|
|
State Report Cards
For Workers’ Compensation
As workers’ compensation costs continue their upward spiral, it becomes
increasingly important to identify those factors contributing to the cost increases,
especially those that may be controlled.
History has shown that there are major differences in costs from state
to state. In the past the ratio of
workers’ compensation insurance costs from state to state has been over 4 times
from the upper to the lower end.[1] These cost differences can play a major role
in the competitiveness of companies operating in these various states, and also
on decisions to expand or relocate in those states.
There are two major drivers of these workers’ compensation costs. The first is outcomes, specifically the
success within a state in preventing injuries, and when they occur, the success
in returning the injured worker to health and productive endeavor, thus
avoiding prolonged absence and medical treatment costs. The second driver of these costs is
administrative burden, sometimes referred to as the “friction” inherent in that
state’s workers’ compensation system.
“Friction” is the accumulation of rules, procedures, disputes, delays,
discretionary charges and patterns of practice, including lawsuits, that press
upon the resolution of claims.[2]
This report will focus on the first driver of costs, outcomes, and this
is the second edition of “State Report
Cards for Workers' Compensation”. The
first edition was published in March of 2003 and covered data from a single
year, 2000. This report, published in
July 2004, is based on data over a three-year period, 2000 through 2002. Due to accelerated reporting of the 2002
OSHA BLS data, the time lag before publication has been reduced. Because of the amount of data included in
this report, unlike the first edition, this edition does not include all of the
data as tables within the report, but instead includes links from within the report
(as a Microsoft Word document) to the data in spread sheet files (using
Microsoft Excel) for the detailed files providing outcomes by ICD9 code for
each state for each year. Besides
keeping this report document to a manageable size, this offers the additional
benefit of providing the raw data to users in a format that can easily be
manipulated for additional analysis.
A key requirement for production of this report was the proprietary
crosswalk program that has been developed by Work Loss Data Institute, which
converts OSHA-reported data into an ICD9 code format. This allows condition adjusted analysis and comparison among
different states. This is also a
requirement for the use of techniques to improve outcomes, such as evidence
based treatment and disability duration guidelines, since these guidelines
cannot be applied without a correct diagnosis.
WLDI developed this program for use in publishing guidelines used to
improve outcomes in workers’ compensation, including Official Disability
Guidelines, ODG Treatment in Workers’ Comp, the ACOEM
Occupational Medicine Practice Guidelines, and the CCGPP Chiropractic
Practice Guidelines. More
information on these is in Appendix C.
Although this report looks only at outcomes, history has shown that
over time, costs, as well as other system attributes, will follow these
outcomes. For example, the current
troubled state of the California workers’ compensation system could have been
predicted years earlier by reviewing the types of outcomes presented in this
report.
This WLDI report is based on the Survey of Occupational Injuries and
Illnesses from the Bureau of Labor Statistics OSHA Form 200, for 2000 and for
2001, and the OSHA Form 300[3]
for 2002, the most recent year for which complete state-by-state data is
available. The Survey of OII is a
Federal/State program in which employer reports are collected annually from
private industry establishments and processed by state agencies in cooperation
with BLS, the principal fact-finding agency for the Federal Government in the
field of labor economics and statistics.
The Survey of OII, Form 200 and the latest Form 300 also serve in part
as a foundation for Official Disability Guidelines, which is published
by Work Loss Data Institute.
This OSHA database covers all OSHA recordable cases within those states
in the program. For the year 2002 there
were 45 participating states and territories, and 8 states did not
participate. Among those states not
participating for the year 2002 were Colorado, Idaho, Mississippi, New
Hampshire, North Dakota, Ohio, Pennsylvania, and South Dakota.
This special report, “State Report Cards for Workers' Compensation”, is
unique in comparing outcomes among different states using comparable measures,
putting each state on a level playing field.
For example, there has been significant deterioration in return-to-work
outcomes for the country as whole from 2000 to 2002. Some of this deterioration has been caused by national trends
that may be out of the control of individual states, such as overall trends in
medical costs and utilization, or the aging of the workforce. There may also be the effect of changes in
the reporting mechanism, i.e., going from workdays with the OSHA Form 200 to
calendar-days with the OSHA Form 300.
