A Review of Pathology Claims from 1995–1997

The Doctors Company is a physician-owned professional liability insurance company located in Napa, California. We insure almost 10 percent of U.S. pathologists, who practice in nearly every state. Consequently, our claims experience should be a representative sample of all pathology claims, and an analysis of the problems leading to these claims should be representative of the problems occurring in all pathology practices.

Pathology claims involving 344 cases reported to The Doctors Company in 1995, 1996, and 1997 were reviewed. When we exclude Pap smear claims (17 percent of total claims, of which 93 percent involved false negatives), nuisance claims (13 percent of total claims; these are “frivolous” claims with no basis or “John Doe” claims naming everyone involved in a case), and claims involving clinical pathology (3 percent), autopsies (2 percent), and non-gyn cytology (1 percent), we are left with 218 claims involving surgical pathology (including fine needle aspiration [FNA] biopsies). These claims are presented in the table below.

Forty-six percent of the surgical pathology claims in the table showed no particular pattern relating to specimen type, diagnostic category, or diagnostic error. These appear to represent “random errors,” which are most appropriately addressed through intradepartmental quality assurance programs. Within this Miscellaneous Surgical Pathology category, however, there were several noteworthy observations:

1. Three claims involved cervical lymph nodes containing metastatic squamous carcinoma, which were misdiagnosed as branchial cleft cysts.
2. Three claims involved diagnostic errors made by locum tenens pathologists.
3. Two claims involved pathologists held liable for diagnostic errors made by the expert consultants to whom they referred the case. This is called vicarious liability; i.e., you can be held responsible for choosing a negligent consultant—so choose your consultant carefully!
4. Three claims involved a “missed” micrometastasis of breast carcinoma in axillary lymph nodes. It was claimed that these errors contributed to disease recurrence, which may have been prevented had appropriate chemotherapy been given.
5. Four claims resulted from misidentified pathology reports or mislabeled blocks or specimens, resulting in patients being told they had a malignant diagnosis when in fact they did not (and vice versa).

However, 54 percent of surgical pathology claims fell into six repetitive patterns of specimen type or diagnostic category—suggesting “systematic errors.” These “high-risk areas” are most appropriately addressed by heightening the practicing pathologist’s awareness of them and by spending more time in our residency training programs and continuing medical education programs on these problem categories. Overall, 52 percent of these claims involved the false-negative diagnosis of cancer, and 33 percent involved the false-positive diagnosis of cancer. These “high-risk” areas of surgical pathology, which together account for 54 percent of total claims, are listed below:

1. Breast Biopsy: 14 percent of total claims
2. Fine Needle Aspiration: 13 percent of total claims, of which 44 percent are breast FNA and 56 percent are “other” FNA
3. Malignant Melanoma: 11 percent of total claims
4. Malignant Lymphoma: 7 percent of total claims
5. Frozen Section: 5 percent of total claims
6. Prostate Biopsy: 4 percent of total claims

 
Click to view a larger image.

Breast Fine Needle Aspiration and Breast Biopsy are discussed in our article “Breast Biopsy and Fine Needle Aspiration.”

Malignant Melanoma is discussed in our article “Melanoma Expert Panel Report.”

Malignant Lymphoma: The failure to diagnose lymphoma in extra-nodal locations (skin, nasal cavity, pancreas, mediastinum, and stomach) accounted for seven of the 12 false-negative claims reviewed (58 percent), and three of these seven involved unrecognized gastric lymphomas. We need to have a high index of suspicion for lymphoma when confronted with extra-nodal lymphoproliferative processes, particularly if there is a tumor mass, evidence of destruction of surrounding structures, systemic symptoms, immunocompromise, or prior biopsies showing a lymphoproliferative process. The fact that extra-nodal lymphomas may be polymorphous contributes to the diagnostic difficulty, and definitive diagnosis often requires immunohistochemical studies and expert consultation. Two of the three false-positive diagnoses were malignant melanomas metastatic to lymph node. This is a recurrent problem that can be prevented by routinely including melanoma, lymphoma, and carcinoma in the differential diagnosis of every undifferentiated malignant neoplasm and by obtaining immunostains for S100 protein, CD45, and keratin.

