Differential Use of Screening Mammography in Older Women Initiating Metformin versus Sulfonylurea

Jin Liern Hong, Louise M. Henderson, Michele Jonsson Funk, Jennifer L. Lund, John B. Buse, Virginia Pate, Til Stürmer

Research output: Research - peer-reviewArticle

Abstract

Purpose: Differential use of screening mammography may lead to biased detection of breast cancer. This study aimed to compare receipt of screening mammography and the incidence of screen-detected breast cancer between metformin and sulfonylurea initiators. Methods: We used 2006-2014 US Medicare claims to identify initiators of metformin or sulfonylurea aged 65+ years continuously enrolled in Parts A/B for ≥2 years pre-initiation and ≥2 years post-initiation. We reported frequencies of screening mammograms and screen-detected breast cancer in 1 year pre-initiation among all cohort members and in 1 year post-initiation among cancer-free cohort members. Weighted screening risk differences (RDs) were estimated comparing metformin to sulfonylurea group. Results: We identified 41,436 and 13,367 initiators of metformin and sulfonylurea, 35% and 24% of which had ≥1 screening mammogram in 1 year pre-initiation (weighted RD: 6 percentage points; 95% CI: 5 to 7), respectively. The weighted RD for screen-detected breast cancer associated with metformin was 0.00 percentage points (95% CI: -0.09 to 0.09). Among cancer-free cohort members, metformin initiators had 5 percentage points (95% CI: 4 to 6) and 0.11 percentage points (95% CI: -0.02 to 0.23) absolute risk excess of screening mammography and screen-detected breast cancer in 1 year post-initiation, compared with sulfonylurea initiators, respectively. Conclusions: Metformin initiators were more likely to receive screening mammograms than sulfonylurea initiators pre- and post-initiation, indicating possible detection bias due to differential screening mammography. Researchers should be aware of the potential for more screening mammograms pre- and post-initiation when interpreting the findings of metformin on breast cancer incidence.

LanguageEnglish (US)
Pages666-675
Number of pages10
JournalPharmacoepidemiology and Drug Safety
Volume26
Issue number6
DOIs
StatePublished - Jun 1 2017

Fingerprint

Metformin
Mammography
Breast Neoplasms
Incidence
Neoplasms
Medicare
Research Personnel
cyclo(Arg-Pro)

Keywords

  • Breast Cancer
  • Diabetes
  • Metformin
  • Screening mammography

ASJC Scopus subject areas

  • Epidemiology
  • Pharmacology (medical)

Cite this

Differential Use of Screening Mammography in Older Women Initiating Metformin versus Sulfonylurea. / Hong, Jin Liern; Henderson, Louise M.; Jonsson Funk, Michele; Lund, Jennifer L.; Buse, John B.; Pate, Virginia; Stürmer, Til.

In: Pharmacoepidemiology and Drug Safety, Vol. 26, No. 6, 01.06.2017, p. 666-675.

Research output: Research - peer-reviewArticle

@article{d9ca3665aad64b2fbfb882cddf4b94f9,
title = "Differential Use of Screening Mammography in Older Women Initiating Metformin versus Sulfonylurea",
abstract = "Purpose: Differential use of screening mammography may lead to biased detection of breast cancer. This study aimed to compare receipt of screening mammography and the incidence of screen-detected breast cancer between metformin and sulfonylurea initiators. Methods: We used 2006-2014 US Medicare claims to identify initiators of metformin or sulfonylurea aged 65+ years continuously enrolled in Parts A/B for ≥2 years pre-initiation and ≥2 years post-initiation. We reported frequencies of screening mammograms and screen-detected breast cancer in 1 year pre-initiation among all cohort members and in 1 year post-initiation among cancer-free cohort members. Weighted screening risk differences (RDs) were estimated comparing metformin to sulfonylurea group. Results: We identified 41,436 and 13,367 initiators of metformin and sulfonylurea, 35% and 24% of which had ≥1 screening mammogram in 1 year pre-initiation (weighted RD: 6 percentage points; 95% CI: 5 to 7), respectively. The weighted RD for screen-detected breast cancer associated with metformin was 0.00 percentage points (95% CI: -0.09 to 0.09). Among cancer-free cohort members, metformin initiators had 5 percentage points (95% CI: 4 to 6) and 0.11 percentage points (95% CI: -0.02 to 0.23) absolute risk excess of screening mammography and screen-detected breast cancer in 1 year post-initiation, compared with sulfonylurea initiators, respectively. Conclusions: Metformin initiators were more likely to receive screening mammograms than sulfonylurea initiators pre- and post-initiation, indicating possible detection bias due to differential screening mammography. Researchers should be aware of the potential for more screening mammograms pre- and post-initiation when interpreting the findings of metformin on breast cancer incidence.",
keywords = "Breast Cancer, Diabetes, Metformin, Screening mammography",
author = "Hong, {Jin Liern} and Henderson, {Louise M.} and {Jonsson Funk}, Michele and Lund, {Jennifer L.} and Buse, {John B.} and Virginia Pate and Til Stürmer",
year = "2017",
month = "6",
doi = "10.1002/pds.4195",
volume = "26",
pages = "666--675",
journal = "Pharmacoepidemiology and Drug Safety",
issn = "1053-8569",
publisher = "John Wiley and Sons Ltd",
number = "6",

