CAP Laboratory Improvement Programs. Surgical Pathology Extradepartmental Consultation Practices

Transcription

1 CAP Laboratory Improvement Programs Surgical Pathology Extradepartmental Consultation Practices A College of American Pathologists Q-Probes Study of 2746 Consultations From 180 Laboratories Muhammad Azam, MD; Raouf E. Nakhleh, MD Objectives. To document the practice parameters and case characteristics associated with personal (expert) consultations. We also examine the value, level of participant (customer) satisfaction, turnaround time, and rate of personal consultations. Design. We asked participants in the College of American Pathologists Q-Probes program to document cases sent for consultation during 4 months or up to 20 cases. They documented patient and specimen characteristics, the turnaround times, and the participants levels of satisfaction with the consultation experience. Participants. One hundred eighty laboratories/surgical pathology practices. One hundred seventy-two (9.6%) were from the United States; the remainder were located in Canada and Australia. Main Outcome Measures. Rate and turnaround time of consultations and participant level of satisfaction. Results. A total of 2746 consultation cases were examined for an aggregate consultation rate of 0.% (median, 0.7%). Institutions with a higher occupied bed size and a greater number of surgical pathology cases both had lower consultation rates (P.0). The median turnaround time (defined as the interval from the date on which the case was sent to the date on which the diagnosis was received) was 6 days. Twenty-nine percent and 68% of cases had a turnaround time within 3 and 7 days, respectively. Fifty-two percent of cases were sent to nationally known experts, and 32% were sent to local experts. Skin (18.0%), hematolymphoid (11.6%), and breast (9.6%) specimens were most commonly sent for consultation. In 70.% of cases, the consultant confirmed the referring pathologist s original diagnosis, but in 1.9% of cases, the consultant also added significant information. rates were higher with faster turnaround times and verbal reporting. rates were lower for cases in which the patient or the clinician requested the consultation and in which the consultant s diagnosis was ambiguous. Conclusions. This study establishes a multi-institutional consultation rate of 0.%, defines the nature of surgical pathology consultations, and demonstrates that satisfaction with consultations is associated with a faster turnaround time and receipt of additional, clinically meaningful information. (Arch Pathol Lab Med. 2002;126:40 412) Surgical pathology cases are frequently sent for consultation to other institutions. Typically, extradepartmental consultations are sought to resolve diagnostic uncertainty or to obtain input on a case from an expert. These consultations are termed personal consultations by the Association of Directors of Anatomic and Surgical Pathology (ADASP). 1 Although their numbers may be small relative to accessioned cases, personal consultations represent a significant part of pathology practice by virtue of their problematic nature. Clinicians or patients can also seek this type of expert consultation. 2 This issue of second opinion in surgical pathology has recently received public scrutiny with media attention from ABC World News (May 9, 2001) 3 and the Wall Street Journal (April 13, 2001). 4 Accepted for publication vember 28, From the Department of Pathology, rth Ottawa Community Hospital, Grand Haven, Mich (Dr Azam), and the Department of Pathology, Henry Ford Hospital, Detroit, Mich (Dr Nakhleh). Reprints: Raouf E. Nakhleh, MD, Department of Pathology, Henry Ford Hospital, 2799 West Grand Blvd, Detroit, MI ( rnakhle1@hfhs.org). The Q-Probes program of the College of American Pathologists (CAP) has attempted to measure and define quality indicators in anatomic and clinical pathology.,6 This Q-Probes study is the first multi-institutional study to measure and document the practice parameters and case characteristics associated with personal consultations. We also examined the value, level of participant (customer) satisfaction, turnaround times, and rate of personal consultations. MATERIALS AND METHODS Laboratories enrolled in the CAP s voluntary Q-Probes quality improvement program participated in this study in October December of The Q-Probes program and format for data collection and handling have been previously described in detail. 6,7 Laboratories were asked to prospectively document any surgical pathology case that was sent for extradepartmental personal consultation. For the sake of consistency, certain pertinent terms were defined (Table 1), including the definition of personal consultation. This study included any surgical pathology case that was sent for extradepartmental personal consultation. The pathologist, the clinician, or the patient may have sought this consultation. Also included were hematopathology cases (lymph node and bone mar- Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh 40

2 Table 1. Definitions of Terms Discordant diagnosis: a difference in diagnoses when a comparison is made between the original or preliminary diagnoses and the consultant s diagnosis. The discordance is due to differences in the main diagnosis such as benign to malignant or other changes such as the type of malignancy. This does not include modifications or clarifications made to the original diagnosis (eg, malignant lymphoma modified to large cleaved cell lymphoma, B-cell type). Expert pathologist: a pathologist whose diagnostic acumen in a particular field is recognized by his/her peers by virtue of his/her experience. An expert pathologist may be locally or nationally known. Extradepartmental consultation: a surgical pathology case that is sent out to a particular pathologist or to another institution. Institutional consultation: a case sent out primarily because of the patient s referral to a different hospital or clinic with a need for tissue diagnosis (case review) by pathologists at the new institution. Personal (expert) consultation: a case sent for a second opinion to a specific pathologist or pathology department. The pathologist, the clinician, or the patient may seek this consultation. Preliminary diagnosis: a written diagnosis that is issued with the intent of following up with a supplemental or final diagnosis after the consultation. time for diagnostic material return: for the purposes of this study, this is the time frame between when a surgical pathology case is sent for extradepartmental consultation and when the diagnostic material is returned to the referring pathologist. time for first contact: for the purposes of this study, this is the time between when a surgical pathology case is sent for extradepartmental consultation and when the referring pathologist first receives a verbal or written report from the consultant. row) that were sent primarily to seek expert consultation. The following types of cases were excluded from this study: institutional consultation cases resulting from a patient s referral to a different hospital or clinic (definition provided in Table 1); cytopathology consultations; muscle, renal, or other tissue biopsies that were sent in their entirety for histologic interpretation and ancillary studies (no diagnostic impression was rendered in the primary laboratory); cases sent for special tests (eg, flow cytometry, polymerase chain reaction, cytogenetics, hormone receptor assay, in situ hybridization, liver iron studies); and cases sent for research purposes only. For each case, the following specific information was collected: the patient s age and gender, the date on which the case was sent, the date on which a verbal report was given, the date on which a written report was received (including faxed reports), the reason for extradepartmental consultation, the reason for choosing the consultant, the referring pathologist s impression of the nature of the disease process before seeking consultation, the type of procedure used to obtain the specimen, the nature of the specimen, the type of diagnostic material that was sent, whether the consulting pathologist notified the referring pathologist(s) of delays (where applicable), the reason for delay (where applicable), agreement of the consultant s (expert) diagnosis with the referring pathologist s original impression, and the referring pathologist s overall satisfaction rating of the consultation process for the case (excellent, good, fair, or poor). The study concluded when 20 extradepartmental consultation cases were documented or 4 months had passed, whichever came first. An additional 1 month was allowed for receipt of reports and return of diagnostic materials. The total number of accessioned cases during 1998, the number of pathologists involved in the sign-out, and the availability of an expert pathologist within the group were also documented. In addition, each laboratory completed a questionnaire regarding laboratory practices and protocols related to extradepartmental consultation cases. The consultation rate (expressed as a percentage) was calculated for the aggregate data and for each institution that participated in this study. This was defined as the number of consultations divided by the number of accessioned cases during the study, multiplied by 100. Certain measures of consultation turnaround time were calculated for this study. The turnaround time for first contact was defined as the number of days it took to receive a diagnosis, verbal or written, from the day on which the specimen was sent to the consultant. The turnaround time for diagnostic material return was calculated in a similar fashion (date on which the case was sent to the consultant to the date on which the diagnostic materials were returned to the participant). The correlation of the main indicator variables (consultation rate and turnaround time) were assessed for each of the predictor variables separately