#安装包运行包
library(BUGSnet)##
## This version of BUGSnet is distributed under the Creative Commons
## Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
## You are responsible for conforming to the terms of this license.
## Commercial use is strictly prohibited. Contact the authors for more information.library(readr)
library(rmarkdown)
library(knitr)
library(caTools)
library(rticles)#读取数据
setwd("D:\\黄老师META分析\\减量网状META")
data1 <- read_csv("dedosenetmeta.csv",show_col_types = FALSE)#输出数据
data <- data.prep(arm.data = data1,
varname.t = "T",
varname.s = "study") #将表格data1输入成参数data
network.char <- net.tab(data = data,
outcome = "R",
N = "N",
type.outcome = "binomial",
time = NULL) #实现参数的输出
print(data)## $arm.data
## # A tibble: 25 × 4
## study T N R
## <chr> <chr> <dbl> <dbl>
## 1 san ctxslow 39 15
## 2 san tacslow 42 16
## 3 si ctxslow 24 5
## 4 si mmfslow 23 6
## 5 ba ctxslow 20 1
## 6 ba mulslow 20 10
## 7 jiu ctxslow 20 6
## 8 jiu mmfslow 20 9
## 9 jiu tacslow 20 9
## 10 shisan ctxfast 50 25
## # ℹ 15 more rows
##
## $treatments
## # A tibble: 7 × 1
## T
## <chr>
## 1 ctxfast
## 2 ctxslow
## 3 mmffast
## 4 mmfslow
## 5 mulslow
## 6 tacslow
## 7 vocfast
##
## $studies
## # A tibble: 12 × 1
## study
## <chr>
## 1 ba
## 2 ershier
## 3 ershisi
## 4 jiu
## 5 san
## 6 shiba
## 7 shijiu
## 8 shiliu
## 9 shiqi
## 10 shisan
## 11 shisi
## 12 si
##
## $n.arms
## # A tibble: 12 × 2
## study n.arms
## <chr> <int>
## 1 ba 2
## 2 ershier 2
## 3 ershisi 2
## 4 jiu 3
## 5 san 2
## 6 shiba 2
## 7 shijiu 2
## 8 shiliu 2
## 9 shiqi 2
## 10 shisan 2
## 11 shisi 2
## 12 si 2
##
## $varname.t
## [1] "T"
##
## $varname.s
## [1] "study"
##
## attr(,"class")
## [1] "BUGSnetData"??network.char #教学## 打开httpd帮助服务器… 好了#画网状图
net.plot(data)
#选择效应模型
ranmodel <- nma.model(data = data,#随机效应模型
outcome = "R",
N = "N",
reference="ctxslow",
family="binomial",
link="logit",
effects="random") #建立模型## Warning: Unknown or uninitialised column: `sd`.??nma.model
ranresults <- nma.run(ranmodel,
n.adapt=1000,
n.burnin=10000,
n.iter=50000) #运行模型## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 25
## Unobserved stochastic nodes: 32
## Total graph size: 569
##
## Initializing modelfixmodel <- nma.model(data=data,#固定效应模型
outcome = "R",
N = "N",
reference="ctxslow",
family="binomial",
link="logit",
effects="fixed",
sd = NULL)## Warning: Unknown or uninitialised column: `sd`.fixresults <- nma.run(fixmodel,
n.adapt=1000,
n.burnin=10000,
n.iter=50000)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 25
## Unobserved stochastic nodes: 18
## Total graph size: 526
##
## Initializing model??nma.run#输出诊断图（Diagnostic plot），选择DIC最小的模型
par(mfrow=c(1,2))
nma.fit(ranresults, main= "random model")## Warning in nma.fit(ranresults, main = "random model"): Leverage cannot be
