Validation of data quality in the Swedish quality register of gynecologic cancer for cervical cancer and vulvar cancer—a Swedish gynecologic cancer group (Swe‐GCG) study

Abstract

Abstract Introduction Population‐based registers provide an important source of real‐world data. The growing number of large cohort studies using data from cancer registers makes validation of such registers important. The Swedish Quality Register of Gynecologic Cancer (SQRGC) is a nationwide population‐based register containing data on patient and tumor characteristics, treatment, and follow‐up. To ensure that the results from research and quality assurance reports using SQRGC data are robust and reliable, the accuracy and completeness of the register need to be validated. The aim of this study was to evaluate the quality of data on cervical cancer and vulvar cancer in the SQRGC. Material and Methods Quality of data in the SQRGC was investigated by evaluating completeness, timeliness, comparability, and validity in accordance with recommendations from the International Agency for Research on Cancer and the national Swedish guidelines on validation of cancer registers. Completeness was evaluated by coverage relative to the Swedish National Cancer Register, and timeliness as the time from diagnosis until entry into the SQRGC. We randomly selected 276 women diagnosed with cervical cancer (n = 138) and vulvar cancer (n = 138) between 2014 and 2019 for validation. An external monitor manually re‐abstracted data on 10 core variables per sub‐register from the patients' medical records. Comparability was assessed by reviewing the adherence to international standards regarding coding. Validity was evaluated by the agreement between re‐abstracted data and original data in the SQRGC. Correlations were estimated using Pearson's correlation coefficient and Cohen's kappa coefficient. Results For cervical cancer, the completeness was 99% and the timeliness was 87.1% within 12 months. The corresponding figures for vulvar cancer were 100% and 87.9%, respectively. Adherence to international coding standards was satisfactory. The median degree of agreement between re‐abstracted data and data in the SQRGC was 90.8% (range 73.2%–100%) for cervical cancer, and 85.4% (range 59.6%–98.2%) for vulvar cancer. Conclusions The data on cervical and vulvar cancer in the SQRGC are of adequate quality and may well be used for research and clinical purposes.

Keywords

• cancer register
• cervical cancer
• data quality
• quality register
• validation
• vulvar cancer