using nonparametric Wilcoxon rank sum and Kruskal-Wallis analysis to determine if they were associated with the consultation rate. The correlations with continuous predictor variables were assessed by fitting linear regression models with the rank of consultation rate as the dependent variable. Statistically significant associations are defined at the P.0 level. Some participating institutions did not answer all of the questions on the demographics form and/or on the input forms. These institutions were excluded only from the tabulations and analyses that required the missing data element. One institution with a 100% consultation rate was excluded from these analyses. RESULTS A total of 180 institutions submitted data for this study. Most institutions (9.6%) were located in the United States; the remainder were located in Canada and Australia. Of the participating institutions, 28.% were teaching hospitals, and 12.4% had a pathology residency program. Most institutions (92.3%) participating in this study were accredited by the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), while the CAP accredited 88.3% of laboratories participating in the study. The participant institutional sizes were as follows; 3.9% had fewer than 10 beds; 39.1%, 11 to 300 beds; 14.8%, 301 to 40 beds; 4.7%, 41 to 600 beds; and.%, more than 600 beds. Private nonprofit institutions comprised 64.8% of institutions, 14.% were governmental, 4.8% were private for profit, 4.1% represented independent laboratories, and 11.8% were other. Fifty-one percent of institutions were within a city, 29% were suburban, and 20% were rural. The numbers of surgical pathology cases accessioned during 1998 and during the study period were (median, 9673) and (median, 219), respectively. Of the 180 institutions that submitted data for this study, 1 participants reported their number of surgical pathology cases accessioned ( ) during the study. Of these cases, 2428 were sent for extradepartmental consultation for an aggregate consultation rate of 0.%. The median institution had a consultation rate of 0.7% the 10th and 90th percentile institutions had 0.2% and 2.0% consultation rates, respectively. The median turnaround times for first contact and diagnostic material return were 6 and 13 days, respectively. For all cases in this study, 29.3% had a turnaround time of 3 days or less, and 67.7% had a turnaround time of 7 days or less. Diagnostic material was returned within 14 days in 8.7% of cases. The distributions of median turnaround times and the percentage of case that met various turnaround time criteria are shown in Figures 1 and 2, respectively. By the end of this study, diagnostic material 406 Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh

3 Figure 1. Distribution of individual institutional median turnaround times for first contact and diagnostic material return. Dashed line indicates median; hinge, 2th to 7th percentiles; and fence, 10th to 90th percentiles. was returned for 67.9% of cases. A verbal report was given in 38.0% of cases. The satisfaction rate for the overall consultation process (defined as the percentage of cases with an excellent or good satisfaction rating) was 94.3%. Several aspects of practice and their effect on surgical pathology consultations were studied. A higher occupied bed size and a greater number of surgical pathology cases accessioned in 1998 were both found to have statistically significant associations (P.0) with lower consultation rates. Several characteristics of surgical pathology cases sent for consultation were found to have statistically significant associations with measures for surgical pathology consultation. These findings are as follows. When either Figure 2. Distribution of the percentage of cases at individual institutions that met turnaround time goals for first contact within 3 days, first contact within 7 days, and diagnostic material return within 14 days. Dashed line indicates median; hinge, 2th to 7th percentiles; and fence, 10th to 90th percentiles. the patient or clinician sought these consultations, the satisfaction rate was lower. Moreover, turnaround times were longer when patients requested consultations. The main reasons for personal consultation and the corresponding satisfaction rates are shown in Table 2. The participants reasons for selecting the consultant pathologist and the corresponding satisfaction rates are shown in Table 3. The use of local experts and low-cost consultants tended to be associated with lower satisfaction rates. Specimens from a variety of sources were sent for consultation during the study (Table 4). Consultations on female genital tract spec- Table 2. Associations Between Reasons for Personal (Expert) Consultations and Consultation Measures Main Reason for Personal (Expert) Consultation Diagnostic uncertainty of the referring pathologist(s) Disagreement of opinion between 2 or more pathologists To seek additional information/recommendation from an expert Clinician s request Patient s request Time for First Contact Median Time, d of cases with a satisfaction rating of excellent or good With Overall Consultation Process Table 3. Association Between Reasons for Choosing Consultants and Ratings Reason for Choosing Consultant Nationally known expert Local expert in field Overall satisfaction with prior service Administrative ease of billing (eg, direct billing of insurance company, etc) Low cost of the consultant charges Better turnaround time Quality of consultant s report Business contract with a particular pathology group of cases with a satisfaction rating of excellent or good. Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh

4 Table 4. Nature of the Specimen Skin Lymph node, bone marrow, spleen Breast Gastrointestinal tract Male genital tract Female genital tract Pulmonary system Neuropathology Head and neck Soft tissue Liver Endocrine glands Bone and joints Urinary tract Pleura, pericardium, mediastinum, peritoneum, omentum, retroperitoneum Placenta, products of conception Cardiovascular system Thymus Associations Between the Nature of the Specimen and Consultation Measures of All Consultations Reported During Study* Time for First Contact Within 7 d With Time 7 d of cases with a satisfaction rating of excellent or good. With Overall Consultation Process Table. Associations Between Diagnostic Material Sent to Consultant and Diagnostic Material Return Measures Diagnostic Material Sent to Consultant* Representative H&E-stained slides All H&E-stained slides Specially stained slides Paraffin-embedded tissue blocks Fresh or formalin-fixed tissue (gross specimen) Cases Diagnostic Material Return Time Median Time With Time 14 d * H&E indicates hematoxylin-eosin. Some participants did not provide the necessary data to calculate all measures for all characteristics. Diagnostic Material Return by End of Study Cases Returned by End of Study imens tended to have higher satisfaction rates. Consultations on neuropathology specimens tended to have a poorer chance of a turnaround time for first contact within 7 days. An assortment of diagnostic material was sent for consultation (Table ). Cases for which all hematoxylineosin stained slides or paraffin-embedded tissue blocks were sent tended to have a longer turnaround time for diagnostic material return and a poorer chance of diagnostic material return within 14 days. The former also had a poorer chance of return by the end of the study. Cases in which the reports took longer than 1 week and the consultant did not notify the referring pathologist of the delay were associated with lower satisfaction rates (Table 6). Cases in which the consultant s diagnosis confirmed the referring pathologist s original diagnosis but added significant information tended to have a higher satisfaction rate (Table 7). However, when the consultant s diagnosis was ambiguous and not helpful, the satisfaction rate was comparatively lower (Table 7). The participants overall satisfaction with the consultation process was associated with shorter turnaround times for first contact and provision of a verbal report (Table 8). Table 9 shows other characteristics of surgical pathology consultation cases that were not associated with consultation measures. In most of the cases (69.8%), there was no delay in the report. But of those cases that were delayed, the delay was usually due to immunohistochemistry (11.1%) or factors not explained in the study (12.6%). Study participants also provided information about their institutions. ne of the institutional practices studied were found to have statistically significant associations with the surgical pathology consultation measures. At the median institution, 4 pathologists were involved in sign-out of surgical pathology cases; at the 10th and 90th percentile institutions, 2 and 8 pathologists were involved, respectively. The average number of pathologists involved in sign-out was.2. Most of the institutions that participated in the study (9.9%) did not have a nationally or locally known 408 Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh

5 Table 6. tification of the Referring Pathologist(s) if Consultation Took More Than 1 wk t applicable no delay Associations Between tification of Delays and Consultation Measures Time for First Contact Median Time, d With Time 3 d With Time 7 d of cases with a satisfaction rating of excellent or good. With Overall Consultation Process Table 7. Association Between Consultant s Diagnosis and Ratings Consultant s Diagnosis of All Consultations Reported During Study* With Overall Consultation Process Was discordant from the referring pathologist s original impression Was 1 of the referring pathologist s several differential diagnoses Confirmed the referring pathologist s original diagnostic impression Confirmed the referring pathologist s original diagnosis but added significant information Was ambiguous and not helpful of cases with a satisfaction rating of excellent or good Table 8. Rating Excellent Good Fair Poor Associations Between Rating and Surgical Pathology Consultation Measures Time for First Contact Median Time, d of cases in which a verbal report was given Cases in Which a Verbal Report Was Given Verbal Report expert in a branch of surgical pathology in their pathology groups (Table 10). The most frequently represented experts in participant pathology groups were hematopathologists (23.6%) or dermatopathologists (22.3%). Most of the study participants (6.0%) rarely or never contacted a consultant to check availability before sending a case (Table 11). The majority of institutions (96.8%) routinely provided background information in the cover letter to the consultant, including the reasons for consultation, the working or differential diagnosis, and any pertinent clinical history including previously diagnosed tumors. Most institutions (80.9%) also sent the consultant a copy of an incomplete or complete surgical report for the site, a gross description, and a section summary of the case. Most institutions (74.2%) reported that they have a written protocol or guideline for sending consultation cases. institutional protocols for sending consultation cases are described in Table 12. The median institution reported using a total of 6 consultants during this study; the 10th and 90th percentile institutions reported using 2 and 13 consultants, respectively. COMMENT Consultations are widely used in the field of medicine and are routinely used by nearly all medical specialties when the treating physician is faced with a complex or unfamiliar clinical situation. Likewise, in diagnostic pathology, personal (expert) consultations are widely sought and represent a significant part of pathology practice. Pathology consultations are perhaps more widely used because of the ability to easily send and/or duplicate the diagnostic material. In a recent CAP Q-Probes study of amended reports, Nakhleh and Zarbo 8 reported extradepartmental consultation as the most common cause for a change in provisional diagnosis and the second most common cause of amended reports. Recently, the public sector has focused much attention on medical errors, and second opinions have been touted as possible remedies. The Institute of Medicine 9 published a comprehensive report on medical errors and public safety in the United States. After outlining these issues, the report contained several recommendations to prevent er- Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh 409

6 Table 9. Characteristics of Surgical Pathology Consultation Cases That Were t Associated With Consultation Measures Patient sex Male Female Characteristic Cases (%)* 124 (46.6) 1426 (3.4) Pathologist s impression of the nature of the disease process before consultation Malignant neoplasm 1149 (42.1) Inflammation, immunologic disease, infections 376 (13.8) Benign proliferation/hyperplasia/atypia 334 (12.2) Benign neoplasm 272 (10.0) Low malignant potential/borderline neoplasms 233 (8.) Premalignant neoplasia (eg, carcinoma in situ) 191 (7.0) Hereditary/congenital disease Metabolic disease Therapy-induced changes Unknown Type of procedure used to obtain the specimen Excisional biopsy Needle core biopsy Incisional biopsy Resection Endoscopic biopsy Curettings Segmental resection Forceps Unknown/unspecified 20 (0.7) 16 (0.6) 12 (0.4) 113 (4.1) 42 (1.) 103 (37.8) 472 (17.2) 348 (12.7) 307 (11.2) 246 (9.0) 108 (3.9) 49 (1.8) 18 (0.7) 134 (4.9) 43 (1.6) Factors to which a delay in the report could be attributed Immunohistochemistry 284 (11.1) Electron microscopy 13 (0.) Fluorescence in situ hybridization (FISH) 4 (0.2) Any other molecular biology procedure 20 (0.8) Evaluation of additional material 62 (2.4) 11 (.9) The delay was not attributed to any of the above 321 (12.6) t applicable no delay experienced 1779 (69.8) Cases 0th Percentile (10th 90th Percentile) Patient age, y (24 78) * Some participants did not provide the necessary data to calculate all measures for all characteristics. rors and improve public safety. This document reported a staggering number of to deaths per year resulting directly from many kinds of medical errors. Consequently, it has received strong media attention in an attempt to raise public awareness. 10 The proportion of these errors attributable to diagnostic pathology is unknown, but is likely very small. However, a small series of recently published articles have discussed the routine review of surgical pathology diagnostic material by a second pathologist, with the aim of reducing diagnostic errors. 2,11 13 Despite all of the discussion, very little is known about personal (expert) consultations. This Q-Probes study is the first multi-institutional study to measure and document Table 10. Areas in Which Nationally or Locally Known Experts in a Branch of Surgical Pathology Were Employed by the Participating Pathology Groups Branch of Surgical Pathology Hematopathologist Dermatopathologist Breast pathologist Gynecologic pathologist Participants (%) 37 (23.6) 3 (22.3) 24 (1.3) 24 (1.3) Gastrointestinal tract and liver pathologist 19 (12.1) Neuropathologist Nephropathologist Genitourinary pathologist Soft tissue pathologist Transplant pathologist 18 (11.) 