Prostate Biopsy: Most of these claims involved the false-positive diagnosis of adenocarcinoma on needle biopsy, discovered when no carcinoma was found in the radical prostatectomy specimen and the biopsy slides were reviewed by another pathologist. A prostate biopsy false-positive rate of 1.3 percent is reported. There is a considerable amount of recent literature on the mimics of carcinoma on needle biopsy, e.g., partial atrophy, postatrophic hyperplasia, adenosis, and prostatic intraepithelial neoplasm (PIN). Every pathologist needs to be familiar with the cytohistologic features of these entities, and each has to be considered before making a diagnosis of carcinoma. The false-negative claims involved local recurrence of carcinoma following prostatectomy when the pathologist had stated that the surgical margins were not involved; one of these cases resulted from sampling error (the sections submitted showed clear margins), and the other resulted from failure to ink the prostate margin and to take a sufficient number of sections. Interestingly, claims involving failure to recognize residual carcinoma following androgen-deprivation therapy have not yet been seen, but this may reflect the recent use of this treatment modality.

Frozen Section: There were 10 claims for misdiagnosed frozen sections, and most appear to represent random errors. However, four (40 percent) involved breast biopsies, and two of these involved lesions less than 1 cm in size—suggesting a lack of familiarity with published guidelines on the management of breast biopsy frozen sections.

1. Many diagnostic errors could be prevented if we routinely asked ourselves the following questions:
   
   • What is the differential diagnosis for this lesion?
   • How can I rule out these differential diagnostic possibilities? Put the case aside, and take time to review a journal article, refer to a textbook, consult with a colleague, send the case to an expert consultant, or use special stains; e.g., when looking at an “undifferentiated” malignant neoplasm, routinely obtain a panel of immunohistochemical stains to rule out melanoma, lymphoma, and carcinoma.
   • Do the clinical history, patient age, location of the lesion, and gross appearance “fit” with the diagnosis I’m about to make? For example, if this is a recurrent lesion, then it isn’t nodular fasciitis; or if the patient is an adult, be hesitant to make a diagnosis of epithelioid Spitz nevus; or when making a diagnosis of pseudomembranous colitis involving the splenic flexure of an elderly patient, always consider the possibility of ischemic bowel disease. Remember, all pathologic diagnoses are ultimately clinicopathologic diagnoses and require correlation of gross and microscopic pathologic findings with clinical information and sometimes special studies, e.g., immunohistochemical stains.
   • What are the clinical and management consequences of the diagnosis I’m considering? For example, failure to diagnose atypical duct hyperplasia on a breast biopsy will result in less frequent “follow-up” and therefore less opportunity to detect a subsequent carcinoma; or the failure to advise the clinician that breast FNA has a 3–5 percent false-negative rate (which can be reduced by applying the triple test strategy) may result in the failure to diagnose breast carcinoma; or failure to recommend that a “worrisome” melanocytic lesion be completely excised may result in the recurrence of a “missed” melanoma.
   • Do I have reservations about my diagnosis? If you do, share these with the clinician in your report, coupled with a recommendation for appropriate follow-up or additional diagnostic studies—stressing the importance of clinicopathologic correlation. If the implications for patient management are significant, show the case to a colleague, and if you can’t agree, send it to an expert consultant.
   • Am I too rushed, stressed, or fatigued to give this case the concentration it requires? Am I about to make a hasty diagnosis for an anxious clinician, or am I in a rush to get a case off my desk so I can get to a meeting? If the answer is yes, put the case aside or give it to a colleague. Examples of claims that may have resulted from loss or disruption of concentration include the failure to recognize carcinoma in situ in a bladder biopsy diagnosed as chronic cystitis with numerous von Brunn nests; failure to notice tumor emboli in lymphatic spaces in a breast biopsy diagnosed as ductal carcinoma in situ (DCIS); and failure to notice a cluster of signet ring cells in a biopsy diagnosed as chronic gastritis.
   