Since the Report Cards are based on relative performance among states
within the same year, they represent a fair evaluation of the success or
failure of individual state workers’ compensation systems, to the extent that
these national trends are having the same relative effect on each state.
Unlike insurance company claims data, the data on which this report is
based includes outcomes from self-insured employers, as well as outcomes from
employers who have workers’ compensation insurance. This is important because the percentage of business that is
self-insured may fluctuate significantly from year to year. As insurance premiums go up, large employers
tend to self-insure, and when rates go down with the next cycle, they may
choose to go with an insurance company again.
Furthermore, there are many options for partial self-insurance. Many in the industry think the outcomes from
self-insured employers are better than insured employers. Texas represents a unique opportunity to
study this because they offer the ultimate in self-insurance, becoming a
non-subscriber, which allows employers to opt out of all the requirements of
the Texas Workers’ Compensation Commission.
A recent WLDI study found that the median disability durations
experienced by non-subscribers may be less than 20% of those reporting to TWCC.[4]
In comparing outcomes, six key variables were looked at in depth for
each state.
Of course, good workers’ compensation outcomes start with prevention;
proper attention to safety can minimize the chances of a claim ever happening
in the first place. Specifically, we
looked at the incidence of claims involving days away from work for each
state. Table A-2002 (OSHA rates &
counts) shows both OSHA Incidence Rates per 100 full-time workers, and OSHA
Counts in thousands, for 2002. Tables
A-2001 and A-2000 provide the same information for the years 2000 and
2001. These tables give rates for Total
Cases, Cases Without Lost Work Days, and Lost Work Day Cases, and within the
last category it provides rates for Total Lost Work Day Cases (including
restricted activity), and for Cases With Days Away From Work. Finally, it provides a state ranking in red
based on the incidence of Cases With Days Away From Work for each state.
The national incidence for 2002 was 1.6 cases per 100 full-time
workers, compared to 1.7 in 2001 and 1.8 in 2000, reflecting a steady decline
in the incidence of cases requiring time out of work. On a national basis, initiatives to improve safety and prevention
seem to be working. For 2002 the rate
varied from a low of 1.1 in North Carolina (NC) to highs of 3.5 in Puerto Rico
(PR) and 3.1 in West Virginia (WV). In
other words, if the “worst” states had controlled incidence as well as the
“best” states, their number of lost time cases could be cut to one-third of
what they are.
When an injury happens, many cases do not require any time off from
work and these cases place minimal burden on the system. Tables A-2002, A-2001, and A-2000 also show
OSHA Counts in thousands for each year within each state using the above
categories, and calculates a percent of total cases missing work as well as a
state ranking in red.
For the U.S. as a whole, 31% of OSHA recordable cases required time off
from work in 2002, compared to 29% in 2001 and 2000. On a national basis, we are beginning to see some deterioration
in this statistic. This percentage
varied from a low of 22% in Iowa (IA) and Utah (UT) to a high of 76% in Puerto
Rico (PR).
Since Table A-2002 also provides number of cases, it can be used to
identify the relative impact of various states. In 2002 there were a total of 1,436,200 cases requiring lost work
for the country as a whole, compared to 1,537,600 in 2001 and 1,664,000 in
2000, again reflecting improvements made in safety and prevention. The top five states in 2002 represented
about 35% of total cases in the U.S.
These states were California (CA) with 181,400 cases (12.6% of total),
New York (NY) with 96,100 cases (6.7%), Texas (TX) with 86,200 cases (6.0%),
Florida (FL) with 71,400 cases (5.0%), and Illinois (IL) with 62,600 cases
(4.4%).
When a case requires missed work, the longer the case is out the higher
the indemnity costs. Table B-2002
(disability durations) shows Days Away From Work by State for 2002, and Percent
of Total Cases by Days for 1 day, 2 days, 3 to 5 days, 6 to 10 days, 11 to 20
days, 21 to 30 days, and over 30 days, along with Median disability duration
for 2002. Tables B-2001 and B-2000 show
similar information for the years 2001 and 2000. The OSHA data from the Form 200 in 2000 and 2001 used workdays,
with an average of 5 workdays counted per calendar week. This changed beginning with the 2002 data,
because the new OSHA Form 300 uses calendar days.
