Meta-analysis thrombectomy latam
Niels Pacheco
2024-05-14
——————Meta analysis of proportions for mRS 0-2 at 3 months n(%) ———————————————
# Define the data as a text variable
data_text <- "First Author\tsample\tmRS
Yang 2020\t327\t119
Kiriinya 2023\t84\t25
Yang 2019\t60\t30
Cabral 2016\t31\t17
Jovin 2022\t110\t43
Luo 2023\t120\t57
Kurniawan 2024\t32\t9
Li 2023\t207\t79
Li 2023\t113\t48
DeCastroAlfonso 2019\t30\t15
NAkiri 2017\t161\t58
Zi 2017\t698\t304
Sun 2019\t187\t68
Long 2017\t139\t34
Bai 2018\t472\t194
Martins 2020\t111\t39
Antunes 2019\t27\t10
DeCastroAlfonso 2021\t645\t266
Huo 2023\t230\t69
Alet 2020\t27\t10
Pazuello 2021\t80\t21
Cirio 2020\t139\t66
Lucena 2016\t20\t7
Fornazari 2022\t565\t483
Phuong 2022\t49\t22
Huded 2014\t45\t29
Gurkas 2017\t31\t19
Ding 2024\t268\t132
Sang 2019\t48\t27
Quan 2019\t89\t47
Zhao 2018\t89\t43
Tao 2022\t228\t75
Liu 2020\t66\t22
Hao 2017\t632\t281
Castro-Afonso 2017\t15\t9
Beckhauser 2020\t54\t18
Zhao 2018\t30\t16
Zi-Chang 2019\t26\t12
Li 2022 b\t186\t78
Castro-Afonso 2012\t21\t13
Sun 2019\t187\t68
Ramazanoglu 2023\t359\t169"
# Read the data using read.table, specifying the header and delimiter
df <- read.table(text = data_text, header = TRUE, sep = "\t")
# Print the DataFrame to view the data
print(df)## First.Author sample mRS
## 1 Yang 2020 327 119
## 2 Kiriinya 2023 84 25
## 3 Yang 2019 60 30
## 4 Cabral 2016 31 17
## 5 Jovin 2022 110 43
## 6 Luo 2023 120 57
## 7 Kurniawan 2024 32 9
## 8 Li 2023 207 79
## 9 Li 2023 113 48
## 10 DeCastroAlfonso 2019 30 15
## 11 NAkiri 2017 161 58
## 12 Zi 2017 698 304
## 13 Sun 2019 187 68
## 14 Long 2017 139 34
## 15 Bai 2018 472 194
## 16 Martins 2020 111 39
## 17 Antunes 2019 27 10
## 18 DeCastroAlfonso 2021 645 266
## 19 Huo 2023 230 69
## 20 Alet 2020 27 10
## 21 Pazuello 2021 80 21
## 22 Cirio 2020 139 66
## 23 Lucena 2016 20 7
## 24 Fornazari 2022 565 483
## 25 Phuong 2022 49 22
## 26 Huded 2014 45 29
## 27 Gurkas 2017 31 19
## 28 Ding 2024 268 132
## 29 Sang 2019 48 27
## 30 Quan 2019 89 47
## 31 Zhao 2018 89 43
## 32 Tao 2022 228 75
## 33 Liu 2020 66 22
## 34 Hao 2017 632 281
## 35 Castro-Afonso 2017 15 9
## 36 Beckhauser 2020 54 18
## 37 Zhao 2018 30 16
## 38 Zi-Chang 2019 26 12
## 39 Li 2022 b 186 78
## 40 Castro-Afonso 2012 21 13
## 41 Sun 2019 187 68
## 42 Ramazanoglu 2023 359 169- Thrombectomy proportion mRS
# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}## Loading required package: meta## Loading required package: metadat## Loading 'meta' package (version 7.0-0).
## Type 'help(meta)' for a brief overview.
## Readers of 'Meta-Analysis with R (Use R!)' should install
## older version of 'meta' package: https://tinyurl.com/dt4y5drs# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis1 <- metaprop(event = mRS, n = sample, data = df,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = df$`First.Author`
)
# Summary of the meta-analysis
summary(meta_analysis1)## proportion 95%-CI %W(random)
## Yang 2020 0.3639 [0.3117; 0.4186] 2.8
## Kiriinya 2023 0.2976 [0.2027; 0.4073] 2.4
## Yang 2019 0.5000 [0.3681; 0.6319] 2.3
## Cabral 2016 0.5484 [0.3603; 0.7268] 1.9
## Jovin 2022 0.3909 [0.2993; 0.4886] 2.6
## Luo 2023 0.4750 [0.3831; 0.5682] 2.6
## Kurniawan 2024 0.2812 [0.1375; 0.4675] 1.8
## Li 2023 0.3816 [0.3152; 0.4515] 2.7
## Li 2023 0.4248 [0.3323; 0.5213] 2.6
## DeCastroAlfonso 2019 0.5000 [0.3130; 0.6870] 1.9
## NAkiri 2017 0.3602 [0.2862; 0.4395] 2.7
## Zi 2017 0.4355 [0.3984; 0.4732] 2.9
## Sun 2019 0.3636 [0.2947; 0.4370] 2.7
## Long 2017 0.2446 [0.1757; 0.3247] 2.6
## Bai 2018 0.4110 [0.3662; 0.4569] 2.9
## Martins 2020 0.3514 [0.2631; 0.4477] 2.6
## Antunes 2019 0.3704 [0.1940; 0.5763] 1.8
## DeCastroAlfonso 2021 0.4124 [0.3741; 0.4515] 2.9
## Huo 2023 0.3000 [0.2415; 0.3637] 2.7
## Alet 2020 0.3704 [0.1940; 0.5763] 1.8
## Pazuello 2021 0.2625 [0.1704; 0.3729] 2.3
## Cirio 2020 0.4748 [0.3895; 0.5612] 2.6
## Lucena 2016 0.3500 [0.1539; 0.5922] 1.6
## Fornazari 2022 0.8549 [0.8231; 0.8829] 2.8
## Phuong 2022 0.4490 [0.3067; 0.5977] 2.2
## Huded 2014 0.6444 [0.4878; 0.7813] 2.1
## Gurkas 2017 0.6129 [0.4219; 0.7815] 1.9
## Ding 2024 0.4925 [0.4312; 0.5541] 2.8
## Sang 2019 0.5625 [0.4118; 0.7052] 2.2
## Quan 2019 0.5281 [0.4194; 0.6349] 2.5
## Zhao 2018 0.4831 [0.3759; 0.5916] 2.5
## Tao 2022 0.3289 [0.2684; 0.3941] 2.7
## Liu 2020 0.3333 [0.2220; 0.4601] 2.3
## Hao 2017 0.4446 [0.4054; 0.4843] 2.9
## Castro-Afonso 2017 0.6000 [0.3229; 0.8366] 1.4
## Beckhauser 2020 0.3333 [0.2109; 0.4747] 2.2
## Zhao 2018 0.5333 [0.3433; 0.7166] 1.9
## Zi-Chang 2019 0.4615 [0.2659; 0.6663] 1.8
## Li 2022 b 0.4194 [0.3476; 0.4938] 2.7
## Castro-Afonso 2012 0.6190 [0.3844; 0.8189] 1.6
## Sun 2019 0.3636 [0.2947; 0.4370] 2.7
## Ramazanoglu 2023 0.4708 [0.4182; 0.5238] 2.8
##
## Number of studies: k = 42
## Number of observations: o = 7038
## Number of events: e = 3151
##
## proportion 95%-CI
## Random effects model 0.4340 [0.3931; 0.4758]
##
## Quantifying heterogeneity:
## tau^2 = 0.2507 [0.1449; 0.4376]; tau = 0.5007 [0.3806; 0.6615]
## I^2 = 90.3% [87.8%; 92.3%]; H = 3.21 [2.87; 3.60]
##
## Test of heterogeneity:
## Q d.f. p-value
## 422.81 41 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis1, layout = "RevMan5")
Meta analysis of proportions for mRS 0-2 at 3 months n(%)