## calculated for zero cells.## $DIC
## [1] NaN
##
## $Dres
## [1] 27.58237
##
## $pD
## [1] NaN
##
## $leverage
##
## 1 0.3559842 0.7924677 NaN 0.5456982 0.8121302 1.009352 0.7722612 0.7807798
##
## 1 0.9016629 0.9490217 0.8852822 0.7357541 0.919962 0.7149068 0.9995687
##
## 1 0.6941508 0.7986344 0.9879383 0.7479476 0.9077212 0.9343481 0.946188
##
## 1 0.9210682 0.7233903 0.6508653
##
## $w
## r.1.1. r.2.1. r.3.1. r.4.1. r.5.1. r.6.1. r.7.1.
## -1.4628762 0.9439790 -1.8361361 -0.7503795 1.0064615 1.0566477 0.8788279
## r.8.1. r.9.1. r.10.1. r.11.1. r.12.1. r.1.2. r.2.2.
## 0.8869761 -0.9642400 -0.9741855 -0.9424244 0.9006083 1.2692334 -0.9215205
## r.3.2. r.4.2. r.5.2. r.6.2. r.7.2. r.8.2. r.9.2.
## 1.4697377 -0.8400963 -0.9942089 -1.0325077 -0.8648826 -0.9536726 0.9783885
## r.10.2. r.11.2. r.12.2. r.4.3.
## 0.9727325 0.9608061 -0.8840046 0.8390891
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 2.1400069 0.8910964 3.3713958 0.5630694 1.0129647 1.1165043
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 0.7723386 0.7867266 0.9297587 0.9490375 0.8881637 0.8110954
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 1.6109535 0.8491999 2.1601289 0.7057617 0.9884514 1.0660722
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 0.7480220 0.9094914 0.9572441 0.9462085 0.9231484 0.7814642
## dev_a.4.3.
## 0.7040705nma.fit(fixresults, main ="fixed model" )## Warning in nma.fit(fixresults, main = "fixed model"): Leverage cannot be
## calculated for zero cells.
## $DIC
## [1] NaN
##
## $Dres
## [1] 28.94215
##
## $pD
## [1] NaN
##
## $leverage
##
## 1 0.1941098 0.7115702 NaN 0.3835879 0.6163187 0.9928508 0.6767383 0.5570787
##
## 1 0.8241577 0.9220939 0.7742487 0.6255326 0.7510127 0.6025514 0.9097314
##
## 1 0.5896783 0.6218012 0.9684799 0.646007 0.8382775 0.8775562 0.9191788
##
## 1 0.8527868 0.6404246 0.4905574
##
## $w
## r.1.1. r.2.1. r.3.1. r.4.1. r.5.1. r.6.1. r.7.1.
## -1.7662739 0.9326049 -1.9693365 -0.6427294 1.0805132 1.0879559 -0.8227106
## r.8.1. r.9.1. r.10.1. r.11.1. r.12.1. r.1.2. r.2.2.
## 0.7628487 -0.9560651 -0.9602674 -0.8849978 0.8545428 1.4444345 -0.9036061
## r.3.2. r.4.2. r.5.2. r.6.2. r.7.2. r.8.2. r.9.2.
## 1.5778177 -0.7733471 -1.0498429 -1.0521666 0.8038066 -0.9199851 0.9736959
## r.10.2. r.11.2. r.12.2. r.4.3.
## 0.9587732 0.9272975 -0.8524616 0.7454072
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 3.1197236 0.8697519 3.8782861 0.4131010 1.1675087 1.1836481
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 0.6768528 0.5819382 0.9140605 0.9221134 0.7832211 0.7302435
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 2.0863909 0.8165039 2.4895088 0.5980657 1.1021701 1.1070545
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 0.6461050 0.8463726 0.9480838 0.9192460 0.8598807 0.7266907
## dev_a.4.3.
## 0.5556319#算不出DIC不知道怎么解决，若是依据第二个参数决定，则选择随机效应模型
nma.diag(ranresults)
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## Press [ENTER] to continue plotting trace plots (or type 'stop' to end plotting)>
