ProPublica’s slanted “reporting” is a public disservice

After several attempts to get the ProPublica and NPR reporters working the work comp beat to see both sides of a very complex “system”, better understand the drivers and dynamics, and correct their errors publicly, I’m giving up.

Clearly, they don’t want to hear anything that doesn’t agree with their superficial and clearly biased view of the work comp system, are quite willing to distort and misuse industry data to support their position, and have an agenda they are determined to promote, damn the facts.

The latest screed from PP continues what is by no measure “reporting”, but rather a bald-faced advocacy effort using anecdotes in an attempt to indict an entire system, to simplistically and cynically post pictures of terribly injured workers to get readers to blame the awful insurance companies, to distort and misuse data to promote their ideological position.

I’m going to elaborate on their misuse of NCCI and NAIC data to tout the high financial “returns” of the work comp system next week; for now I’ll focus on the employer direction issue.

In the original Demolition article, reporter Michael Grabell said:

In 37 states, workers can’t pick their own doctor or are restricted to a list provided by their employers. [italics added for emphasis]

That is categorically false.  I sent a detailed explanation of why and where this is false, provided the background documents from WCRI, and followed up.

Did he correct the original article?  No.

Did he update it and note the mistake?  No.

Did he admit his initial statement was wrong, even in a private message? No.

However, in his latest polemic, he states:

37 states now restrict injured workers’ ability to choose their own doctors. Green states [citing a graphic] allow employers and insurers to choose workers’ doctors, at least initially. Blue states restrict workers to doctors approved by their employer, state, or insurer [italics added for emphasis] or to those in their employers’ managed care plans.

This isn’t a minor oversight or wordsmithing; Grabell’s reporting has been an attempt to show how the work comp system is slanted in favor of employers, and his mischaracterization of WCRI reports clearly is intended to support his position.

There’s something more insidious here.  Grabell is attempting to show that somehow workers choosing their doctors is better, that those awful employers and insurers are just out to screw the worker.

In the vast majority of cases, that’s utter bullshit. (sorry, no other way to describe it). And if Grabell had actually wanted to get the story right, and not just promote his ideological position, he would have talked with employers who are doing everything they can to identify the best docs and get their injured workers to those docs.

In emails and conversations I pointed out to Grabell that the reason employers want control is to prevent injured workers from being harmed by profit-seeking physicians, and suggested he look into physician dispensing of opioids, especially in Florida; the Drobot case where hundreds of workers reportedly received horrible care; and other examples (Illinois prison guards) where profiteering physicians have gamed the system, harmed workers, and cost employers and taxpayers millions.

And I heard…crickets.

What does this mean for you?

This isn’t reporting, it is advocacy.

37 States Now Restrict Injured Worke Ability to Choose Their Doctor

Think about this…

There’s a driverless car on the way from California to the East Coast - today.  Right now.

As of now, the computerized Audi is three days into the trek; using “four short-range radars, three vision-based cameras, six lidars, a localization system, intelligent software algorithms and a full suite of Advanced Drive Assistance Systems” to navigate safely and efficiently on city streets, highways, parking lots, and all manner of other paved surfaces.

This isn’t just a “wow that’s cool” thing.

It portends huge changes in employment in this country – and others.

Traffic accidents kill about 32,000 people a year and injure over 2 million more. Property damage is in the scores of billions of dollars.

So, the computers driving vehicles don’t have to be perfect – they just have to be better than we texting, drinking, tired, distracted, angry, dumb, oh-so-human humans.  Doesn’t sound like much of a challenge for devices that crush we humans in chess, medicine, and Jeopardy

There’s no question – none at all – that automated transport will dominate within two decades – and will be common far before then.  Here’s my superficial sense for what that means.

  • millions of jobs are going away – drivers, body shop techs, spare part manufacturers, auto adjusters, claims personnel. There are 2.4 million truck drivers…
  • some jobs will be created – programmers, radar experts, robot polishers, things I can’t even conceive of.

What does this mean for you?

What’s going to happen to those newly-unemployed? Where are they going to get new jobs?  What will those jobs pay? If a worker is about to be computerized out of a job, are they going to be more likely to file a work comp claim?

Poor quality of care in work comp and the impact on SSDI and medicare

The estimable Gary Franklin MD, Medical Director of Washington Labor and Industry, weighed in with passion and precision at CWCI; Gary’s message was simple and straightforward:

Workers injured on the job get worse and worse and worse, unlike people with the same injured hurt outside employment.  Why? They become chronic pain patients.  Preventing the transition from acute and sub-acute pain management to chronic pain and the inevitable disability that accompanies chronic pain is critical; that is “secondary prevention.”

Causes of disability and more specifically causes of the decade-long increase in the duration of disability include the:

  • use of harmful treatments including spinal surgery (especially lumbar fusion) and opioids
  • creeping diagnoses that seem to evolve as claims age
  • severely damaging effects of bad providers

A key data point – a case-mix adjusted study indicates patients getting opioids for more than 7 days double their risk of disability for a year.

The takeaway is if patients receiving opioids are not showing improvements in pain and functionality, the treatment should be modified – and not with an increase in opioids.

L&I is working diligently to address the issue of secondary and tertiary disability with an eye towards intervening and re-directing claims before they head too far down the disability “curve”. There are a relatively few docs who are “bad” (perhaps some may be the high IMR filers noted in a previous post).

And that work is paying off; among L&I’s claimants, the percentage of workers receiving opioids 6-12 weeks from injury has decreased from 4.9% in 2012 to 1.1% in Q4 2013.

In contrast, a recent study by PBM Express Scripts indicated that fully have of all patients taking drugs for more than 90 days will still be on opioids in three years.

If anything Dr Franklin is even more passionate, and more compelling, when speaking about lumbar fusion and the disasters associated therewith.  Fully 64% of those getting lumbar fusion are disabled 2 years out, with many of those on SSDI. What’s scarier is California’s outcomes for spinal fusion surgery are significantly and categorically worse.

WA has been able to reduce the percentage of workers who end up on SSDI via their COHE program; clearly their focus on appropriate medical care has been the leading factor.

What does this mean for you?

Crappy medical care by crappy docs looking to make money off the easily-abused workers comp system leads to unnecessary disability, hugely damaged people and families, unknown but certainly very high societal costs.



California’s IMR program – clarity at last.

Much clarity was brought to the IMR process, costs, efficiency, and IMR filers by the next speaker, CWCI SVP and CFO Rena David.  Speaking about 2014 data, Rena noted:

  • 4.1 percent of treatment requests were “IMR eligible”
  • There were about 261,000 IMR service decisions made in 2014
  • Days from application to final determination was 134 in January 2014l that dripped to 50 in December
  • 91% of the IMR reviews upheld the UR decision, and thereby the denial was deemed as appropriate.
  • Overall, 5.4% of all treatment requests are being denied or modified

Pharmacy accounted for almost 45% of requests, with decisions upheld 92% of the time.  And;

  • 12% of IMRs were for compounds, consistent with overall figures; 98% of the time compound requests were deemed to be not medically necessary.
  • 23% of IMRs for anti-depressants overturned the UR decision, the highest percentage for any drug category.
  • Only 2 percent of anti-anxiety medication denials were overturned
  • Oftentimes, information available AFTER the original UR decision was made was used in the IMR process to overturn the UR decision.

There were ten individual providers who represented fully 15% of claims; the top two individual providers each submitted the equivalent of 10 requests a day…

A key takeaway – Rena noted that for judges and other non-medical people to be in a position to determine what medical treatments are and are not appropriate is “unfathomable”.

She also noted that “without oversight, injured workers may receive deleterious care”, citing:

  • A worker anxious abut an an epidural injection was denied propophyl
  • A complete spinal fusion was requested for a 76 year old man with evidence of arthritis.

What does this mean for you?

IMR works, and would be working a lot better if bad actors weren’t jamming the system with spurious requests for damaging, or at best inappropriate, medical services.

A quick take on the state of California work comp

One of the best places to visit on business is the SF Bay Area – great people, great wine, great food, a fun vibe, and, of course, the site of one of my favorite meetings – CWCI’s annual meeting.

The most-of-the day meeting started with State Fund Chair Vern Steiner’s quick business review, followed by CWCI Alex Swedlow’s review of the rollout and trends following the rollout of California’s most recent reform – namely SB 863.

Alex highlighted changes in Ambulatory Surgery Center fees after SB 863 implementation, perhaps the most significant change being the 29 percent decrease in amount paid.  Closely tied to the ASC cost was the notorious double-charge for surgical implants.  Once that double-charge was removed, something surprising happened.

The volume of hospital inpatient discharges has been declining overall, driven by – in large part – a rather dramatic drop in the number of discharges associated with spinal implants. CWCI estimates savings of around $60 million for 2014 alone.

Overall, spinal surgery discharges dropped by over ten percent from 2008 – 2013.

Alex also reviewed MPNs and physician networks; here are my key takeaways:

  • 81% of “first year” visits have been to network providers – and this figure has been steady since 2010.
  • That’s a big jump from about 55% back in pre-MPN days
  • However – and it’s a big “however”, is the savings attributed to care within networks in MPNs have decreased from 16% below non-network to just 3 percent.

On to a discussion of drug costs – pharma costs were 13.2% of all medical benefits in 2012, a rather dramatic increase from 6.7 percent just seven year previously.

This 13.2 percent equates to $1.2 billion in direct spend – of which 30 percent is Schedule II and III drugs – the most potent opioids.

With drug costs escalating rapidly, one has to look for answers, and one answer may be a formulary.  There’s a lot of detail here related to drugs that are within and outside formulary, how formularies are set up, and motivations thereof. That said, the potential savings are rather amazing.  To wit:

  • 96 percent of CA approved brand drugs would have been blocked by WA formulary along with 27 percent of generics
  • The total savings is somewhere north of $400 million…

Next up, a discussion of the “medical cost containment” cost change, which have climbed to $2,330 in 2012, up from $1004 in 2005.

One of the primary drivers may well be the IMR program.  Fortunately CWCI just published a research paper describing payor efficiency in complying with Division of Workers Comp UR audit standards.  This is critical, as the data show if payors are complying with requirements for content, timeliness, and correct communication regarding UR requests.

97 percent of the time payers are compliant – far above the 85 percent standard set by DWC. And this has been consistent since 2009.

Off to CWCI…

It’s the middle of conference season, with WCRI behind us, CWCI coming Thursday, and RIMS on the horizon.

David Deitz MD PhD will be giving the keynote at CWCI, and it is one I am eagerly anticipating.  David’s a keen observer of the workers compensation space, one of the nation’s leading experts on work comp medical issues, and the perfect choice to provide much-needed clarity on medical cost drivers, evidence-based medicine and the application thereof.

It is also particularly timely as there’s been much discussion of issues surrounding the Independent Medical Review process, the provision (or lack thereof) of medical records, and related issues.

I’m thinking not so much about IMR specifically, but rather the larger question of why the IMR process was put in place, how it has worked or not to date, and what the real issues are. And, of course, how IMR should be modified – if at all.

Unfortunately, there has been a good bit of confusion surrounding IMRs, and the latest WorkCompCentral reporting provides a good example of how complex and multi-faceted this is.  After an initial report that apparently indicated thousands of medical records had not been received on a timely basis, subsequent reports have revealed that many of these missing records likely pertained to cases that are closed.

I’d note that the furor created by the initial report was, while predictable, also premature. As with many “incidents”, both within and outside workers’ compensation, a full understanding of the background, contributing factors, processes involved, and facts of the case often result in a very different view than the original reports seem to provide.

Right now I’m in an airport listening to a Congressional inquiry into the alleged incident involving Secret Service agents, alcohol, and barriers in front of the White House.  Our elected officials are pillorying the head of the Secret Service for not firing/disciplining the alleged offenders, and the Secret Service head continues to try to impress upon our representatives that he can’t do that unless and until the investigative process is completed.

Sorry to see there’s a predilection to rush to judgment in a lot of different settings…

Should be an interesting week in California.

What does this mean for you?

Always good to wait until all the relevant facts are provided before coming up with conclusions…


I was wrong.

In my post last week related to ProPublica’s Demolition article, I said:

Informed sources indicated that reporters Michael Grabell of ProPublica and NPR’s Howard Berkes failed to contact WCRI before this week; a rather stunning oversight on the part of the reporters.

I was wrong, and I apologize for the error.  Turns out at least one of the reporters attended WCRI’s 2014 meeting, and they did email WCRI multiple times requesting various reports and research documents.  The error is mine and mine alone and is solely due to my misinterpretation of conversation with sources.

Speaking about ProPublica/NPR’s recent work comp reporting at last week’s WCRI conference, WCRI President and CEO Rick Victor PhD. noted it is very difficult to write a balanced article using anecdotes as the basis.

Dr Victor’s point assumes one is looking to write a balanced article.

While there are numerous examples of what I see as bias in the initial article, here’s one that makes the case. It pertains to choice of physician; the authors infer this is a “limitation of benefits.”  I disagree with that inference for multiple reasons, but let’s focus on this statement:

“In 37 states, workers can’t pick their own doctor or are restricted to a list provided by their employers.”

This sentence is factually inaccurate.

There are two reports from WCRI that speak to choice of physician; neither provides support for that statement.  It appears the authors added two different and distinct categories – states with employer initial choice, and states restricting employee choice – to arrive at the 37 number.

Considering the first category, states with employer initial choice, many states allow employees to change doctors via various mechanisms and without employer approval.  ID, ME, MI, NC, NM, UT, and VT are examples.

Re the second category, restrictions on employee choice are many and varied; certainly they aren’t limited to requiring the worker to pick from a “list provided by their employers.”

Some states allow employees to make one change on their own, others allow a judicial entity to authorize a change, others have different mechanisms.  In fact, there are only 15 states that allow the employer to “select [the initial treating] provider without limitation”; several of those allow the employee to change via various mechanisms.

I’d suggest that it would be just as inaccurate to say “35 states allow employees to choose their physician” as it was for the authors to write “In 37 states, workers can’t pick their own doctor or are restricted to a list provided by their employers.”

Put another way, it would be equally “accurate”.

I’m not splitting hairs here - this as an example of what I believe to be a bias that pervades this and other articles published to date.  There is much nuance and complexity in workers’ comp regulation, nuance that, if not supportive of the authors’ apparent bias, they apparently chose to ignore or misrepresent.

I don’t make this statement lightly. Grabell and Berkes are highly experienced, well-thought-of professional reporters.  I’m just a blogger, albeit one who happens to know a lot about work comp.

There’s an awful lot more to this, but you – and I – have work to do and don’t have time to fact-check every statement.

What does this mean for you?

If you see this differently, if you think I’m being unfair, if, as one subscriber said in an email they inadvertently sent me that I’m a “shill for the insurance industry”, I’d like to hear how and why I’m getting this wrong.

What happened while we were at WCRI wrap-up

Just digging out after WCRI; between built-up work and a desire to give you, dear reader, a break after filling your inbox late last week, MCM has been on a brief holiday.

We’re back!

While we were buried in all things work comp-related, the real world kept a-spinning. First up, what’s been going on with health reform, the costs thereof, and the impact on budgets.

A lot.  The most recent federal budget projections show a decline of around $300 billion in future costs for health insurance.

That’s huge.  Gigantic. Monumental.  Unprecedented.

In comparison, cutting NASA completely – $77 billion.  Ending Amtrak subsidies – $14 billion. Eliminating the deduction for all charitable giving? $214 billion.

There are two ways the federal government “pays” for health insurance – subsidies for folks insured via their employer; the portion of their “pay” isn’t taxed.  Second, the feds subsidize premiums for individuals buying insurance via the Exchanges on a sliding scale based to income. A more detailed discussion of the changes and impact thereof is here.

Note this does NOT include Medicare – although those projections are decreasing as well.

The latest budget estimate has Medicare costs over ten years coming in about $700 billion below 2010 projections.  

One of the reasons Medicare cost projections are declining – Between January 2012 and December 2013 there have been 150,000 fewer readmissions among Medicare patients—an 8% decline.” – hat tip to the Economist, and don’t forget to credit PPACA!

As if that’s gonna happen…

From a couple weeks back  is this news that California’s docs aren’t likely to be overwhelmed with patients due to ACA.  This from California Healthline:

  • 13.04 additional emergency department visits per week by newly insured individuals; and
  • 1.16 additional primary care visits per week by newly insured individuals.

 Back to work comp…

Good piece by PRIUM CEO Michael Gavin in WorkCompWire on work comp drug formularies; my only suggested add would be to make sure formularies aren’t rigid.  Need to allow payers, and their PBMs, to okay and reject meds based on the claimant’s diagnosis, medical treatment, other drug regimen, and other clinical factors.

Medata has promoted long-time COO Tom Herndon to president.  Tom’s not only a black belt in all things bill review and IT, he’s also a good guy, knows ops inside and out, and casts a mean dry fly.  Congratulations to Tom and his colleagues at Medata.

Enjoy hump day!

The state of the States’ work comp systems

The variation in the cost to manage work comp clams varies wildly across states – from a low of less than $4000 in Wisconsin to more than twice that in Louisiana.

And, these costs are steadily increasing at annual rates betwen 4% and 11%.

Litigation is a big cost driver in several states, accounting for more than half of expenses in CA MN FL and LA.  WCRI’s Carol Telles described the factors contributing to those differences, noting bill review, networks, and other medical cost management efforts are a factor particularly in non-fee schedule states.  Other drivers are choice of medical provider and mandated medical management tools including UR and guidelines.

Telles contrasted Texas and New Jersey, states that have quite different approaches to medical managment and the regulation thereof.  Somewhat counter-intuitively, Texas is highly regulated, while NJ is most assuredly not. NJ has no fee schedule, treatment guidelines or UR, but does have employer direction and has allowed networks since 2011.

TX has required medical guidelines and UR, a drug formulary, and eliminated “informal” networks several years ago.  Networks are highly regulated and there is a medicare-based fee schedule.

To paraphrase Ms Telles, “Medical management may reduce costs, but there is a cost associated with that effort.”  She made that statement while discussing Texas – the data indicates medical management costs have stablized at 21% of medical costs per claim since 2011 (however medical costs per claim continue to increase).

Telles provided some excellent insights into litigation rates, costs, and drivers as well; factors contributing to expenses include:

  • complexity and length of dispute resolution process
  • process for obtaining and challenging medicl opinions on MMI
  • settlement usage

Notably, defense attorney expense was the largest single component in most states; however as plaintiff attorney fees are usually included in the indemnity fees we don’t know what the plaintiff attorney expenses were.



What Maryland SHOULD be studying

Three weeks ago a group of stakeholders in Maryland decided physician dispensing wasn’t that bad [scroll down in link].

These stakeholders agreed to not do anything legislatively to address doc dispensing for another two years because their own analysis had indicated physician dispensing in MD was not changing.  Now, a lobbyist for physician dispensing “technology” firm Automated Healthcare Solutions has penned an opinion piece that can only be described as a hit job on WCRI, a highly respected research organization.

There are two related problems here.

  • It’s obvious the doc dispensers’ strategy is to try to discredit WCRI – no other reason to publish an editorial in a paper in a state that you’ve already won.
  • The stakeholders that signed the letter agreeing to forgo any legislation ignored research from Johns Hopkins University (located in Maryland) proving physician dispensing is associated with much worse patient outcomes.

I won’t dignify the lobbyist’s moronic prattling with a point-by-point rebuttal; WCRI already has in the measured, professional, and very precise way that is the hallmark of academic research. Suffice it to say that the lobbyist’s own writing shows he is even less knowledgeable about statistics, research standards, and data analytics than our Newfies are.

This guy calling out WCRI on statistical analysis is akin to me telling Blake Shelton he doesn’t know the music business.

Next, in a letter citing the Maryland Workers’ Compensation Commission, the stakeholders asserted “contrary to previous trends reported by the Workers’ Compensation Research Institute, Maryland claimants received a smaller proportion of prescription drugs dispensed directly from their physicians, as compared with prescriptions dispensed from pharmacies.”  After much review, my conclusion is this – there are differences in the methodologies used by the MWCC and WCRI – but those differences do NOT mean WCRI’s work is wrong.

First, the data collection process the stakeholders used to come up with their conclusion is not as rigorous as it could – and likely should – have been. For example, they asked multiple sources for data on physician dispensing, but failed to provide tight criteria or definitions for these sources to categorize the data. As a result, the findings are questionable because the sources may well have:

  • used different criteria to identify “physician dispensers”
  • used different definitions of “repackaged” drugs
  • differing ability to identify what entity dispensed a drug
  • differing ability to differentiate between physician-owned “pharmacies” and retail pharmacies
  • different definitions of “generic” and “branded” drugs

Second, the MWCC analysis used an entirely different methodology than WCRI, a methodology that, among other factors, included different time periods and a different set of claims.  It is NOT surprising that different data sets, different methodologies, different time lines yield different results.

On its own initiative, WCRI used the stakeholders’ methodology in an attempt to understand the discrepancy, with the following result:

When we replicate the data and methods used by the Commission on the data used in our Maryland draft study, we get 16.7 percent where the Commission reported that 15.7 percent of prescriptions were dispensed at physicians’ offices. Hence, when we use similar methods on different data sets, we get similar results.

Ignored in the lobbyist’s “editorial”, and by the stakeholders as well, is this:

In the last published WCRI study on this topic, Maryland was compared to 20 other larger than average states. We found that physician dispensing in Maryland was more frequent than in 17 of these 20 states—twice as common as in the median state, [emphasis added] and four times more frequent than in the neighboring state of Virginia. 

Rather than get into a “mine’s better than your’s” conversation, here’s what we know.

There’s no question Maryland has a very large physician dispensing problem – one that all the research indicates is likely driving worse outcomes for patients and higher costs for employers and taxpayers.  The really troubling thing here is the stakeholders know, or should have known outcomes may be significantly and adversely affected by doc-dispensed drugs, yet went along with the deal.

In conversations with stakeholders, I asked why they didn’t consider this, and got no answer.  When I pressed and asked if they were going to work with JHU’s researchers to look at outcomes, I was told they “may have to think about that.”

Think about…what?

I don’t think these stakeholders are bad people or ill-intentioned; they do have a lot on their collective plate.

I do think they have – for their own reasons, which may make sense to them – given up the fight against physician dispensing.

In so doing, they are missing out on an opportunity to help Maryland employers, taxpayers, and injured workers.

They are also empowering the dispensers in other states.

What does this mean for you?

All the research indicates physician dispensing increases disability duration, indemnity expense and medical costs.  THAT is what Maryland should study.

Note – in the interest of full disclosure, I am (as most of you already know) president of CompPharma LLC, a consortium of workers’ compensation PBMs. It’s also important for readers to know that it matters not one iota to me financially if physician dispensing increases or decreases.

It does matter to me personally as it is flat out wrong. It is bad policy that is damaging the many to enrich a very few.