   
2. Some problems could be prevented or minimized by accurately and effectively communicating our findings to clinicians in our surgical pathology reports:
   
   • Diagnoses should be based on accepted “diagnostic criteria” and, in problem or difficult cases, these should be stated in the report. An incorrect diagnosis is not necessarily negligent and is easier to defend if the written report reflects the thinking of a prudent, careful, and informed pathologist.
   • Do not be hesitant to communicate to the clinician that you are using your judgment to arrive at one diagnosis chosen from other differential diagnostic possibilities. For difficult diagnostic problems, the report should include a discussion of the “differential diagnostic” considerations. This assures that they have been considered by the pathologist and read by the clinician, who may have clinical information unknown to the pathologist that leads to the correct diagnosis.
   • The report should include recommendations for appropriate “follow-up studies” or “additional diagnostic studies” when appropriate; e.g., always recommend breast biopsy if the breast FNA diagnosis, mammography findings, and clinical breast examination are nonconcordant.
   • For tumors frequently encountered in everyday practice, pathologic features of prognostic or therapeutic importance should be specified—either in the body of the report or preferably in a synoptic format (see Appendix). Not only does this make the report more clinically relevant, but it also assures that each important pathologic characteristic of the tumor has been methodically considered and not overlooked. For example, a pathologist correctly diagnosed endometrial carcinoma with superficial myometrial invasion. One year later metastases occurred, and a review of the original slides showed myometrial lymphatic space invasion, which had been overlooked. The surgical pathology report content recommendations for common malignant tumors developed by the Association of Directors of Anatomic and Surgical Pathology are recommended.
   • Always issue a written report to document verbal consultations. When you are asked to look at a slide and give an opinion, you are giving a consultation for which you can be held liable. Without a written report, the only record is the clinician’s recollection of the conversation or a handwritten note made in the office record or chart, which is often incomplete or inaccurate. This is especially important when consulting with dermatologists, since they are apt to show you slides from melanocytic lesions they are clinically concerned about.
3. 
   
4. An effective quality assurance program for surgical pathology is an essential component of any strategy for minimizing diagnostic errors.
   One important quality assurance activity is the intradepartmental review by another pathologist of all provisional diagnoses made on difficult or unusual cases and on all cases with a malignant, atypical, or suspicious diagnosis. If the “second opinion” is concordant with the provisional diagnosis, this should be documented in the final report or on the file copy. If the review is nonconcordant, then the case should be reviewed by another pathologist or sent for consultation to an expert. This program is easy to implement, effective, and reassuring to clinicians. Examples of claims that may have been prevented by this process include the misdiagnosis of:
   
   • serous papillary endometrial carcinoma as endometrioid carcinoma
   • embryonal carcinoma of the testes as seminoma
   • renal angiomyolipoma as sarcomatoid renal cell carcinoma
   • appendiceal mucinous cystadenocarcinoma as benign mucocele
   • bronchial carcinoid as small cell carcinoma
   • pleomorphic xanthoastrocytoma as glioblastoma
   • sellar meningioma as pituitary adenoma
   
   A good quality assurance policy is to send all cases for which there is intradepartmental disagreement to an expert consultant. If there are any reservations about a case that is sufficiently challenging to require intradepartmental review, then get an expert’s opinion. Neither seniority nor plebiscite assures a correct diagnosis. If you have a choice, it is preferable to use an expert in your geographic area, since his or her expertise may well contribute to the “standard of care” perceived for your community.
   
   Adequate sampling of small biopsies is important, since diagnostic findings may be seen in only one or two sections taken from one of multiple levels. This is especially true for prostate needle biopsies. A recommended approach is to serial section the entire biopsy and examine multiple sections from each of three levels. Save the intervening unstained sections in case you subsequently need to examine them or perform special stains. Discard the unused sections after one month.
   