—————— Meta analysis of proportions of mTICI 2b-3 post thrombectomy n(%) —————————
# Define the data as a text variable
data_text2 <- "First Author\tsample\tmTICI
Antunes 2017\t19\t15
Yang 2020\t2821\t240
Kiriinya 2023\t84\t52
Yang 2019\t60\t47
Cabral 2016\t31\t22
Luo 2023\t113\t61
Marquez-Romero 2020\t49\t34
Kurniawan 2024\t32\t20
Li 2023\t113\t100
DeCastroAlfonso 2019\t30\t27
NAkiri 2017\t158\t120
Zi 2017\t698\t579
Sun 2019\t187\t158
Long 2017\t139\t124
Bai 2018\t472\t407
Li 2022 a\t622\t510
Martins 2020\t111\t91
Antunes 2019\t27\t23
DeCastroAlfonso 2021\t645\t549
Huo 2023\t230\t169
Alet 2020\t27\t21
Pazuello 2021\t80\t69
Constantino 2023\t31\t24
Cirio 2020\t139\t104
Lucena 2016\t20\t17
Fornazari 2022\t565\t246
Phuong 2022\t49\t49
Gurkas 2017\t31\t20
Ding 2024\t268\t256
Sang 2019\t48\t45
Quan 2019\t89\t87
Zhao 2018\t89\t73
Tao 2022\t223\t208
CollaMachado 2016\t11\t8
Liu 2020\t63\t45
Churojana 2017\t41\t38
Hao 2017\t632\t534
Castro-Afonso 2017\t15\t12
Beckhauser 2020\t54\t50
Zhao 2018\t30\t18
Zi-Chang 2019\t26\t23
Li 2022 b\t186\t141
Castro-Afonso 2012\t21\t19
Geng 2021\t7674\t5152
Sun 2019\t187\t158
Ramazanoglu 2023\t359\t172"
# Read the data using read.table, specifying the header and delimiter
df2 <- read.table(text = data_text2, header = TRUE, sep = "\t")
# Print the DataFrame to view the data
print(df2)## First.Author sample mTICI
## 1 Antunes 2017 19 15
## 2 Yang 2020 2821 240
## 3 Kiriinya 2023 84 52
## 4 Yang 2019 60 47
## 5 Cabral 2016 31 22
## 6 Luo 2023 113 61
## 7 Marquez-Romero 2020 49 34
## 8 Kurniawan 2024 32 20
## 9 Li 2023 113 100
## 10 DeCastroAlfonso 2019 30 27
## 11 NAkiri 2017 158 120
## 12 Zi 2017 698 579
## 13 Sun 2019 187 158
## 14 Long 2017 139 124
## 15 Bai 2018 472 407
## 16 Li 2022 a 622 510
## 17 Martins 2020 111 91
## 18 Antunes 2019 27 23
## 19 DeCastroAlfonso 2021 645 549
## 20 Huo 2023 230 169
## 21 Alet 2020 27 21
## 22 Pazuello 2021 80 69
## 23 Constantino 2023 31 24
## 24 Cirio 2020 139 104
## 25 Lucena 2016 20 17
## 26 Fornazari 2022 565 246
## 27 Phuong 2022 49 49
## 28 Gurkas 2017 31 20
## 29 Ding 2024 268 256
## 30 Sang 2019 48 45
## 31 Quan 2019 89 87
## 32 Zhao 2018 89 73
## 33 Tao 2022 223 208
## 34 CollaMachado 2016 11 8
## 35 Liu 2020 63 45
## 36 Churojana 2017 41 38
## 37 Hao 2017 632 534
## 38 Castro-Afonso 2017 15 12
## 39 Beckhauser 2020 54 50
## 40 Zhao 2018 30 18
## 41 Zi-Chang 2019 26 23
## 42 Li 2022 b 186 141
## 43 Castro-Afonso 2012 21 19
## 44 Geng 2021 7674 5152
## 45 Sun 2019 187 158
## 46 Ramazanoglu 2023 359 172# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis2 <- metaprop(event = mTICI, n = sample, data = df2,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = First.Author
)
# Summary of the meta-analysis
summary(meta_analysis2)## proportion 95%-CI %W(random)
## Antunes 2017 0.7895 [0.5443; 0.9395] 2.0
## Yang 2020 0.0851 [0.0750; 0.0960] 2.4
## Kiriinya 2023 0.6190 [0.5066; 0.7229] 2.3
## Yang 2019 0.7833 [0.6580; 0.8793] 2.3
## Cabral 2016 0.7097 [0.5196; 0.8578] 2.2
## Luo 2023 0.5398 [0.4435; 0.6340] 2.3
## Marquez-Romero 2020 0.6939 [0.5458; 0.8175] 2.3
## Kurniawan 2024 0.6250 [0.4369; 0.7890] 2.2
## Li 2023 0.8850 [0.8113; 0.9373] 2.3
## DeCastroAlfonso 2019 0.9000 [0.7347; 0.9789] 1.9
## NAkiri 2017 0.7595 [0.6852; 0.8238] 2.3
## Zi 2017 0.8295 [0.7995; 0.8567] 2.4
## Sun 2019 0.8449 [0.7849; 0.8936] 2.3
## Long 2017 0.8921 [0.8283; 0.9383] 2.3
## Bai 2018 0.8623 [0.8279; 0.8921] 2.4
## Li 2022 a 0.8199 [0.7874; 0.8494] 2.4
## Martins 2020 0.8198 [0.7355; 0.8863] 2.3
## Antunes 2019 0.8519 [0.6627; 0.9581] 2.0
## DeCastroAlfonso 2021 0.8512 [0.8213; 0.8777] 2.4
## Huo 2023 0.7348 [0.6728; 0.7906] 2.4
## Alet 2020 0.7778 [0.5774; 0.9138] 2.1
## Pazuello 2021 0.8625 [0.7673; 0.9293] 2.2
## Constantino 2023 0.7742 [0.5890; 0.9041] 2.1
## Cirio 2020 0.7482 [0.6676; 0.8179] 2.3
## Lucena 2016 0.8500 [0.6211; 0.9679] 1.9
## Fornazari 2022 0.4354 [0.3941; 0.4774] 2.4
## Phuong 2022 1.0000 [0.9275; 1.0000] 1.0
## Gurkas 2017 0.6452 [0.4537; 0.8077] 2.2
## Ding 2024 0.9552 [0.9231; 0.9767] 2.3
## Sang 2019 0.9375 [0.8280; 0.9869] 1.9
## Quan 2019 0.9775 [0.9212; 0.9973] 1.8
## Zhao 2018 0.8202 [0.7245; 0.8936] 2.3
## Tao 2022 0.9327 [0.8915; 0.9619] 2.3
## CollaMachado 2016 0.7273 [0.3903; 0.9398] 1.8
## Liu 2020 0.7143 [0.5865; 0.8211] 