Effect of changing breast cancer incidence rates on the calibration of the Gail model

Sara J. Schonfeld, David Pee, Robert T. Greenlee, Patricia Hartge, James V. Lacey, Yikyung Park, Arthur Schatzkin, Kala Visvanathan, Ruth M. Pfeiffer

Research output: Contribution to journalArticle

Abstract

Purpose: The Gail model combines relative risks (RRs) for five breast cancer risk factors with age-specific breast cancer incidence rates and competing mortality rates from the Surveillance, Epidemiology, and End Results (SEER) program from 1983 to 1987 to predict risk of invasive breast cancer over a given time period. Motivated by changes in breast cancer incidence during the 1990s, we evaluated the model's calibration in two recent cohorts. Methods: We included white, postmenopausal women from the National Institutes of Health (NIH) -AARP Diet and Health Study (NIH-AARP, 1995 to 2003), and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO, 1993 to 2006). Calibration was assessed by comparing the number of breast cancers expected from the Gail model with that observed. We then evaluated calibration by using an updated model that combined Gail model RRs with 1995 to 2003 SEER invasive breast cancer incidence rates. Results: Overall, the Gail model significantly underpredicted the number of invasive breast cancers in NIH-AARP, with an expected-to-observed ratio of 0.87 (95% CI, 0.85 to 0.89), and in PLCO, with an expected-to-observed ratio of 0.86 (95% CI, 0.82 to 0.90). The updated model was well-calibrated overall, with an expected-to-observed ratio of 1.03 (95% CI, 1.00 to 1.05) in NIH-AARP and an expected-to-observed ratio of 1.01 (95% CI: 0.97 to 1.06) in PLCO. Of women age 50 to 55 years at baseline, 13% to 14% had a projected Gail model 5-year risk lower than the recommended threshold of 1.66% for use of tamoxifen or raloxifene but ≥ 1.66% when using the updated model. The Gail model was well calibrated in PLCO when the prediction period was restricted to 2003 to 2006. Conclusion: This study highlights that model calibration is important to ensure the usefulness of risk prediction models for clinical decision making.

Original languageEnglish (US)
Pages (from-to)2411-2417
Number of pages7
JournalJournal of Clinical Oncology
Volume28
Issue number14
DOIs
StatePublished - May 10 2010
Externally publishedYes

Fingerprint

Calibration
Breast Neoplasms
National Institutes of Health (U.S.)
Incidence
SEER Program
Tamoxifen
Early Detection of Cancer
Ovarian Neoplasms
Colorectal Neoplasms
Lung Neoplasms
Prostatic Neoplasms
Epidemiology
Diet
Mortality
Health

ASJC Scopus subject areas

  • Cancer Research
  • Oncology
  • Medicine(all)

Cite this

Schonfeld, S. J., Pee, D., Greenlee, R. T., Hartge, P., Lacey, J. V., Park, Y., ... Pfeiffer, R. M. (2010). Effect of changing breast cancer incidence rates on the calibration of the Gail model. Journal of Clinical Oncology, 28(14), 2411-2417. https://doi.org/10.1200/JCO.2009.25.2767

Effect of changing breast cancer incidence rates on the calibration of the Gail model. / Schonfeld, Sara J.; Pee, David; Greenlee, Robert T.; Hartge, Patricia; Lacey, James V.; Park, Yikyung; Schatzkin, Arthur; Visvanathan, Kala; Pfeiffer, Ruth M.

In: Journal of Clinical Oncology, Vol. 28, No. 14, 10.05.2010, p. 2411-2417.

Research output: Contribution to journalArticle

Schonfeld, SJ, Pee, D, Greenlee, RT, Hartge, P, Lacey, JV, Park, Y, Schatzkin, A, Visvanathan, K & Pfeiffer, RM 2010, 'Effect of changing breast cancer incidence rates on the calibration of the Gail model', Journal of Clinical Oncology, vol. 28, no. 14, pp. 2411-2417. https://doi.org/10.1200/JCO.2009.25.2767
Schonfeld SJ, Pee D, Greenlee RT, Hartge P, Lacey JV, Park Y et al. Effect of changing breast cancer incidence rates on the calibration of the Gail model. Journal of Clinical Oncology. 2010 May 10;28(14):2411-2417. https://doi.org/10.1200/JCO.2009.25.2767
Schonfeld, Sara J. ; Pee, David ; Greenlee, Robert T. ; Hartge, Patricia ; Lacey, James V. ; Park, Yikyung ; Schatzkin, Arthur ; Visvanathan, Kala ; Pfeiffer, Ruth M. / Effect of changing breast cancer incidence rates on the calibration of the Gail model. In: Journal of Clinical Oncology. 2010 ; Vol. 28, No. 14. pp. 2411-2417.
@article{5d379898de3f4361a5fb68cc490ba8e9,
title = "Effect of changing breast cancer incidence rates on the calibration of the Gail model",
abstract = "Purpose: The Gail model combines relative risks (RRs) for five breast cancer risk factors with age-specific breast cancer incidence rates and competing mortality rates from the Surveillance, Epidemiology, and End Results (SEER) program from 1983 to 1987 to predict risk of invasive breast cancer over a given time period. Motivated by changes in breast cancer incidence during the 1990s, we evaluated the model's calibration in two recent cohorts. Methods: We included white, postmenopausal women from the National Institutes of Health (NIH) -AARP Diet and Health Study (NIH-AARP, 1995 to 2003), and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO, 1993 to 2006). Calibration was assessed by comparing the number of breast cancers expected from the Gail model with that observed. We then evaluated calibration by using an updated model that combined Gail model RRs with 1995 to 2003 SEER invasive breast cancer incidence rates. Results: Overall, the Gail model significantly underpredicted the number of invasive breast cancers in NIH-AARP, with an expected-to-observed ratio of 0.87 (95{\%} CI, 0.85 to 0.89), and in PLCO, with an expected-to-observed ratio of 0.86 (95{\%} CI, 0.82 to 0.90). The updated model was well-calibrated overall, with an expected-to-observed ratio of 1.03 (95{\%} CI, 1.00 to 1.05) in NIH-AARP and an expected-to-observed ratio of 1.01 (95{\%} CI: 0.97 to 1.06) in PLCO. Of women age 50 to 55 years at baseline, 13{\%} to 14{\%} had a projected Gail model 5-year risk lower than the recommended threshold of 1.66{\%} for use of tamoxifen or raloxifene but ≥ 1.66{\%} when using the updated model. The Gail model was well calibrated in PLCO when the prediction period was restricted to 2003 to 2006. Conclusion: This study highlights that model calibration is important to ensure the usefulness of risk prediction models for clinical decision making.",
author = "Schonfeld, {Sara J.} and David Pee and Greenlee, {Robert T.} and Patricia Hartge and Lacey, {James V.} and Yikyung Park and Arthur Schatzkin and Kala Visvanathan and Pfeiffer, {Ruth M.}",
year = "2010",
month = "5",
day = "10",
doi = "10.1200/JCO.2009.25.2767",
language = "English (US)",
volume = "28",
pages = "2411--2417",
journal = "Journal of Clinical Oncology",
issn = "0732-183X",
publisher = "American Society of Clinical Oncology",
number = "14",

}

TY - JOUR

T1 - Effect of changing breast cancer incidence rates on the calibration of the Gail model

AU - Schonfeld, Sara J.

AU - Pee, David

AU - Greenlee, Robert T.

AU - Hartge, Patricia

AU - Lacey, James V.

AU - Park, Yikyung

AU - Schatzkin, Arthur

AU - Visvanathan, Kala

AU - Pfeiffer, Ruth M.

