Meta analysis of Neurocisticercosis surgical management
Niels Pacheco
2024-05-19
—————– 1. Meta-analysis of proportions of patients cured by endoscopic treatment —————————
metaprop1 <- read.table(text = "
Study_ID number_ENDOS ENDOS_cured
'Bergsneider, 1999' 5 3
'Bergsneider, 2000' 10 7
'Torres-Corzo, 2006' 7 7
'Husain, 2007' 10 10
'Husain, 2007' 21 21
'Ranjan, 2014' 17 14
'Sharma, 2019' 5 0
'Aggarwal, 2020' 26 26
", header = TRUE, stringsAsFactors = FALSE)
print(metaprop1)## Study_ID number_ENDOS ENDOS_cured
## 1 Bergsneider, 1999 5 3
## 2 Bergsneider, 2000 10 7
## 3 Torres-Corzo, 2006 7 7
## 4 Husain, 2007 10 10
## 5 Husain, 2007 21 21
## 6 Ranjan, 2014 17 14
## 7 Sharma, 2019 5 0
## 8 Aggarwal, 2020 26 26# 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 = ENDOS_cured, n = number_ENDOS, data = metaprop1,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = metaprop1$Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis1)## proportion 95%-CI %W(random)
## Bergsneider, 1999 0.6000 [0.1466; 0.9473] 15.1
## Bergsneider, 2000 0.7000 [0.3475; 0.9333] 17.9
## Torres-Corzo, 2006 1.0000 [0.5904; 1.0000] 9.6
## Husain, 2007 1.0000 [0.6915; 1.0000] 9.7
## Husain, 2007 1.0000 [0.8389; 1.0000] 9.9
## Ranjan, 2014 0.8235 [0.5657; 0.9620] 18.5
## Sharma, 2019 0.0000 [0.0000; 0.5218] 9.5
## Aggarwal, 2020 1.0000 [0.8677; 1.0000] 9.9
##
## Number of studies: k = 8
## Number of observations: o = 101
## Number of events: e = 88
##
## proportion 95%-CI
## Random effects model 0.8305 [0.6077; 0.9394]
##
## Quantifying heterogeneity:
## tau^2 = 1.4578 [0.2118; 13.6229]; tau = 1.2074 [0.4602; 3.6909]
## I^2 = 57.7% [7.2%; 80.7%]; H = 1.54 [1.04; 2.28]
##
## Test of heterogeneity:
## Q d.f. p-value
## 16.55 7 0.0206
##
## 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_analysis1, layout = "JAMA")
—————– 2. Meta-analysis of proportions of patients improved by endoscopic treatment —————————
metaprop2<- read.table(text = "
Study_ID number_ENDOS ENDOS_improved
'Bergsneider, 1999' 5 5
'Bergsneider, 2000' 10 10
'Zhang, 2000' 8 8
'Anandh, 2001' 9 9
'Torres-Corzo, 2006' 7 7
'Husain, 2007' 10 10
'Husain, 2007' 21 21
'Suri, 2008' 6 6
'Goel, 2008' 22 22
'Rangel-Castilla, 2009' 4 4
'Ranjan, 2014' 17 17
'Zhenye, 2017' 21 18
'Sharma, 2019' 5 5
'Aggarwal, 2020' 26 26
", header = TRUE, stringsAsFactors = FALSE)
print(metaprop2)## Study_ID number_ENDOS ENDOS_improved