   The use and correct interpretation of immunohistochemical stains as an adjunct to arriving at a correct diagnosis are now part of the standard of care we are judged by. If a diagnostic error could have been prevented by their use, we will be found negligent. Examples include the misdiagnosis of lymphocyte-rich thymoma as lymphoma on thoracoscopic biopsy because of failure to consider thymoma in the differential diagnosis and to obtain keratin stains and the misdiagnosis of nasal melanoma as lymphoma because of failure to stain the “undifferentiated” neoplasm for CD45 and HMB45. It is equally important to be familiar with the full spectrum of immunoreactivity of the various immunostains, e.g., an intracranial schwannoma was misdiagnosed as astrocytoma based on a positive glial fibrillary acidic protein (GFAP) stain, and an atypical fibroxanthoma containing S100 positive Langerhans cells was misdiagnosed as desmoplastic melanoma.
   
   
5. Procedures to follow if you think a diagnostic error may have been made or you suspect that a malpractice claim may be filed:
   Promptly report to us any claim or incident that might result in a claim, e.g., diagnostic errors that you become aware of, verbal or written threats of a lawsuit, requests for records or slides by an attorney or a patient, requests for deposition or interview regarding a case that you were involved with, and any formal paper served in which you are named as a defendant or witness. It is important to notify us first—not the hospital risk manager or legal counsel. The hospital risk manager’s responsibility is to minimize the hospital’s liability risk—not the pathologist’s. Furthermore, if the hospital gets involved first, they may impound the microscopic slides and prevent their review by our claims experts. This makes it difficult for us to assess whether or not negligence has occurred, which, in turn, compromises our ability to protect you in the event that a claim is filed.
   
   Original slides, blocks, and fixed tissues are irreplaceable and may become important evidence in a case. They are the property of the laboratory and are in your custody, and only a court order (subpoena) can force you to turn them over to a lawyer or the patient. For this reason, do not send original slides or blocks if they are requested by an attorney or a patient. Send recuts (clearly labeled as recuts), but only after you have reviewed them to be certain that they show the lesion in question.
   
   If a written request for pathology reports and slides is received from an attorney, respond by sending a letter: 1) asking for the reason for the request and stating the need for a properly executed release form; 2) requesting a list of the specific case accession number(s) needed; and 3) stating your policy that only recuts of slides will be sent. If the attorney insists on receiving only the original slides or blocks, respond in writing that the requested original materials will be made available for inspection and review only on your premises and under your direct supervision. Absent a subpoena, an attorney cannot compel you to send original slides or blocks. If a subpoena is received in a case where a formal lawsuit has been filed, notify us immediately. We will then contact the attorney to again determine if recuts will suffice. If original materials are still demanded or if only original slides are available (e.g., cytology smears), we may file a motion in court to limit discovery to examination of the original slides on your premises. If the court orders that you turn over the materials, you must comply, but if the slides and blocks are subsequently broken or lost, you will be operating under the protection of the court and will be protected from charges of spoliation of evidence.
   
   
6. Retention of reports, slides, and blocks:
   Except for educational or research purposes, pathology slides and blocks should not be retained longer than required by applicable laws and regulatory agencies. The following guidelines fulfill the requirements of the College of American Pathologists Laboratory Accreditation Program, the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), and the Clinical Laboratory Improvement Act (CLIA ’88). State laws and regulations may differ.
   
   
   1. Surgical pathology, cytology, and autopsy reports must be retained for a minimum of 10 years.
   2. Surgical pathology slides must be retained for a minimum of 10 years.
   3. Cytology slides (both normal and abnormal) must be retained for a minimum of five years.
   4. Surgical pathology, cytology, and autopsy blocks must be retained for a minimum of five years (College of American Pathologists). JCAHO and CLIA ’88 require only two years.
   
   
7. How to review a pathology case if you are asked to assess a claim for an insurance company or asked to be an expert witness:
   When you are asked to review a case by an insurance company, it is usually because the company has received a claim from a patient’s attorney or an incident report from an insured pathologist. The insurance company wants an objective professional assessment about whether or not the standard of care for pathology practice has been met and, if not, whether the patient’s injury was the consequence of the pathologist’s negligence. Since the company’s decision to settle a case or defend it in court is often at stake, the reviewer assumes a great responsibility—for the involved pathologist, the patient, and the company!
   
   When you are asked to be an expert witness, it is generally because a claim is being litigated, and either the insurance company wants an expert to defend the pathologist or the plaintiff’s (patient’s) attorney wants an expert to testify against the pathologist.
   