2.3
## Churojana 2017 0.9268 [0.8008; 0.9846] 1.9
## Hao 2017 0.8449 [0.8143; 0.8723] 2.4
## Castro-Afonso 2017 0.8000 [0.5191; 0.9567] 1.9
## Beckhauser 2020 0.9259 [0.8211; 0.9794] 2.0
## Zhao 2018 0.6000 [0.4060; 0.7734] 2.2
## Zi-Chang 2019 0.8846 [0.6985; 0.9755] 1.9
## Li 2022 b 0.7581 [0.6900; 0.8177] 2.4
## Castro-Afonso 2012 0.9048 [0.6962; 0.9883] 1.8
## Geng 2021 0.6714 [0.6607; 0.6819] 2.4
## Sun 2019 0.8449 [0.7849; 0.8936] 2.3
## Ramazanoglu 2023 0.4791 [0.4264; 0.5322] 2.4
##
## Number of studies: k = 46
## Number of observations: o = 17599
## Number of events: e = 10937
##
## proportion 95%-CI
## Random effects model 0.8000 [0.7339; 0.8529]
##
## Quantifying heterogeneity:
## tau^2 = 1.4939 [0.7523; 2.4996]; tau = 1.2222 [0.8673; 1.5810]
## I^2 = 98.5% [98.3%; 98.7%]; H = 8.16 [7.68; 8.67]
##
## Test of heterogeneity:
## Q d.f. p-value
## 2993.59 45 0
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies
## - Continuity correction of 0.5 in studies with zero cell frequencies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis2, layout = "RevMan5")
Meta analysis of proportions of mTICI 2b-3 post thrombectomy n(%)
————–Meta analysis of proportions of Symptomatic intracerebral hemorrhage —————————-
# Define the data as a text variable
data_text3 <- "First Author\tsample\tSymp_intrac_hem
Antunes 2017\t19\t1
Yang 2020\t327\t14
Yang 2019\t60\t7
Cabral 2016\t31\t3
Jovin 2022\t110\t6
Luo 2023\t120\t9
Marquez-Romero 2020\t49\t2
Li 2023\t204\t18
Li 2023\t113\t8
DeCastroAlfonso 2019\t30\t2
NAkiri 2017\t161\t11
Zi 2017\t698\t108
Sun 2019\t187\t8
Bai 2018\t472\t81
Li 2022 a\t622\t27
Martins 2020\t111\t5
Antunes 2019\t27\t1
DeCastroAlfonso 2021\t645\t54
Huo 2023\t230\t14
Alet 2020\t27\t2
Pazuello 2021\t80\t7
Constantino 2023\t31\t5
Cirio 2020\t139\t11
Lucena 2016\t20\t1
Fornazari 2022\t565\t46
Phuong 2022\t32\t5
Huded 2014\t45\t2
Ding 2024\t268\t16
Sang 2019\t48\t2
Quan 2019\t89\t4
Zhao 2018\t89\t9
Tao 2022\t228\t12
Liu 2020\t66\t5
Churojana 2017\t41\t5
Hao 2017\t632\t101
Castro-Afonso 2017\t15\t0
Beckhauser 2020\t54\t6
Zhao 2018\t30\t4
Zi-Chang 2019\t26\t3
Castro-Afonso 2012\t21\t3
Wickly 2009\t17\t1
Sun 2019\t187\t8
Ramazanoglu 2023\t359\t29"
# Read the data using read.table, specifying the header and delimiter
df3 <- read.table(text = data_text3, header = TRUE, sep = "\t")
# Print the DataFrame to view the data
print(df3)## First.Author sample Symp_intrac_hem
## 1 Antunes 2017 19 1
## 2 Yang 2020 327 14
## 3 Yang 2019 60 7
## 4 Cabral 2016 31 3
## 5 Jovin 2022 110 6
## 6 Luo 2023 120 9
## 7 Marquez-Romero 2020 49 2
## 8 Li 2023 204 18
## 9 Li 2023 113 8
## 10 DeCastroAlfonso 2019 30 2
## 11 NAkiri 2017 161 11
## 12 Zi 2017 698 108
## 13 Sun 2019 187 8
## 14 Bai 2018 472 81
## 15 Li 2022 a 622 27
## 16 Martins 2020 111 5
## 17 Antunes 2019 27 1
## 18 DeCastroAlfonso 2021 645 54
## 19 Huo 2023 230 14
## 20 Alet 2020 27 2
## 21 Pazuello 2021 80 7
## 22 Constantino 2023 31 5
## 23 Cirio 2020 139 11
## 24 Lucena 2016 20 1
## 25 Fornazari 2022 565 46
## 26 Phuong 2022 32 5
## 27 Huded 2014 45 2
## 28 Ding 2024 268 16
## 29 Sang 2019 48 2
## 30 Quan 2019 89 4
## 31 Zhao 2018 89 9
## 32 Tao 2022 228 12
## 33 Liu 2020 66 5
## 34 Churojana 2017 41 5
## 35 Hao 2017 632 101
## 36 Castro-Afonso 2017 15 0
## 37 Beckhauser 2020 54 6
## 38 Zhao 2018 30 4
## 39 Zi-Chang 2019 26 3
## 40 Castro-Afonso 2012 21 3
## 41 Wickly 2009 17 1
## 42 Sun 2019 187 8
## 43 Ramazanoglu 2023 359 29# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis3 <- metaprop(event = Symp_intrac_hem, n = sample, data = df3,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = First.Author
)
# Summary of the meta-analysis
summary(meta_analysis3)## proportion 95%-CI %W(random)
## Antunes 2017 0.0526 [0.0013; 0.2603] 0.7
## Yang 2020 0.0428 [0.0236; 0.0708] 3.2
## Yang 2019 0.1167 [0.0482; 0.2257] 2.4
## Cabral 2016 0.0968 [0.0204; 0.2575] 1.6
## Jovin 2022 0.0545 [0.0203; 0.1149] 2.4
## Luo 2023 0.0750 [0.0349; 0.1376] 2.7
## Marquez-Romero 2020 0.0408 [0.0050; 0.1398] 1.3
## Li 2023 0.0882 [0.0531; 0.1359] 3.3
## Li 2023 0.0708 [0.0311; 0.1347] 2.6
## DeCastroAlfonso 2019 0.0667 [0.0082; 0.2207] 1.2
## NAkiri 2017 0.0683 [0.0346; 0.1190] 2.9
## Zi 2017 0.1547 [0.1287; 0.1837] 4.0
## Sun 2019 0.0428 [0.0186; 0.0826] 2.7
## Bai 2018 0.1716 [0.1387; 0.2087] 3.9
## Li 2022 a 0.0434 [0.0288; 0.0625] 3.6
## Martins 2020 0.0450 [0.0148; 0.1020] 2.2
## Antunes 2019 0.0370 [0.0009; 0.1897] 0.7
## DeCastroAlfonso 2021 0.0837 [0.0635; 0.1078] 3.9
## Huo 2023 0.0609 [0.0337; 0.1000] 3.1
## Alet 2020 0.0741 [0.0091; 0.2429] 1.2
## Pazuello 2021 0.0875 [0.0359; 0.1720] 2.5
## Constantino 2023 0.1613 [0.0545; 0.3373] 2.0
## Cirio 2020 0.0791 [0.0402; 0.1372] 2.9
## Lucena 2016 0.0500 [0.0013; 0.2487] 0.7
## Fornazari 2022 0.0814 [0.0602; 0.1071] 3.8
## Phuong 2022 0.1562 [0.0528; 0.3279] 2.0
## Huded 2014 0.0444 [0.0054; 0.1515] 1.3
## Ding 2024 0.0597 [0.0345; 0.0951] 3.2
## Sang 2019 0.0417 [0.0051; 0.1425] 1.3
## Quan 2019 0.0449 [0.0124; 0.1111] 1.9
## Zhao 2018 0.1011 [0.0473; 0.1833] 2.7
## Tao 2022 0.0526 [0.0275; 0.0901] 3.0
## Liu 2020 0.0758 [0.0251; 0.1680] 2.1
## Churojana 2017 0.1220 [0.0408; 0.2620] 2.1
## Hao 2017 0.1598 [0.1321; 0.1907] 4.0
## Castro-Afonso 2017 0.0000 [0.0000; 0.2180] 0.4
## Beckhauser 2020 0.1111 [0.0419; 0.2263] 2.3
## Zhao 2018 0.1333 [0.0376; 0.3072] 1.8
## Zi-Chang 2019 0.1154 [0.0245; 0.3015] 1.6
## Castro-Afonso 2012 0.1429 [0.0305; 0.3634] 1.5
## Wickly 2009 0.0588 [0.0015; 0.2869] 0.7
## Sun 2019 0.0428 [0.0186; 0.0826] 2.7
## Ramazanoglu 2023 0.0808 [0.0548; 0.1140] 3.6
##
## Number of studies: k = 43
## Number of observations: o = 7325
## Number of events: e = 666
##
## proportion 95%-CI
## Random effects model 0.0795 [0.0667; 0.0946]
##
## Quantifying heterogeneity:
## tau^2 = 0.2231 [0.0217; 0.3041]; tau = 0.4723 [0.1472; 0.5514]
## I^2 = 74.3% [65.4%; 80.9%]; H = 1.97 [1.70; 2.29]
##
## Test of heterogeneity:
## Q d.f. p-value
## 163.53 42 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies
## - Continuity correction of 0.5 in studies with zero cell frequencies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis3, layout = "RevMan5")
Meta analysis of proportions of Symptomatic intracerebral hemorrhage
———— Meta analysis of proportions Mortality rate within 3 months after discharge n(%) ————————–
# Define the data as a text variable
data_text4 <- "First Author\tsample\tMortality
Antunes 2017\t19\t8
Yang 2020\t327\t58
Kiriinya 2023\t84\t29
Yang 2019\t60\t12
Cabral 2016\t31\t8
Jovin 2022\t110\t34
Luo 2023\t120\t18
Li 2023\t207\t26
Li 2023\t113\t21
DeCastroAlfonso 2019\t30\t9
NAkiri 2017\t161\t37
Zi 2017\t698\t177
Sun 2019\t187\t38
Long 2017\t139\t20
Bai 2018\t472\t122
Li 2022 a\t622\t78
Martins 2020\t111\t27
Antunes 2019\t27\t10
DeCastroAlfonso 2021\t645\t157
Huo 2023\t230\t50
Pazuello 2021\t80\t31
Cirio 2020\t139\t26
Lucena 2016\t20\t4
Fornazari 2022\t565\t125
Huded 2014\t45\t8
Gurkas 2017\t31\t5
Ding 2024\t268\t47
Quan 2019\t89\t21
Zhao 2018\t89\t19
Tao 2022\t228\t83
CollaMachado 2016\t11\t3
Liu 2020\t66\t22
Hao 2017\t632\t166
Beckhauser 2020\t54\t11
Zhao 2018\t30\t0
Zi-Chang 2019\t26\t4
Li 2022 b\t186\t39
Castro-Afonso 2012\t21\t2
Sun 2019\t187\t38
Ramazanoglu 2023\t359\t105"
# Read the data using read.table, specifying the header and delimiter
df4 <- read.table(text = data_text4, header = TRUE, sep = "\t")
# Print the DataFrame to view the data
print(df4)## First.Author sample Mortality
## 1 Antunes 2017 19 8
## 2 Yang 2020 327 58
## 3 Kiriinya 2023 84 29
## 4 Yang 2019 60 12
## 5 Cabral 2016 31 8
## 6 Jovin 2022 110 34
## 7 Luo 2023 120 18
## 8 Li 2023 207 26
## 9 Li 2023 113 21
## 10 DeCastroAlfonso 2019 30 9
## 11 NAkiri 2017 161 37
## 12 Zi 2017 698 177
## 13 Sun 2019 187 38
## 14 Long 2017 139 20
## 15 Bai 2018 472 122
## 16 Li 2022 a 622 78
## 17 Martins 2020 111 27
## 18 Antunes 2019 27 10
## 19 DeCastroAlfonso 2021 645 157
## 20 Huo 2023 230 50
## 21 Pazuello 2021 80 31
## 22 Cirio 2020 139 26
## 23 Lucena 2016 20 4
## 24 Fornazari 2022 565 125
## 25 Huded 2014 45 8
## 26 Gurkas 2017 31 5
## 27 Ding 2024 268 47
## 28 Quan 2019 89 21
## 29 Zhao 2018 89 19
## 30 Tao 2022 228 83
## 31 CollaMachado 2016 11 3
## 32 Liu 2020 66 22
## 33 Hao 2017 632 166
## 34 Beckhauser 2020 54 11
## 35 Zhao 2018 30 0
## 36 Zi-Chang 2019 26 4
## 37 Li 2022 b 186 39
## 38 Castro-Afonso 2012 21 2
## 39 Sun 2019 187 38
## 40 Ramazanoglu 2023 359 105# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis4 <- metaprop(event = Mortality, n = sample, data = df4,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = First.Author
)
# Summary of the meta-analysis
summary(meta_analysis4)## proportion 95%-CI %W(random)
## Antunes 2017 0.4211 [0.2025; 0.6650] 1.3
## Yang 2020 0.1774 [0.1375; 0.2232] 3.4
## Kiriinya 2023 0.3452 [0.2448; 0.4569] 2.7
## Yang 2019 0.2000 [0.1078; 0.3233] 2.0
## Cabral 2016 0.2581 [0.1186; 0.4461] 1.5
## Jovin 2022 0.3091 [0.2245; 0.4043] 2.9
## Luo 2023 0.1500 [0.0914; 0.2267] 2.5
## Li 2023 0.1256 [0.0837; 0.1786] 2.9
## Li 2023 0.1858 [0.1189; 0.2699] 2.6
## DeCastroAlfonso 2019 0.3000 [0.1473; 0.4940] 1.6
## NAkiri 2017 0.2298 [0.1673; 0.3026] 3.1
## Zi 2017 0.2536 [0.2217; 0.2876] 3.9
## Sun 2019 0.2032 [0.1480; 0.2681] 3.1
## Long 2017 0.1439 [0.0902; 0.2134] 2.6
## Bai 2018 0.2585 [0.2195; 0.3005] 3.7
## Li 2022 a 0.1254 [0.1004; 0.1540] 3.6
## Martins 2020 0.2432 [0.1668; 0.3338] 2.8
## Antunes 2019 0.3704 [0.1940; 0.5763] 1.6
## DeCastroAlfonso 2021 0.2434 [0.2108; 0.2784] 3.8
## Huo 2023 0.2174 [0.1659; 0.2764] 3.3
## Pazuello 2021 0.3875 [0.2806; 0.5030] 2.7
## Cirio 2020 0.1871 [0.1260; 0.2619] 2.8
## Lucena 2016 0.2000 [0.0573; 0.4366] 1.0
## Fornazari 2022 0.2212 [0.1877; 0.2578] 3.8
## Huded 2014 0.1778 [0.0800; 0.3205] 1.6
## Gurkas 2017 0.1613 [0.0545; 0.3373] 1.2
## Ding 2024 0.1754 [0.1318; 0.2263] 3.3
## Quan 2019 0.2360 [0.1524; 0.3378] 2.5
## Zhao 2018 0.2135 [0.1337; 0.3131] 2.4
## Tao 2022 0.3640 [0.3015; 0.4301] 3.5
## CollaMachado 2016 0.2727 [0.0602; 0.6097] 0.7
## Liu 2020 0.3333 [0.2220; 0.4601] 2.4
## Hao 2017 0.2627 [0.2287; 0.2988] 3.9
## Beckhauser 2020 0.2037 [0.1063; 0.3353] 1.9
## Zhao 2018 0.0000 [0.0000; 0.1157] 0.2
## Zi-Chang 2019 0.1538 [0.0436; 0.3487] 1.0
## Li 2022 b 0.2097 [0.1536; 0.2753] 3.1
## Castro-Afonso 2012 0.0952 [0.0117; 0.3038] 0.6
## Sun 2019 0.2032 [0.1480; 0.2681] 3.1
## Ramazanoglu 2023 0.2925 [0.2459; 0.3425] 3.7
##
## Number of studies: k = 40
## Number of observations: o = 7519
## Number of events: e = 1698
##
## proportion 95%-CI
## Random effects model 0.2267 [0.2059; 0.2490]
##
## Quantifying heterogeneity:
## tau^2 = 0.0943 [0.0571; 0.2434]; tau = 0.3071 [0.2389; 0.4934]
## I^2 = 74.6% [65.5%; 81.3%]; H = 1.99 [1.70; 2.31]
##
## Test of heterogeneity:
## Q d.f. p-value
## 153.69 39 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies
## - Continuity correction of 0.5 in studies with zero cell frequencies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis4, layout = "RevMan5")
Meta analysis of proportions Mortality rate within 3 months after discharge n(%)
——————Meta analysis of proportions for mRS 0-2 at 3 months n(%) in posterior circulation ———————————————
mRS_posterior <- read.table(text = "
Study_ID Sample mRS
Jovin_2022 110 43
Sun_2019 187 68
Long_2017 139 34
Pazuello_2021 80 21
Phuong_2022 49 22
Sang_2019 48 27
Quan_2019 89 47
Sun_2019 187 68
Liu_2020 66 22
", header = TRUE, stringsAsFactors = FALSE)
print(mRS_posterior)## Study_ID Sample mRS
## 1 Jovin_2022 110 43
## 2 Sun_2019 187 68
## 3 Long_2017 139 34
## 4 Pazuello_2021 80 21
## 5 Phuong_2022 49 22
## 6 Sang_2019 48 27
## 7 Quan_2019 89 47
## 8 Sun_2019 187 68
## 9 Liu_2020 66 22# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis5 <- metaprop(event = mRS, n = Sample, data = mRS_posterior,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis5)## proportion 95%-CI %W(random)
## Jovin_2022 0.3909 [0.2993; 0.4886] 11.9
## Sun_2019 0.3636 [0.2947; 0.4370] 13.1
## Long_2017 0.2446 [0.1757; 0.3247] 11.8
## Pazuello_2021 0.2625 [0.1704; 0.3729] 10.2
## Phuong_2022 0.4490 [0.3067; 0.5977] 9.3
## Sang_2019 0.5625 [0.4118; 0.7052] 9.2
## Quan_2019 0.5281 [0.4194; 0.6349] 11.4
## Sun_2019 0.3636 [0.2947; 0.4370] 13.1
## Liu_2020 0.3333 [0.2220; 0.4601] 10.0
##
## Number of studies: k = 9
## Number of observations: o = 955
## Number of events: e = 352
##
## proportion 95%-CI
## Random effects model 0.3799 [0.3178; 0.4461]
##
## Quantifying heterogeneity:
## tau^2 = 0.1263 [0.0411; 0.6764]; tau = 0.3553 [0.2028; 0.8224]
## I^2 = 74.5% [50.6%; 86.8%]; H = 1.98 [1.42; 2.76]
##
## Test of heterogeneity:
## Q d.f. p-value
## 31.39 8 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis5, layout = "RevMan5")
—————— Meta analysis of proportions of mTICI 2b-3 post thrombectomy n(%) in posterior circulation —————————
mTICI_posterior <- read.table(text = "
Study_ID Sample mTICI
Antunes_2017 19 15
Sun_2019 187 158
Long_2017 139 124
Pazuello_2021 80 69
Phuong_2022 49 49
Sang_2019 48 45
Quan_2019 89 87
Sun_2019 187 158
Liu_2020 63 45
", header = TRUE, stringsAsFactors = FALSE)
print(mTICI_posterior)## Study_ID Sample mTICI
## 1 Antunes_2017 19 15
## 2 Sun_2019 187 158
## 3 Long_2017 139 124
## 4 Pazuello_2021 80 69
## 5 Phuong_2022 49 49
## 6 Sang_2019 48 45
## 7 Quan_2019 89 87
## 8 Sun_2019 187 158
## 9 Liu_2020 63 45# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis6 <- metaprop(event = mTICI, n = Sample, data = mTICI_posterior,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis6)## proportion 95%-CI %W(random)
## Antunes_2017 0.7895 [0.5443; 0.9395] 8.3
## Sun_2019 0.8449 [0.7849; 0.8936] 16.5
## Long_2017 0.8921 [0.8283; 0.9383] 14.7
## Pazuello_2021 0.8625 [0.7673; 0.9293] 13.4
## Phuong_2022 1.0000 [0.9275; 1.0000] 2.0
## Sang_2019 0.9375 [0.8280; 0.9869] 7.8
## Quan_2019 0.9775 [0.9212; 0.9973] 6.2
## Sun_2019 0.8449 [0.7849; 0.8936] 16.5
## Liu_2020 0.7143 [0.5865; 0.8211] 14.6
##
## Number of studies: k = 9
## Number of observations: o = 861
## Number of events: e = 750
##
## proportion 95%-CI
## Random effects model 0.8696 [0.8139; 0.9104]
##
## Quantifying heterogeneity:
## tau^2 = 0.2397 [0.1251; 2.6230]; tau = 0.4896 [0.3537; 1.6196]
## I^2 = 69.3% [38.7%; 84.6%]; H = 1.81 [1.28; 2.55]
##
## Test of heterogeneity:
## Q d.f. p-value
## 26.06 8 0.0010
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies
## - Continuity correction of 0.5 in studies with zero cell frequencies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis6, layout = "RevMan5")
Meta analysis of proportions of mTICI 2b-3 post thrombectomy n(%) in posterior circulation
————–Meta analysis of proportions of Symptomatic intracerebral hemorrhage in posterior circulation —————————-
sympt_posterior <- read.table(text = "
Study_ID Sample Symp_intrac_hem
Antunes_2017 19 1
Jovin_2022 110 6
Sun_2019 187 8
Pazuello_2021 80 7
Phuong_2022 32 5
Sang_2019 48 2
Quan_2019 89 4
Sun_2019 187 8
Liu_2020 66 5
", header = TRUE, stringsAsFactors = FALSE)
print(sympt_posterior)## Study_ID Sample Symp_intrac_hem
## 1 Antunes_2017 19 1
## 2 Jovin_2022 110 6
## 3 Sun_2019 187 8
## 4 Pazuello_2021 80 7
## 5 Phuong_2022 32 5
## 6 Sang_2019 48 2
## 7 Quan_2019 89 4
## 8 Sun_2019 187 8
## 9 Liu_2020 66 5# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis7 <- metaprop(event = Symp_intrac_hem, n = Sample, data = sympt_posterior,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis7)## proportion 95%-CI %W(random)
## Antunes_2017 0.0526 [0.0013; 0.2603] 2.4
## Jovin_2022 0.0545 [0.0203; 0.1149] 13.2
## Sun_2019 0.0428 [0.0186; 0.0826] 17.2
## Pazuello_2021 0.0875 [0.0359; 0.1720] 14.7
## Phuong_2022 0.1562 [0.0528; 0.3279] 10.1
## Sang_2019 0.0417 [0.0051; 0.1425] 4.8
## Quan_2019 0.0449 [0.0124; 0.1111] 9.2
## Sun_2019 0.0428 [0.0186; 0.0826] 17.2
## Liu_2020 0.0758 [0.0251; 0.1680] 11.0
##
## Number of studies: k = 9
## Number of observations: o = 818
## Number of events: e = 46
##
## proportion 95%-CI
## Random effects model 0.0605 [0.0448; 0.0813]
##
## Quantifying heterogeneity:
## tau^2 = 0.0213 [0.0000; 0.6283]; tau = 0.1461 [0.0000; 0.7927]
## I^2 = 9.0% [0.0%; 68.0%]; H = 1.05 [1.00; 1.77]
##
## Test of heterogeneity:
## Q d.f. p-value
## 8.79 8 0.3602
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis7, layout = "RevMan5")
Meta analysis of proportions of Symptomatic intracerebral hemorrhage in posterior circulation
———— Meta analysis of proportions Mortality rate within 3 months after discharge n(%) in posterior circulation ————————–
Mortality_posterior <- read.table(text = "
Study_ID Sample Mortality
Antunes_2017 19 8
Jovin_2022 110 34
Sun_2019 187 38
Long_2017 139 20
Pazuello_2021 80 31
Quan_2019 89 21
Sun_2019 187 38
Liu_2020 66 22
", header = TRUE, stringsAsFactors = FALSE)
print(Mortality_posterior)## Study_ID Sample Mortality
## 1 Antunes_2017 19 8
## 2 Jovin_2022 110 34
## 3 Sun_2019 187 38
## 4 Long_2017 139 20
## 5 Pazuello_2021 80 31
## 6 Quan_2019 89 21
## 7 Sun_2019 187 38
## 8 Liu_2020 66 22# Install the meta package if not already installed
if (!require(meta)) {
install.packages("meta")
library(meta)
}
# Assuming 'df' is your DataFrame loaded as previously shown
# Meta-analysis of proportions with study labels
# Calculate the pooled proportion using metaprop function, labeling each study by the First Author
meta_analysis8 <- metaprop(event = Mortality, n = Sample, data = Mortality_posterior,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis8)## proportion 95%-CI %W(random)
## Antunes_2017 0.4211 [0.2025; 0.6650] 7.3
## Jovin_2022 0.3091 [0.2245; 0.4043] 13.7
## Sun_2019 0.2032 [0.1480; 0.2681] 14.4
## Long_2017 0.1439 [0.0902; 0.2134] 12.7
## Pazuello_2021 0.3875 [0.2806; 0.5030] 13.0
## Quan_2019 0.2360 [0.1524; 0.3378] 12.4
## Sun_2019 0.2032 [0.1480; 0.2681] 14.4
## Liu_2020 0.3333 [0.2220; 0.4601] 12.1
##
## Number of studies: k = 8
## Number of observations: o = 877
## Number of events: e = 212
##
## proportion 95%-CI
## Random effects model 0.2607 [0.2036; 0.3273]
##
## Quantifying heterogeneity:
## tau^2 = 0.1546 [0.0401; 0.9080]; tau = 0.3931 [0.2002; 0.9529]
## I^2 = 74.6% [48.6%; 87.4%]; H = 1.98 [1.39; 2.82]
##
## Test of heterogeneity:
## Q d.f. p-value
## 27.52 7 0.0003
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Jackson method for confidence interval of tau^2 and tau
## - Logit transformation
## - Clopper-Pearson confidence interval for individual studies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis8, layout = "RevMan5")
————————————– Meta analysis of odds ratios ———————————–
————— Meta-analysis of odds ratios for the risk of mRS 0-2 in thrombectomy vs bridge therapy ————-
mrs_odds <- read.table(text = "
Study_ID Sample_brid event_mRs_bridge no_event_mRS_brigde Sample_thromb event_mRs_throm no_event_mRS_throm
Furlan_2022 15 6 9 17 7 10
Du_2021 54 18 36 57 20 37
Tong_2021 405 174 231 568 251 317
Zi_2019 118 55 63 116 65 51
Sun_2019 36 15 21 151 53 98
Ramazanoglu_2023 149 83 66 210 86 124
", header = TRUE, stringsAsFactors = FALSE)
print(mrs_odds)## Study_ID Sample_brid event_mRs_bridge no_event_mRS_brigde