16 (10.2) 14 (8.9) 13 (8.3) 13 (8.3) Head and neck (including oral and thyroid) pathologist 11 (7.0) Pediatric pathologist 11 (7.0) Pulmonary pathologist 10 (6.4) Cardiovascular pathologist 9 (.7) Endocrine pathologist (3.2) 16 (10.2) ne 94 (9.9) Table 11. Communication With Consultants by Participating Institutions Participants (%) How often referring pathologist called the consultant before sending out a case (to check availability) Always ( 9% of the time) 13 (8.3) Usually (76% 9% of the time) 1 (9.6) Often (1% 7% of the time) 9 (.7) Sometimes (26% 0% of the time) 18 (11.) Rarely (% 2% of the time) 61 (38.9) Never ( % of the time) 41 (26.1) Referring pathologist routinely stated in the cover letter the reasons for consultation, working diagnosis or differential diagnosis, and pertinent clinical history with previously diagnosed tumors 410 Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh 13 (96.8) (3.2) Referring pathologist routinely sent a copy of an incomplete or complete surgical report for the site, gross description, and section summary of the case to the consultant 127 (80.9) 30 (19.1) Billing information was routinely provided to the consultant when sending out cases 80 (1.0) 77 (49.0) the practice parameters and case characteristics associated with personal consultations. In the present study, the aggregate rate of personal consultations was 0.% (median, 0.7%; 90th percentile, 2.0%). Institutions with a higher occupied bed size and a greater number of surgical pathology cases accessioned were both found to have significantly lower consultation rates. The use of personal (expert) consultation depends on multiple factors. The consensus conference of the American Society of Clinical Pathologists states that a pathologist should

7 Table 12. Protocols for Sending Out Cases for Consultation Among Participating Institutions Before sending out cases for consultation duplicate sets of recut slides are made for the institution s files for the consultant Participants (%) 36 (23.1) 47 (30.1) but only if the paraffin blocks are sent 24 (1.4) if the lesion is small (present on only one slide) 8 (.1) recuts are not usually done 48 (30.8) How institutions keep track of consultation cases Each pathologist in the group keeps track of his or her own cases 32 (20.4) Designated secretary/clerk keeps track of all of the extradepartmental surgical pathology consultations (centralized handling) 120 (76.4) specific method of tracking (3.2) 1 (0.6) How institutions keep track of the diagnostic material sent out A log book 110 (69.6) Periodical computer-generated list in the department 40 (2.3) 10 (6.3) Don t keep track of material sent out 3 (1.9) 6 (3.7) 32 (20.4) 34 (21.7) If the pathologist is sending out a case, the consultant is usually paid by: Hospital Pathology department Insurance company Pathology department and insurance company (shared expense) 14 (8.9) Pathologist himself/herself Patient 8 (.1) 19 (12.1) Don t know (3.2) Institution is under any constraint with regard to extradepartmental consultations, obligated by business contract to send cases to a particular consultant 2 (1.3), monetary cap for extradepartmental consultations 4 (2.), other institutional limitations or restrictions 8 (.1), other Institution includes surgical pathology extradepartmental consultations in quality assurance program as a case review mechanism (3.2) 139 (88.0) 134 (8.4) 23 (14.7) nominate for second opinion cases that are problem prone, as defined by the individual, the group, or the literature. 2 The availability of an expert within a department may mitigate the need for consultation, and larger groups are more likely to have these experts on staff. Most of the institutions that participated in the study (9.9%) did not have an expert in a branch of surgical pathology in their pathology group (Table 10). Of those that did, the most common areas of specialty of these experts were hematopathology (23.6%) and dermatopathology (22.3%). The bulk of consultations were for specimens of skin, hematopoietic tissues, breast, gastrointestinal tract, male and female genital tracts, and the pulmonary system (Table 4). Consultation rates depend on the expertise of the pathology staff and the case mix of that institution; therefore, it is not possible to deduce an acceptable or benchmark rate of consultation. A low or high consultation rate does not necessarily indicate an appropriate or inappropriate level. In a 4-month time frame, the median institution reported using a total of 6 consultants during this study, and the 10th and 90th percentile institutions reported using 2 and 13 consultants, respectively. These facts point to the breadth of pathology and indicate a rather liberal use of consultants. This finding is significant in that these numbers