Context.— Original surgical pathology materials from external sources are reviewed by our pathologists for referred patients before their clinical evaluation and treatment.
Objective.— To identify the rate of major disagreements with diagnoses from external institutions and to characterize the nature and impact of discordant diagnoses on patient care.
Design.— We identified and reviewed all surgical pathology cases, except for medical liver, medical renal, and cardiac pathology cases, for the period between January 1, 2005, and December 31, 2010, to determine the overall frequency of major disagreements, defined as any change in diagnosis having a substantial impact on patient management.
Results.— Our review of 71 811 cases initially examined between 2005 and 2010 identified 457 major disagreements (0.6%). The most frequent areas of disagreement were gastrointestinal (80 cases; 17.5%), lymph node (73; 16.0%), bone/soft tissue (47; 10.3%), and genitourinary (43; 9.4%). For a subset of 166 cases reviewed between July 1, 2009, and December 31, 2010, follow-up data were available for 140 (84.3%). Treatment was affected by a changed diagnosis in 126 cases (90.0%), and prognosis was affected in 129 cases (92.1%). For 86 (51.8%) of the 166 cases, additional tissue was obtained. Revised diagnoses concurred with follow-up tissue diagnosis in 84.9% (73 cases), whereas they differed from follow-up tissue diagnosis in 15.1% (13 cases).
Conclusions.— Our findings demonstrate the value of outside case review of pathology materials for referred patients, and suggest that it decreases the likelihood of diagnostic errors and provides better protection for patients.
With the current focus on controlling escalating health care costs, the value of the outside case review of externally acquired pathology material for referred patients has been questioned within Mayo Clinic. Although outside case review is recommended by the Association of Directors of Anatomic and Surgical Pathology,1 and has been found to result in a change in diagnosis in 2% to 5% of cases,2 a survey of community hospitals and tertiary care medical centers found that only 50% of institutions report having implemented a mandatory policy of outside case review.2
The Division of Anatomic Pathology at our tertiary care academic medical center reviews externally acquired surgical pathology materials of referred patients before their treatment at Mayo Clinic. Although this policy is not mandatory, it is highly encouraged and supported by our clinical and surgical colleagues. We therefore sought to identify the rate of major disagreements between our findings and those of external entities, to categorize them by organ system, to determine how these differences in diagnosis affected clinical management, and to compare our interpretations with any subsequent tissue diagnoses.
MATERIALS AND METHODS
Institutional review board approval was obtained for this study. As part of the care provided at Mayo Clinic, patients are asked to have their previously obtained pathology materials submitted to us for review. These materials are reviewed by surgical pathologists in a general sign-out area. If there is a difference in opinion from the original diagnosis, the case is shared with a member of the specialty working group for that case. Disagreements are designated as minor (no significant impact on patient care) or major (significant potential impact on therapy or prognosis), as determined by the primary and specialty pathologists. The level of agreement is entered into a laboratory information system (CoPath; Cerner Corp, Kansas City, Missouri). The clinicians are contacted, and in the case of major disagreements, the external pathologist is contacted.
In this study, all cases with major and minor disagreements were retrieved from the laboratory database for the period between January 1, 2005, and December 31, 2010. All cases were included, with the exception of medical liver, medical renal, and cardiac pathology cases because of an alternative workflow review process. Limitations of the CoPath system precluded the retrieval of detailed information about cases reviewed by the subspecialty group between January 1, 2005, and June 30, 2009, but information was available that allowed determination of the frequency of minor and major disagreements, identification of organ systems, and determination of diagnoses. Beginning in July 2009, more detailed information regarding major disagreements was collected, including subspecialty classification, original diagnosis, revised diagnosis, reason for disagreement, nature of change (eg, benign to malignant), impact on prognosis, and impact on treatment. A chart review was also performed to determine whether follow-up tissue was obtained to document or clarify the diagnosis. The overall frequency of major disagreements was determined and then categorized by organ system. The nature of the change in diagnosis (eg, benign to malignant or malignant to benign) was also recorded. Finally, the diagnosis at review was compared with the diagnosis after additional tissue was obtained at Mayo Clinic.