## $gelman.rubin
## $psrf
## Point est. Upper C.I.
## d[2] 1.000944 1.001806
## d[3] 1.366634 2.069017
## d[4] 1.000681 1.002122
## d[5] 1.000422 1.001509
## d[6] 1.000390 1.001038
## d[7] 1.366568 2.068667
## sigma 1.001539 1.004704
##
## $mpsrf
## [1] 1.234939
##
## attr(,"class")
## [1] "gelman.rubin.results"
##
## $geweke
## $stats
## Chain 1 Chain 2 Chain 3
## d[2] -2.0203711 -1.479560625 -0.9312995
## d[3] -0.3784787 -0.325031220 2.0805119
## d[4] -1.8235731 -1.002071371 -1.9261887
## d[5] -0.5743604 0.044772307 -0.4557182
## d[6] -2.9024467 0.801797645 -0.8450174
## d[7] -0.3703438 -0.325448015 2.0423271
## sigma -0.3097959 0.005800912 -0.8304517
##
## $frac1
## [1] 0.1
##
## $frac2
## [1] 0.5
##
## attr(,"class")
## [1] "geweke.results"#选择一致性模型
inconmodel <- nma.model(data = data,
outcome = "R",
N = "N",
reference="ctxslow",
family="binomial",
link="logit",
type = "inconsistency",
effects="random")## Warning: Unknown or uninitialised column: `sd`.inconresults<- nma.run(inconmodel,
n.adapt=5000,
n.burnin=3,
n.iter=20000)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 25
## Unobserved stochastic nodes: 47
## Total graph size: 533
##
## Initializing modelconresults<- nma.model(data = data,
outcome = "R",
N = "N",
reference="ctxslow",
family = "binomial",
link = "logit",
effects = "random",
type="consistency",)## Warning: Unknown or uninitialised column: `sd`.conresults <- nma.run(conresults,
n.adapt=5000,
n.burnin=3,
n.iter=20000)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 25
## Unobserved stochastic nodes: 32
## Total graph size: 569
##
## Initializing modelpar(mfrow=c(1,2))
nma.fit(conresults,main = "consistency" )## Warning in nma.fit(conresults, main = "consistency"): Leverage cannot be
## calculated for zero cells.## $DIC
## [1] NaN
##
## $Dres
## [1] 27.52796
##
## $pD
## [1] NaN
##
## $leverage
##
## 1 0.3807674 0.7925824 NaN 0.5476298 0.8275581 1.005241 0.7512784 0.779917
##
## 1 0.9097601 0.9219575 0.8757634 0.7385188 0.9356071 0.7104301 1.001939
##
## 1 0.6988889 0.802747 0.9773404 0.7339876 0.9206257 0.9472211 0.9310996
##
## 1 0.9201437 0.7304123 0.657638
##
## $w
## r.1.1. r.2.1. r.3.1. r.4.1. r.5.1. r.6.1. r.7.1.
## -1.4656820 0.9395430 -1.8350514 -0.7508497 1.0269248 1.0552742 0.8667783
## r.8.1. r.9.1. r.10.1. r.11.1. r.12.1. r.1.2. r.2.2.
## 0.8853111 -0.9628331 -0.9604073 -0.9374066 0.8952143 1.2667678 -0.9148861
## r.3.2. r.4.2. r.5.2. r.6.2. r.7.2. r.8.2. r.9.2.
## 1.4741742 -0.8396718 -0.9995027 -1.0285979 -0.8567358 -0.9604769 0.9797496
## r.10.2. r.11.2. r.12.2. r.4.3.
## 0.9650082 0.9601159 -0.8867539 0.8367713
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 2.1482238 0.8827410 3.3674138 0.5637752 1.0545745 1.1136036
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 0.7513046 0.7837758 0.9270475 0.9223822 0.8787311 0.8014086
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 1.6047006 0.8370166 2.1731897 0.7050487 0.9990056 1.0580136
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 0.7339962 0.9225159 0.9599093 0.9312408 0.9218225 0.7863325