## 1 Bergsneider, 1999 5 5
## 2 Bergsneider, 2000 10 10
## 3 Zhang, 2000 8 8
## 4 Anandh, 2001 9 9
## 5 Torres-Corzo, 2006 7 7
## 6 Husain, 2007 10 10
## 7 Husain, 2007 21 21
## 8 Suri, 2008 6 6
## 9 Goel, 2008 22 22
## 10 Rangel-Castilla, 2009 4 4
## 11 Ranjan, 2014 17 17
## 12 Zhenye, 2017 21 18
## 13 Sharma, 2019 5 5
## 14 Aggarwal, 2020 26 26# 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 = ENDOS_improved, n = number_ENDOS, data = metaprop2,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = metaprop2$Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis2)## proportion 95%-CI %W(random)
## Bergsneider, 1999 1.0000 [0.4782; 1.0000] 5.3
## Bergsneider, 2000 1.0000 [0.6915; 1.0000] 5.5
## Zhang, 2000 1.0000 [0.6306; 1.0000] 5.4
## Anandh, 2001 1.0000 [0.6637; 1.0000] 5.4
## Torres-Corzo, 2006 1.0000 [0.5904; 1.0000] 5.4
## Husain, 2007 1.0000 [0.6915; 1.0000] 5.5
## Husain, 2007 1.0000 [0.8389; 1.0000] 5.6
## Suri, 2008 1.0000 [0.5407; 1.0000] 5.3
## Goel, 2008 1.0000 [0.8456; 1.0000] 5.6
## Rangel-Castilla, 2009 1.0000 [0.3976; 1.0000] 5.2
## Ranjan, 2014 1.0000 [0.8049; 1.0000] 5.6
## Zhenye, 2017 0.8571 [0.6366; 0.9695] 29.5
## Sharma, 2019 1.0000 [0.4782; 1.0000] 5.3
## Aggarwal, 2020 1.0000 [0.8677; 1.0000] 5.6
##
## Number of studies: k = 14
## Number of observations: o = 171
## Number of events: e = 168
##
## proportion 95%-CI
## Random effects model 0.9350 [0.8811; 0.9654]
##
## Quantifying heterogeneity:
## tau^2 = 0; tau = 0; I^2 = 0.0% [0.0%; 55.0%]; H = 1.00 [1.00; 1.49]
##
## Test of heterogeneity:
## Q d.f. p-value
## 4.76 13 0.9800
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - 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 = "JAMA")
—————– 3. Meta-analysis of proportions of patients who needed VPS after
endoscopic treatment —————————
data_vps <- read.table(text = "
Study_ID number_ENDOS ENDOS_VPS
'Apuzzo 1984' 1 0
'Bergsneider, 1999' 5 2
'Bergsneider, 2000' 10 3
'Zhang, 2000' 8 0
'Anandh, 2001' 9 0
'Torres-Corzo, 2006' 7 0
'Husain, 2007' 10 0
'Husain, 2007' 21 0
'Suri, 2008' 6 0
'Goel, 2008' 22 0
'Jimenez-Vazquez' 9 0
'Kumar, 2008' 1 0
'Rangel-Castilla, 2009' 4 0
'Torres-Corzo, 2010' 86 8
'Ranjan, 2014' 17 2
'Zhenye, 2017' 21 7
'Sharma, 2019' 5 0
'Kaif, 2019' 30 1
'Singh, 2019' 12 3
'Konar, 2020' 61 1
'Aggarwal, 2020' 26 3
", header = TRUE, stringsAsFactors = FALSE)
print(data_vps)## Study_ID number_ENDOS ENDOS_VPS