}

TY - JOUR

T1 - Differential Use of Screening Mammography in Older Women Initiating Metformin versus Sulfonylurea

AU - Hong,Jin Liern

AU - Henderson,Louise M.

AU - Jonsson Funk,Michele

AU - Lund,Jennifer L.

AU - Buse,John B.

AU - Pate,Virginia

AU - Stürmer,Til

PY - 2017/6/1

Y1 - 2017/6/1

N2 - Purpose: Differential use of screening mammography may lead to biased detection of breast cancer. This study aimed to compare receipt of screening mammography and the incidence of screen-detected breast cancer between metformin and sulfonylurea initiators. Methods: We used 2006-2014 US Medicare claims to identify initiators of metformin or sulfonylurea aged 65+ years continuously enrolled in Parts A/B for ≥2 years pre-initiation and ≥2 years post-initiation. We reported frequencies of screening mammograms and screen-detected breast cancer in 1 year pre-initiation among all cohort members and in 1 year post-initiation among cancer-free cohort members. Weighted screening risk differences (RDs) were estimated comparing metformin to sulfonylurea group. Results: We identified 41,436 and 13,367 initiators of metformin and sulfonylurea, 35% and 24% of which had ≥1 screening mammogram in 1 year pre-initiation (weighted RD: 6 percentage points; 95% CI: 5 to 7), respectively. The weighted RD for screen-detected breast cancer associated with metformin was 0.00 percentage points (95% CI: -0.09 to 0.09). Among cancer-free cohort members, metformin initiators had 5 percentage points (95% CI: 4 to 6) and 0.11 percentage points (95% CI: -0.02 to 0.23) absolute risk excess of screening mammography and screen-detected breast cancer in 1 year post-initiation, compared with sulfonylurea initiators, respectively. Conclusions: Metformin initiators were more likely to receive screening mammograms than sulfonylurea initiators pre- and post-initiation, indicating possible detection bias due to differential screening mammography. Researchers should be aware of the potential for more screening mammograms pre- and post-initiation when interpreting the findings of metformin on breast cancer incidence.

AB - Purpose: Differential use of screening mammography may lead to biased detection of breast cancer. This study aimed to compare receipt of screening mammography and the incidence of screen-detected breast cancer between metformin and sulfonylurea initiators. Methods: We used 2006-2014 US Medicare claims to identify initiators of metformin or sulfonylurea aged 65+ years continuously enrolled in Parts A/B for ≥2 years pre-initiation and ≥2 years post-initiation. We reported frequencies of screening mammograms and screen-detected breast cancer in 1 year pre-initiation among all cohort members and in 1 year post-initiation among cancer-free cohort members. Weighted screening risk differences (RDs) were estimated comparing metformin to sulfonylurea group. Results: We identified 41,436 and 13,367 initiators of metformin and sulfonylurea, 35% and 24% of which had ≥1 screening mammogram in 1 year pre-initiation (weighted RD: 6 percentage points; 95% CI: 5 to 7), respectively. The weighted RD for screen-detected breast cancer associated with metformin was 0.00 percentage points (95% CI: -0.09 to 0.09). Among cancer-free cohort members, metformin initiators had 5 percentage points (95% CI: 4 to 6) and 0.11 percentage points (95% CI: -0.02 to 0.23) absolute risk excess of screening mammography and screen-detected breast cancer in 1 year post-initiation, compared with sulfonylurea initiators, respectively. Conclusions: Metformin initiators were more likely to receive screening mammograms than sulfonylurea initiators pre- and post-initiation, indicating possible detection bias due to differential screening mammography. Researchers should be aware of the potential for more screening mammograms pre- and post-initiation when interpreting the findings of metformin on breast cancer incidence.

KW - Breast Cancer

KW - Diabetes

KW - Metformin

KW - Screening mammography

UR - http://www.scopus.com/inward/record.url?scp=85016740129&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016740129&partnerID=8YFLogxK

U2 - 10.1002/pds.4195

DO - 10.1002/pds.4195

M3 - Article

VL - 26

SP - 666

EP - 675

JO - Pharmacoepidemiology and Drug Safety

T2 - Pharmacoepidemiology and Drug Safety

JF - Pharmacoepidemiology and Drug Safety

SN - 1053-8569

IS - 6

ER -