indirect_PAF_Sjolander_onesimulation {causalPAF}R Documentation

Calculation of Population Attributable Fraction (PAF), with a decomposition of the total PAF into direct and indirect components.

Description

Calculation of Population Attributable Fraction (PAF), with a decomposition of the total PAF into direct and indirect components. It performs one simulation which can be combined with a bootstrap approach to perform multiple simulations. If we think of Y_0 as the potential outcome for an individual if they were never exposed to the risk factor, can be directly interpreted as the relative change in disease prevalence if an exposure was absent from the population. Sjolander introduced the ideas of mediation into the literature for PAF, defining a decomposition of the total PAF into direct and indirect components: and CausalDAG.jpg options: width=100 alt="Causal Bayesian network DAG"

Usage

indirect_PAF_Sjolander_onesimulation(
  data_frame,
  exposure,
  mediator,
  response,
  mediator_model,
  response_model,
  response_model_2,
  weights
)

Arguments

data_frame

Data frame containing the data. The data frame has cases in rows and variables in columns.

exposure

The exposure name in the form of character string e.g. “phys” for physical exercise.

mediator

The mediator name in the form of character string e.g. “whr” for waist hip ratio.

response

The outcome name in the form of character string e.g. “case” for a stroke case.

mediator_model

A list containing each of the fitted mediator regression models e.g. mediator_model=list(model_list[[6]],model_list[[7]],model_list[[8]]).

response_model

is a regression for the outcome on all mediators together with all parents and confounders of the mediators in a Markov causal Bayesian network DAG e.g.

response_model_2

is a regression for the outcome on the exposure together with all parents and confounders of the exposure in a Markov causal Bayesian network DAG along with other risk factors at the same level of the causal Bayesian network DAG. E.G. If physical exercise (“exer”) in the example given in the diagram is the exposure. Then the regression would include all parents of “exer” (i.e. sex, region, educ, age) as well as risk factors at the same level of the causal Bayesian network (i.e. stress, smoke, diet, alcoh).

weights

A numeric n x 1 vector where n is the number of patients in the case control data frame. Different weighting approaches can be applied as per the literature, Pathway-specific population attributable fractions (PS-PAF) O'Connell and Ferguson (2022), Revisiting sequential population attributable fractions Ferguson, O'Connell and O'Donnell (2020). For more information on weighting, a tutorial paper will be published and linked here when it is published. For example, O'Connell and Ferguson (2022), for a case-control study design, when prevalence pi is known, and the sampled disease cases and controls are randomly selected from their respective populations. We assume for simplicity that the case to control matching ration is 1 to r, for some r greater than or equal to 1. Under assumptions outlined in O'Connell and Ferguson (2022), the components of the PAF can be found as the corresponding empirical expectations and distributions in the re-weighted dataset where cases are assigned the weights, 1, and controls are assigned the weights of (((1 divided by pi) minus 1) all divided by r). Effectively, then we can think of the re-weighted population as a random sample. A tutorial paper will be linked here when published for more information which will show how to apply the weighting for different study designs.

Value

totalPAF

total PAF

directPAF

direct PAF

indirectPAF

indirect PAF


[Package causalPAF version 1.2.5 Index]