Consultation rates in cervical screening non-attenders: opportunities to increase screening uptake in GP primary care.

Abstract

OBJECTIVE: To estimate the proportion of cervical screening non-attenders presenting to general practice (GP) primary care over one year. SETTING: 137 practices in East London, UK. METHODS: Anonymous primary care records were downloaded using EMIS web (clinical software). Cervical screening nonattendance was defined as no recorded smear in the last 3.5 years (women aged 25-49) or 5.5 years (women aged 50-64). The last three consultation entries were used to estimate the proportion of non-attenders who consulted in GP over 3 months and 1 year using the Kaplan-Meier method. Newly registered women were assessed separately. Results were calculated for each practice and the median and interquartile range (IQR) across practices are presented. Heterogeneity was assessed using funnel plots. RESULTS: Of 261,810 women, 224,313 (86%) had been registered for >1 year. The proportion classified as non-attenders differed between those registered for >1 year (30%, IQR 27%--35%) and within the last year (49%, IQR 40%--57%), suggesting that screening records were less up-to-date in newly registered women. A median of 32% (IQR: 27%--37%) of non-attenders presented over 3 months, and 60% (IQR: 52%--67%) over 1 year. Funnel plots of the proportion of non-attenders presenting by the number of non-attenders showed substantial variation between practices. CONCLUSIONS: Over half of cervical screening non-attenders present to their GP at least once a year, in over 75% of practices. This represents a good opportunity for improving coverage by offering an alternative form of screening, such as self-sampling for human papillomavirus testing.