## dev_a.4.3.
## 0.7001862nma.fit(inconresults,main = "inconsistency")## Warning in nma.fit(inconresults, main = "inconsistency"): Leverage cannot be
## calculated for zero cells.
## $DIC
## [1] NaN
##
## $Dres
## [1] 28.09926
##
## $pD
## [1] NaN
##
## $leverage
##
## 1 0.4261908 0.8233271 NaN 0.5540792 0.8562468 1.013849 1.03266 0.8244538
##
## 1 0.9110295 1.021298 0.9184864 0.7807206 0.9590641 0.7553314 1.028758 0.7553367
##
## 1 0.8411754 0.9790015 1.051092 0.9330515 0.9354141 1.015662 0.9414785 0.7578589
##
## 1 0.706121
##
## $w
## r.1.1. r.2.1. r.3.1. r.4.1. r.5.1. r.6.1. r.7.1.
## -1.3989091 0.9546478 -1.7928626 -0.7536378 1.0153448 1.0490591 1.0162206
## r.8.1. r.9.1. r.10.1. r.11.1. r.12.1. r.1.2. r.2.2.
## 0.9108063 -0.9627123 -1.0106059 -0.9599298 0.9193610 1.2405317 -0.9391352
## r.3.2. r.4.2. r.5.2. r.6.2. r.7.2. r.8.2. r.9.2.
## 1.4350745 -0.8719611 -1.0023062 -1.0238485 -1.0253724 -0.9668896 0.9734387
## r.10.2. r.11.2. r.12.2. r.4.3.
## 1.0078482 0.9712304 -0.8998323 0.8625290
##
## $pmdev
## dev_a.1.1. dev_a.2.1. dev_a.3.1. dev_a.4.1. dev_a.5.1. dev_a.6.1.
## 1.9569466 0.9113525 3.2143563 0.5679700 1.0309251 1.1005250
## dev_a.7.1. dev_a.8.1. dev_a.9.1. dev_a.10.1. dev_a.11.1. dev_a.12.1.
## 1.0327043 0.8295681 0.9268149 1.0213243 0.9214652 0.8452247
## dev_a.1.2. dev_a.2.2. dev_a.3.2. dev_a.4.2. dev_a.5.2. dev_a.6.2.
## 1.5389188 0.8819748 2.0594388 0.7603162 1.0046177 1.0482658
## dev_a.7.2. dev_a.8.2. dev_a.9.2. dev_a.10.2. dev_a.11.2. dev_a.12.2.
## 1.0513886 0.9348755 0.9475829 1.0157580 0.9432886 0.8096982
## dev_a.4.3.
## 0.7439562#选择一致性#绘制森林图
options(scipen = 1)
nma.forest(nma = ranresults,
comparator = "ctxslow",
)
#不知道为什么横坐标是
??nma.forest
??data.prep
try1 <- nma.model(
data = data,
outcome = "R",
N = "N",
reference = "ctxslow",
family = "binomial",
link = "logit",
effects = "random",
type = "consistency",
time = NULL
)## Warning: Unknown or uninitialised column: `sd`.try.run <- nma.run(
model = try1,
n.adapt = 100,
n.burnin = 0,
n.iter = 100
)## Compiling model graph
## Resolving undeclared variables
## Allocating nodes
## Graph information:
## Observed stochastic nodes: 25
## Unobserved stochastic nodes: 32
## Total graph size: 569
##
## Initializing model??nma.run
#make forest plot
nma.forest(nma = try.run,comparator = "vocfast")
#两两对比森林图
#ctx缓慢减量对比mmf缓慢减量
pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "ctxslow",
name.trt2 = "mmfslow",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 ctxslow vs. mmfslow 4 0 0.5932529 0.2857987 1.231458
## fe.estimate fe.lci fe.uci
## 1 0.5441868 0.2696855 1.098091
##
## $forest
## $forest$xlim
## [1] -6.028214 6.028214
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "si" "jiu" "ershier" "ershisi"
