To The Editor:
I read with interest "Long-Term Disability and Return to Work
Among Patients Who Have a Herniated Lumbar Disc: The Effect of Disability Compensation"
(82-A: 4-15, Jan. 2000), by Atlas et al.
Were the authors able to distinguish between temporary and permanent
state Workers' Compensation disability in assessing the patients' Workers'
Compensation status at the four-year follow-up? The study indicates
that the authors asked more detailed questions about Workers' Compensation
status if a patient was on disability compensation at the follow-up
assessment.
I am unfamiliar with the Workers' Compensation system in Maine.
In New Jersey, temporary disability compensation is awarded until
curative treatment is concluded. Thereafter, a permanent award of
disability compensation may be made if the patient meets the statutory
requirements. If a patient meets the requirements, a total fixed
amount of permanent disability compensation is determined and paid
out on a weekly basis with use of a formula set by statute. For
example, an award of total permanent disability in New Jersey results
in weekly payments for 600 weeks until the entire permanent disability
award is paid. In New Jersey, patients can be 100 percent disabled
under the state Workers' Compensation system, be receiving weekly
permanent disability payments for over ten years after the permanency
award, and be working full-time. Using a patient's Workers' Compensation
status as an outcome measure may merely represent continued payments
for permanent disability compensation and not reflect ongoing disability.
Was Workers' Compensation status at follow-up a marker for continued
temporary disability and continued difficulty working?
Were the authors able to distinguish if a patient had ongoing
litigation at each assessment point over the course of the four-year
follow-up? It is not unreasonable to expect, human nature being what
it is, that individuals anticipating a permanency award as a result
of the herniated disc may exaggerate or inflate their symptoms in
the expectation that it may result in a higher monetary award. In
New Jersey, a patient's complaints, restrictions, and limitations
at work and at home are given great weight in the determination
of permanent disability and can often outweigh any objective evidence
of impairment. Litigation typically can drag on for a year or more
after the conclusion of treatment. After the conclusion of litigation,
a patient's complaints, limitations, and restrictions may abate.
Were the authors able to determine the status of any ongoing litigation
over the course of follow-up and whether complaints or symptoms
changed after the conclusion of litigation?
Did the authors create a multivariate model to determine whether
the likelihood of a patient working at the time of the four-year
follow-up was in any way related to smoking at the time of the baseline
assessment? There are data suggesting that smoking may be a risk
factor for back pain1,3. Smoking
may also have an effect on the outcome of treatment. With regard
to spine fusion, there is controversial evidence to suggest that the
pseudarthrosis rate may be related to smoking status2,4,6,7. If there is a relationship
between smoking and outcome in patients with a herniated disc, it
provides another piece of evidence that smoking is detrimental to
treatment outcomes. There are interesting societal implications
in addition to obvious treatment implications if this is true. If smoking
predisposes an employee to back pain, and if smoking affects the
outcome of treatment for a disc herniation, should the employer
have the right to discriminate against employees who smoke by not
hiring them, especially for strenuous work?
The study indicates that individuals on compensation at baseline
will have a higher number of residual complaints at the four-year
follow-up. Eighty-seven percent of the individuals who were receiving
Workers' Compensation at baseline returned to work at least once
in the four-year follow-up period. The individuals on Workers' Compensation
at the baseline assessment worked at more physically intensive jobs than
the noncompensation group. Was the higher number of residual complaints
in the Workers' Compensation group due to the more strenuous work?
This question has important implications.
The type of study that Atlas et al. performed is sometimes dismissed
by critics claiming that the increased number of residual subjective
complaints in the Workers' Compensation group is due to return to
more strenuous labor, resulting in higher residual symptomatology.
A study of total knee-replacement patients on Workers' Compensation
showed that the non-Workers' Compensation patients returned to more
strenuous levels of activity and had lower levels of residual complaints
and limitations than the Workers' Compensation patients5. Is the group of baseline non-Workers'
Compensation patients who returned to strenuous work large enough
to draw a comparison with the Workers' Compensation group?
The authors are to be congratulated for their fine work in this
area.
Kenneth C. Peacock, M.D.
Laurel Orthopedic Evaluations
Moorestown Office Center, Suite 305
110 Marter Avenue
Moorestown, New Jersey 08057
S. J. Atlas and R. B. Keller reply:
We appreciate Dr. Peacock's thoughtful questions and comments.
