Title: Pool Literature-Based and Individual Participant Data Based Spline Estimates
Version: 0.1.0
Author: Tommi Härkänen [aut, cre]
Maintainer: Tommi Härkänen <tommi.harkanen@thl.fi>
Depends: R (≥ 4.2.0)
Description: Pooling estimates reported in meta-analyses (literature-based, LB) and estimates based on individual participant data (IPD) is not straight-forward as the details of the LB nonlinear function estimate are not usually reported. This package pools the nonlinear IPD dose-response estimates based on a natural cubic spline from lm or glm with the pointwise LB estimates and their estimated variances. Details will be presented in Härkänen, Tapanainen, Sares-Jäske, Männistö, Kaartinen and Paalanen (2025) "Novel pooling method for nonlinear cohort analysis and meta-analysis estimates: Predicting health outcomes based on climate-friendly diets" (under revision) https://journals.lww.com/epidem/pages/default.aspx.
License: GPL (≥ 3)
Encoding: UTF-8
RoxygenNote: 7.3.2
Imports: rlang, dplyr, tidyr, tibble, stringr, meta, optimization
Suggests: knitr, rmarkdown, splines2
VignetteBuilder: knitr
NeedsCompilation: no
Packaged: 2025-07-17 05:14:38 UTC; thah
Repository: CRAN
Date/Publication: 2025-07-21 08:50:02 UTC

Title Pool meta-analysis estimates and estimates from a regression model.

Description

Title Pool meta-analysis estimates and estimates from a regression model.

Usage

pool_all_splines(v, meta.df, glm.res)

Arguments

v

Name of the covariate, which is modeled using an nsk spline.

meta.df

Meta-analysis estimates: dataframe with columns variable (covariate name), est (log HR estimate), est.var (estimated variance) and cov.value (covariate values where est and est.var were reported).

glm.res

Regression analysis result object.

Value

List containing pooled estimates of the spline parameters.

Examples

# Estimate a linear regression model using an individual participant data (IPD):
library(metasplines)
library(splines2)
res <- lm(
  Petal.Width ~
    Species +
    nsk(Sepal.Length, Boundary.knots = c(4.5, 7.5), knots = c(5, 6, 6.5)),
  data=iris)
# "Literature-based" (LB) estimates:
lb.df <- read.table(text=
"variable,     cov.value,  est,  est.var
Sepal.Length,  4.5,	       0,     0
Sepal.Length,  5,	         0.15,  0.01
Sepal.Length,  5.5,	       0.25,  0.01
Sepal.Length,  6,	         0.4,   0.01
Sepal.Length,  6.5,	       0.5,   0.01
Sepal.Length,  8,          0.25,  0.04
", sep=",", header=TRUE)
# Output table with the point estimates and the estimated variances:
pool_splines(v="Sepal.Length", meta.df=lb.df, glm.res=res)

Title Pool meta-analysis estimates and estimates from a regression model.

Description

Title Pool meta-analysis estimates and estimates from a regression model.

Usage

pool_splines(
  v,
  meta.df,
  glm.res,
  cor.m = NULL,
  x.range = NULL,
  full.output = FALSE
)

Arguments

v

Name of the covariate, which is modeled using an nsk spline (see package splines2).

meta.df

Meta-analysis estimates: dataframe with columns est (e.g. log HR estimate), est.var (estimated variance), variable (name of the covariate used in the spline) and cov.value (covariate value at which est and est.var were reported).

glm.res

Regression analysis result object.

cor.m

Assumed correlation matrix. If NULL (default) or NA then use correlation matrix from glm.res.

x.range

If NULL (default), then take the range from meta.df, otherwise give range as a vector with two components.

full.output

If TRUE then output also the log HR values and 95% confidence intervals over a grid of covariate values.

Value

List containing pooled estimates of the spline parameters.