Thursday, May 28, 2015

Let’s not wait to address variation in prostate cancer treatment

Today’s Managing Health Care Costs Number is 80.1%

JAMA Internal Medicine had a disquieting article this fall about the treatment of prostate cancer in the United States.   Researchers reviewed the 2006-2009 SEER (Surveillance, Epidemiology, and End Results) database for men 66 and over to see evaluate variation in prostate cancer .

The first disheartening number was that 80.1% of those diagnosed with low risk prostate cancer received treatment within 12 months of diagnosis; only a fifth were treated with watchful waiting.  55% of those over 80 received definitive treatment; this is the group least likely to gain quality adjusted life years by having radiation therapy or surgery.  Few of those treated were treated hormonally; the vast majority received surgery or radiation therapy.

Perhaps even more disheartening was that self-referral played a big role in decision-making. Urologists who owned surgery centers and radiation therapy equipment were most likely to prescribe the services for which they billed, all of which are associated with high margins.   Other posts on self-referral here.

Remember that Medscape survey on physician compensation?  Urologists are among the highest paid specialties (average income, $340,000), and 27% say they “have offered new ancillary services in the last three years.”  That’s on top of all the urologists who have had ancillary services for longer.

The study authors suggest full public disclosure of ancillary ownership and robust public reporting about physician propensity to treat.   But the evidence that this will change behavior is scant.   We have an enormous bias toward treating cancer when we find it.  The thought of waking up every morning knowing that there are cancer cells in your prostate would be worse than unsettling. However, autopsy studies suggest that the prevalence of cancer in men from ages 71-80 in the US is over 80!.  The vast majority of older men have prostate cancer cells when they wake up each morning – and few of them will die of the disease.  

An important step is to stop doing prostate specific antigen tests for screening purposes.  The USPSTF has issued thoughtful guidelines, and it would be nice to see the rate of low risk prostate cancer diagnosis go down – since aggressive screening in the last few decades hasn’t lowered the risk of death of prostate cancer.   If fewer people have their low risk prostate cancer diagnosed, we will cause less medical harm. 

A commentary published on this study this week in JAMA suggests that it’s premature to do full public reporting, and we need to “strengthen the evidence base” for the best treatment, and better understand patient preferences. 

[We should] foster regional collaborations that facilitate comparative feedback to clinicians on their patterns of care relative to existing guidelines and their peers, as well as learning opportunities that can reduce variation in the delivery and outcomes of prostate cancer care.

It’s hard to oppose physician collaboration or getting more evidence or listening to what patients want – but physicians are playing the major role in treatment variation here.   We should publish data now on physician decision-making and we should more strictly regulate self-referral.  As more physicians are employed, we should be careful that their compensation is not tied to ordering more expensive (and high margin) treatment.

We don’t need decades’ more evidence that physician ancillary ownership leads to discretionary overutilization. 

Tuesday, May 26, 2015

It’s hard to change minds, and it’s hard to do good social science research.

Today’s Managing Health Care Costs Number is Zero

There was huge excitement about a December report in Science that described in-person visits to those who opposed marriage equality – and found that these highly-scripted motivational interviews really changed minds. Further, the impact was substantially greater when the canvasser was gay.

How much excitement?   The study was picked up by major media outlets including the New York Times, the Washington Post, the Wall Street Journal and This American Life.

This study was especially exciting because it contradicted a large body of research that shows that arguing with people doesn’t work.  They just hunker down, and become even more convinced that they were right in the first place. This has real-life implications in politics.  Political campaigns call those who are already convinced, and we knock on doors of the acolytes or rarely the undecided. We never try to contact those with opposing ideology.  This also means that there’s no point to debating my wife’s right-wing cousin.   He’s just convinced that I’m a limousine liberal and that his point of view is correct.

It was so heartening to hear that a sincere canvasser could be let into the home of someone who opposed gay marriage, and the interviewee would at least sometimes change his or her mind.  This American Life played tapes of these conversions –so the canvassing really happened.   A political science researcher surveyed the interviewees before and after the canvassing and found that the impact paradoxically increased over time.

The research element of this study, conducted by a UCLA graduate student, appears to have been faked.  Other researchers couldn’t replicate his high rate of survey response, and queried the research firm that he said had sent out the surveys.  The research firm said, essentially, “huh?” and denied any knowledge of the study.  The PhD student, whose Princeton job offer sounds like it’s being retracted just as this article will be, said that he destroyed the original data to protect the privacy of the subjects.  Right!

This is sad on many levels.   The canvassing really happened, and we know there was some anecodotal effectiveness.   The canvassers poured their hearts and souls into the project, and the Los Angeles LGBT Center spend hundreds of thousands of dollars on the study. The publication of this study in an especially prestigious journal is a black eye for peer-reviewed research.  We’re reminded that peer reviewers are unpaid and rarely if ever have access to the primary data.

This is also a reminder that when we see studies that contradict our current understanding, we should treat them with scepticism.   It’s easy to get public attention for a new surprising study; it’s much harder to get attention for a study that simply confirms what we already know.   What about the newest study suggesting there is a single answer to diet, or that a brand new drug is substantially better than existing drugs?   Most new “ground breaking” studies are never successfully replicated.  These articles are usually not fraudulent, but more likely to represent small numbers and garden variety randomness. 

New, ground-breaking research might prove to be correct – but best to wait for confirmation before we apply the findings to clinical care or to public policy.

Compendium of articles about the apparent fraud and likely retraction

Wednesday, May 20, 2015

Get your cancer screening advice from the American College of Physicians, not Taylor Swift!

Today’s Managing Health Care Costs Number is 5

This week’s Annals of Internal Medicine showcases “Screening for Cancer:  A Guide to High Value Care…” The article has no paywall- which is great, since this is important material for the general public.  

The American College of Physicians reviewed the literature on screening for 5 cancers, and here is a summary

I’m not sure the general public is ready to say “I only want high-value care.”  I’m not even sure that what is considered “low value” from a population perspective is necessarily low value from the point of view of an individual.   But I did actively choose not to have my prostate specific antigen tested, and have recommended to relatives over 75 to skip those mammograms! If others have the information to make informed decisions, we can save substantial resources AND save a lot of complications and pain.    

We need new language to express the fact that extra screening isn’t merely low value- but often exposes patients to excessive risks. Those risks can include surgery or radiation therapy or chemotherapy that they don’t need – and they could suffer genuine, not just theoretical, harm.
Gilbert Welch of Dartmouth has a great metaphor to explain why we often should not screen for cancer:

What’s clear is that cancers fall into a few general behavior patterns, which Welch and others have compared to animals that must be kept in the barnyard to prevent a deadly rampage. Papillary tumors are like turtles — they move very slowly and never pose an escape risk. They don’t need screening, because they will never cause trouble. Then there are rabbits, which are eager to hop away to other parts of the body, but can be confined if they’re found and fenced. These are the cancers that can be helped by early detection and treatment. Birds, on the other hand, are so flighty and quick that they can’t be confined. Screening makes no difference for bird cancers, because they’re so aggressive that they can’t be detected before they’ve begun their deadly course.
(This quote is from an article on thyroid cancer in South Korea)

Two Vox articles worth looking at on this topic as well.

Tuesday, May 19, 2015

Medscape Physician Comp: Primary Care Continues to Lag

Today’s Managing Health Care Costs Number is 45%

Medscape just published its annual physician compensation survey. [Requires free registration]   The survey includes almost 20,000 physicians and is performed on line – so it’s probably not entirely representative.  The survey results are self-reported, and physicians generally underreport their income.  But the comparative reported income among specialties is informative. This survey is among the largest available, and does not require an expensive paid subscription.

The results are no surprise.   Specialists make 45% more than primary care physicians, and orthopedists make 224% as much as pediatricians.  The majority of respondent physicians were employed, and men consistently make more than women in the same specialty. Women have the largest representation in specialties with the lowest incomes.    Physician income was a bit lower in the northeast but higher in the northwest.  Massachusetts’ physicians report that their income is 46th in the nation.

Internists are the least satisfied in their job (47%), and the least likely to choose their specialty if they could choose again (25%), but high in the rankings of specialties where the respondent would choose medicine again (71%). Family physicians were only slightly more likely to choose the same specialty again as internists (31%), yet they were the most likely to say they would choose to go into medicine again (74%).  Pediatrician income is among the lowest of all specialties, yet they are twice as likely to say that they would choose the same specialty.

This compensation survey speaks to the challenges of attracting and retaining physicians in the core primary care specialties, especially those which deal with adults.    Internists and family physicians would go into medicine again, but they would go into subspecialties, and not do general primary care.   The high cost of health care in the US is in part due not to a shortage of primary care physicians, but also due to a surplus of specialists.  

The average pediatrician in this survey was paid $189,000 per year, the lowest amount of any of the surveyed specialties.  This is in the top 6% of all earners in the US.   The Primary care perception of being underpaid is related to the high earnings of specialists, and this will only change if there are large decreases in specialist income.  Modest increases in primary care income just won’t change this perception, and huge increases in primary care income would be hard to afford.

Recent efforts to train more physicians are not likely to lead to more practicing primary care physicians unless we are more successful at improving the lifestyle and status of those who practice primary care.   We’ll just likely have more specialists, and our costs of health care will continue to be the highest in the world.

Monday, May 18, 2015

Let’s not use bake sales to fund medical care

Today’s Managing Health Care Costs Number is 194,000

The case of Luis Lang of Fort Mill, South Carolina is a cautionary tale.

Last week, an article in the Charlotte Observer pointed out that this 49 year old diabetic refused to purchase health insurance.  His state legislators didn’t only fail to expand Medicaid, the House passed a law making Affordable Care Act implementation a crime.

Lang has developed retinal hemorrhages and needs an expensive procedure.  His visual problems have meant he is functionally disabled, so his income is below 133% of the federal poverty level and he therefore isn’t able to get any federal subsidy to purchase health insurance. 

Lang went to purchase health insurance on the Affordable Care Act (ACA) exchange, since South Carolina didn’t set up its own exchange.   

Lang learned two things: First, 2015 enrollment had closed earlier that month. And second, because his income has dried up, he earns too little to get a federal subsidy to buy a private policy.
Lang, a Republican, says he knew the act required him to get coverage, but he chose not to do so. But he thought help would be available in an emergency. He and his wife blame President Barack Obama and congressional Democrats for passing a complex and flawed bill.

So Lang has posted on “Go Fund Me,” and he’s so far raised over $21,000 toward the $30,000 he needs for surgery.

This is a terrible way to fund needed medical care.  

Mr Lang’s situation also points out some of the necessary albeit unpopular elements of the ACA.  Obamacare must not allow enrollment throughout the year, since that would encourage people to wait until they were sick to obtain insurance. This would make health insurance more expensive, increasing the rate of uninsurance and selecting for “adverse risk” among those insured through the exchange.  That’s why there is an individual mandate. 

There are other ways that the ACA could have expanded those covered by insurance, but the combination of defined enrollment periods and an individual mandate supported the private market-based approach.   The ACA also provides substantial choice, including choice to not be insured and pay a penalty, and the Supreme Court added the choice of states to refuse Medicaid expansion.

There are 194,000 South Carolinians with income below 133% of the federal poverty level who are ineligible for Medicaid or Obamacare premium support. The FPL is $24,250 for a  family of 4

Choice is good – but it’s terrible to be paying for health care with bake sales.

Tuesday, May 12, 2015

“A proverb in Norway says necessity teaches the naked woman how to knit”

Today’s Managing Health Care Costs Number is 13

The New York Times reports today that the oil extraction industry, buffeted by plummeting prices, has has been shedding workers and costs.  It reports that “once bustling road and hotels [are] now empty” in Texas, which is suffering from painful unemployment.  The article is titled “Drillers Answer Low Oil Prices With Cost-Saving Innovations”   Even facing these huge layoffs and scarce capital, the oil companies are getting more petroleum products out of the ground.

How are they doing this?   They have cut the cost of drilling a well from $4.5 million to 3.5 million , and in one oil field are extracting 1/3 more petroleum with 1/3 fewer wells.  Here’s more on cost saving techniques:

… trying out different grades of sand to blast along with water and chemicals to better loosen the hard rock deep underground and increase a well’s production, and varying the depths of wells to squeeze out even more oil.

…using new well chokes that technicians can operate remotely from a computer or even a smartphone to quickly adjust flows to maximize production without overtaxing pipelines

…new speed controls on well pumps reduced equipment repair downtime by 48 percent and helped increase the productivity of its wells by 28 percent

We all hope that decreases in hospital rate increases mandated by the Affordable Care Act will have the same impact on hospitals and physicians that the falling oil prices have had on oil and gas companies.

The Washington Post reported yesterday about Sedasys, a new Johnson and Johnson machine that could replace anesthesiologists  to sedate patients in (initially) low risk procedures.   J&J has been working on Sedasys for years, and the anesthesiology community has vehemently objected.  The FDA rejected the first application in 2000, and ultimately approved Sedasys in 2013.   Thirteen years is a long time!

We clearly need approaches to systematize “routine” care – and Sedasys could be the self driving care of the operating room.   We want to do more colonoscopies because they save lives, but we need to look hard at the expenses. It simply doesn’t make sense to pay an anesthesiologist more than the gastroenterologist for this procedure. Further, routine anesthesia in very low risk patients might be safer administered by a computer.  Experienced anesthesiologists will continue to be absolutely necessary for care of sicker patients, and will continue to play a substantial role in oversight of care delivered to those at lower risk.

We should see many low risk medical services provided via app with less professional-patient interaction in other areas too.  The time is long overdue for all young women to have access to a UTI app (One Medical has this). Young healthy women know when they have UTIs – and they shouldn’t have to go to an office at a cost that can reach $200 (including culture, though we don’t wait for the result).   Some rashes require biopsies, but most simply require a look.   You don’t need the latest iphone to get good enough resolution for a dermatologist to make a diagnosis in many cases.

Physicians are pretty worried about these type of disruptive innovations.  There’s some fidelity lost when a doctor is further away.  But the main worry is more financial than clinical.   Removing physicians from direct low-value patient interactions will leave a hole in many doctors’ schedules, and could leave them with only the difficult cases where their training and expertise is really necessary.  These easy cases make many days bearable, and have high margin in a fee for service world.

Planet Money had two podcasts on machines replacing humans late last week.  The first recounted the history of Ned Ludd and the Luddites, who took sledgehammers to the machines that took away their well-paying jobs, while making it possible for the population to have more than two sets of clothes.  The high-paying jobs in the weaving industry essentially all disappeared over a few decades.  The second recounts three examples of trying to teach machines to replace humans (folding towels, cognitive behavioral therapy for veterans with PTSD, and writing financial news stories.) 

Replacing humans with machines isn’t painless, and causes enormous social disruption.   The displacement of weavers during the Industrial Revolution couldn’t be prevented by the Luddites, and physician professional societies which will ultimately not succeed in preventing adoption of new technologies that systematize care, decrease variation, and decrease cost.   Pressure to lower costs has helped force the petroleum extraction industry adopt more cost-effective processes.  I’m guardedly optimistic that the pressures to lower medical costs will drive adoption of new technology in health care too.  

[Quote in title from NYT article referenced in first paragraph]

Friday, May 8, 2015

Medicare Overpays for Surgical Global Fees

Today’s Managing Health Care Costs Number is 92

The health policy world is looking to bundle services – and to get away from fee for service. 
But the Centers for Medicare and Medicaid Services proposed last fall to eliminate global fees for thousands of surgical procedures in 2017 and 2018 – because they believe that the global payments represent substantial overpayments, and their best chance to decrease those payments would be to do away with the bundles altogether and return to fee for service payments for postoperative and some preoperative care.   

Minor surgical procedures have a global payment time of 11 days (the day of surgery plus 10 postoperative days), while major surgical procedures have a global payment period of 92 days (One postoperative day, the day of surgery, and 90 days afterward.   The resource value unit (RVU) equivalent of these global payments was established in 1992 - and since then surgical practice has changed.  Patients who used to have 4 or 5 postoperative visits now have just 1 or 2, so the global payment should be adjusted to reflect this new reality.  

I’d argue that the CMS approach is flawed.  CMS should have rebased (devalued) the global payments, as opposed to doing away with them altogether.    However, rebasing the value of thousands of global payments is a daunting and unfunded task.

Surgeons also argued against eliminating the bundles, perhaps as vigorously as they argued against imposing these bundles in the first place in the late 1980s and early 1990s.  They said that if they get more efficient they should be able to maintain the benefit of their improved processes.   Eliminating global payments will also increase patient financial responsibility, as patients are responsible for coinsurance on followup office visits (but not on global fees).

Surgeons did argue otherwise.   The repeal of the Sustainable Growth Rate (SGR) rules that would have called for large Medicare physician pay cuts included language to overturn the CMS decision to move away from global surgical fees.

Medicare is more effective than private payers in imposing prices on providers. Even high leverage facilities must accept the Medicare fee schedule.  However, the Medicare fee schedule is subject to Congressional oversight – and lobbying has paid off in continued overpayments for surgeons in these global payments.

Fee for service is not going to go away altogether – it’s a good way to pay for many services, and many providers are not arrayed in large enough groups to accept alternative payment methods.    Medicare fee for service payment rates are important, as they drive reimbursement in Medicaid and commercial health plans.  So it’s especially important to get Medicare fee for service payment levels right.   We should provide CMS the funding and the runway to improve its fee for service payment. 

H/T to Bob Berenson of the Urban Institute, who talked about this topic at a Catalyze Payment Reform meeting earlier this week.