This paper is a very interesting economic analysis of workplace
mammography screening programs. Especially important is the discussion
of the effect of disease prevalence on the cost-effectiveness
of workplace screening programs.
Workplace screening programs would be most beneficial for populations
with a low rate of screening among employees through existing
programs and for locations where there are a sufficient number
of women old enough to require annual screening. In workplaces
where screening rates are already adequate, workplace screening
may provide little incremental benefit to the employed population.
Similarly, workplace screening will provide little benefit in
workplaces that have a high proportion of female employees who
are unlikely to require mammographic screening. These issues should
be addressed directly in the evaluation of breast cancer screening
programs by each employer.
Some aspects of the analysis of workplace screening programs are
not fully developed by the authors and require further elaboration.
These include: (1) the quality of the screening mammograms, (2)
the quality of the follow-up for abnormal mammograms, and (3)
the continuity between the program and the employees' primary-care
providers. Each of these additional topics will be discussed in
Quality of Screening Mammograms
The authors discuss the quality of mammography examinations as
a variable to be considered in an assessment of a workplace screening
program. Obviously, one should be interested in whether the program
meets the quality guidelines established by the American College
of Radiology and the federal Mammography Quality Standards Act.
However, these are not the only issues related to the quality
of the mammography readings that need to be addressed.
Diagnostic quality of the examinations becomes a critical variable
in assessing these programs. This includes not only the technical
quality of the images, but the characteristics of the radiologist
reading the images. For example, radiologists in a worksite screening
program are not involved in the usual care of the patient and
may be concerned about the potential for malpractice suits from
employees. Would radiologists be more likely to overread images
in such a program, increasing the rate of false-positive results
for female employees? This potential for overreading mammograms
may be increased by the fact that these programs suffer from a
lack of access to prior examinations, which can help in the interpretation
of mammograms with questionable findings. These issues need to
be tested empirically.
Follow-up of Abnormal and Suspicious Mammograms
Mammography is not a simple screening test, but rather, is part
of a diagnostic evaluation that requires careful follow-up of
abnormal and suspicious examinations. These follow-up issues need
to be carefully considered in establishing a worksite screening
program. Is there a mechanism for follow-up of abnormal or suspicious
mammograms? What are the costs of implementing the follow-up program?
An additional concern in developing a program is whether privacy
concerns are addressed. For example, are employees to be contacted
through their employer to notify them of abnormal results or to
report for follow-up examinations for these screening programs?
Given the employer's knowledge of results of the test, are employees
then at risk for losing their position or insurance benefits?
Finally, are employers liable for poor follow-up provided by physicians
participating in these programs?
A last issue to consider is how the program will support the efforts
of the primary-care physicians who see these employees on a regular
basis. Mammography is often part of annual physical examinations.
Thus, women may receive several services simultaneously during
primary-care visits, decreasing the lost productivity incurred
when receiving the isolated mammograms reported by Griffiths et
Employers must pay for worksite screening programs after they
have already provided employees with a mammography benefit through
their insurance programs. Thus, even though worksite programs
may have a lower cost per screening examination, employers may
actually be paying twice for these services under certain insurance
programs. If the worksite program is not integrated into primary-care
practice, primary-care physicians may not have records of the
results of these mammograms, either for patient-care purposes
or for reporting purposes within managed-care organizations (ie,
HMO report cards).
Finally, women whose mammograms are abnormal may prefer to receive
counseling from a primary-care physician they trust rather than
from a physician in a worksite program.
Mammography screening rates in the United States are far below
optimal levels. Thoughtful worksite mammography screening programs
may help increase the current rate of mammography screening. Careful
study of new means of providing this service to patients, such
as worksite screening programs, can help improve these efforts.
Economic analysis of these programs must address the issues raised
by Dr. Griffiths and colleagues, as well as those included in