   When reviewing a case, you will be provided with a summary of the incident or claim, the pathology report and slides, and other documents relative to the case, e.g., surgical operative reports, x-ray/mammogram/ultrasound reports, and depositions. It is extremely important to review this material objectively and impartially, keeping in mind that the question being asked is whether or not the diagnosis or behavior was that of a careful, prudent, and informed practicing pathologist (who is usually not an expert). This is very difficult to do, since you will be biased by the knowledge that either the pathologist (incident report) or the patient (claim) thinks an “error” has been made.
   There is a natural tendency to ask oneself if this would be how “I” would have diagnosed or handled the case. Beware of this mindset, because most of us find it difficult to accept that we too make errors (particularly if we are experienced pathologists, which most reviewers are—or, worse yet, an “expert”). Furthermore, the reviewer often has information that the pathologist did not have—bestowing a hindsight bias of 20/20.
   
   Whenever possible, a useful approach is to review the slides and the pathology report before reviewing the entire case in order to let the case “unfold” the way it did for the pathologist who initially saw it. You must then ask yourself if the manner in which the pathologist diagnosed or handled the case was “reasonable” given the information available—or if it was below the performance standard expected of a well-trained, careful pathologist.
   
   Sometimes this is an easy decision; e.g., an obvious nodular melanoma is called a compound nevus, or a breast FNA is called malignant when the mammogram and physical examination suggest a fibroadenoma, and there is no mention of the triple test strategy. But more often, the decision is subjective and falls within the range of interobserver error; e.g., extension of lobular carcinoma in situ into interlobular ducts is misinterpreted as low-grade ductal carcinoma in situ, resulting in re-excision lumpectomy and breast radiation; or a patient with a minimally invasive low-grade endometrial carcinoma develops early recurrence, and a review of the slides reveals a single focus of myometrial lymphatic invasion, which was overlooked, and the patient alleges that had this been recognized, she would have received treatment that would have prevented the recurrence.
   
   Still other cases involve negligent diagnoses that do not result in patient injury; e.g., a frozen section of a renal mass is interpreted as sarcomatoid renal cell carcinoma; a nephrectomy is performed, and permanent sections reveal that the tumor is an angiomyolipoma. The misdiagnosis did not injure the patient because even if the tumor had been correctly diagnosed, a nephrectomy would have been performed.
   
   Some claims result from failure to thoroughly work up a case; e.g., an undifferentiated malignant tumor of the nose is diagnosed as large cell lymphoma, which is a reasonable hematoxylin and eosin (H & E) diagnosis; it decreased in size following radiation and was re-biopsied at another institution where S100, CD45, and cytokeratin stains revealed that the tumor was a malignant melanoma. While failure to obtain the immunostains was negligent, the misdiagnosis did not result in patient injury since the melanoma was initially unresectable and would have been radiated anyway to shrink it to a surgically resectable size.
   
   

Appendix

Breast Biopsy Synoptic Report:

• Site:
• Histologic Type:
• Grade (Bloom & Richardson):
  • Tubule Formation Score:
  • Nuclear Pleomorphism Score:
  • Mitosis Score:
• Tumor Size:
• Lymphatic/Vascular Invasion?
• Ductal Carcinoma In Situ?
  • Nuclear Grade:
  • Comedo Necrosis Present?
• Status of Biopsy Margins:
• Sent for Estrogen/Progesterone Receptor?
• Sent for HER-2/neu?
• Sent for Ki-67?
• pTNM:

Breast Carcinoma Synoptic Report:

• Site:
• Histologic Type:
• Grade (Bloom & Richardson):
  • Tubule Formation Score:
  • Nuclear Pleomorphism Score:
  • Mitosis Score:
• Tumor Size:
• Lymphatic/Vascular Invasion?
• Ductal Carcinoma In Situ?
  • Nuclear Grade:
  • Comedo Necrosis Present?
• Lymph Node Number and Status:
  • Micrometastasis?
  • Transcapsular Extension?
• Status of Final Margins:
• Estrogen/Progesterone Receptor Status
• HER-2/neu Status
• Ki-67 Score
• pTNM: