Practitioner compression force variation in mammography: A 6-year study

The application of breast compression in mammography may be more heavily influenced by the practitioner rather than the client. This could affect image quality and will affect client experience. This study builds on previous research to establish if mammography practitioners vary in the compression force they apply over a six-year period.This longitudinal study assessed 3 consecutive analogue screens of 500 clients within one screening centre in the UK. Recorded data included: practitioner code, applied compression force (daN), breast thickness (mm), BI-RADS® density category and breast dose. Exclusion criteria included: previous breast surgery, previous/ongoing assessment, breast implants. 344 met inclusion criteria. Data analysis: assessed variation of compression force (daN) and breast thickness (mm) over 3 sequential screens to determine whether compression force and breast thickness were affected by practitioner variations.Compression force over the 3 screens varied significantly; variation was highly dependent upon the practitioner who performed the mammogram. Significant thickness and compression force differences over the 3 screens were noted for the same client (<0.0001). The amount of compression force applied was highly dependent upon the practitioner. Practitioners fell into one of three practitioner compression groups by their compression force mean values; high (mean 12.6 daN), intermediate (mean 8.9 daN) and low (mean 6.7 daN).For the same client, when the same practitioner performed the 3 screens, maximum compression force variations were low and not significantly different (p > 0.31). When practitioners from different compression force groups performed 3 screens, maximum compression force variations were higher and significantly different (p < 0.0001).The amount of compression force used is highly dependent upon practitioner rather than client. This has implications for radiation dose, patient experience and image quality consistency.