Pap test perils
Ken Gatter, MD, JD
Most pathologists know the story of how Pap smears have reduced
the mortality and morbidity rates of cervical cancer. Cervical cancer
once ranked as the No. 1 cancer killer of women in the United States.
It now ranks 13th, although African-American women are still 2.2 times
as likely to die from cervical cancer as are white women. This dramatic
reduction in mortality is mostly due to the use of Pap smears.1
Pap smears differ in two important ways from the biopsies of the
breast, skin, prostate, endometrium, and other organs that pathologists
typically examine. First, Pap smears are a screening test. They
are routine, unlike breast or prostate biopsies. As a screening
test, Pap smears are better than some but far from ideal. Even for
high-grade lesions the sensitivity for Pap smears is 70 to 80 percent.2
Moreover, interobserver agreement is poor for low-grade intraepithelial
lesions and moderate for higher grade lesions, with one study showing
kappa values of 0.172 and 0.712 respectively.3
Recent American Cancer Society guidelines recommend that screening
by conventional cytology smears be done every year until a woman
is at least 30 and has had three consecutive, technically satisfactory
normal/negative results, after which conventional Pap smears should
be performed every two or three years. Repeated screening means
that invasive cervical cancer is likely to be recognized despite
the test’s relatively low sensitivity rate.4
The drawback is that the risk of false-positives also increases
with increasing frequency of the test.
Pap smears differ, too, in that most are seen only by cytotechnologists.
The typical Pap smear differs from breast or prostate biopsies because
a pathologist does not make the diagnosis. This does not mean that
cytotechnologists don’t do as good a job as pathologists in interpreting
Pap smears, but it does mean that the legal issues often have an
additional layer of complexity. Not only does CLIA regulate many
aspects of how cytotechnologists work and how cytology laboratories
operate, but also complex legal issues of agency and negligence
law often determine the extent of a pathologist’s liability. Generally,
if the pathologist does not examine the Pap slides in question herself,
she won’t be successfully sued for direct personal negligence of
misreading the slides. However, she can be sued under agency law
for the negligent acts of the cytotechnologist or for negligence
in supervising, hiring, or implementing policies and procedures.
This article’s primary objective is to improve pathologists’ awareness
of some of the legal issues that arise in Pap smear litigation.
A second objective is to improve communication between pathologists,
and lawmakers (courts and legislatures), in the hope of promoting
an understanding and legal recognition of the unique attributes
Unlike the cases on prostate biopsies, skin biopsies for melanoma,
and breast biopsies reported previously in CAP TODAY (April, May,
August, and November 2002), the Pap smear case presented here is
based mostly on a single reported appellate decision. A few caveats
are important. First, the names have been changed. Second, nothing
said here is intended as legal advice. Third, all of the cases,
and most medical malpractice law, involve state courts applying
state law, so your state law may be different.
The ‘routine’ Pap smear case
Facts. A woman we’ll call Ms. Kelsy had been going
to the same OB since 1977. She had Pap smears in 1977, 1978, 1980
(two in 1980), 1984, 1986, and 1987. All of the Paps were evaluated
by Cyto-lab, a professional association organized under state law.
Cyto-lab processed about 18,000 Paps per year and employed two cytotechnologists.
All Pap smear reports included a pathologist’s signature, even if
the pathologist had not looked at the slides. The lab reviewed at
least 10 percent of all "normals." One pathologist per day rotated
through the cytology lab and functioned as
All of Ms. Kelsy’s Pap smears except the last one, the 1987 Pap,
were signed out as "class I negative." The 1987 Pap was diagnosed
as "class IIB" or, as the court interpreted the diagnosis, "suggestive
of mild to moderate dysplasia."
The 1986 Pap report noted that "moderate inflammation" was present.
One of the pathologists, whom we’ll call Dr. Jones, made it a practice
to review all slides with inflammation, and he reviewed the 1986
Pap. However, he looked only at the inflammation and the court concluded
that "he did not look at the entire slide nor did he look for abnormal
In 1987, Ms. Kelsy, 34 years old, returned to her OB because she
had pelvic pain and erratic periods. The OB did a Pap smear during
this visit that was diagnosed as "class IIB," as previously mentioned,
and referred Ms. Kelsy to a gynecologist. The gynecologist took
biopsies that showed invasive squamous cell carcinoma. Her cervical
cancer was stage IIIB with an estimated 20 to 25 percent chance
of five-year survival. Ms. Kelsy died
Two events took place in the time between her diagnosis and death.
