sens_plot: Plots for sensitivity analyses

Description Usage Arguments Details Examples

Description

Produces line plots (.type=="line") showing the bias factor on the relative risk (RR) scale vs. the proportion of studies with true RRs above .q (or below it for an apparently preventive relative risk). The plot secondarily includes a X-axis scaled based on the minimum strength of confounding to produce the given bias factor. The shaded region represents a 95% pointwise confidence band. Alternatively, produces distribution plots (.type=="dist") for a specific bias factor showing the observed and true distributions of RRs with a red line marking exp(.q).

Usage

1
2
3
sens_plot(.type, .q, .muB = NULL, .Bmin = log(1), .Bmax = log(5),
  .sigB = 0, .yr, .vyr = NULL, .t2, .vt2 = NULL, breaks.x1 = NULL,
  breaks.x2 = NULL, CI.level = 0.95)

Arguments

.type

dist for distribution plot; line for line plot (see Details)

.q

True effect size that is the threshold for "scientific significance"

.muB

Single mean bias factor on log scale (only needed for distribution plot)

.Bmin

Lower limit of lower X-axis on the log scale (only needed for line plot)

.Bmax

Upper limit of lower X-axis on the log scale (only needed for line plot)

.sigB

Standard deviation of log bias factor across studies (length 1)

.yr

Pooled point estimate (on log scale) from confounded meta-analysis

.vyr

Estimated variance of pooled point estimate from confounded meta-analysis

.t2

Estimated heterogeneity (tau^2) from confounded meta-analysis

.vt2

Estimated variance of tau^2 from confounded meta-analysis

breaks.x1

Breaks for lower X-axis (bias factor) on RR scale

breaks.x2

Breaks for upper X-axis (confounding strength) on RR scale

CI.level

Poitnwise confidence level as a proportion

Details

Arguments .vyr and .vt2 can be left NULL, in which case no confidence band will appear on the line plot.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
# with variable bias and with confidence band
sens_plot( .type="line", .q=log(1.1), .Bmin=log(1), .Bmax=log(4), .sigB=0.1,
           .yr=log(1.3), .vyr=0.005, .t2=0.4, .vt2=0.03 )

# with fixed bias and without confidence band
sens_plot( .type="line", .q=log(1.1), .Bmin=log(1), .Bmax=log(4),
           .yr=log(1.3), .t2=0.4 )

# apparently preventive
sens_plot( .type="line", .q=log(0.90), .Bmin=log(1), .Bmax=log(4),
           .yr=log(0.6), .vyr=0.005, .t2=0.4, .vt2=0.04 )

# distribution plot: apparently causative
# commented out because takes 5-10 seconds to run
# sens_plot( .type="dist", .q=log(1.1), .muB=log(2),
#           .yr=log(1.3), .t2=0.4 )
           
# distribution plot: apparently preventive
# commented out because takes 5-10 seconds to run
# sens_plot( .type="dist", .q=log(0.90), .muB=log(1.5),
#           .yr=log(0.7), .t2=0.2 )

Example output

There were 50 or more warnings (use warnings() to see the first 50)

ConfoundedMeta documentation built on May 2, 2019, 9:44 a.m.