Estimating the workload associated with symptoms-based ovarian cancer screening in primary care: an audit of electronic medical records.

Abstract

BACKGROUND: Ovarian cancer is the most lethal gynaecological malignancy in the United Kingdom (UK). Studies have found that many women with ovarian cancer have symptoms for several months before diagnosis. Using a symptoms-based tool to diagnose ovarian cancer (OC) earlier is appealing, but may increase general practitioner (GP) workload because the symptoms are typically vague and non-specific. This study aimed to provide estimates of the GP workload associated with offering symptoms-based ovarian cancer screening. METHODS: A cross-sectional analysis of electronic records from four general practices in England, UK. We downloaded anonymous data on women aged 45-74 who consulted over one week to estimate the proportion who would be offered 'screening' according to the UK National Institute for Health and Care Excellence (NICE) guidelines and a symptoms index (Index 2) over one year. We used previous consultations (censoring women with no prior symptom at the date of their last recorded consultation) to estimate the proportion of women presenting with a new (not recorded in previous 12 months) NICE symptom each year. RESULTS: Data were obtained from 19,558 women. The proportion presenting over one week varied between practices (5%-14%), however, the proportion with an OC symptom was similar (17% overall). Over one year, an estimated 51.8% (95% CI 44.0%-59.7%) would present with an OC symptom, 26.6% (95% CI 19.3%-35.1%) with a NICE symptom and 20.3% (95% CI 13.7%-28.5%) with an Index 2 symptom. Each year, an estimated 11.9% (95% CI 5.0%-18.3%) of women would present with a new NICE symptom. CONCLUSION: One in two women aged 45-74 present to primary care at least once a year with an OC symptom, 11.9% with a new NICE symptom. This would be comparable to 2 to 8 yearly screening (depending on what symptoms triggered testing).