##
## $forest$TE.format
## [1] "" "" "" "-0.2935" "-0.6466" "-0.2048" "-2.9348"
##
## $forest$seTE.format
## [1] "" "" "" "0.6915" "0.6634" "0.6406" "1.5783"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "0.54" "0.59" "" "0.75" "0.52" "0.81" "0.05"
##
## $forest$ci.format
## [1] "[0.27; 1.10]" "[0.29; 1.23]" "" "[0.19; 2.89]" "[0.14; 1.92]"
## [6] "[0.23; 2.86]" "[0.00; 1.17]"
##
## $forest$effect.ci.format
## [1] "0.54 [0.27; 1.10]" "0.59 [0.29; 1.23]" ""
## [4] "0.75 [0.19; 2.89]" "0.52 [0.14; 1.92]" "0.81 [0.23; 2.86]"
## [7] "0.05 [0.00; 1.17]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 2.8 14 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci" "col.w.common" "col.w.random"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of studies: k = 4
## Number of observations: o = 147 (o.e = 78, o.c = 69)
## Number of events: e = 50
##
## OR 95%-CI z p-value
## Common effect model 0.5442 [0.2697; 1.0981] -1.70 0.0894
## Random effects model 0.5933 [0.2858; 1.2315] -1.40 0.1611
##
## Quantifying heterogeneity (with 95%-CIs):
## tau^2 = 0; tau = 0; I^2 = 0.0% [0.0%; 84.7%]; H = 1.00 [1.00; 2.56]
##
## Test of heterogeneity:
## Q d.f. p-value
## 2.78 3 0.4267
##
## Details of meta-analysis methods:
## - Mantel-Haenszel method (common effect model)
## - Inverse variance method (random effects model)
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)
## - Calculation of I^2 based on Q
## - Continuity correction of 0.5 in studies with zero cell frequencies#不知道为什么改变trt1和trt2的位置森林图里的实验组和对照组不改变位置
#异质性小无统计学意义
??pma
#ctx缓慢减量对比tac缓慢减量
pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "ctxslow",
name.trt2 = "tacslow",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 ctxslow vs. tacslow 3 0 0.6389882 0.429002 0.9517578
## fe.estimate fe.lci fe.uci
## 1 0.6391702 0.4297211 0.9507062
##
## $forest
## $forest$xlim
## [1] -1.946891 1.946891
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "san" "jiu" "shiqi"
##
## $forest$TE.format
## [1] "" "" "" "0.0155" "-0.6466" "-0.5506"
##
## $forest$seTE.format
## [1] "" "" "" "0.4575" "0.6634" "0.2415"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "0.64" "0.64" "" "1.02" "0.52" "0.58"
##
## $forest$ci.format
## [1] "[0.43; 0.95]" "[0.43; 0.95]" "" "[0.41; 2.49]" "[0.14; 1.92]"
## [6] "[0.36; 0.93]"
##
## $forest$effect.ci.format
## [1] "0.64 [0.43; 0.95]" "0.64 [0.43; 0.95]" ""
## [4] "1.02 [0.41; 2.49]" "0.52 [0.14; 1.92]" "0.58 [0.36; 0.93]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 2.6 13 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci" "col.w.common" "col.w.random"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of studies: k = 3
## Number of observations: o = 420 (o.e = 201, o.c = 219)
## Number of events: e = 161
##
## OR 95%-CI z p-value
## Common effect model 0.6392 [0.4297; 0.9507] -2.21 0.0271
## Random effects model 0.6390 [0.4290; 0.9518] -2.20 0.0276
##
## Quantifying heterogeneity (with 95%-CIs):
## tau^2 = 0; tau = 0; I^2 = 0.0% [0.0%; 89.6%]; H = 1.00 [1.00; 3.10]
##
## Test of heterogeneity:
## Q d.f. p-value
## 1.30 2 0.5229
##
## Details of meta-analysis methods:
## - Mantel-Haenszel method (common effect model)
## - Inverse variance method (random effects model)
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)
## - Calculation of I^2 based on Q#异质性小有统计学差异
#mmf快速减量对比voc快速减量
pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "mmffast",
name.trt2 = "vocfast",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 mmffast vs. vocfast 2 0 0.5288887 0.3611352 0.7745667
## fe.estimate fe.lci fe.uci
## 1 0.5294187 0.3613511 0.7756562
##
## $forest
## $forest$xlim
## [1] -1.198777 1.198777
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "shisi" "shiliu"
##
## $forest$TE.format
## [1] "" "" "" "-0.6721" "-0.5795"
##
## $forest$seTE.format
## [1] "" "" "" "0.2471" "0.3160"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "0.53" "0.53" "" "0.51" "0.56"
##
## $forest$ci.format
## [1] "[0.36; 0.78]" "[0.36; 0.77]" "" "[0.31; 0.83]" "[0.30; 1.04]"
##
## $forest$effect.ci.format
## [1] "0.53 [0.36; 0.78]" "0.53 [0.36; 0.77]" ""
## [4] "0.51 [0.31; 0.83]" "0.56 [0.30; 1.04]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 2.4 12 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci" "col.w.common" "col.w.random"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of studies: k = 2
## Number of observations: o = 622 (o.e = 266, o.c = 356)
## Number of events: e = 163
##
## OR 95%-CI z p-value
## Common effect model 0.5294 [0.3614; 0.7757] -3.26 0.0011
## Random effects model 0.5289 [0.3611; 0.7746] -3.27 0.0011
##
## Quantifying heterogeneity:
## tau^2 = 0; tau = 0; I^2 = 0.0%; H = 1.00
##
## Test of heterogeneity:
## Q d.f. p-value
## 0.05 1 0.8173
##
## Details of meta-analysis methods:
## - Mantel-Haenszel method (common effect model)
## - Inverse variance method (random effects model)
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)
## - Calculation of I^2 based on Q#异质性小有统计学意义
#ctx缓慢减量对比多靶点缓慢减量
pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "ctxslow",
name.trt2 = "mulslow",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 ctxslow vs. mulslow 2 0.6884099 0.1948402 0.0285524 1.32958
## fe.estimate fe.lci fe.uci
## 1 0.3557971 0.2318916 0.5459085
##
## $forest
## $forest$xlim
## [1] -5.138051 5.138051
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "ba" "shiba"
##
## $forest$TE.format
## [1] "" "" "" "-2.9444" "-0.9105"
##
## $forest$seTE.format
## [1] "" "" "" "1.1192" "0.2267"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "0.36" "0.19" "" "0.05" "0.40"
##
## $forest$ci.format
## [1] "[0.23; 0.55]" "[0.03; 1.33]" "" "[0.01; 0.47]" "[0.26; 0.63]"
##
## $forest$effect.ci.format
## [1] "0.36 [0.23; 0.55]" "0.19 [0.03; 1.33]" ""
## [4] "0.05 [0.01; 0.47]" "0.40 [0.26; 0.63]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 2.4 12 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci" "col.w.common" "col.w.random"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of studies: k = 2
## Number of observations: o = 402 (o.e = 201, o.c = 201)
## Number of events: e = 140
##
## OR 95%-CI z p-value
## Common effect model 0.3558 [0.2319; 0.5459] -4.73 < 0.0001
## Random effects model 0.1948 [0.0286; 1.3296] -1.67 0.0951
##
## Quantifying heterogeneity (with 95%-CIs):
## tau^2 = 1.4405; tau = 1.2002; I^2 = 68.8% [0.0%; 93.0%]; H = 1.79 [1.00; 3.77]
##
## Test of heterogeneity:
## Q d.f. p-value
## 3.21 1 0.0732
##
## Details of meta-analysis methods:
## - Mantel-Haenszel method (common effect model)
## - Inverse variance method (random effects model)
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)
