Measuring the nature and duration of symptoms of cervical cancer in young women: developing an interview-based approach

BACKGROUND Some young women experience delays in diagnosis of cervical cancer, but little research about ways of studying these delays has been published. A major challenge is that gynaecological symptoms are common in young women, but cervical cancer is rare. This study describes the development and testing of a measure for studying delays in diagnosis in young women with cervical cancer. METHODS Prospective development of an interview measure and testing of its ability to reliably and systematically collect relevant data in two large hospitals in London, UK using 27 women aged 18-40 diagnosed with cervical cancer in the previous two years. We developed a semi-structured interview schedule and data extraction form to systematically collect data on symptoms (including nature and duration) and risk factors for delayed diagnosis from young women with cervical cancer. We piloted the measure among young women with cervical cancer (audiorecording it with their permission), refining it iteratively. To complete the measure, we developed a database for managing the data and a manual for using the schedule. Two researchers extracted data from the recorded interviews to assess inter-rater reliability. RESULTS The final interview schedule yielded quantitative data on the nature and duration of symptoms and risk factors for delayed diagnosis. Inter-rater reliability was high. In the pilot, 12 of the 27 women were diagnosed via symptomatic presentation. Median time from the symptom triggering presentation to presentation was one month (interquartile range 0-4 months). Median time from presentation to diagnosis was three months (interquartile range 1-8.5 months). CONCLUSIONS We have developed a reliable tool for measuring the nature and duration of symptoms in young women with cervical cancer. Pilot data suggest that a substantial proportion of women experience delay between first presentation and diagnosis.