-service, the entrenched practice of paying doctors for the time and skill they put into a procedure rather than its benefit to patients. This custom incentivizes doctors to indiscriminately peddle profitable crap like ESI. The obvious alternative is fee-for-value, where providers get paid in proportion to the value of patient-outcomes.

Unfortunately, genuine fee-for-value is unlikely to be implemented in models of "Value Based Care" (VBC) promoted under Obama-Care. That's because VBC programs measure the performance of doctors and hospitals in terms of "quality indicators" that are not based on patient outcomes. Furthermore, as creatures of Medicare and big insurers, VBC programs tend to equate value with cost-savings. And cost cutting invariably translates into less income for doctors, hospitals, and drug companies - a prospect that is sure to be resisted. Doctors will never relinquish fee-for-service entirely.

Harnessing Market Forces to Achieve Fee-for-Value?

To economists, the best measure of a product's value-to-consumers is the price they are willing to pay for it (not the price charged by the supplier).

This is just as true for medical services as it is for restaurant meals. But unlike the doctor, the restaurateur makes a profit only if the menu items are more valuable to the customers than the listed prices. If the incomes of medical providers were equally dependent on the customers' willingness to pay, doctors and hospitals would be offering high-value menus just like successful restaurateurs.

But that won't happen without massive radical reforms of the healthcare system. To appreciate the magnitude of the problem I suggest John Cochrane's controversial essay or this short audio version.


* For example, by controlling cholesterol levels statins are supposed to prevent heart attacks. If you already have heart disease, they lower your chance of a heart attack by 2.6% ; if not, your risk of heart attack is reduced by only 0.5%. Also, physicians still recommend acetaminophen for all sorts of pain even though it doesn't work. And they are more likely to prescribe an expensive brand-name drug over a generic, but it's dose dependent on gifts from drug companies.

** Links to meta-analyses and randomized control studies published in years 2009, 2009, 2013, 2014, 2015, 2015, 2015.

*** American Pain Society, American College of Physicians, American Academy of Neurology and their counterparts in Canada and Australia.



bottom of page