Abstract
Background:
Clear, accurate, and complete reporting of postsurgical pathology is crucial for the correct evaluation and management of thyroid cancer patients. This study aimed to describe the completeness, as defined by international guidelines, of pathology reporting in a cohort of newly diagnosed thyroid cancer patients in New South Wales (NSW) and to identify factors associated with the completeness of reports.
Methods:
Postsurgical pathology reports, held by the NSW Central Cancer Registry, for 448 thyroid cancer patients were reviewed. Presence or absence of recommended key features (tumor histology type, maximum dimension, focality, completeness of excision, extrathyroidal extension, lymphovascular invasion, and lymph node involvement) was recorded. Associations between the number of key items reported and several patient characteristics were investigated.
Results:
For 285 (63.6%) patients one or more key pathological features were missing, with 177 (39.5%) missing one only, 88 (19.6%) missing two, and 20 (4.5%) missing three or more. Extrathyroidal extension was the most poorly reported key feature, being present in only 228 (50.9%) reports [95% confidence interval 46.2, 55.6]. Pathology reports were less complete for patients with small tumor size (p<0.001) or localized spread (p<0.001). Synoptic reports were significantly more complete than narrative-style reports (98.3% vs. 27.1%, p<0.001).
Conclusions:
Postsurgical pathology reporting of differentiated thyroid cancer in NSW was found to be far from complete, with 64% of reports missing information on at least one feature that is considered internationally to be a critical factor in the prognosis and treatment of thyroid cancer patients. Synoptic reporting reduces the number of key features missing from pathology reports.
Introduction
Surgery is usually the primary treatment for thyroid cancer (7), with further treatment and follow-up decisions then informed by the postsurgical pathology report and disease staging. The accuracy and completeness of this report are therefore crucial for effective disease management. However, it has been found that postsurgical pathology reports for many cancer types are often missing information that is central to informed treatment decision making (8 –11). With an increasing number of patients diagnosed with thyroid cancer, the effective and timely treatment of this disease, and therefore the accurate and complete reporting of pathology findings, is of utmost importance.
Several international pathology committees have developed recommended formats for thyroid cancer pathology reporting (12,13), and, recently, a new structured reporting protocol was released by the Royal College of Pathologists of Australasia (14). As the use of these guidelines is recommended, but voluntary, assessing the current completeness of thyroid cancer pathology reporting in NSW may be useful in encouraging adherence to these new guidelines. Further to this, by identifying any patterns in the quality of reporting it may be possible to identify particular areas for improvement.
Methods
This pathology review was part of a larger study—Pathways to the Diagnosis of Thyroid Cancer in NSW (15). This was a population-based cross-sectional study of newly diagnosed cases of thyroid cancer in NSW, Australia. The Pathways study and this pathology audit were approved by the Cancer Council NSW and Cancer Institute NSW's Human Research Ethics Committees.
Subjects
Patients with thyroid cancer were ascertained directly from notifications to the population-based NSW Central Cancer Registry (CCR). Patients were eligible to participate if they were diagnosed in the period 1 May 2006–31 August 2008; aged between 18 and 79 years at diagnosis; had a pathologically confirmed diagnosis of papillary, follicular, or medullary cancer; and were able to provide written, informed consent. Four hundred fifty-two patients were recruited and 448 pathology reports were reviewed. Four of the participants in the Pathways study were excluded because they did not provide consent (n=1), or their postsurgical pathology report was not available (n=3).
The pathology review form
Relevant information was extracted from the pathology reports onto an audit form based on the 2006 guidelines for thyroid cancer pathology reporting from the Royal College of Pathologists (13). The form captured available information regarding the key features of tumor histology type, focality, maximum dimension, completeness of excision, extrathyroidal extension, lymphovascular invasion, and lymph node involvement (number taken, number involved, and extranodal spread). Additional information on other thyroid disease in the sample, location of tumors, and number of blocks were also recorded. The format of the report was recorded as synoptic (structured checklist) or narrative (free text).