## - Calculation of I^2 based on Q#I2=68.8，异质性偏大P=0.07，普通模型有意义，随机效应模型无意义
#mmf快速减量对比ctx缓慢减量；ctx快速减量对比缓慢减量，都仅一个研究，都没意义
pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "mmfslow",
name.trt2 = "ctxfast",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 mmfslow vs. ctxfast 1 NA 0.8518519 0.3884504 1.868067
## fe.estimate fe.lci fe.uci
## 1 0.8518519 0.3884504 1.868067
##
## $forest
## $forest$xlim
## [1] -0.9455897 0.9455897
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "shisan"
##
## $forest$TE.format
## [1] "" "" "" "-0.1603"
##
## $forest$seTE.format
## [1] "" "" "" "0.4006"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "" "" "" "0.85"
##
## $forest$ci.format
## [1] "" "" "" "[0.39; 1.87]"
##
## $forest$effect.ci.format
## [1] "" "" ""
## [4] "0.85 [0.39; 1.87]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 1.2 6 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of observations: o = 100 (o.e = 50, o.c = 50)
## Number of events: e = 52
##
## OR 95%-CI z p-value
## 0.8519 [0.3885; 1.8681] -0.40 0.6890pma(data = data,
type.outcome="binomial",
sm="OR",
name.trt1 = "ctxslow",
name.trt2 = "ctxfast",
outcome = "R",
N = "N")
## $summary
## comparison n.studies i.squared re.estimate re.lci re.uci
## 1 ctxslow vs. ctxfast 1 NA 0.64 0.1278929 3.202679
## fe.estimate fe.lci fe.uci
## 1 0.64 0.1278929 3.202679
##
## $forest
## $forest$xlim
## [1] -2.056562 2.056562
##
## $forest$addrows.below.overall
## [1] 0
##
## $forest$colgap
## [1] 2mm
##
## $forest$colgap.left
## [1] 2mm
##
## $forest$colgap.right
## [1] 2mm
##
## $forest$colgap.studlab
## [1] 2mm
##
## $forest$colgap.forest
## [1] 2mm
##
## $forest$colgap.forest.left
## [1] 2mm
##
## $forest$colgap.forest.right
## [1] 2mm
##
## $forest$studlab
## [1] "shijiu"
##
## $forest$TE.format
## [1] "" "" "" "-0.4463"
##
## $forest$seTE.format
## [1] "" "" "" "0.8216"
##
## $forest$cluster.format
## [1] "" "" ""
##
## $forest$cycles.format
## [1] "" "" ""
##
## $forest$effect.format
## [1] "" "" "" "0.64"
##
## $forest$ci.format
## [1] "" "" "" "[0.13; 3.20]"
##
## $forest$effect.ci.format
## [1] "" "" ""
## [4] "0.64 [0.13; 3.20]"
##
## $forest[[18]]
## NULL
##
## $forest[[19]]
## NULL
##
## $forest$figheight
## total_height total_rows height_per_row spacing
## 1 1.2 6 0.2 1
##
## $forest$leftcols
## [1] "col.studlab" "col.event.e" "col.n.e" "col.event.c" "col.n.c"
##
## $forest$leftlabs
## NULL
##
## $forest$rightcols
## [1] "col.effect" "col.ci"
##
## $forest$rightlabs
## NULL
##
##
## $raw
## Number of observations: o = 27 (o.e = 13, o.c = 14)
## Number of events: e = 18
##
## OR 95%-CI z p-value
## 0.6400 [0.1279; 3.2027] -0.54 0.5870#累计疗效排序图（SUCRA plot）
sucra <- nma.rank(try.run,
largerbetter=T,
sucra.palette= "Set1")## Warning: `separate_()` was deprecated in tidyr 1.2.0.
## ℹ Please use `separate()` instead.
## ℹ The deprecated feature was likely used in the BUGSnet package.
## Please report the issue at <https://github.com/audrey-b/BUGSnet/issues>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.??nma.rank
#形式1
sucra$rankogram
#形式2
sucra$sucraplot
??pdflatex