## 1 Apuzzo 1984 1 0
## 2 Bergsneider, 1999 5 2
## 3 Bergsneider, 2000 10 3
## 4 Zhang, 2000 8 0
## 5 Anandh, 2001 9 0
## 6 Torres-Corzo, 2006 7 0
## 7 Husain, 2007 10 0
## 8 Husain, 2007 21 0
## 9 Suri, 2008 6 0
## 10 Goel, 2008 22 0
## 11 Jimenez-Vazquez 9 0
## 12 Kumar, 2008 1 0
## 13 Rangel-Castilla, 2009 4 0
## 14 Torres-Corzo, 2010 86 8
## 15 Ranjan, 2014 17 2
## 16 Zhenye, 2017 21 7
## 17 Sharma, 2019 5 0
## 18 Kaif, 2019 30 1
## 19 Singh, 2019 12 3
## 20 Konar, 2020 61 1
## 21 Aggarwal, 2020 26 3# 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 = ENDOS_VPS, n = number_ENDOS, data = data_vps,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = data_vps$Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis3)## proportion 95%-CI %W(random)
## Apuzzo 1984 0.0000 [0.0000; 0.9750] 2.0
## Bergsneider, 1999 0.4000 [0.0527; 0.8534] 5.3
## Bergsneider, 2000 0.3000 [0.0667; 0.6525] 7.7
## Zhang, 2000 0.0000 [0.0000; 0.3694] 2.5
## Anandh, 2001 0.0000 [0.0000; 0.3363] 2.5
## Torres-Corzo, 2006 0.0000 [0.0000; 0.4096] 2.5
## Husain, 2007 0.0000 [0.0000; 0.3085] 2.5
## Husain, 2007 0.0000 [0.0000; 0.1611] 2.6
## Suri, 2008 0.0000 [0.0000; 0.4593] 2.4
## Goel, 2008 0.0000 [0.0000; 0.1544] 2.6
## Jimenez-Vazquez 0.0000 [0.0000; 0.3363] 2.5
## Kumar, 2008 0.0000 [0.0000; 0.9750] 2.0
## Rangel-Castilla, 2009 0.0000 [0.0000; 0.6024] 2.4
## Torres-Corzo, 2010 0.0930 [0.0410; 0.1751] 13.6
## Ranjan, 2014 0.1176 [0.0146; 0.3644] 6.9
## Zhenye, 2017 0.3333 [0.1459; 0.5697] 11.6
## Sharma, 2019 0.0000 [0.0000; 0.5218] 2.4
## Kaif, 2019 0.0333 [0.0008; 0.1722] 4.5
## Singh, 2019 0.2500 [0.0549; 0.5719] 8.0
## Konar, 2020 0.0164 [0.0004; 0.0880] 4.6
## Aggarwal, 2020 0.1154 [0.0245; 0.3015] 8.8
##
## Number of studies: k = 21
## Number of observations: o = 371
## Number of events: e = 30
##
## proportion 95%-CI
## Random effects model 0.1196 [0.0775; 0.1802]
##
## Quantifying heterogeneity:
## tau^2 = 0.3035 [0.0000; 1.3545]; tau = 0.5509 [0.0000; 1.1638]
## I^2 = 28.2% [0.0%; 57.9%]; H = 1.18 [1.00; 1.54]
##
## Test of heterogeneity:
## Q d.f. p-value
## 27.86 20 0.1128
##
## 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 = "JAMA")
3. Meta-analysis of proportions of patients who needed VPS after endoscopic treatment
—————– 4. Meta-analysis of proportions of mortality in patients who received endoscopic treatment —————————
data_mortality <- read.table(text = "
Study_ID number_ENDOS ENDOS_mortality
'Apuzzo 1984' 1 0
'Zhang, 2000' 8 0
'Anandh, 2001' 9 0
'Torres-Corzo, 2006' 7 0
'Husain, 2007' 10 0
'Husain, 2007' 21 0
'Suri, 2008' 6 0
'Goel, 2008' 22 0
'Kumar, 2008' 1 0
'Rangel-Castilla, 2009' 4 0
'Torres-Corzo, 2010' 86 3
'Ranjan, 2014' 17 0
'Zhenye, 2017' 21 1
'Nash, 2018' 3 0
'Sharma, 2019' 5 0
'Singh, 2019' 12 0
'Konar, 2020' 61 1
'Aggarwal, 2020' 26 0
", header = TRUE, stringsAsFactors = FALSE)