Pathology report review
It is a statutory requirement that all pathology laboratories in NSW provide a copy of reports that diagnose cancer to the population-based CCR. Thus, once study participants provided written consent, a trained researcher could review the pathology reports on site at the CCR. All relevant and recorded information was collected using our audit form. For features that could be present or absent, such as extrathyroidal extension or complete excision, the absence of the feature had to be specifically described in the pathology report to be recorded as absent in the audit form. This allowed us to distinguish between reports that did not provide any information on this feature from those that reported a negative result.
Additional data
Participants of the Pathways study completed a questionnaire that captured demographic information and details of their diagnostic pathway. Additional data were obtained from the CCR for all study participants, including age at diagnosis, date of diagnosis, local government area and area health service of residence at time of diagnosis, tumor morphology code, and spread of disease. Whether patients underwent surgery in a public or private facility was derived from Medicare claims data, provided with the individual's consent by Medicare Australia.
Sample size
Sample size calculations for the Pathways (15) study were based on 2001 incidence rates of thyroid cancer in NSW (16). It was estimated that ∼663 new cases of papillary, follicular, or medullary thyroid cancer would occur in 18 months. Experience from previous studies using a similar recruitment process predicted that 60%–70% of all new eligible cases would be recruited (17), resulting in at least 400 participants. This number of participants would provide at least 80% power to detect small differences in dichotomous comparisons. With 448 pathology forms, proportions are estimated to within 4.6% of the true proportion.
Statistical analysis
The completeness of reporting for each data item was described using frequencies and proportions (with 95% confidence intervals [CIs]). A summary variable for the completeness of each report was calculated by summing the number of missing core data items (key features) as defined by the Royal College of Pathologists (listed in Table 2) (13), with the exclusion of lymph node data due to concerns about its accuracy. Associations between completeness of reporting (summary variable) and participants' characteristics (tumor size, disease spread, cancer type, patients' region of residence and level of health insurance, and surgery in a public or private facility) were examined using chi-square tests. For this analysis we used disease spread (not specified, localized to the thyroid only, or spread beyond the thyroid) as recorded by the CCR rather than that derived from the pathology forms so that it was independent of the completeness of reporting. Initial statistical analysis was performed using the statistical software package SPSS v15.0 (SPSS for Windows, release 15.0.1, 2006; SPSS Inc., Chicago, IL), while exact CIs for proportions were obtained using STATA 11 software (StataCorp 2010 Statistical Software, release 11.0; STATA, College Station, TX).
Results
Characteristics of the sample
Of the participants, 341 (76.1%) were women and 107 (23.9%) were men. The median age at diagnosis was 48 years for women (interquartile range [IQR] 39–57.5) and 53 years for men (IQR 42–63). There were 381 patients (85.0%) who were diagnosed with papillary thyroid cancer, 56 (12.5%) had follicular thyroid cancer, and 11 (2.5%) had medullary thyroid cancer (Table 1).
Nonparticipants' data for these variables are not available.
TAFE, technical and further education institution.
When the study sample was compared with eligible nonparticipants (n=572) using aggregate data provided by the CCR, we found that there were no significant differences for gender, age at diagnosis, cancer type, or disease spread at diagnosis. There was, however, a significant difference in place of residence (p<0.001), with an overrepresentation of residents from nonmetropolitan areas (rural and “other urban”) in the study sample (Table 1).
Completeness of pathology reports
The majority (285, 63.6%) of the pathology reports were missing at least one of the key items assessed. There were 163 (36.4%) pathology reports that were complete for all key features, while 177 (39.5%) were missing one key item, 88 (19.6%) were missing two, and 20 (4.5%) were missing three or more key items. Extrathyroidal extension was the least frequently reported feature, missing from nearly half of the reports (Table 2). The other items (tumor histology type, maximum dimension, focality, completeness of excision, and lymphovascular invasion) were recorded in 75%–100% of reports. For reports missing two key items the most common combination of missing information was extrathyroidal extension and lymphovascular invasion (59, 13.2% of all reports). And of reports missing more than two key items 16 (3.6% of all reports) were missing information on extrathyroidal extension, lymphovascular invasion, and completeness of excision (Table 3).