Between January 1, 2005, and December 31, 2010, a total of 71 811 cases were reviewed. Minor disagreements were identified in 3906 cases (5.4%) and major disagreements in 457 (0.6%) (Table 1). Major disagreements occurred mostly in the areas of gastrointestinal (GI) (80 cases; 17.5%), lymph node (73; 16.0%), bone/soft tissue (BST) (47; 10.3%), and genitourinary (GU) (43; 9.4%) pathology (Figure 1). Examination of the detailed information available for a subset (n = 166) of the 20 079 cases reviewed between July 1, 2009, and December 31, 2010, showed that the distribution of major disagreements was similar then to that which was observed in previous years. Most of the major disagreements for this shorter period occurred in lymph node (23 cases; 13.9%), GU (19; 11.4%), GI (18; 10.8%), cytology (16; 9.6%), and BST (15; 9.0%). In this subset of 166 cases, follow-up data were available for 140 cases (84.3%). Treatment was affected by a changed diagnosis in 126 cases (90.0%), and prognosis was affected in 129 of the 140 cases (92.1%) (Figure 2).
The nature of the diagnostic changes for the 166-case subset is detailed in Table 2. The most common changes were from a benign to a malignant neoplasm (37 cases; 23%) or from a malignant to a benign neoplasm (37 cases; 23%). The original diagnosis (from an external institution), the revised diagnosis (upon outside case review at Mayo Clinic), and the follow-up biopsy or resection diagnosis (obtained at Mayo Clinic) for each disagreement by specimen type are listed for GI (Table 3), lymph node (Table 4), BST (Table 5), and GU (Table 6). Lymph node pathology was the most common area with a major disagreement. Nearly half (11 of 23; 48%) of the disagreements were a change in diagnosis from an inflammatory condition to lymphoma or vice versa, or a change in diagnosis from lymphoma to atypical or suspicious for lymphoma. A change in classification of lymphoma subtype was the second most common change (8 of 23 cases; 35%). Gastrointestinal and GU pathology were the second most common areas with a major disagreement. In GI pathology, major disagreements occurred mainly in the diagnosis of reactive versus dysplasia and in the diagnosis of malignancy in a background inflammatory lesion or adenoma (7 of 18; 38%). In GU pathology, major disagreements occurred most often in differentiating reactive urothelial atypia from carcinoma (7 of 19; 37%) and in differentiating benign prostatic tissue or atypical small acinar proliferation from adenocarcinoma (5 of 19; 26%). In BST pathology, various tumors were reclassified, notably 3 cases with an external diagnosis of benign cartilaginous tumors (2 chondroblastomas and 1 enchondroma) that were revised and confirmed by subsequent surgery as osteosarcoma (2 cases) and chondrosarcoma (1 case). One interesting and challenging case was an S100-positive tumor initially diagnosed as malignant peripheral nerve sheath tumor for which the diagnosis was later changed to malignant melanoma. Subsequent surgery revealed a desmoplastic small round cell tumor. Surprisingly, amyloidosis was a category not related to a specialty group, and it accounted for 7% of all major disagreements. The change in diagnosis was based solely on the execution and interpretation of an ancillary study, the Congo red stain (Figure 1).
In 86 (51.8%) of the subset of 166 cases, additional tissue was obtained. The revised diagnosis concurred with the findings upon examination of the follow-up tissue in 73 cases (84.9%), whereas for 13 cases (15.1%), the revised diagnosis differed from the follow-up tissue diagnosis and was often in agreement with the original diagnosis from the external institution (Table 7).
As interest in determining ways to control or reduce health care costs has grown at our institution, the outside case review of external pathology materials for referral patients has come under scrutiny. Some consider outside case review to be redundant, and thus not cost-effective. However, outside case review of surgical pathology material from external institutions appears to be an increasingly common practice, and it is recommended by the Association of Directors of Anatomic and Surgical Pathology.1 In a recent survey of 5 community hospitals, 7 nonacademic tertiary care hospitals, and 61 academic tertiary care hospitals, Gupta and Layfield2 found that 50% of the institutions (63 of 126) had a mandatory policy of outside case review and that 75% of the academic centers (46 of 61) detailed such a policy. An additional 37% encouraged internal review but did not have a policy, and the remaining 13% had no policy. Our objective was to determine whether secondary pathology review plays a role in optimal patient care at Mayo Clinic by examining the frequency of diagnostic disagreement during outside case review, the distribution of disagreement by organ site, the type of disagreement, and the impact of outside case review findings on patient management.
