This month Harvard Business Review (HBR) published the dramatic results of Walmart’s innovative centers-of-excellence (COE) care-bundling initiative. Walmart’s third-party administrator, Health Designs Plus (HDP), claims that this approach saved Walmart $86 million on spine surgery alone over 3 years, and could “address the cost-and-quality dilemma and drive change nationally.”
But will it?
FixingUSHealthcare blog agrees that Walmart’s “experiment of nature” does validate key claims made by this blog.
But in this post, we will discuss why Walmart’s approach alone will not fix the whole U.S. healthcare system, and will not itself bend the curve of relentlessly rising national health spending.
Walmart’s Centers-of-Excellence Program
In 1991, Sam Walton famously railed to his managers, “These people are skinnin’ us alive not just here in Bentonville but everywhere else, too….They’re charging us five and six times what they ought to charge us….So we need to work on a program where we’ve got hospitals and doctors…saving our customers money and our employees money.”
Twenty years later, Walmart – the largest private employer in the world, with over 1 million associates – rolled out its first bundled-care program for spine surgery. HBR recounts,
Building on its experience with a long-term relationship with the Mayo Clinic for organ transplants, the company set out to develop similar arrangements with other providers for an expanded set of conditions. Early in the discussions its benefits plan leadership zeroed in on the procedures with the greatest opportunity for improvement: common and expensive surgeries (those costing more than $20,000, on average) with high variation in cost and clinical outcomes across providers. Heart and spine surgeries meet those criteria… Walmart launched its heart and spine surgery programs in 2013. It went live with joint replacements (hip and knee) in 2014, certain cancer evaluations in 2015, and bariatric, or weight loss, surgery in 2016.
More detailed information is available at Walmart’s website.
These programs, according the HBR, involve
bundled payment arrangements that cover the cost of an employee’s care for certain episodes from start to finish — all the procedures, devices, tests, drugs, and services needed for, say, a knee replacement or a back surgery. They’re also, in most instances, picking up the tab for any necessary travel, lodging, and meals for the employee and a caregiver, thus democratizing destination care programs that have historically been reserved as an executive perk.
HBR reports that Lowe’s, McKesson, GE, and Boeing are in the process of emulating these programs.
Now, HDP’s founder Ruth Coleman and co-authors from Geisinger Health System spine COE and Walmart’s benefits manager have published in HBR early results of these COE programs. This report was featured by fellow WordPress blogger Henry Kotula at his Leaders in Healthcare blogsite.
Here is a summary of what Coleman and co-authors report:
1. Spine surgery:
From 2015 to 2018, Geisinger Health System in central Pennsylvania performed 2300 spine surgeries for Walmart employees at a cost averaging $32,177 apiece. This was slightly higher than the average cost at community hospitals otherwise used by Walmart employees.
However, of 5,000 patients initially recommended to have spine surgery, Geisinger treated 2,700 non-surgically, avoiding an $86 million cost.
They cite “Bill,” whose local neurosurgeon noted worsening neck pain and new hand tremors. On the basis of an abnormal MRI, the surgeon offered spine surgery. But at Geisinger the team also noticed his shuffling gait, a dead giveaway for Parkinson’s disease. With the correct diagnosis, Bill ignored the red-herring MRI, avoided surgery, and got his proper drug treatment instead.
For the 2,300 Walmart employees who did have surgery at Geisinger, only 14 needed a rehab stay in a nursing home (instead of the usual 49 per 1,000 cases elsewhere). Meticulous checklist-driven expert care reduced complications and readmissions by 95%. Employees returned to work in 10 weeks instead of the usual 13 elsewhere. This was the result of intensive case management and best-practice “care pathway” strategy.
2. Joint Replacement Surgery
Eighteen percent of Walmart associates having joint replacement surgery between 2015 and 2018, totaling 1,836, traveled to Johns Hopkins Medical Center in Baltimore. Another 367 were “guided” to non-surgical treatments because they only needed physical therapy or because of “health reasons that rendered surgery inadvisable.” These COE patients spent 32% less time in the hospital post-surgery and had 70% fewer readmissions. None required rehab in a nursing home, compared with 50 per 1,000 receiving usual care. The overall cost at Hopkins was $23,505 compared with usual of $27,721. They returned to work in 11.3 weeks compared with 12.8 weeks for usual care patients.
The cost savings was $44.5 million, not counting the costs avoided by not needing nursing home rehab stays.
What the Walmart Experiment Proves
Academics will rightly point out that the results published by HBR are “anecdotal,” that is, are not scientifically rigorous. The research was “observational,” namely not randomized. Patients self-selected whether to receive local care or COE care. It is not hard to imagine how the COE patients might have been different in clinically important regards from the comparison “usual care” local patients. For one, the COE patients needed to be able to tolerate a long airplane ride and shuttle ride from the airport to the Geisinger Medical Center or Johns Hopkins Hospital.
Furthermore, these results would not be publishable in HBR’s cousin medical journal, the New England Journal of Medicine (NEJM). Since 2004 NEJM and other premier research journals have required advance registration of all trials at ClinicalTrials.gov to prevent cherry-picking only favorable study results for publication. Naturally, Walmart’s COE initiative was not registered in advance as a rigorous scientific clinical trial.
Regards potential conflicts of interest, Coleman and co-authors have been up-front about theirs in publishing this article. Each has explicitly detailed how they could personally profit from the favorable testimonials in their report.
Taken together, this means that we cannot assume that Walmart’s programs are scalable or that the same dramatic results could be expected from other COE programs.
Nevertheless, we can take the report at face value. Walmart is a business, not a research university. Walmart’s “experiment of nature” does give us useful business data.
And, in my opinion, the report does validate some important claims made by FixUSHealthcare.blog:
- Healthcare costs in the U.S. are “skinnin’ us alive”
- Small area analysis (comparing various health center’s overall results to each other) works to identify “best practices” (See Segment 5)
- Using “best practice” benchmarks works to improve processes of care
- The principle of diminishing marginal benefit applies to medical treatments – spending more and doing more doesn’t necessarily get more value, and is sometimes harmful, such as spine surgery or joint replacement performed on high-risk surgical candidates, with resultant complications or death
- Fee-for-service payment induces over-treatment (note that the surgeons at both Geisinger and Johns Hopkins are salaried, thus not susceptible to monetary incentives)
- Insurance and other administrative costs add significantly to healthcare spending. Just by self-insuring, Walmart is reducing these costs considerably.
- There is waste in the whole healthcare system (See previous post)
Why COE Bundling Won’t Fix U.S. Healthcare
So, if Walmart’s COE bundling approach addresses all these problems in the system, won’t scaling up this initiative to a national level fix U.S. healthcare, as claimed by HDP?
First of all, let’s agree that the COE Bundling innovation is great. It will save Walmart and other big companies money, will get their employees better health outcomes, and will make their employees healthier, happier and more productive. All well and good.
But here are the main reasons why this alone won’t fix U.S. healthcare.
1. The COE bundling approach applies only to big-ticket procedures
HDP admits that other approaches are needed for chronic disease. Health administration leaders are brainstorming ideas for value-based healthcare payment systems rather than volume-based. They are also seeking ways to allow health systems that manage health risk to also bear the risk (and be rewarded for reducing risk and improving health status) of large populations.
NEJM Catalyst is one example of a forum for brainstorming ideas in these realms. There is much interest by both private companies and Medicare in accountable care organizations (ACOs). But ACOs have so far yielded mixed results. (See, for example, Medicare Administrator Seema Verma’s status report.)
2. COE bundling will apply only to large populations
Simply put, smaller companies don’t have enough employees to make a COE program worth their while and worth the risk. There is an irreducible minimum “third party administrator” cost. If this cost cannot be spread over thousands of employee-patients, it eats up the per-patient savings. Likewise, actuarial advantages break down below a certain number of insured lives, when there are not enough individuals across whom to spread risk. Sooner or later, there will be one catastrophic case. Walmart’s large program can absorb the risk, but a smaller self-insured company would be bankrupted by it.
3. Squeezing balloons
Guess what: In Bill’s community (he’s the employee who avoided spine surgery after his second opinion at Geisinger) the neurosurgeon who proposed Bill’s surgery figured out a way to maintain his income – increasing volume (by lowering criteria to justify surgeries), raising prices, increasing “intensity” (adding extra services), or diversifying (many doctors are getting into the lucrative cosmetic market). Even the nursing home, which will now have fewer Walmart “customers,” can increase its volume (stretch other patients’ length of stay), raise its prices, increase its “intensity”, or diversify. (See my previous post on how healthcare is not a “pure market,” since to a large degree the producers control both supply and demand.)
This phenomenon of healthcare professionals and organizations gaming the system to constantly drive up revenue has been termed “squeezing balloons.” When you squeeze down costs one place, costs increase somewhere else.
Do you want proof? Then, get the data from each individual provider, or from each small area (local healthcare market), or from the nation. Walmart may save some money. But unless the whole system is reformed, total spending will rise at each level of the system, with no detectable effect of the COE bundling initiative, no matter how many big employers sign on.
4. Pricing paradox
And guess what: Since Geisinger charges more for its spines, everyone else will feel justified in doing so, too. Funny how that works.
In fact, that’s how it works in the whole system. Elisabeth Rosenthal, a Harvard-trained MD who became a healthcare reporter for the New York Times, has laid out extensive case examples in her exposé, An American Sickness.
Here are some of her Economic Rules of the Dysfunctional Medical Market:
- Economies of scale don’t translate to lower prices. With their market power, big providers can simply demand more.
- As technologies age, prices can rise rather than fall.
- There is no such thing as a fixed price for a procedure or test.
- There are no standards for billing. There’s money to be made in billing for anything and everything
- More competitors vying for business doesn’t mean better prices; it can drive prices up, not down.
- More treatment is always better. Default to the most expensive option.
Only a single-payer with the full faith and clout of the federal government can encircle the whole system, take efficiencies up to national scale, and apply the small area analysis, benchmarking, and best-practice process improvement without squeezing up price balloons elsewhere in the system.
Now, Take Action.
Title: Walmart Exterior, Laredo, Texas
License: This file is licensed under the Creative Commons Attribution-Share Alike 3.0 Unported license. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation