Barry Lipton PhD did a quick review of a few ways work comp and group health differ – and how they are sometimes comparable.

The biggest difference is that workers’ comp is very focused on return to work, and more broadly, we care more about functionality.  In group health, not so much.

Differences – other than the obvious e.g. WC=ortho and trauma; group health= everything

• price differences are 12% higher for work comp than group health,
• while utilization is 60% higher for workers’ comp
• so total costs are 77% higher.

This differs by state, with Alabama, Missouri, and Virginia showing the biggest cost difference compared to national averages, and Colorado and South Dakota with the lowest cost compared to that average.

It also differs by type of service – not surprisingly physical medicine is used much more in workers’ comp, driven almost entirely by more utilization.

That’s not surprising, as comp conditions are predominantly musculoskeletal injuries and focused on return to functionality, while group health deals with many more conditions and RTW is irrelevant.

There’s also a big difference in the cost of MRIs…

You can see that Medicare pays way less than work comp (WC is blue, Medicare is green-ish). You can also see the impact of Medicare’s change in reimbursement in 2013; it started to really impact workers comp in the years after 2014 as regulators adopted/factored in/used Medicare’s rate for reimbursement in their state.

What does this mean for you?

Price + Utilization = Cost. But one has to factor in RTW in comp, a focus that is nonexistent in group health.

 

NCCI Chief Actuary Kathy Antonello’s presentation on the state of the industry has just too much information for a single post – so here’s three key details.

Reserves

Private carriers are over-reserved. That means there’s several billion dollars of excess cash on carrier books. I’d note that this ASSUMES the projections are accurate, and that losses for already-incurred claims don’t get worse (or, in insurance-speak, develop upward).

Frequency

Lost time claim frequency declined by 1 percent last year – significantly less than we’ve seen over the decade. My take is this is related to several factors.

• Employers aren’t focusing on work comp issues – e.g. safety and loss prevention – as premiums are so low that they aren’t a problem.
• Hiring standards have been relaxed as we’re at full employment
• More overtime is being worked, leading to higher injuries due to tired workers

Severity

Medical costs went up a mere 1 percent in 2018, continuing a trend of relatively low increases that’s persisted since 2008 (the jump of 4.1% in 2017 looks like an aberration).

What does this mean for you?

Workers’ comp is not a problem for employers – which means it will get little attention from legislators.

When buyers aren’t experiencing pain, they have little reason to buy.

This year’s Annual Issues Symposium was the best I’ve attended – and I’ve been to 20 or so.

The hotel and conference center were excellent – great service, everything was right on site, food was very good, all around best conference site experience in memory.

The opening day’s content was rich and mostly very well done. Kathy Antonello’s discussion of results continues to improve. I would have liked a bit deeper dive into cost drivers, but that’s a very minor quibble; you can’t cover everything in an hour. Graphic presentation was helpful, and Kathy is clearly comfortable on stage and enjoys presenting.

For me, after Kathy’s State of the Line the highlight was the discussion of AI and human decision making. Jim Guszcza of Deloitte was brilliant, laying out a compelling case for the joint use of both AI and humans in decision making.

A discussion of TRIA renewal was – I’m sure – of keen interest to many, but the speaker’s impact suffered a bit as he read his talk.  David Priebe of Guy Carpenter is clearly expert in all things TRIA and knows his stuff.

Moments into David Deitz’ physician panel, the hotel lost power and all went dark.  Staff responded quickly, using social media to keep all of us informed – they handled the unexpected with aplomb.

The physician panel is up there somewhere…

When things got started, I had the sense the blackout was a metaphor for payers’ views of treating providers – there’s little visibility into what docs have to deal with when serving work comp patients.

In fact, the physicians had pointed comments about the problems docs face trying to do the right thing, many of which are caused by well-intentioned but ultimately dumb “requirements”. Takeaway – if we want good care, we need to make sure the people delivering it like to work with us. We have a long way to go to make the occ docs who care for our patients true partners.

Barry Lipton quickly ran thru three research foci, I particularly liked Barry’s insights into ways work comp and group health are different.

Alas I won’t be attending the second day; the boy’s annual mountain bike trip conflicted.  It’s off to Moab, Utah, for four days of back-to-boyhood.

What does this mean for you?

This is a must go. Sign up early so you don’t get locked out of 2020.

What passes for predictive modeling today is like Google maps, except the app tells you when to turn a half-hour after you’ve passed the intersection.

Adjuster finding out a claimant had spinal surgery 6 weeks ago…

By the time adjusters figure out a claim has gone off the tracks, it’s often too late to do anything but increase reserves. That’s because there’s no real-time monitoring, no way to clearly and definitively identify when – exactly – that happens.

A promising approach is in the works at Gallagher Bassett.