print(data_mortality)## Study_ID number_ENDOS ENDOS_mortality
## 1 Apuzzo 1984 1 0
## 2 Zhang, 2000 8 0
## 3 Anandh, 2001 9 0
## 4 Torres-Corzo, 2006 7 0
## 5 Husain, 2007 10 0
## 6 Husain, 2007 21 0
## 7 Suri, 2008 6 0
## 8 Goel, 2008 22 0
## 9 Kumar, 2008 1 0
## 10 Rangel-Castilla, 2009 4 0
## 11 Torres-Corzo, 2010 86 3
## 12 Ranjan, 2014 17 0
## 13 Zhenye, 2017 21 1
## 14 Nash, 2018 3 0
## 15 Sharma, 2019 5 0
## 16 Singh, 2019 12 0
## 17 Konar, 2020 61 1
## 18 Aggarwal, 2020 26 0# 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 = ENDOS_mortality, n = number_ENDOS, data = data_mortality,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = data_mortality$Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis4)## proportion 95%-CI %W(random)
## Apuzzo 1984 0.0000 [0.0000; 0.9750] 3.2
## Zhang, 2000 0.0000 [0.0000; 0.3694] 4.0
## Anandh, 2001 0.0000 [0.0000; 0.3363] 4.1
## Torres-Corzo, 2006 0.0000 [0.0000; 0.4096] 4.0
## Husain, 2007 0.0000 [0.0000; 0.3085] 4.1
## Husain, 2007 0.0000 [0.0000; 0.1611] 4.2
## Suri, 2008 0.0000 [0.0000; 0.4593] 4.0
## Goel, 2008 0.0000 [0.0000; 0.1544] 4.2
## Kumar, 2008 0.0000 [0.0000; 0.9750] 3.2
## Rangel-Castilla, 2009 0.0000 [0.0000; 0.6024] 3.8
## Torres-Corzo, 2010 0.0349 [0.0073; 0.0986] 24.7
## Ranjan, 2014 0.0000 [0.0000; 0.1951] 4.1
## Zhenye, 2017 0.0476 [0.0012; 0.2382] 8.1
## Nash, 2018 0.0000 [0.0000; 0.7076] 3.7
## Sharma, 2019 0.0000 [0.0000; 0.5218] 3.9
## Singh, 2019 0.0000 [0.0000; 0.2646] 4.1
## Konar, 2020 0.0164 [0.0004; 0.0880] 8.4
## Aggarwal, 2020 0.0000 [0.0000; 0.1323] 4.2
##
## Number of studies: k = 18
## Number of observations: o = 320
## Number of events: e = 5
##
## proportion 95%-CI
## Random effects model 0.0449 [0.0258; 0.0769]
##
## Quantifying heterogeneity:
## tau^2 = 0; tau = 0; I^2 = 0.0% [0.0%; 50.0%]; H = 1.00 [1.00; 1.41]
##
## Test of heterogeneity:
## Q d.f. p-value
## 6.47 17 0.9894
##
## Details on meta-analytical method:
## - Inverse variance method
## - DerSimonian-Laird estimator for tau^2
## - 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 = "JAMA")
—————– 5. Meta-analysis of proportions of complications in patients who received endoscopic treatment —————————
data_complic <- read.table(text = "
Study_ID number_ENDOS ENDOS_Complic
'Apuzzo 1984' 1 0
'Bergsneider, 1999' 5 1
'Bergsneider, 2000' 10 2
'Zhang, 2000' 8 0
'Anandh, 2001' 9 3
'Torres-Corzo, 2006' 7 0
'Husain, 2007' 10 0
'Husain, 2007' 21 0
'Suri, 2008' 6 0
'Goel, 2008' 22 2
'Kumar, 2008' 1 0
'Rangel-Castilla, 2009' 4 0
'Torres-Corzo, 2010' 86 8