Core data items defined by the Royal College of Pathologists, Standards and Datasets for Reporting Cancers, dataset for thyroid cancer histopathology reports, February 2006, Coordinators: Dr. Anne Marie McNicol and Dr. Sarah Johnson.
CI, confidence interval.
We could not accurately assess the completeness of recording for lymph node involvement because a clear distinction between reports that recorded that no nodes were available for assessment and reports that did not provide any information on the assessment of lymph nodes was not made in the data extraction. However, for those that specifically referred to lymph node assessment, the reporting of number of nodes resected (97.8%; 95% CI 95.0–99.0) and the number of nodes involved (99.6%; [95% CI 97.6–100.0]) was nearly complete (Table 2).
The completeness of pathology reporting, as defined by the number of core data items reported, was significantly associated with tumor size (largest dimension of the primary tumor, with cases missing this information excluded from the analysis, p<0.001) and disease spread (as recorded by the CCR, p<0.001) (Table 4). Cancer type, region of residence, and treatment in a public or private facility were not significantly associated with the completeness of pathology reporting.
n=448 unless otherwise specified.
Twelve cases where tumor size was not recorded in the pathology report were excluded from this analysis. Therefore, for this analysis, the summary variable of missing core data items includes only five core data items.
Three participants did not answer this question on the questionnaire.
An exact definition of form type was not possible in one case.
The type of pathology report (synoptic or narrative) was significantly associated with its completeness (p<0.001). Of the synoptic reports, 98.3% (58 of 59) were complete and one report (1.7%) was missing one key item. In comparison 27.8% of narrative reports (108 of 388) were missing two or more key features and only 105 (27.1%) narrative reports were complete (Table 4).
Discussion
Clear, complete, and accurate reporting of postoperative pathology findings is central to the successful treatment of people with thyroid cancer. When we assessed postsurgical pathology reports for a cohort of patients with differentiated thyroid cancer we found that in general the pathology reporting was far from complete. Approximately two-thirds of the reports assessed were missing information on at least one key feature, and nearly half of all reports were missing any information on extrathyroidal extension. The reports for small cancers, and cancers localized to the thyroid tended to be less complete than the reports for larger or more advanced tumors. Synoptic reports were significantly more complete than narrative-style reports.
Surgery is the primary treatment for all types and stages of differentiated thyroid cancer (7). Subsequent treatment generally consists of radioiodine ablation or therapy, with the aim of destroying any residual thyroid tissue and any potential metastatic disease (18). The decision to treat patients with radioiodine is dependent on many factors, such as the size of the primary tumor, completeness of excision, the age of the patient, the presence of any lymph node involvement and/or distant metastases, histological type, and extrathyroidal extension (18 –20), most of which should be investigated in the postoperative pathology assessment. With some concerns about the potential long- and short-term side effects of radioiodine treatment (20), and the cost, stress, and discomfort associated with the thyroxine replacement withdrawal, low iodine diet and time spent in isolation that are required for successful radioiodine therapy, the decision to recommend this treatment can be complicated and must be informed by all relevant information. Thus, the completeness of the postoperative pathology report is of particular importance in providing best practice treatment for thyroid cancer patients.
Our finding that pathology reports for small or localized tumors tended to be less complete than those for larger tumors is perhaps not surprising. Small tumor size is generally associated with less advanced cancer, and it is therefore likely that the missing information may indicate unreported negative results. This may be due to an underlying assumption that when the tumor is small or localized the reporting of further absent features is unnecessary, and that these negative findings are implied. The nonreporting of absent features is problematic, especially as, with regards to tumor size, it has been shown that for differentiated thyroid cancer there are other features that are of possibly equal importance in predicting recurrence and mortality (7,21,22). Further, it has been found that papillary microcarcinomas (tumors of <1 cm) can display aggressive features and levels of recurrence similar to that of larger papillary thyroid cancers (21,23), indicating that other prognostic factors in addition to tumor size need to considered (24). Thus, when assessed, even the absence of all other aggressive features should be explicitly recorded in the postsurgical pathology report.
