Qualitative analysis of the coding of pathological data of cancer registry centers: A study in North of Iran.

BACKGROUND: The cancer registry system is an important part of the cancer control program. Improper coding of cancers leads to misclassification and incorrect statistical information about cancer. Therefore, in this study, the main objective of the qualitative analysis was the accuracy in assigning the codes to the pathological reports in the centers responsible for cancer registry. METHODS: This study was descriptive, retrospective and applied. The data source in this study included 15,659 pathology reports received during the years 2017-2019 in the population-based cancer registry centers of Mazandaran province. Out of 1800 reports, 1765 samples of reports were selected and analysis was done on them by stratified random sampling method. A researcher-made checklist was used to collect data, and the Kappa agreement coefficient and Cohen's agreement percentage were presented to check the accuracy of the reports. STATA13 was used for data analysis. RESULTS: 1150 of 1765 pathology reports (65.0%), did not have a topographic, morphological and behavioral codes and 410 (23.2%) had grade codes. The Kappa coefficient in reports with a topography code was 0.916 and with a morphology code it was 0.929, respectively. In behavior coding, the highest agreement is in the category of benign cancers at 65.2% and in grade coding in the category without grade is 100%. CONCLUSION: The most reports were on carcinoma morphology, and the Kappa coefficient in morphology codes has almost complete reliability. In terms of behavior coding, there was the most agreement in the category of benign cancers. The Kappa coefficient in given behavior codes has low reliability.

Quality assessment of pathologic data in cancer registry centers based on ICD-O-3.

Background: Prerequisite for achieving the goals of the registration program is the existence of valid and accurate data, and the usability of this data is possible if they are coded correctly. This study assets the quality of pathological data of the population-based cancer registration centers based on ICD-O-3. Methods: This applied study was performed descriptively and retrospectively. The study population included 20129 pathology reports sent to the population-based cancer registration center of Mazandaran Province during 2018-2020. A total of 2015 out of, 2050 samples of the received reports were examined according to stratified random sampling method. A researcher checklist was made to collect the data, and STATA 13 and Cohen's Kappa agreement coefficient were used to analyze the data. Results: Among the 2015 reports of pathology, 1114 (55.3%) pathology reports were related to government centers, (42.9%) 865 cases were registered with their topographic code, morphology and behavior. Based on the registration of the exact topographic code, the kappa coefficient and the total agreement were 0.266 and 27.70%, respectively. Kappa coefficient in all received reports and reports with topographic code was 0.346 and 0.906, respectively. In the reports with topographic code, the most reports of cancers were related to cancers of the gastrointestinal organs (97.6%) 246. Conclusion: The accuracy of the codes given in the pathology centers in terms of topographic, morphological, behavioral and grade codes based on the percentage of agreement with the coding was above average, which were higher in governmental centers and also in some cancers.