In general, studies regarding second opinions are often confined to histology involving a single organ or tissue type, with some conducted by a panel of experts from many institutions,3,4 and others by experts at a single institution.5–11 We are aware of only 4 studies that evaluated the impact of outside case review on general surgical pathology,12–15 with a rate of discordance ranging from 1.4% to 7.8%.
In this series, which we believe to be the largest series published to date, the 0.6% (457 of 71 811) rate of major disagreement with the external diagnosis was quite low. In a study by Abt et al,15 51 of 657 (7.8%) cases were discordant. In comparison, studies at the University of Iowa13 and at Johns Hopkins Medical Institutions12 found a 2.3% (132 of 5629) and a 1.4% (86 of 6171) rate of disagreement, respectively. Tsung14 found 42 of 715 cases (5.9%) with major disagreements. Although the frequency of major disagreements at Mayo Clinic is low, the changes in diagnosis were significant, and in many cases they had a major impact on patient management. The most common change was from a benign to a malignant diagnosis or vice versa. The change in diagnosis affected treatment and prognosis in 90.0% (126 of 140) and 92.1% (129 of 140) of cases, respectively. This finding is similar to results reported by Kronz et al12 from Johns Hopkins that rereviewing pathology findings led to a change in diagnosis and a subsequent change in treatment plans for 80 (93.0%) of 86 patients and to a change in prognosis for 81 (94.2%). Our findings and those of Kronz et al12 are higher than the 51.5% frequency of “documented changes in patient management” reported by Manion et al.13 However, only 68 of 132 cases in the study by Manion et al13 had follow-up, so a meaningful comparison is not possible. Tsung14 reported an effect on therapy in 31 of 42 (73.8%).
The subspecialties with the largest number of cases with disagreements in our study differ from those in previous reports. In our cases, the organ systems affected most frequently by a change in diagnosis were GI, lymph node, GU, and BST. In contrast, the previous report from the University of Iowa13 identified gynecologic, GI, head and neck, and skin as the most frequent sites involved, whereas the Johns Hopkins report12 found changes in diagnosis upon outside case review more likely to involve the skin, central nervous system, breast, and GU. These differences may be due to the differences in patient population and institutional expertise among the 3 referral centers.
For our 86 cases with additional tissue samples, the diagnoses rendered at Mayo Clinic upon outside case review agreed with the additional tissue diagnosis in most cases (84.9%; 73 of 86). However, in the remaining 15.1% of cases (13 of 86), our revised diagnosis was in disagreement with a subsequent tissue diagnosis (Table 7). This variation in pathology findings serves as a reminder that a reclassification does not necessarily denote an original misclassification by the external pathologist.16 In many instances, these are significant disagreements that affect patient management. Several are the result of an equivocal diagnosis on review that returns to unequivocal once additional material is obtained. It is our hypothesis that based on these findings, outside case review by any institution (tertiary and nontertiary) would result in diagnostic changes that would benefit patient care and improve patient safety, and we have begun to work with a number of institutions to test this hypothesis. The frequency of review misclassification in our experience is higher than that reported by Manion et al13 (3 of 59 cases with follow-up; 5.1%) and Tsung14 (3 of 42 cases with follow-up; 7.1%).
On the one hand, the low overall rate of major disagreements (0.6%) should be reassuring for pathologists in all clinical settings. Given the large volume of shared cases, pathologists appear to be in agreement most of the time. However, on the other hand, this percentage represents nearly 2 (1.85) patients per week who would potentially receive inappropriate treatment. We therefore believe that outside case review has protective benefit for patients.
We have deliberately avoided examining the potential financial savings inherent in correcting a missed diagnosis by use of secondary pathology review. It is next to impossible to calculate the potential financial (and other) savings in such cases because of the heterogeneity of the cases and the speculative nature of treatment and prognosis. From a systems and risk management perspective, outside case review would likely confer considerable benefit; thus, consideration should be given as to whether it justifies reimbursement from third parties. Nonetheless, it is our opinion that increasing financial pressures represent a challenge to widespread implementation of outside case review, despite the recommendation of the Association of Directors of Anatomic and Surgical Pathology.1 Although we cannot estimate a net monetary value, our findings show a benefit of outside case review of pathology materials for referred patients, and they suggest that widespread use of outside case review would be a mechanism to decrease diagnostic errors and better protect patients.
The authors have no relevant financial interest in the products or companies described in this article.