For the U.S. as a whole, the median disability duration in 2002 was 7
days, compared to 6 days in 2001 and 2000.
This varied from a low of 5 days in Minnesota (MN), Virginia (VA),
Georgia (GA), Nevada (NV), Montana (MT), Florida (FL), Iowa (IA), North
Carolina (NC), and Guam (GU) to highs of 20 in Puerto Rico (PR), 12 in Texas
(TX), and 11 in California (CA). In
other words, if the “worst” states and territories had performed as well as the
“best” states, their number of lost days could be cut to one-quarter of what
they are.
A key driver of workers’ compensation costs is cases that fail to
resolve in a relatively short period of time.
The frequency of long-term injuries has a huge impact on workers’ comp
costs. When the injured worker is able to stay at work, or can return within a
few days, the average cost of an injury is far less than $1,000. But injuries
that extend beyond as little as 30 days have enormous pressure to increase in
costs. They average more than $50,000,
and they consume a vast share of money spent on injured workers.[5]
Tables B-2002, B-2001, and B-2000 also show the percentage of cases out
of work for more than 30 days for each state, and provide a ranking by
state. For the total U.S., 25.1% of
cases were out of work for 31 days or longer in 2002, compared to 22.0% in 2001
and 21.0% in 2000. This is an ominous
trend that has a significant impact on workers’ compensation costs as well as
the ability of workers to return to good health and productive endeavor. This statistic ranged from a low of 15.3% in
Minnesota (MN), to highs of 39.3% in Puerto Rico (PR), 34.6% in Texas (TX),
34.2% in California (CA), and 29.1% in New York (NY).
To investigate in depth the different variables in state-by-state workers’
compensation outcomes, it is necessary to analyze each condition within each
state. Some states may have worse
outcomes because certain conditions are more prevalent, but if we compare
outcomes for different states for the same condition, we can focus more
directly on the success or failure of the workers’ compensation systems in each
state.
The standard classification system for healthcare conditions in the
U.S. is the ICD9 diagnostic coding system. The OSHA BLS system, based on the
OSHA Form 200 and Form 300, does not use the ICD9 coding system, but Work Loss
Data Institute has developed a crosswalk program to convert the OSHA claims to
an ICD9 based system. OSHA captures
Nature of Injury, which is similar to the ICD9 coding structure, and in many
cases maps directly to the proper ICD9 code.
For example, the OSHA Nature code 1241 is “carpal tunnel syndrome”, and
it maps directly to the ICD9 code 354.0, also “carpal tunnel syndrome”. For other codes WLDI uses a combination of
the OSHA Nature Code with the OSHA Body Part code to determine ICD9 code. For example, the OSHA Nature code 021,
“sprains and strains”, when used in combination with the OSHA Body Part code 23
“back”, is used to define ICD9 code 847 “back sprains and strains”. And, when combined with the OSHA Body Part
code 231 “lumbar region” (a subset of 23 “back”), defines ICD9 code 847.2
“lumbar sprains and strains.”
Using ICD9 code, the Excel file below provides disability duration
outcome information for each condition in the U.S. for each year 2000 through
2002 in three separate worksheets.
US total ICD9 2000-2002.xls
Separate Excel files provide the same information for each of the 45
states and territories where data is available for the years 2000, 2001, and
2002. These files include a total of
133 spreadsheet file worksheets, and there is a link to each one under the
discussion for each state. With these
spreadsheets it is possible to identify the number of cases within a state for
any condition, and determine the outcomes from the cases, including the median
durations and the delayed recovery rates.
In developing State Report Cards we need to focus on a few significant
conditions. The number one condition is
back sprains and strains (ICD9 847), resulting in over 279,688 cases in the
U.S. with lost workdays in the year 2002.
In addition to being the most common condition, this is also a condition
with a great deal of variability in length of disability and utilization of
medical services. Back strain is a
condition for which there are many commonly used treatment modalities, many of
which are not supported by the medical evidence.[6]
Table C-2002 (back sprains), shows Days Away From Work by State for
Back Sprains & Strains (ICD9 847), for the year 2002, and compares state by
state outcomes for back sprains and strains, and a state ranking is provided
based on median disability duration, plus a ranking for delayed recovery rate
(percent of cases out more than 30 days).