First, the radiation oncologist asked Cyto-lab to review the previous
Paps. A pathologist and a cytopathologist re-reviewed the 1986 Pap,
the one with inflammation, and admitted that the Pap was at least
suspicious for high-grade or invasive cancer. Dr. Jones, the pathologist
who initially reviewed the 1986 Pap for inflammation, also re-reviewed
the slides and concluded that it should have been "class IIB," which
was mild to moderate dysplasia.
Second, the pathologists at the university asked to see the Paps.
The university pathologists received the 1984, 1986, and 1987 Pap
smears (those prior to 1984 had been destroyed) and upgraded all
of them. They diagnosed the 1986 and 1987 smears as malignant, with
no equivocation, and the 1984 smear as "class IIB" according to
The 1984 Pap smear
The first issue centered on the admissibility of the 1984 Pap smear.
Recall that the university pathologists had upgraded the diagnosis
of the 1984 Pap to a "class IIB" under the Cyto-lab classification
(mild or moderate dysplasia).
Cyto-lab argued that the statute of limitations had run on the
1984 slide, so any evidence about inaccurately interpreting the
slide should not be allowed. In contrast, the plaintiffs argued
that there was a "continuing treatment relationship" between the
plaintiff and the lab, and, therefore, the 1984 Pap smear should
Why is this important? Interpretation and application of statutes
of limitation are important issues because they can often significantly
increase the amount of damages and in some states be the difference
between getting to the jury or not. The plaintiffs argue that if
an "accurate" diagnosis had been made in 1984, Ms. Kelsy’s chances
of survival would have been 90 percent or higher. Damages are significantly
greater when the loss of chance of survival is the difference between
90 percent and the 20 to 25 percent five-year survival she had when
her cancer was first diagnosed as a stage IIIb. Compare this with
the much smaller reduction in survival and lower damages represented
by the one-year difference in survival between 1986 and 1987.
The appellate court concluded that the continuing-treatment exception
to the statute of limitations applied to the 1984 Pap smear. The
court emphasized that the relationship between the lab and the patient
"was not sporadic. Rather it was routine." The key for the court
was that the relationship was ongoing and dependent. It was a dependent
relationship because the OB, and implicitly the patient, relied
on the lab "for detecting potential cancer and relied upon its reports
to facilitate his choice of therapy or further work up." The relationship
was ongoing because, the court wrote, the patient did not seek other
testing because she trusted the reports and, implicitly, the pathologists.
It is not surprising that the court would consider a series of
routine screening exams enough to establish a continuing relationship
between the patient and the cytology lab. The result, however, increases
exposure to litigation for Pap smears. Since Pap smears are routine
screening tests, most women will have many more Pap smears and have
them over a longer time than, for example, a single breast biopsy.
In contrast, a court may find a cause of action based on a misdiagnosed
single breast biopsy barred by the statute of limitations even though
the sensitivity for interpreting the breast biopsy is better, meaning
the odds of a false-negative are lower.
The risk of being sued for Pap smears, therefore, is higher because
of the continuing-treatment exception to the statute of limitations.
The 1987 Pap smear
The second issue in Ms. Kelsy’s case was whether the 1987 Pap smear
should be admitted into evidence. Cyto-lab argued that even if the
1987 Pap was negligently misread, the lack of damages prevents its
introduction into evidence. The defendants argued that the 1987
Pap made no difference as to damages, since the biopsy results were
known and no significant time elapsed between the Pap and the biopsy.
The trial court allowed the 1987 Pap smear into evidence to help
the jury decide whether the screener had "the appropriate level
of skill" and for "determining whether or not there was a pattern
of misreading Pap smears." Information about the 1987 Pap was not
to help the jury decide proximate cause of the harm to Ms. Kelsy.
