S L Savaridas1, J Brook2, J P Codde3, M Bulsara3, E Wylie4. 1. BreastScreen WA, Eastpoint Plaza, 233 Adelaide Terrace, Perth, WA, 6000, Australia. Electronic address: slsavaridas@doctors.org.uk. 2. BreastScreen WA, Eastpoint Plaza, 233 Adelaide Terrace, Perth, WA, 6000, Australia. 3. Institute for Health Research, The University of Notre Dame Australia, 19 Mouat Street, P.O Box 1225, Fremantle, WA, 6959, Australia. 4. BreastScreen WA, Eastpoint Plaza, 233 Adelaide Terrace, Perth, WA, 6000, Australia; University of Western Australia Medical School, 35 Stirling Highway, Crawley, WA, 6009, Australia.
Abstract
AIM: To establish whether individual radiographers had significantly different rescreening rates whilst controlling for other known confounding factors. MATERIALS AND METHODS: Women aged 50-69 years were identified from a state-wide screening database at their first screening attendance during the study period (2007-2013). The radiographer performing this index screen and potential confounding factors were recorded and subsequent screening behaviour was assessed. Clients with abnormal screens and those known to have died during the time period were excluded. A univariate analysis of the data from 160,028 women was assessed using the chi-square test to compare those women who attended their next mammography with non-re-attenders. Logistic regression was used to calculate the likelihood of "re-attendance success" across a range of variables. The probability of re-attendance for 11 randomly selected radiographers was determined from the logistic regression model, whilst controlling for other variables. RESULTS: Comparison of non-re-attenders (n=49,698) with 110,330 (69%) women attending the next round of screening revealed significant differences, including radiographer (Wald statistics=1188, p<0.000) even when all other known factors were controlled. CONCLUSION: This large, population-level study demonstrates that individual radiographer factors appear to influence a women's decision to return for their next screening round. Further research is required to identify reasons for differing rescreen rates and provide education and retraining of individual radiographers as appropriate.
AIM: To establish whether individual radiographers had significantly different rescreening rates whilst controlling for other known confounding factors. MATERIALS AND METHODS:Women aged 50-69 years were identified from a state-wide screening database at their first screening attendance during the study period (2007-2013). The radiographer performing this index screen and potential confounding factors were recorded and subsequent screening behaviour was assessed. Clients with abnormal screens and those known to have died during the time period were excluded. A univariate analysis of the data from 160,028 women was assessed using the chi-square test to compare those women who attended their next mammography with non-re-attenders. Logistic regression was used to calculate the likelihood of "re-attendance success" across a range of variables. The probability of re-attendance for 11 randomly selected radiographers was determined from the logistic regression model, whilst controlling for other variables. RESULTS: Comparison of non-re-attenders (n=49,698) with 110,330 (69%) women attending the next round of screening revealed significant differences, including radiographer (Wald statistics=1188, p<0.000) even when all other known factors were controlled. CONCLUSION: This large, population-level study demonstrates that individual radiographer factors appear to influence a women's decision to return for their next screening round. Further research is required to identify reasons for differing rescreen rates and provide education and retraining of individual radiographers as appropriate.