The overall U.S. trend for back sprains mirrors the trend for all
workers’ comp conditions together, in part because it is the most common
diagnosis. The median in 2002 was 7
days, compared to 6 in 2001 and 2000.
The delayed recovery rate was 23.7% in 2002, 22.0% in 2001, and 20.9% in
2000, showing a steady worsening. The
five states with the best outcomes for back strain in 2002 (in order with the
“best” listed first) were Minnesota (MN), Wisconsin (WI), Oregon (OR), Indiana
(IN), and Virginia (VA). The seven
worst states (in order with the “worst” listed first) were Texas (TX),
California (CA), Louisiana (LA), Delaware (DE), West Virginia (WV), New York
(NY), and Illinois (IL).
While not as common as low back sprains, carpal tunnel syndrome also
has a significant impact on workers’ compensation costs. According to the analysis of cases by ICD9
code, carpal tunnel syndrome is ranked 9th in frequency, but has the
longest disability durations among conditions in the top 10. Furthermore, carpal tunnel syndrome is also
seen as a proxy for identification of a successful disability management
program. There is considerable
variability in outcomes for carpal tunnel cases. Appropriate, timely management of diagnostic testing, treatment
and vocational rehabilitation is significant.
There are many symptomatic cases
that, when using evidence based diagnostic methods, would be determined
not to have the condition at all.[7] In addition, evidence based treatment
guidelines dictate a treatment and return-to-work plan that would result in
less lost work-days than the average for even the most severe cases.[8]
Table D-2002 (carpal tunnel) shows Days Away From Work by State for
Carpal Tunnel Syndrome (ICD9 354.0) for the year 2002, compares state-by-state
outcomes, and provides state rankings based on median disability duration and
delayed recovery rate (percent of cases out more than 30 days). Tables D-2001 and D-2000 provide the same
information for the years 2001 and 2000.
Nationally the median was 30 days in 2002, 25 days in 2001, and 27 days
in 2000. The outlier percentage was
49.1% in 2002, 44.5% in 2001, and 44.9% in 2000. The five best performing states for carpal tunnel syndrome were
Alabama (AL), Virginia (VA), Iowa (IA), Alaska (AK), and Missouri (MO), and the
five worst states were Nevada (NV), Vermont (VT), California (CA), Delaware
(DE), and New York (NY).
Please note that outcomes for these conditions are also included in the
overall outcomes that are also used in grading the different states, so in a
sense we are “double counting” for these conditions. This is deliberate. As a
developer of evidence based guidelines, Work Loss Data Institute has analyzed
treatment and return to work outcomes for every condition seen in workers’
comp. For many conditions, such as broken
bones, burns, and minor cuts and bruises, there is little variability in
treatment and return-to-work because the medical decision-making is fairly
clear, and therefore there is little abuse of the system. On the other hand, back pain and carpal tunnel
syndrome are not like this at all, and they have been responsible for much of
the trend for worsening outcomes as well as abuses. So while there is double counting in the grading system used in
this report, this results in increased focus on the drivers of successful
outcomes in workers’ comp, and early identification of state systems that are
on their way “up” or “down”.
In preparing an overall report card for each state we looked at each of
the above six key outcomes measures, and the state’s rank in that outcome. A low rank is good and a high one is
poor. (A state ranking number 1 is the
best in that category.) Then we
combined the six rankings for each state into an overall ranking, and assigned
5 grades (A, B, C, D, & F) based on where the overall ranking fell. These grades are summarized in Table E-2002
(state rankings), which shows State Rankings by Measure with Final Grade for
2002, and Tables E-2001 and E-2000 for the years 2001 and 2000. Table F-2002 (increases) shows Ranking
Increases (2002 over 2000) in the Overall Ranking. Also see the U.S. Map Showing Grades by State. The results are described in a chapter for
each state.
A table and supporting spreadsheet analyzes the associations of specific managed care programs to outcomes by comparing the average ranking for those states with the program with the average ranking for those states without the program. It should be noted that this is a retrospective, observational, non-controlled analysis, and therefore its predictive ability is not the same as if it were a prospective controlled study. In other words, there could be many other factors that have not been controlled that are causing these differences, and users should be careful in interpreting these results.
