fitOutcomeModel: Create an outcome model, and compute the relative risk

In escott12/CohortMethod: New-User Cohort Method with Large Scale Propensity and Outcome Models

Description

Create an outcome model, and computes the relative risk

Usage

fitOutcomeModel(
  population,
  cohortMethodData = NULL,
  modelType = "logistic",
  stratified = FALSE,
  useCovariates = FALSE,
  inversePtWeighting = FALSE,
  estimator = "ate",
  maxWeight = 0,
  interactionCovariateIds = c(),
  excludeCovariateIds = c(),
  includeCovariateIds = c(),
  profileGrid = NULL,
  profileBounds = c(log(0.1), log(10)),
  prior = createPrior("laplace", useCrossValidation = TRUE),
  control = createControl(cvType = "auto", seed = 1, startingVariance = 0.01, tolerance
    = 2e-07, cvRepetitions = 10, noiseLevel = "quiet")
)

Arguments

population	A population object generated by createStudyPopulation(), potentially filtered by other functions.
cohortMethodData	An object of type CohortMethodData as generated using getDbCohortMethodData(). Can be omitted if not using covariates and not using interaction terms.
modelType	The type of outcome model that will be used. Possible values are "logistic", "poisson", or "cox".
stratified	Should the regression be conditioned on the strata defined in the population object (e.g. by matching or stratifying on propensity scores)?
useCovariates	Whether to use the covariates in the cohortMethodData object in the outcome model.
inversePtWeighting	Use inverse probability of treatment weighting (IPTW)? See details.
estimator	for IPTW: the type of estimator. Options are estimator = "ate" for the average treatment effect, and estimator = "att"for the average treatment effect in the treated.
maxWeight	for IPTW: the maximum weight. Larger values will be truncated to this value. maxWeight = 0 means no truncation takes place.
interactionCovariateIds	An optional vector of covariate IDs to use to estimate interactions with the main treatment effect.
excludeCovariateIds	Exclude these covariates from the outcome model.
includeCovariateIds	Include only these covariates in the outcome model.
profileGrid	A one-dimensional grid of points on the log(relative risk) scale where the likelihood for coefficient of variables is sampled. See details.
profileBounds	The bounds (on the log relative risk scale) for the adaptive sampling of the likelihood function. See details.
prior	The prior used to fit the model. See Cyclops::createPrior() for details.
control	The control object used to control the cross-validation used to determine the hyperparameters of the prior (if applicable). See Cyclops::createControl() for details.

Details

IPTW estimates either the average treatment effect (ate) or average treatment effect in the treated (att) using stabilized inverse propensity scores (Xu et al. 2010).

For likelihood profiling, either specify the profileGrid for a completely user- defined grid, or profileBounds for an adaptive grid. Both should be defined on the log effect size scale. When both profileGrid and profileGrid are NULL likelihood profiling is disabled.

Value

An object of class OutcomeModel. Generic function print, coef, and confint are available.

References

Xu S, Ross C, Raebel MA, Shetterly S, Blanchette C, Smith D. Use of stabilized inverse propensity scores as weights to directly estimate relative risk and its confidence intervals. Value Health. 2010;13(2):273-277. doi:10.1111/j.1524-4733.2009.00671.x

escott12/CohortMethod documentation built on Dec. 20, 2021, 6:37 a.m.