We are unable to provide direct answers to many of his questions
because of numerous difficulties encountered when studying these
patients. Outcomes for those receiving Workers' Compensation may
be available in the distinct disability and medical systems that
provide services to these patients. Reports in the disability literature
are generally derived from analyses of large private or state Workers'
Compensation data systems, and they may have detailed information
about work sites, job specifications, disability days, medical costs,
and indemnity payments. However, they often lack detailed clinical
information about physical examination and radiographic findings; medical
treatments; and patient-reported symptoms, functional status, and
satisfaction with care. Additionally, these studies are usually
limited to patients receiving Workers' Compensation, without a comparison
group.
On the other hand, studies in the medical literature generally
focus on treatments and outcomes and contain much more detailed
clinical information, but they often lack specific information about
work and disability. Few studies have examined a broad range of
clinical, work, and disability outcomes. When work and disability outcomes
are considered, they are often included as a multidimensional, unvalidated
outcome measure. In our study, we attempted to collect a broad range
of outcomes with use of validated measures. In order to create a
follow-up questionnaire that would not be too long (and would lower
response rates), we had to compromise between breadth and detail
in choosing questions. This accounts for our inability to answer all
of Dr. Peacock's questions and reinforces our view that additional
research is needed.
As Dr. Peacock notes, Workers' Compensation systems are state-based,
and laws governing eligibility and the type and extent of payments
vary considerably. Even within individual states, Workers' Compensation
legislation has been changing at a rapid pace, mainly in response
to escalating costs during the 1980s. In Maine, major Workers' Compensation
legislation was enacted in the early 1990s during the time frame
of our study, making it difficult to address the questions about
the nature of disability eligibility and payments that Dr. Peacock
comments upon.
Information about work and disability at baseline and follow-up
was reported by the patient. State Workers' Compensation records
were used to verify Workers' Compensation status at baseline only
when there was uncertainty between the patient and physician responses
(16 percent of patients). We did not have access to detailed medical
claims records or employer files, and we did not ask the patient
to differentiate between temporary and permanent disability. We
did not inquire about ongoing litigation or collect detailed work
information in follow-up surveys. Future studies would benefit from patient-reported
and employer and insurer-provided data. However, this could adversely
influence patient participation. Our informed-consent process stated
that the patient and physician information would be kept strictly
confidential and would not be collected from or shared with the
employer or the insurance company. We believe that this assurance
of confidentiality helped to foster patient participation, high
follow-up rates, and unbiased responses.
Though the Workers' Compensation system may create financial
incentives not to work, we believe that patients receiving disability
compensation at four years were symptomatically and functionally
impaired. Regardless of the Workers' Compensation status at baseline,
patients receiving disability compensation had worse symptoms and
functional status and were less satisfied with their current state
than patients who were not receiving disability compensation. Additionally,
patients receiving disability compensation were less likely to be working
than those not receiving disability compensation (51 versus 91 percent;
p < 0.001). Among patients who were receiving Workers' Compensation
at baseline and receiving any disability compensation at four years,
53 percent (nineteen of thirty-six) reported working at four years.
We collected information about litigation only at baseline. As
stated in the article, retaining an attorney at baseline was an
independent predictor of receiving disability compensation at four years.
It is unclear whether the factor of continued involvement with an
attorney during the follow-up period would provide additional information
beyond ascertaining the patient's current disability compensation
status.
In multivariate models of baseline factors associated with disability
and work status at four years, we assessed a variety of potential
predictors, including cigarette-smoking status. Specifically, we
considered patients to be smokers if they reported current smoking
(or cessation within the past six months) at the baseline evaluation.
With use of this definition, smoking was neither a univariate nor
a multivariate predictor of disability or work status at four years.
Notwithstanding our results, smoking cessation should be encouraged
by physicians for many other reasons.
Finally, Dr. Peacock inquires about the relationship between
residual complaints and the physical intensity of one's job. Unfortunately,
we did not collect detailed information about the job that patients
returned to. At the baseline evaluation (Tables I and II), there
were important sociodemographic and job-related differences among
patients who were receiving Workers' Compensation and those who
were not. Prior studies reporting outcomes of patients receiving and
not receiving Workers' Compensation have inadequately controlled
for these differences between groups. Our multivariate models were
an attempt to address some of the heterogeneity among patients receiving
and not receiving Workers' Compensation. However, Dr. Peacock is
correct that larger studies with more detailed questions at baseline
and follow-up are needed to understand the various factors that
may account for the differences in outcomes that were observed.
Steven J. Atlas, M.D., M.P.H.
Robert B. Keller, M.D.
Corresponding author: Steven J. Atlas, M.D., M.P.H.
General Medicine Division
Massachusetts General Hospital
50 Staniford Street
Boston, Massachusetts 02114