Extrathyroidal extension was missing in 49% of the pathology reports. It is likely that the lack of information on extrathyroidal extension indicates the unreported absence of this feature. This is a concerning result given that extrathyroidal extension is one of the most important factors for predicting recurrent or persistent disease, and even in cases where the tumor is considered low risk in terms of tumor size and histopathology, the presence of extrathyroidal extension is indicative of a higher risk of recurrence (25). Extrathyroidal extension is also an important factor in the decision to administer postsurgical radioiodine to patients, even in the absence of any other high-risk features (18,26).
The type of pathology report (narrative or synoptic) was also recorded during data extraction, and it is interesting to note that all except one of the synoptic reports were complete, while the remaining one report was missing only one core data item. In comparison 73% of the narrative forms were missing core data items. In addition to the completeness of the reports it was found that the extraction of information from synoptic reports was noticeably easier and faster than for narrative reports, although this assessment is based purely on the experience of one researcher, and was not recorded routinely. A crucial difference in the process of extracting data from the synoptic reports was that unlike narrative reports it is never necessary to search for information, or to extrapolate from the available data. These results and observations add support to the findings of several other studies that have called for the implementation of structured reporting to improve pathology reporting practices (8,11,27), and should further encourage pathologists to adopt the available guidelines.
A strength of this study is that it was population based, with eligible patients identified through the NSW CCR. Unfortunately, the complex three-tiered consent process mandated by ethics committees to recruit through the Cancer Registry resulted in a low participation rate, with 43.8% of all eligible thyroid cancer patients included (15). We are, however, confident that the study sample is representative of all eligible patients, as there were no significant differences between the study sample and eligible nonparticipants for age, gender, cancer type, and disease spread. There was, however, a significant overrepresentation of residents from nonmetropolitan areas (rural and “other urban”) in the study sample. As our ethics approval did not allow us to record the details of the actual pathology company or reporting pathologist we do not know whether this overrepresentation of nonmetropolitan residents is also reflected in the patients' place of treatment and pathology assessment. Nonetheless, as 79% of patients lived in metropolitan or other urban areas we are still satisfied that the sample is reasonably representative of all eligible thyroid cancer patients resident in NSW.
Another limitation of the study was that our data extraction form did not accurately capture information regarding the reporting of lymph node assessment. We would recommend that any further studies make a distinction between reports that recorded that no nodes were available for assessment and reports that simply did not provide any information on the assessment of lymph nodes at all.
While audits of postsurgical pathology reports for other types of cancer have been published (8,10,11,28 –30), to our knowledge this is the first such audit of thyroid cancer pathology reports. In general, studies have found that without the systematic use of synoptic reports or checklists, pathology reports for surgical cancer specimens are often missing important information (8,10,28,30,31). Our results suggest that this is also true for thyroid cancer pathology reporting in NSW, Australia, and adds further support to the recommendation for the use of synoptic reports for cancer pathology.
Our finding that the majority of postsurgical pathology reports were missing information on at least one of the key features required for clinicians to provide patients with the best possible treatment suggests that the recent publication of a set of Australian guidelines for the pathology reporting of thyroid cancer (14) is timely. Adherence to these guidelines would result in accurate, complete, and clear reports that will benefit patients, clinicians, and the greater community.
Footnotes
Acknowledgments
The authors are very grateful to the NSW CCR at the Cancer Institute NSW for all their help with recruitment, data extraction from the pathology forms, and for providing data for participants and nonparticipants. We would like to thank Sam Egger (Cancer Council NSW) for his assistance with statistical analyses.
Disclosure Statement
The authors declare that no competing financial interests exist.