Instead, the information about the 1987 Pap smear helped the jury
determine whether the standard of care was breached in 1984 and
1986. Moreover, the appellate court concluded that the 1987 Pap
smear was relevant in determining the plaintiff’s claim that the
lab was negligent in failing to monitor, supervise, select, and
review its medical staff.
A third issue discussed by the appellate court was the admissibility
of the personnel records of the cytotechnologist who looked at the
1984 and 1986 Pap smears. We’ll call him Mr. CT. A cytotechnologist
coworker had informed a supervisor that Mr. CT was sloppy and often
hurried through his screening of Paps.
The defendant Cyto-lab argued that this was prejudicial and had
nothing to do with the issue of negligence for the particular Pap
smears. The risk from the defendant’s perspective was that the jury
would not look at the Pap smears at issue, but would be biased to
conclude that Mr. CT must have made a mistake and must be generally
The trial court allowed the evidence. It found that the personnel
record evidence had probative value, since the information addressed
the question of whether Cyto-lab supervised its employees properly
and implemented procedures to ensure accuracy in the interpretation
of Pap smears.
The appellate court, though it acknowledged that allowing the
evidence harmed the defendant, held that the trial court had not
abused its discretion. It was up to the trial court to weigh the
probative value of the evidence of Mr. CT’s hasty Pap screening
against the prejudicial value of the evidence. The appellate court
could overturn the trial court only if it found that the trial court
abused its discretion.
The appellate court allowed into evidence all of Ms. Kelsy’s available
Pap smears, including the 1984 and the 1987 Pap smears. This suggests
that the court was willing to look at the broader context by allowing
evidence from all the Pap smears and not just the 1986 missed Pap.
Had the court accepted the defendant’s theories and allowed only
the 1986 Pap smear into evidence, the plaintiff’s case would have
been significantly more difficult. Damages would have been less,
and the jury would have heard about only one missed Pap. The result
is that for the jury it looks like what is on trial is not a single
potentially negligent occurrence of misreading the 1986 Pap smear
but instead whether the lab generally operated below the standard
in its interpretation of all of Ms. Kelsy’s Pap smears.
In a sense the court followed the suggestion of Richard DeMay,
MD, that in Pap smear litigation it should be "the track record,
not the individual case that is important."5
Notwithstanding, this case demonstrates that routine screening
Paps have greater exposure to liability when courts apply the continuing-treatment
exception to the statute of limitations because all of the Pap smears
the lab reviews for a single patient might be available as evidence.
This result is problematic. The continuing-relationship exception
makes sense as long as courts also understand that the standard
of care for reviewing routine Pap smears should incorporate the
test’s inherent shortcomings, namely the moderately good sensitivity
rate. One missed Pap smear should not necessarily represent incontrovertible
evidence of negligence. Cases of a single Pap with frequent clusters
of malignant cells on the slide may be enough for a finding of negligence,
but many cases are not so straightforward. One of the reasons for
routine Pap smears is that the frequency of the test makes up for
its less than perfect sensitivity.
(Interestingly, the defendant’s claim that a patient who failed
to get routine Pap smears was contributorily negligent was rejected
in a 1998 case, in which the court wrote that even if the standard
of care required yearly Pap smears, this standard applied to physicians
and not to patients.6)
The case also illustrates why plaintiffs typically bring more
than one cause of action and include more than one defendant. For
example, the plaintiffs included, in addition to the routine medical
malpractice claims against the pathologist and cytopathologist,
an allegation that Cyto-lab was negligent in supervising and selecting
its medical staff. This means that the jury hears testimony it might
not otherwise hear, such as information about the 1987 Pap smear
and information about the cytotechnologist’s hasty screening habits.
It is often difficult for the jury to compartmentalize information
despite the trial judge’s efforts to instruct the jury to consider
some information only for certain purposes.
Although not clearly an issue in the appellate opinion, the pathologist
who reviewed the 1986 Pap only for inflammation made a mistake.
Once he reviewed the slide, he should have reviewed all of it. Moreover,
one may have a tendency to not review the slide as carefully after
one arrives at a diagnosis and an explanation for the clinical question.
Pap smears, like surgical specimens, may have more than one diagnosis.
