Evaluating costs with unmeasured confounding

A sensitivity analysis for the treatment effect

Elizabeth A. Handorf, Justin E. Bekelman, Daniel F. Heitjan, Nandita Mitra

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Estimates of the effects of treatment on cost from observational studies are subject to bias if there are unmeasured confounders. It is therefore advisable in practice to assess the potential magnitude of such biases. We derive a general adjustment formula for loglinear models of mean cost and explore special cases under plausible assumptions about the distribution of the unmeasured confounder. We assess the performance of the adjustment by simulation, in particular, examining robustness to a key assumption of conditional independence between the unmeasured and measured covariates given the treatment indicator. We apply our method to SEER-Medicare cost data for a stage II/III muscle-invasive bladder cancer cohort.We evaluate the costs for radical cystectomy vs. combined radiation/chemotherapy, and find that the significance of the treatment effect is sensitive to plausible unmeasured Bernoulli, Poisson and Gamma confounders.

Original languageEnglish (US)
Pages (from-to)2062-2080
Number of pages19
JournalAnnals of Applied Statistics
Volume7
Issue number4
DOIs
StatePublished - Dec 1 2013

Fingerprint

Confounding
Treatment Effects
Sensitivity analysis
Sensitivity Analysis
Costs
Adjustment
Bladder Cancer
Observational Study
Log-linear Models
Chemotherapy
Conditional Independence
Bernoulli
Muscle
Covariates
Siméon Denis Poisson
Radiation
Robustness
Treatment effects
Evaluate
Estimate

Keywords

  • Censored costs
  • SEER-Medicare
  • Sensitivity analysis

ASJC Scopus subject areas

  • Statistics, Probability and Uncertainty
  • Modeling and Simulation
  • Statistics and Probability

Cite this

Evaluating costs with unmeasured confounding : A sensitivity analysis for the treatment effect. / Handorf, Elizabeth A.; Bekelman, Justin E.; Heitjan, Daniel F.; Mitra, Nandita.

In: Annals of Applied Statistics, Vol. 7, No. 4, 01.12.2013, p. 2062-2080.

Research output: Contribution to journalArticle

Handorf, Elizabeth A. ; Bekelman, Justin E. ; Heitjan, Daniel F. ; Mitra, Nandita. / Evaluating costs with unmeasured confounding : A sensitivity analysis for the treatment effect. In: Annals of Applied Statistics. 2013 ; Vol. 7, No. 4. pp. 2062-2080.
@article{87b6e3a32ba845cdb39129aa3ce026d0,
title = "Evaluating costs with unmeasured confounding: A sensitivity analysis for the treatment effect",
abstract = "Estimates of the effects of treatment on cost from observational studies are subject to bias if there are unmeasured confounders. It is therefore advisable in practice to assess the potential magnitude of such biases. We derive a general adjustment formula for loglinear models of mean cost and explore special cases under plausible assumptions about the distribution of the unmeasured confounder. We assess the performance of the adjustment by simulation, in particular, examining robustness to a key assumption of conditional independence between the unmeasured and measured covariates given the treatment indicator. We apply our method to SEER-Medicare cost data for a stage II/III muscle-invasive bladder cancer cohort.We evaluate the costs for radical cystectomy vs. combined radiation/chemotherapy, and find that the significance of the treatment effect is sensitive to plausible unmeasured Bernoulli, Poisson and Gamma confounders.",
keywords = "Censored costs, SEER-Medicare, Sensitivity analysis",
author = "Handorf, {Elizabeth A.} and Bekelman, {Justin E.} and Heitjan, {Daniel F.} and Nandita Mitra",
year = "2013",
month = "12",
day = "1",
doi = "10.1214/13-AOAS665",
language = "English (US)",
volume = "7",
pages = "2062--2080",
journal = "Annals of Applied Statistics",
issn = "1932-6157",
publisher = "Institute of Mathematical Statistics",
number = "4",

}

TY - JOUR

T1 - Evaluating costs with unmeasured confounding

T2 - A sensitivity analysis for the treatment effect

AU - Handorf, Elizabeth A.

AU - Bekelman, Justin E.

AU - Heitjan, Daniel F.

AU - Mitra, Nandita

PY - 2013/12/1

Y1 - 2013/12/1

N2 - Estimates of the effects of treatment on cost from observational studies are subject to bias if there are unmeasured confounders. It is therefore advisable in practice to assess the potential magnitude of such biases. We derive a general adjustment formula for loglinear models of mean cost and explore special cases under plausible assumptions about the distribution of the unmeasured confounder. We assess the performance of the adjustment by simulation, in particular, examining robustness to a key assumption of conditional independence between the unmeasured and measured covariates given the treatment indicator. We apply our method to SEER-Medicare cost data for a stage II/III muscle-invasive bladder cancer cohort.We evaluate the costs for radical cystectomy vs. combined radiation/chemotherapy, and find that the significance of the treatment effect is sensitive to plausible unmeasured Bernoulli, Poisson and Gamma confounders.

AB - Estimates of the effects of treatment on cost from observational studies are subject to bias if there are unmeasured confounders. It is therefore advisable in practice to assess the potential magnitude of such biases. We derive a general adjustment formula for loglinear models of mean cost and explore special cases under plausible assumptions about the distribution of the unmeasured confounder. We assess the performance of the adjustment by simulation, in particular, examining robustness to a key assumption of conditional independence between the unmeasured and measured covariates given the treatment indicator. We apply our method to SEER-Medicare cost data for a stage II/III muscle-invasive bladder cancer cohort.We evaluate the costs for radical cystectomy vs. combined radiation/chemotherapy, and find that the significance of the treatment effect is sensitive to plausible unmeasured Bernoulli, Poisson and Gamma confounders.

KW - Censored costs

KW - SEER-Medicare

KW - Sensitivity analysis

UR - http://www.scopus.com/inward/record.url?scp=84891552387&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84891552387&partnerID=8YFLogxK

U2 - 10.1214/13-AOAS665

DO - 10.1214/13-AOAS665

M3 - Article

VL - 7

SP - 2062

EP - 2080

JO - Annals of Applied Statistics

JF - Annals of Applied Statistics

SN - 1932-6157

IS - 4

ER -