reflect cases in which the pathologist has sought help and do not include cases in which the patient is referred to other institutions (institutional consultations). The issue of routine or mandatory review by a second pathologist as addressed in some recent studies 2,11 13 is complex. Intradepartmental slide reviews are primarily done as a part of quality assurance to find correctable errors. On the other hand, extradepartmental consultation is a formal consultation process, which necessitates a review of all of the relevant slides in the context of given clinical information and the synthesis of a formal pathologic diagnosis. Therefore, this consultation is a result of the practice of medicine and should be clearly distinguished from the intradepartmental quality assurance and administrative activities that also involve slide reviews. This distinction is necessary to keep the value and reimbursement of this valuable service. 14 The aggregate satisfaction rate (percentage of cases with a rating of excellent or good) for the overall consultation process was 94.3%. The satisfaction rate remained relatively high (88.1%) even when the consultant s diagnosis was discordant from the diagnosis of the referring pathologist (Table 7). In only 0.7% of cases, the consultant s diagnosis was ambiguous and not helpful, and these cases had the lowest satisfaction rates. Higher satisfaction rates were associated with shorter turnaround times and provision of a verbal report. A verbal report, however, was given in only 38.0% of cases. In the majority of consultation cases, turnaround time is important, and a delay in the consultant s diagnosis can be frustrating for the referring pathologist and for the patient. This problem can be at least partially rectified if the consulting pathologist notifies the referring pathologist about possible delays. Among cases in which the provision of the report took longer than 1 week, the consultant did not notify the referring pathologist in 68.% of applicable cases. Most of delayed cases were due to either immunohistochemistry (11.1%) or factors not explained in the study (12.6%). On the other hand, most of the study participants (6.0%) rarely or never contacted a consultant to check availability before sending out a case (Table 11). Considering that the majority of the consultants are active in teaching outside of their institutions, this simple step may eliminate any unnecessary delay. When either the patient or the clinician sought a surgical pathology consultation, the pathologist gave the process a lower satisfaction rating. Moreover, the turnaround time was longer when the patient requested the consultation. Whereas the prolongation of the turnaround time does not have an obvious explanation, the lower satisfaction rating by the pathologist may be related to an apparent loss of control over the diagnostic process. rates also tended to be lower with the use of local area experts and low-cost consultants. Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh 411

8 There are only a handful of reports in the literature regarding the topic of second opinion in anatomic pathology. The study material in these reports was primarily derived from single institutional reviews of institutional consultations, in which cases were sent out primarily because of the patient s referral to a different hospital or clinic with a need for tissue diagnosis (case review) by pathologists at the new institution. The reported incidence of discordant diagnosis in these reports is quite variable, depending on the study design, the institution s level of tertiary care, and the specific organ system for which consultation was obtained. Rates ranging from 0.26% to 1.2% have been found in studies in which a second pathologist comprehensively reviewed the cases. 1,16 On the other hand, in a study of 777 patients by Abt et al, 17 there was some level of disagreement in 9.1% of the cases and a.8% incidence of clinically significant changes in the pathology diagnosis. Similarly, Malhotra et al 18 identified 28 cancer-related cases with disagreement (10.%) in a study of 27 cases in a large university referral center. In a review of prostate biopsies, Epstein et al 19 found significant disagreement in 1.3% of cases, while Selman et al 20 found significant disagreement in 4.7% of gynecologic pathology cases. In a study of sarcoma cases, this discordance rate was 22% as reported by Harris et al. 21 In this first multi-institutional Q-Probes study, the discordance rate was 6.2%. Personal consultation cases, however, are usually not finalized since pathologists are actively seeking input to arrive at an appropriate diagnosis. Therefore, this rate of discordance does not have the same impact on patient care as discrepancies that are noted after cases have been finalized and that may have been acted on. It is interesting to point out that the ADASP estimated an acceptable threshold for clinically significant disagreement following arbitration is 2%, as applied to those cases in which it is decided that the correct interpretation is that from the outside institution. 