[home]
Send mail to our webmaster
with questions or comments about this web site.
Copyright © 2004 Official Disability Guidelines
Last modified: June 27, 2004
[1] See Appendix A of this report for a table with workers' compensation insurance costs in 2000, based on data from Actuarial and Technical Solutions, Inc., and a study by the Oregon Dept. of Consumer & Business Services. This table shows a high of $7.20 per $100 of annual payroll for California (the highest cost state) to a low of $1.62 per $100 of payroll for Virginia (the lowest cost state).
[2] Rousmaniere PF, Denniston PL, “Spiraling workers comp insurance costs: a disturbing trend?”, Risk & Insurance Management, March, 2003
[3] New record keeping rules went into effect on January 1, 2002. OSHA's new record keeping rules were issued on the last day of the Clinton administration. However, the Bush administration put a hold on those rules pending further review. On June 29, 2001, Secretary of Labor Elaine L. Chao announced that the new rules on record keeping would go into effect as originally scheduled on January 1, 2002, with a few minor exceptions. One of the major changes was to the OSHA Form 300 from the Form 200, and now "Days away" and "days restricted or transferred" are counted on a calendar basis rather than scheduled workdays. See Appendix B for more information.
[4] Since Texas is unique in having a variation on self-insured called non subscribers, these employers can completely ignore any state WC "friction", and employ other options, such as the ability to "direct care" and use treatment guidelines. WLDI compared the outcomes from all Texas employers (using BLS OSHA data that comes from OSHA logs and therefore covers all employers) with outcomes from all cases reported to TWCC (which do not include non subscribers). For the year 2000 the data show:
BLS for TX TWCC
median duration 9 days 50 days
% 31+ days 28% well over 50%
Another significant finding seems to be that state variations in workers comp costs are not necessarily relevant to a self-insured employer with a well-managed program. The down side of this is that, with adverse selection, the workers' compensation insurance costs (for non self-insured's) in the poorly performing states will accelerate their upward spiral. (WLDI Custom Report to the Texas ROC, 12-10-02.)
[5] Rousmaniere PF, Denniston PL, “Spiraling workers comp insurance costs: a disturbing trend?”, Risk & Insurance Management, March, 2003
[6] ODG Treatment in Workers’ Comp, Low Back Problems, “The strongest medical evidence regarding potential therapies for low back pain indicates that having the patient return to normal activities has the best long term outcome. There are many therapies, both invasive and noninvasive, whose purpose is to cure the pain, but there is no strong evidence that they accomplish this as successfully as therapies that focus on restoring functional ability, without focusing on the pain.”, June, 2004, www.odgtreatment.com.
[7] ODG Treatment in Workers’ Comp, Carpal Tunnel Syndrome, “After ruling out ‘red flags’, there is considerable value in, first, making sure the diagnosis of CTS is correct, and second, isolating severe CTS. Symptoms of pain, numbness, and tingling in the hands are common in the general population, but based on studies, only about 1 in 5 symptomatic subjects would be expected to have CTS based on clinical examination and electrophysiologic testing. Nerve conduction studies are the ‘gold standard’ in CTS diagnosis, but other tests may also be useful. Identification and treatment of underlying conditions may resolve symptoms of CTS: pregnancy (usually resolves within 6 weeks after delivery), hypothyroidism, diabetes mellitus, rheumatoid arthritis.”, June 2004, www.odgtreatment.com.
[8] ODG Treatment in Workers’ Comp, Carpal Tunnel Syndrome, “Carpal tunnel release is well supported, both open and endoscopic (with proper surgeon training), assuming the diagnosis of CTS is correct. (Unfortunately, many CTR surgeries are performed on patients without a correct diagnosis of CTS, and these surgeries do not have successful outcomes.) Surgical treatment of properly diagnosed moderate or severe CTS may result in faster recovery than prolonged conservative treatment. A recent Cochrane meta-analysis concluded that surgical treatment of carpal tunnel syndrome seems to be better than splinting. ODG Return-To-Work Pathways: Endoscopic surgery, modified work: 3-5 days; Endoscopic surgery, regular work, non-dominant arm: 28 days”, June 2004, www.odgtreatment.com.