PY - 2010/5/10

Y1 - 2010/5/10

N2 - Purpose: The Gail model combines relative risks (RRs) for five breast cancer risk factors with age-specific breast cancer incidence rates and competing mortality rates from the Surveillance, Epidemiology, and End Results (SEER) program from 1983 to 1987 to predict risk of invasive breast cancer over a given time period. Motivated by changes in breast cancer incidence during the 1990s, we evaluated the model's calibration in two recent cohorts. Methods: We included white, postmenopausal women from the National Institutes of Health (NIH) -AARP Diet and Health Study (NIH-AARP, 1995 to 2003), and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO, 1993 to 2006). Calibration was assessed by comparing the number of breast cancers expected from the Gail model with that observed. We then evaluated calibration by using an updated model that combined Gail model RRs with 1995 to 2003 SEER invasive breast cancer incidence rates. Results: Overall, the Gail model significantly underpredicted the number of invasive breast cancers in NIH-AARP, with an expected-to-observed ratio of 0.87 (95% CI, 0.85 to 0.89), and in PLCO, with an expected-to-observed ratio of 0.86 (95% CI, 0.82 to 0.90). The updated model was well-calibrated overall, with an expected-to-observed ratio of 1.03 (95% CI, 1.00 to 1.05) in NIH-AARP and an expected-to-observed ratio of 1.01 (95% CI: 0.97 to 1.06) in PLCO. Of women age 50 to 55 years at baseline, 13% to 14% had a projected Gail model 5-year risk lower than the recommended threshold of 1.66% for use of tamoxifen or raloxifene but ≥ 1.66% when using the updated model. The Gail model was well calibrated in PLCO when the prediction period was restricted to 2003 to 2006. Conclusion: This study highlights that model calibration is important to ensure the usefulness of risk prediction models for clinical decision making.

AB - Purpose: The Gail model combines relative risks (RRs) for five breast cancer risk factors with age-specific breast cancer incidence rates and competing mortality rates from the Surveillance, Epidemiology, and End Results (SEER) program from 1983 to 1987 to predict risk of invasive breast cancer over a given time period. Motivated by changes in breast cancer incidence during the 1990s, we evaluated the model's calibration in two recent cohorts. Methods: We included white, postmenopausal women from the National Institutes of Health (NIH) -AARP Diet and Health Study (NIH-AARP, 1995 to 2003), and the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial (PLCO, 1993 to 2006). Calibration was assessed by comparing the number of breast cancers expected from the Gail model with that observed. We then evaluated calibration by using an updated model that combined Gail model RRs with 1995 to 2003 SEER invasive breast cancer incidence rates. Results: Overall, the Gail model significantly underpredicted the number of invasive breast cancers in NIH-AARP, with an expected-to-observed ratio of 0.87 (95% CI, 0.85 to 0.89), and in PLCO, with an expected-to-observed ratio of 0.86 (95% CI, 0.82 to 0.90). The updated model was well-calibrated overall, with an expected-to-observed ratio of 1.03 (95% CI, 1.00 to 1.05) in NIH-AARP and an expected-to-observed ratio of 1.01 (95% CI: 0.97 to 1.06) in PLCO. Of women age 50 to 55 years at baseline, 13% to 14% had a projected Gail model 5-year risk lower than the recommended threshold of 1.66% for use of tamoxifen or raloxifene but ≥ 1.66% when using the updated model. The Gail model was well calibrated in PLCO when the prediction period was restricted to 2003 to 2006. Conclusion: This study highlights that model calibration is important to ensure the usefulness of risk prediction models for clinical decision making.

UR - http://www.scopus.com/inward/record.url?scp=77952529935&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952529935&partnerID=8YFLogxK

U2 - 10.1200/JCO.2009.25.2767

DO - 10.1200/JCO.2009.25.2767

M3 - Article

C2 - 20368565

AN - SCOPUS:77952529935

VL - 28

SP - 2411

EP - 2417

JO - Journal of Clinical Oncology

JF - Journal of Clinical Oncology

SN - 0732-183X

IS - 14

ER -