A Richardson1. 1. Department of Public Health and General Practice, Christchurch School of Medicine, University of Otago, Christchurch, New Zealand. ann.richardson@chmeds.ac.nz
Abstract
OBJECTIVE: To develop a measure analogous to the "number needed to treat" (NNT), which would be appropriate for population based screening programmes. METHODS: A previous measure analogous to the NNT, the "number needed to screen" (NNS), calculated as the reciprocal of the absolute risk reduction, does not account for participation and selection effects in screening trials. A new measure is proposed; the NNBS (number needed to be screened) derived from the NNT adjusted for participation in screening and for selection effects associated with participation in screening. The NNBS for breast cancer screening and colorectal cancer screening are calculated, based on the results of randomised controlled trials. RESULTS: In the first 8 years from the start of each screening programme, the NNS to prevent one death from breast cancer was 781 women (95% confidence interval (95% CI) 550 to 1560). The NNS to prevent one death from colorectal cancer was 1250 people (95% CI 690 to 9090). The NNBS was 23% lower than the NNS for breast cancer screening (NNBS=601), and 45% lower than the NNS for colorectal cancer (NNBS=688). CONCLUSIONS: The NNT calculations may make population based screening programmes seem expensive and inefficient compared with other interventions. A new measure, the number needed to be screened (NNBS), which takes into account the participation rate adjusted for selection, may be more appropriate, especially when comparing screening programmes with disparate participation rates, or when comparing screening with other interventions.
OBJECTIVE: To develop a measure analogous to the "number needed to treat" (NNT), which would be appropriate for population based screening programmes. METHODS: A previous measure analogous to the NNT, the "number needed to screen" (NNS), calculated as the reciprocal of the absolute risk reduction, does not account for participation and selection effects in screening trials. A new measure is proposed; the NNBS (number needed to be screened) derived from the NNT adjusted for participation in screening and for selection effects associated with participation in screening. The NNBS for breast cancer screening and colorectal cancer screening are calculated, based on the results of randomised controlled trials. RESULTS: In the first 8 years from the start of each screening programme, the NNS to prevent one death from breast cancer was 781 women (95% confidence interval (95% CI) 550 to 1560). The NNS to prevent one death from colorectal cancer was 1250 people (95% CI 690 to 9090). The NNBS was 23% lower than the NNS for breast cancer screening (NNBS=601), and 45% lower than the NNS for colorectal cancer (NNBS=688). CONCLUSIONS: The NNT calculations may make population based screening programmes seem expensive and inefficient compared with other interventions. A new measure, the number needed to be screened (NNBS), which takes into account the participation rate adjusted for selection, may be more appropriate, especially when comparing screening programmes with disparate participation rates, or when comparing screening with other interventions.
Authors: Alvie Ahsan; Eva Zimmerman; Elisa Marie Rodriguez; Christy Widman; Deborah Oates Erwin; Frances Georgette Saad-Harfouche; Martin Christopher Mahoney Journal: J Cancer Treat Res Date: 2019-03-11
Authors: Antonio Gutierrez; Robert Suh; Fereidoun Abtin; Scott Genshaft; Kathleen Brown Journal: Semin Intervent Radiol Date: 2013-06 Impact factor: 1.513
Authors: Anssi Auvinen; Sue M Moss; Teuvo L J Tammela; Kimmo Taari; Monique J Roobol; Fritz H Schröder; Chris H Bangma; Sigrid Carlsson; Gunnar Aus; Marco Zappa; Donella Puliti; Louis J Denis; Vera Nelen; Maciej Kwiatkowski; Marco Randazzo; Alvaro Paez; Marcos Lujan; Jonas Hugosson Journal: Clin Cancer Res Date: 2015-08-19 Impact factor: 12.531
Authors: Joseph Lunyera; John W Stanifer; Prossie Ingabire; Wilson Etolu; Peace Bagasha; Joseph R Egger; Uptal D Patel; Gerald Mutungi; Robert Kalyesubula Journal: BMC Res Notes Date: 2016-02-16