|
AMA Guides Impairment Ratings Erroneously Elevated – Best
Practices Approach is Expert Review of All
Impairment Ratings
Impairment ratings are a multi-billion dollar
cost driver in workers’ compensation systems
throughout the country. These systems rely on
claims professionals, attorneys and triers of
fact to decipher the accuracy of impairment
ratings for purposes of distributing benefits in
their system. The methodology outlined in the
AMA Guides to the Evaluation of Permanent
Impairment requires a level of clinical
experience and specialized training to fully
understand appropriate calculation of
impairment. This is not typically within the
scope of expertise of those professionals who
are tasked with making decisions relative to
impairment. This becomes challenging in an
environment where there is high error rate among
evaluators providing impairment ratings.
Studies have revealed the vast
majority of ratings are incorrect, typically
overrated and the errors are often
undetected. Impairment Resources, LLC has
performed research studies at the request of the
American Medical Association, Workers’
Compensation Insurance Rating Bureau
(California), State of Colorado, State of
Florida, Sedgwick CMS and other entities to
determine impairment ratings based on different
Editions of the Guides and to determine accuracy
of ratings. A 2010 study by Impairment
Resources, which was published in the March -
April 2010 issue of the Guides Newsletter,
provides insights to the scope of the problem.
(Reference: Brigham CR, Uejo C, Dilbeck L,
Uehlein WF. Impairment Ratings: Observations
Based on Review of More Than 6000 Cases. The
Guides Newsletter. March /April 2010.)
Study Process
The 2010 Impairment Resources’
study was based on impairment ratings submitted
for expert review and, therefore, is not a
random sample of all impairment ratings.
However, the data provides insight into some of
the challenges seen with Guides ratings. In the
review process, if the original rating was
judged to be incorrect by the expert reviewer
and if there was adequate clinical information
to rate impairment, then the case was re-rated
using the Guides criteria and the data provided.
Most Ratings Incorrect
and Result in Over-Rating
Impairment Resources’ review of
6,233 impairment ratings between July 2006 and
January 2010 resulted in a 78% disagreement
rate. 97% of the WPI ratings were obtained by
means of the Fifth Edition, and 81% of all cases
reviewed had been rated by California
physicians. Of the 5082 Fifth Edition WPI
ratings, the average rating was 18.3% WPI, and
the revised ratings, as opined by an AMA Guides
expert reviewer, averaged 8.3% WPI. This
reflects a difference of 10.0% WPI. The
relationship between revised and original
ratings is illustrated in Figure 1; if there
were interrater reliability, the ratings would
appear on a diagonal, reflecting a one-to-one
relationship.
Figure 1.
Comparison of Average Original Rating with
Average Revised Rating
A previous study,
completed in 2005, evaluated 2100 cases referred
for impairment rating review and found that 80%
of ratings resulted in different outcomes when
reviewed by an expert reviewer. (Reference:
Brigham CR. Erroneous impairment ratings. The
Guides Newsletter. May/June 2006; 59:1,
July/August 2006; 60:1.) These rating errors are
costly to stakeholders and have been associated
with overpayments exceeding one million dollars
for every one hundred erroneously rated claims.
For employees, such errors send the wrong
message – that they are severely disabled versus
the goal of care to become enabled.
Higher Ratings
More Likely to be Erroneous
Analysis by impairment
ratings, categorizing by the initial rating,
showed that higher original impairment ratings
were more likely to be erroneous. Table 1
presents these findings in California for WPI
ratings. As the value of the original rating
increased, there was increasing error.
Table 1. Case
Observations by Value of Original Rating,
California
| |
Average WPI % |
|
|
| Criteria |
Original |
Revised |
Difference |
Error Rate % |
% of Cases |
| Original <5% WPI |
2.1 |
1.8 |
0.3 |
38 |
12 |
| Original ≥5% and <10% WPI |
6.7 |
3.3 |
3.4 |
73 |
19 |
| Original ≥ 10% and < 15% WPI |
12.0 |
5.4 |
6.6 |
82 |
16 |
| Original ≥ 15% and < 20% WPI |
16.9 |
6.3 |
10.6 |
95 |
13 |
| Original ≥20% WPI |
34.0 |
14.8 |
19.2 |
94 |
39 |
Insights from
Study
The results of this 2010
study provide several insights, including:
|
There are significant problems with
interrater reliability with Fifth
Edition ratings. |
|
The majority of the ratings reviewed
were performed incorrectly and resulted
in ratings that averaged more than twice
what was appropriate. |
|
Problems are more common in
jurisdictions where the impairment
rating plays a greater role in the
determination of benefits. |
|
Good interrater reliability can be
achieved by ensuring that impairment
ratings are performed by knowledgeable,
skilled, and unbiased examiners, and
that these ratings are independently
reviewed by personnel who are equally
skilled and knowledgeable in impairment
rating. The data obtained may be useful
for total quality improvement. |
Continue to
National Presentations on Driving Accurate
Impairment Ratings
|