To make matters worse, the pathologists re-reviewed the slides
and apparently recorded their "new and improved" diagnosis. CLIA
requires a laboratory to review all normal or negative Pap smears
received and available within the previous five years for every
Pap with a diagnosis of moderate dysplasia or above.7
However, CLIA requires the laboratory to issue an amended report
only if the change in diagnosis will change the patient’s care.
Furthermore, the regulations state that only a review must be conducted.
Certainly the lab should document the review in its quality assurance
records, but it is not required to document the new diagnosis.8
No good comes of recording the new diagnosis made through the
"retroscope." There is no benefit to patient care, and the case
against the pathologist and the laboratory is strengthened.
Another mistake is to talk to colleagues and have them look at
It is one thing to show slides before signing out the case and quite
another to do so after receiving news that there might be a problem.
The latter leads to colleagues enduring longer depositions and increases
the odds of divergent opinions.
The university pathologists arguably also made a mistake. They
reviewed the previous Pap smears retrospectively since they knew
the biopsy results, and they should have reflected this in their
report. Including this information in the report emphasizes the
retrospective character of the review. This is not to say that the
university pathologists should not have provided an accurate diagnosis
to their best ability once requested to do so. However, it raises
the question of whether their diagnosis would have been as unequivocal
were the Paps in their routine workload. The aim of a trial is the
truth, and a retrospective review by a pathologist who knows the
patient has invasive cancer may not get closer to the truth. Certainly
pathologists should be concerned about maintaining quality in their
profession, but if one is concerned about the overall competency
of another pathologist or laboratory, there are methods in place
to determine whether such concerns are justified.
Cyto-lab probably made a mistake in releasing the Pap smears to
the university pathologists. Instead, the pathologists at Cyto-lab
might have invited the university pathologists to come and take
a look at the slides. Sending the unique Pap smears to the university
was risky, and it is unlikely that the review would have altered
patient care. At this point, Cyto-lab probably knew there was a
possibility of litigation, and it was justified in ensuring the
safety and integrity of the evidence.
Lastly, the plaintiff’s use of the lab’s employment records is
a reminder to take all allegations seriously. There was no evidence,
at least in the appellate opinion, that Cyto-lab addressed the issue
related to Mr. CT. It may have helped Cyto-lab’s case to be able
to point to steps taken to assess the validity of the allegations
and to remedy deficiencies.
The case shows some of the unique difficulties inherent in Pap smear
cases. Pap smears are unlike the biopsies pathologists typically
see. One of the differences is that Pap smears are a routine screening
test with only average sensitivity. For Pap smears to remain an
effective and relatively cheap way for pathologists to promote women’s
health, courts must understand this screening test’s inherent limitations.
This is a difficult message to convey to courts and plaintiffs.
The determinative question in Pap smear cases should be whether
the interpretation of the Pap smear breached the standard of care,
with the standard of care defined to include inherent shortcomings.
As this case illustrates, this message is best conveyed with the
help of competent legal counsel invoked earlier rather than later
in the process.
1. Ghafoor A, Jemal A, et al. Cancer
statistics for African Americans. CA Cancer J Clin. 2002;
2. Sherman ME, Schiffman M, Herrero
R, et al. Performance of a semiautomated Papanicolau smear screening
system: results of a population based study conducted in Guanacaste,
Costa Rica. Cancer. 1998;84:273-280.
3. Llewellyn H. Observer variation,
dysplasia grading, and HPV typing: a review. Am J Clin Pathol.
2000;114 (suppl 1):S21-S35.
4. DeMay R. Should we abandon Pap smear
testing? Am J Clin Pathol. 2000;114(suppl 1):S48-S51.
5. DeMay R. To err is human-to sue,
American. Diagn Cytopathol. 1996;15(3):iii-vi.
6. Hawkins v Pathology Associates
of Greenville, P.A., 498 S.E.2d 395 (1998).
7. CFR Section 493.1257
8. McCoy DE. Defending the Pap smear:
a proactive approach to the litigation threat in gynecologic cytology.
Am J Clin Pathol. 2000;114 (suppl 1):S52-S58.