## 1 Furlan_2022 15 6 9
## 2 Du_2021 54 18 36
## 3 Tong_2021 405 174 231
## 4 Zi_2019 118 55 63
## 5 Sun_2019 36 15 21
## 6 Ramazanoglu_2023 149 83 66
## Sample_thromb event_mRs_throm no_event_mRS_throm
## 1 17 7 10
## 2 57 20 37
## 3 568 251 317
## 4 116 65 51
## 5 151 53 98
## 6 210 86 124# Load the necessary library
library(meta)
# Assuming mrs_odds is already loaded with your data
# You might need to check the names of your columns with names(mrs_odds)
# Meta-analysis of odds ratios
meta_analysis9 <- metabin(
event.e = mrs_odds$event_mRs_throm,
n.e = mrs_odds$Sample_thromb,
event.c = mrs_odds$event_mRs_bridge,
n.c = mrs_odds$Sample_brid,
data = mrs_odds,
studlab = paste(mrs_odds$Study_ID),
sm = "OR", # Specify summary measure as Odds Ratio
method.tau = "DL", # DerSimonian-Laird estimator for tau^2
comb.fixed = FALSE, # Random effects model
comb.random = TRUE, # Include random effects
prediction = FALSE # No prediction interval by default
)
# Summary of the meta-analysis
summary(meta_analysis9)## OR 95%-CI %W(random)
## Furlan_2022 1.0500 [0.2553; 4.3189] 4.5
## Du_2021 1.0811 [0.4932; 2.3699] 11.5
## Tong_2021 1.0512 [0.8128; 1.3594] 30.1
## Zi_2019 1.4599 [0.8723; 2.4434] 19.0
## Sun_2019 0.7571 [0.3605; 1.5902] 12.4
## Ramazanoglu_2023 0.5515 [0.3608; 0.8431] 22.5
##
## Number of studies: k = 6
## Number of observations: o = 1896 (o.e = 1119, o.c = 777)
## Number of events: e = 833
##
## OR 95%-CI z p-value
## Random effects model 0.9318 [0.6769; 1.2827] -0.43 0.6648
##
## Quantifying heterogeneity:
## tau^2 = 0.0711; tau = 0.2666; I^2 = 50.3% [0.0%; 80.2%]; H = 1.42 [1.00; 2.25]
##
## Test of heterogeneity:
## Q d.f. p-value
## 10.06 5 0.0736
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis9, layout = "RevMan5")
Meta-analysis of odds ratios for the risk of mRS 0-2 in thrombectomy vs bridge therapy
————— Meta-analysis of odds ratios for the risk of mTICI 2b-3 in thrombectomy vs bridge therapy ————-
mtici_odds <- read.table(text = "
Study_ID Sample_brid event_mTICI_bridg no_event_mTICI_bridge Sample_thromb event_mTICI_throm no_event_mTICI_throm
Furlan_2022 21 21 0 21 16 5
Du_2021 57 42 15 57 46 11
Geng_2021 5605 1343 4262 5605 3809 1796
Tong_2021 600 379 221 600 523 77
Zi_2019 116 102 14 116 100 16
Sun_2019 151 30 121 151 128 23
Ramazanoglu_2023 210 132 78 210 172 38
", header = TRUE, stringsAsFactors = FALSE)
print(mtici_odds)## Study_ID Sample_brid event_mTICI_bridg no_event_mTICI_bridge
## 1 Furlan_2022 21 21 0
## 2 Du_2021 57 42 15
## 3 Geng_2021 5605 1343 4262
## 4 Tong_2021 600 379 221
## 5 Zi_2019 116 102 14
## 6 Sun_2019 151 30 121
## 7 Ramazanoglu_2023 210 132 78
## Sample_thromb event_mTICI_throm no_event_mTICI_throm
## 1 21 16 5
## 2 57 46 11
## 3 5605 3809 1796
## 4 600 523 77
## 5 116 100 16
## 6 151 128 23
## 7 210 172 38# Load the necessary library
library(meta)
# Assuming mrs_odds is already loaded with your data
# You might need to check the names of your columns with names(mrs_odds)
# Meta-analysis of odds ratios
meta_analysis10 <- metabin(
event.e = mtici_odds$event_mTICI_throm,
n.e = mtici_odds$Sample_thromb,
event.c = mtici_odds$event_mTICI_bridg,
n.c = mtici_odds$Sample_brid,
data = mtici_odds,
studlab = paste(mtici_odds$Study_ID),
sm = "OR", # Specify summary measure as Odds Ratio
method.tau = "DL", # DerSimonian-Laird estimator for tau^2
comb.fixed = FALSE, # Random effects model
comb.random = TRUE, # Include random effects
prediction = FALSE # No prediction interval by default
)
# Summary of the meta-analysis
summary(meta_analysis10)## OR 95%-CI %W(random)
## Furlan_2022 0.0698 [ 0.0036; 1.3535] 3.4
## Du_2021 1.4935 [ 0.6174; 3.6128] 13.3
## Geng_2021 6.7304 [ 6.1936; 7.3138] 18.5
## Tong_2021 3.9606 [ 2.9603; 5.2990] 17.8
## Zi_2019 0.8578 [ 0.3978; 1.8499] 14.3
## Sun_2019 22.4464 [12.3515; 40.7917] 15.7
## Ramazanoglu_2023 2.6746 [ 1.7068; 4.1913] 16.9
##
## Number of studies: k = 7
## Number of observations: o = 13520 (o.e = 6760, o.c = 6760)
## Number of events: e = 6843
##
## OR 95%-CI z p-value
## Random effects model 3.2972 [1.7924; 6.0652] 3.84 0.0001
##
## Quantifying heterogeneity:
## tau^2 = 0.5213; tau = 0.7220; I^2 = 93.2% [88.5%; 96.0%]; H = 3.84 [2.95; 5.00]
##
## Test of heterogeneity:
## Q d.f. p-value
## 88.46 6 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)
## - Continuity correction of 0.5 in studies with zero cell frequencies# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis10, layout = "RevMan5")
Meta-analysis of odds ratios for the risk of mTICI 2b-3 in thrombectomy vs bridge therapy
————— Meta-analysis of odds ratios for the risk of Symptomatic intracerebral hemorrhage in thrombectomy vs bridge therapy ————-
sympt_odds <- read.table(text = "
Study_ID Sample_brid event_sympt_bridge no_event_sympt_bridge Sample_thromb event_sympt_throm no_event_sympt_throm
