Examining the diagnostic process for ovarian cancer in Nova Scotia: A linked administrative data study
Chiara Gottheil, Christy Woolcott, Emily Gard Marshall, Katharina Kieser, Robin UrquhartAbstract
Objective
Ovarian cancer is the deadliest gynecological cancer in the developed world. The absence of screening for early detection of ovarian cancer and the non‐specific symptoms associated with the disease mean that timely diagnosis is difficult. This study examined time to diagnosis (from first presentation to diagnosis) for epithelial ovarian cancer in Nova Scotia, Canada, and factors associated with a long time to diagnosis (> 75th percentile) and healthcare system use prior to diagnosis.
Methods
All Nova Scotians diagnosed with epithelial ovarian cancer between January 1, 2007, and December 31, 2016, were identified from the Nova Scotia Cancer Registry. Independent variables were obtained from linked clinical and administrative databases and census data. Time to diagnosis and healthcare system use were descriptively analyzed. Factors associated with time to diagnosis exceeding the 75th percentile were identified using modified Poisson regression models.
Results
A total of 652 individuals were included. The median time to diagnosis was 30 days. Factors associated with a long time to diagnosis included health zone, stage at diagnosis, year of first presentation, site of first presentation, and continuity of care. Sensitivity analyses examined factors associated with time to diagnosis exceeding 8 weeks and below the 25th percentile. The emergency department was the site of first presentation for 35.7% of individuals. The median number of physician visits between first presentation and diagnosis was four, and the median number of physician specialties seen in the same time interval was three, with the most common specialties seen being primary care, “other” specialties, and obstetrics/gynecology.
Conclusion
Time to diagnosis for ovarian cancer can be long, specifically in some subgroups. Ways to streamline the diagnostic process must be identified.