22 The higher discordance rates of many studies are likely due to the selection bias and complex nature of the cases. In either case, diagnoses in anatomic pathology are essentially judgments dependent on available tissue and clinical information about the patient s condition. 23 Often the diagnosis rendered cannot be classified as right or wrong, and no single expert is an absolute authority. The only available criterion standard to judge the correctness of a diagnosis is good clinical correlation and adequate follow-up. In conclusion we offer the following recommendations. To pathologists sending out cases: (1) choose a consultant who will provide appropriate information needed for patient care, (2) determine the availability of your consultant before sending the case, (3) be as specific as possible in your request to the consultant so that your concerns are addressed, (4) try to anticipate the consultant s needs for blocks or other diagnostic materials, and () communicate to the involved clinicians your intention to seek external consultation. To consultants: (1) in your absence, have your office notify pathologists of your absence and expected return date, (2) try to address the referring pathologist s needs for specific information as cases dictate, and (3) call or fax reports as soon as they are available; calling may help obtain additional or specific information. References 1. Association of Directors of Anatomic and Surgical Pathology. Consultations in surgical pathology. Am J Surg Pathol. 1993;17: Tomaszewski JE, Bear HD, Connally JA, et al. Consensus conference on second opinion in diagnostic anatomic pathology. Am J Clin Pathol. 2000;114: McKenzie J. Misdiagnosing cancer: patients should always consult a second opinion. ABC World News Tonight With Peter Jennings. Available at: abcnews.go.com/sections/wnt/worldnewstonight/wnt01008 misdiagnosing cancer feature.html. Accessed May 8, Parker-Pope T. Risk of error may justify second opinion on pathology reports. Wall Street Journal. May 2001:B1.. Schifman RB, Howanitz PJ, Zarbo RJ. Q-PROBES: a College of American Pathologists benchmarking program for quality management in pathology and laboratory medicine. Adv Pathol. 1996;9: Howanitz PJ, Walker K, Bachner P. Quantification of errors in laboratory reports: a quality improvement study of the College of American Pathologists Q- PROBES program. Arch Pathol Lab Med. 1992;116: Howanitz PJ. Quality assurance measurement in departments of pathology and laboratory medicine. Arch Pathol Lab Med. 1990;114: Nakhleh R, Zarbo RJ. Amended reports in surgical pathology and implications for diagnostic error detection and avoidance: a College of American Pathologists Q-PROBES study of 1,667,47 accessioned cases in 39 laboratories. Arch Pathol Lab Med. 1998;122: Kohn LT, Corrigan JM, Donaldson MS, eds. To Err is Human: Building a Safer Health System. Washington, DC: Institute of Medicine, National Academy Press; Stolberg SG. Do no harm: breaking down medicine s culture of silence. New York Times. December, 1999; sect 4: Kronz JD, Westra WH, Epstein JL. Mandatory second opinion surgical pathology at a large referral hospital. Cancer. 1999;86: Sirota RL. The Institute of Medicine report on medical errors. Implications for pathology. Arch Pathol Lab Med. 2000;124: Woeste S. Second opinions for pathology reports. Lab Med. 2001;32: Fitzgibbons PL, Compton CC. Be careful what you wish for: on calls for mandatory second opinion. Arch Pathol Lab Med. 2001;12: Safrin RE, Bark CJ. Surgical pathology sign out: routine review of every case by a second pathologist. Am J Surg Pathol. 1993;17: Lind AC, Bewtra C, Healy JC, Sims KL. Prospective peer review in surgical pathology. Am J Clin Pathol. 199;104: Abt AB, Abt LG, Olt GJ. The effect of interinstitution anatomic pathology consultation on patient care. Arch Pathol Lab Med. 1998;122: Malhotra R, Massimi G, Woda BA. Interinstitutional surgical pathology consultation and its role in on patient management [abstract]. Mod Pathol. 1996; 9:16A. 19. Epstein JI, Walsh PC, Sanfilippo F. Clinical and cost impact of second pathology review of prostate biopsies prior to radical prostatectomy. Am J Surg Pathol. 1996;20: Selman AE, Niemann TH, Fowler JM, et al. Quality assurance of second opinion in gynecologic oncology. Obstet Gynecol. 1999;94: Harris M, Hartley AL, Blair V, et al. Sarcomas in north west England: histopathological peer review. Br J Cancer. 1991;64: Association of Directors of Anatomic and Surgical Pathology. Recommendations on quality control and quality assurance in anatomic pathology. Am J Surg Pathol. 1991;1: Murphy WM, Rivera Ramirez I, Luciani LG, Wajsman Z. Second opinion of anatomical pathology: a complex issue not easily reduced to matters of right and wrong. J Urol. 2001;16: Arch Pathol Lab Med Vol 126, April 2002 Extradepartmental Consultation Practices Azam & Nakhleh