Building off research conducted by Johns Hopkins, GB has developed a tool that enables real-time monitoring of medical services delivered to its claimants. Using a proprietary platform, alerts are sent when patients’ quality of care is headed in the wrong direction.

The trigger is inappropriate medical treatments. GB matches medical bill data with evidence-based treatment guidelines, with each service, procedure, or medication individually assessed. As the number of inappropriate treatments increases, alarm bells ring.

Of course, that doesn’t mean all treatment that is non-compliant is inappropriate. However, much is, and there’s a clear – and quite strong – correlation between bad medical care and lousy claim outcomes.

Those are clinical words and hide the real import of GB’s approach. Getting claims back on track means patients get better faster, AND the risk of bad outcomes from inappropriate surgeries, injections, drugs, and tests decreases.

What does this mean for you?

There’s lots of data out there – and far too little smart use of data. This is promising indeed.

 

 

 

Our last Bill Review Survey was in 2012; things have changed a lot since then – and mostly for the better.

Overall Industry rating

Six years ago we asked the 24 respondents to rate their overall impression of vendors on our standard 1 – 5 scale: the result as reported in 2012:

In what might be best described as a wake-up call for vendors and application providers, there were no ratings higher than a 3.4, with the average a 3.07 on the 1-5 scale (among respondents who knew of the specific vendor).

Put another way, the people who use BR services view BR providers as generally mediocre/adequate/acceptable.

Today, the average rating was just under 3.3, a small but significant improvement.

However, there were significant shifts for individual companies.

In 2012 the three top vendors – Medata, Mitchell, and StrataCare were all in a statistical dead heat, with Xerox/ACS rated a full point lower. (Xerox/ACS bought Stratacare, converted most/all of its users to the Strataware platform, and now operates under the Conduent brand.)

No longer.

This year, we asked the 30 respondents to provide ratings of ten bill review vendors on four different metrics;

• overall perception
• handling the basics of bill review
• customer service and implementation
• innovation and forward thinking

(All respondents did not give ratings for all categories or companies; the findings below are averages across those respondents who gave a rating for that specific company).

Mitchell (3.75) and Medata (3.71) were in a statistical dead heat for top honors in Handling the basics.

For Customer service and implementation, Medata (3.94) was the clear leader with Mitchell second (3.54).

Medata held serve for Innovation (3.91) and Mitchell (3.5) again took second place honors.

Here’s where it gets interesting. Medata and Mitchell were essentially tied for Overall perception by respondents who rated them.

But across ALL respondents, Medata’s Overall perception rating was 3.4, with Mitchell at 3.08. While Medata has the lowest market share of the top three application providers, it was tied for highest name recognition among our 30 respondents, and thus garnered the top spot in Overall perception amongst ALL respondents.

Here’s where we link back to last week’s posts on the critical importance of customer service;  the data shows Customer Service has the highest statistical correlation with overall perception.

Note – Corvel shared the top spot in name recognition, but got the lowest rating for customer service – and was essentially tied for the lowest score for Overall perception.

What does this mean for you?

Customer service wins.

Note – I’ve received several anonymous comments/emails lately.  I’d remind commenters that anonymous comments on MCM posts are ignored, as are comments with fictitious email addresses.

You know who I am. I and my readers need to know who you are.

 

 

 

 

 

Two important data points hit the news this week, both worthy of your attention.

First, BLS data indicates private industry employers reported 47,000 fewer occupational injuries and illnesses in 2017 compared to the previous year, a decrease of about 1.7 percent.

The rate, or frequency of total reportable cases declined by 0.1 cases per 100 FTE. As we’ve reported in the past, BLS data does not precisely mirror work comp claims – but it’s very close.

(Note this does NOT include public sector employer data)

So, occupational injuries and illnesses, along with work comp claims frequency, both dropped last year.

Next, insurer CNA CEO Dino Robusto said this in CNA’s earnings call:

we’ve been seeing negative sort of mid single-digit frequency trends over the past several quarters, which is less negative than a year ago. Now, while we’ve seen some pockets, where frequency has increased, the negative frequency trend overall is still favorable to our long run trend assumptions, because we did not lower our long run frequency assumptions despite the actual frequency consistently more negative than our assumption. [emphasis added]

(thanks to SeekingAlpha for the transcript)

Recall the Hartford has seen an uptick in claims frequency of late, one their CEO opined is not unique to his company.

I checked on other major workers’ comp insurers, including the Travelers, and  AIG and did not find anything useful pertaining to frequency or claim counts.

So, what does this mean for you?

Watch your claim frequency carefully, especially in geographic areas and business sectors where hiring is very tough. It could be you’ll see an uptick in claims, due probably to compromises in hiring due to the tight labor market.

Well, some posts get a life of their own, and so it is with this discussion of claims frequency and claims counts. After much discussion with colleagues and several back-and-forth emails with WCRI CEO John Ruser PhD about the correlation of OSHA recordable data and work comp claims and why both are declining, I decided the best way to get this to you, dear reader, is via an interview. So, read on.

MCM – I believe that you were responsible for the BLS OSHA-recordable injury data for years. What are a couple key points readers should know about the OSHA-recordable reports?

Dr Ruser – Yes, I was BLS Assistant Commissioner for Occupational Safety and Health Statistics for over 5 years and was a researcher of the BLS OSHA data for many years before that.

While there has been some controversy about the completeness of reporting in the OSHA recordkeeping system (see below), the BLS OSHA-recordable injury rate data are extremely valuable for several reasons.  They are very detailed by State, by industry, by establishment size and by worker characteristics, so that are an important benchmarking tool for risk managers and others seeking to compare their company’s injury rates against their peers.  From the perspective of focusing injury risk reduction efforts, they are important in identifying those groups of workers at higher risk of injury and they are used by OSHA to identify high-risk industries for inspections.  And, with their long relatively-consistent time series and detail, they are a valuable tool for researchers seeking to understand factors that contribute to workplace injuries.

MCM – Where does BLS get the data for the OSHA-recordable reports?

Dr Ruser – BLS’s estimates of OSHA-recordable injuries are based on a very large annual survey of about a quarter-million establishments (that is, specific locations of a company or organization) called the Survey of Occupational Injuries and Illnesses (SOII).  The SOII contains employer-reported data drawn from the OSHA logs that establishments keep throughout the year.  SOII covers non-fatal occupational injuries and those illnesses that can be directly linked to a workplace.  A separate BLS program, the Census of Fatal Occupational Injuries, uses multiple data sources, such as death certificates, OSHA reports and many other sources, to track workplace deaths due to injury.

MCM – there’s been questions about the decline in reportables over the years. Can you comment on these questions?

Dr Ruser – Some skeptics of the declines in the BLS OSHA-recordable injury rates attribute these declines to changes in OSHA-recordkeeping rules and practices or tightening in WC compensability rules, meaning the declines in injury rates are at least in part an artifact of reporting.  External research supported by BLS and other non-BLS-supported research does suggest that the number of injuries captured in SOII undercounts the true number of OSHA-recordable injuries.  (BLS has a very complete webpage on SOII data quality research that you can access here: https://www.bls.gov/iif/soii-bibliography.htm)

But, while the numbers (levels) of injuries and claims may be undercounted, the issue for the observed declines (trends) in injuries (and WC claims) is whether underreporting has grown.  There is little direct research on this.  A study by Washington State comparing SOII data to WC claims found that during the first five years of the study period (2002 – 2006), underreporting decreased, while it increased from 2007 to 2011.  Importantly, the Washington State researchers concluded that the total estimated actual number of SOII-eligible WC time loss injuries decreased over the ten year span, meaning there were real declines in injuries (and some underreporting too).

The Washington State study was excellent, but it focused on one state and a relatively short time span, which included a great recession during the second half of the study period when underreporting was identified.   Another approach to validating the time trends is to compare to other data that should not be susceptible to the concerns raised about reporting.

MCM – what analyses did you do to explore that issue?

Dr Ruser – I compared the SOII data with data from other sources.  First, I looked at how the US injury rate for 3 or more days away from work tracks with the NCCI indemnity claiming rate.  The declines in these two data series track extremely closely.  So, while the OSHA recordkeeping system is technically independent of workers’ compensation, the BLS injury data and the NCCI claims data are telling the same story and the BLS data can be used to try to identify factors associated with the decline in the NCCI WC claiming rate.

Regarding whether the BLS injury rate decline is real, I created an index of the OSHA-recordable case rate for cases with 3 or more days away from work and lined it up with a similar index for 15 EU countries for injuries with 4 or more days away from work (the series most comparable to the US data).  The chart that is attached shows how similar the trends are in the US and in the EU.  The index was set to 100 for injury rate values in 1998 and the other values in the chart are injury rates relative to 1998.  As of 2014, the US injury rate was 54 percent of its value in 1998, while the EU injury rate in 2014 was 49 percent of its value in 1998.

MCM – what does this mean (for our readers)?:

Dr Ruser- The remarkably similar trends in the US and EU data suggest that we need to look beyond US-specific explanations (such as OSHA-recordkeeping rules or WC compensability rules) to understand what is responsible for the long-run aggregate declines in injury rates and WC claims rates.  While there may be some changes in reporting at least over part of the past quarter century, the good news, I believe, is that there has been a remarkable improvement in safety and this improvement is seen in most industries and in many developed countries.