Du_2021 57 9 48 57 7 50
Tong_2021 407 42 365 569 33 536
Zi_2019 115 12 103 115 10 105
Sun_2019 151 1 150 151 7 144
Ramazanoglu_2023 210 11 199 210 18 192
", header = TRUE, stringsAsFactors = FALSE)
print(sympt_odds)## Study_ID Sample_brid event_sympt_bridge no_event_sympt_bridge
## 1 Du_2021 57 9 48
## 2 Tong_2021 407 42 365
## 3 Zi_2019 115 12 103
## 4 Sun_2019 151 1 150
## 5 Ramazanoglu_2023 210 11 199
## Sample_thromb event_sympt_throm no_event_sympt_throm
## 1 57 7 50
## 2 569 33 536
## 3 115 10 105
## 4 151 7 144
## 5 210 18 192# Load the necessary library
library(meta)
# Assuming mrs_odds is already loaded with your data
# You might need to check the names of your columns with names(mrs_odds)
# Meta-analysis of odds ratios
meta_analysis11 <- metabin(
event.e = sympt_odds$event_sympt_throm,
n.e = sympt_odds$Sample_thromb,
event.c = sympt_odds$event_sympt_bridge,
n.c = sympt_odds$Sample_brid,
data = sympt_odds,
studlab = paste(sympt_odds$Study_ID),
sm = "OR", # Specify summary measure as Odds Ratio
method.tau = "DL", # DerSimonian-Laird estimator for tau^2
comb.fixed = FALSE, # Random effects model
comb.random = TRUE, # Include random effects
prediction = FALSE # No prediction interval by default
)
# Summary of the meta-analysis
summary(meta_analysis11)## OR 95%-CI %W(random)
## Du_2021 0.7467 [0.2576; 2.1642] 18.1
## Tong_2021 0.5350 [0.3328; 0.8603] 29.6
## Zi_2019 0.8175 [0.3383; 1.9750] 21.4
## Sun_2019 7.2917 [0.8860; 60.0070] 7.5
## Ramazanoglu_2023 1.6960 [0.7807; 3.6845] 23.4
##
## Number of studies: k = 5
## Number of observations: o = 2042 (o.e = 1102, o.c = 940)
## Number of events: e = 150
##
## OR 95%-CI z p-value
## Random effects model 0.9917 [0.5174; 1.9006] -0.03 0.9799
##
## Quantifying heterogeneity:
## tau^2 = 0.3134; tau = 0.5599; I^2 = 62.6% [0.9%; 85.9%]; H = 1.63 [1.00; 2.66]
##
## Test of heterogeneity:
## Q d.f. p-value
## 10.69 4 0.0303
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis11, layout = "RevMan5")
Meta-analysis of odds ratios for the risk of Symptomatic intracerebral hemorrhage in thrombectomy vs bridge therapy
————— Meta-analysis of odds ratios for the risk of Mortality rate within 3 months in thrombectomy vs bridge therapy ————-
mortality_odds <- read.table(text = "
Study_ID Sample_brid event_mortality_bridge no_event_mortality_bridge Sample_thromb event_mort_throm no_event_mort_throm
Furlan_2022 21 14 7 21 4 17
Du_2021 57 15 42 57 17 40
Tong_2021 405 62 343 568 93 475
Zi_2019 116 21 95 116 20 96
Sun_2019 151 6 145 151 32 119
Ramazanoglu_2023 210 37 173 210 68 142
", header = TRUE, stringsAsFactors = FALSE)
print(mortality_odds)## Study_ID Sample_brid event_mortality_bridge no_event_mortality_bridge
## 1 Furlan_2022 21 14 7
## 2 Du_2021 57 15 42
## 3 Tong_2021 405 62 343
## 4 Zi_2019 116 21 95
## 5 Sun_2019 151 6 145
## 6 Ramazanoglu_2023 210 37 173
## Sample_thromb event_mort_throm no_event_mort_throm
## 1 21 4 17
## 2 57 17 40
## 3 568 93 475
## 4 116 20 96
## 5 151 32 119
## 6 210 68 142# Load the necessary library
library(meta)
# Assuming mrs_odds is already loaded with your data
# You might need to check the names of your columns with names(mrs_odds)
# Meta-analysis of odds ratios
meta_analysis12 <- metabin(
event.e = mortality_odds$event_mort_throm,
n.e = mortality_odds$Sample_thromb,
event.c = mortality_odds$event_mortality_bridge,
n.c = mortality_odds$Sample_brid,
data = mortality_odds,
studlab = paste(mortality_odds$Study_ID),
sm = "OR", # Specify summary measure as Odds Ratio
method.tau = "DL", # DerSimonian-Laird estimator for tau^2
comb.fixed = FALSE, # Random effects model
comb.random = TRUE, # Include random effects
prediction = FALSE # No prediction interval by default
)
# Summary of the meta-analysis
summary(meta_analysis12)## OR 95%-CI %W(random)
## Furlan_2022 0.1176 [0.0285; 0.4855] 10.7
## Du_2021 1.1900 [0.5250; 2.6972] 16.2
## Tong_2021 1.0832 [0.7632; 1.5372] 20.5
## Zi_2019 0.9425 [0.4799; 1.8508] 17.6
## Sun_2019 6.4986 [2.6289; 16.0644] 15.3
## Ramazanoglu_2023 2.2391 [1.4166; 3.5390] 19.6
##
## Number of studies: k = 6
## Number of observations: o = 2083 (o.e = 1123, o.c = 960)
## Number of events: e = 389
##
## OR 95%-CI z p-value
## Random effects model 1.2838 [0.6689; 2.4638] 0.75 0.4527
##
## Quantifying heterogeneity:
## tau^2 = 0.5086; tau = 0.7132; I^2 = 83.5% [65.6%; 92.1%]; H = 2.46 [1.70; 3.56]
##
## Test of heterogeneity:
## Q d.f. p-value
## 30.38 5 < 0.0001
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - Mantel-Haenszel estimator used in calculation of Q and tau^2 (like RevMan 5)# To visualize the results, you can plot a forest plot
meta::forest(meta_analysis12, layout = "RevMan5")
Meta-analysis of odds ratios for the risk of Mortality rate within 3 months in thrombectomy vs bridge therapy