'Ranjan, 2014' 17 2
'Zhenye, 2017' 21 8
'Nash, 2018' 3 2
'Sharma, 2019' 5 0
'Kaif, 2019' 30 3
'Singh, 2019' 12 5
'Konar, 2020' 61 3
'Aggarwal, 2020' 26 5
", header = TRUE, stringsAsFactors = FALSE)
print(data_complic)## Study_ID number_ENDOS ENDOS_Complic
## 1 Apuzzo 1984 1 0
## 2 Bergsneider, 1999 5 1
## 3 Bergsneider, 2000 10 2
## 4 Zhang, 2000 8 0
## 5 Anandh, 2001 9 3
## 6 Torres-Corzo, 2006 7 0
## 7 Husain, 2007 10 0
## 8 Husain, 2007 21 0
## 9 Suri, 2008 6 0
## 10 Goel, 2008 22 2
## 11 Kumar, 2008 1 0
## 12 Rangel-Castilla, 2009 4 0
## 13 Torres-Corzo, 2010 86 8
## 14 Ranjan, 2014 17 2
## 15 Zhenye, 2017 21 8
## 16 Nash, 2018 3 2
## 17 Sharma, 2019 5 0
## 18 Kaif, 2019 30 3
## 19 Singh, 2019 12 5
## 20 Konar, 2020 61 3
## 21 Aggarwal, 2020 26 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_analysis5 <- metaprop(event = ENDOS_Complic, n = number_ENDOS, data = data_complic,
sm = "PLO", method.tau = "DL",
prediction = FALSE, comb.fixed = FALSE,
comb.random = TRUE, studlab = data_complic$Study_ID
)
# Summary of the meta-analysis
summary(meta_analysis5)## proportion 95%-CI %W(random)
## Apuzzo 1984 0.0000 [0.0000; 0.9750] 1.8
## Bergsneider, 1999 0.2000 [0.0051; 0.7164] 3.4
## Bergsneider, 2000 0.2000 [0.0252; 0.5561] 5.5
## Zhang, 2000 0.0000 [0.0000; 0.3694] 2.2
## Anandh, 2001 0.3333 [0.0749; 0.7007] 6.2
## Torres-Corzo, 2006 0.0000 [0.0000; 0.4096] 2.2
## Husain, 2007 0.0000 [0.0000; 0.3085] 2.2
## Husain, 2007 0.0000 [0.0000; 0.1611] 2.2
## Suri, 2008 0.0000 [0.0000; 0.4593] 2.2
## Goel, 2008 0.0909 [0.0112; 0.2916] 5.9
## Kumar, 2008 0.0000 [0.0000; 0.9750] 1.8
## Rangel-Castilla, 2009 0.0000 [0.0000; 0.6024] 2.1
## Torres-Corzo, 2010 0.0930 [0.0410; 0.1751] 10.7
## Ranjan, 2014 0.1176 [0.0146; 0.3644] 5.8
## Zhenye, 2017 0.3810 [0.1811; 0.6156] 9.5
## Nash, 2018 0.6667 [0.0943; 0.9916] 2.9
## Sharma, 2019 0.0000 [0.0000; 0.5218] 2.1
## Kaif, 2019 0.1000 [0.0211; 0.2653] 7.3
## Singh, 2019 0.4167 [0.1517; 0.7233] 7.6
## Konar, 2020 0.0492 [0.0103; 0.1371] 7.5
## Aggarwal, 2020 0.1923 [0.0655; 0.3935] 8.8
##
## Number of studies: k = 21
## Number of observations: o = 365
## Number of events: e = 44
##
## proportion 95%-CI
## Random effects model 0.1558 [0.1047; 0.2257]
##
## Quantifying heterogeneity:
## tau^2 = 0.3686 [0.0000; 1.3087]; tau = 0.6071 [0.0000; 1.1440]
## I^2 = 38.4% [0.0%; 63.5%]; H = 1.27 [1.00; 1.66]
##
## Test of heterogeneity:
## Q d.f. p-value
## 32.47 20 0.0385
##
## 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_analysis5, layout = "JAMA")
5. Meta-analysis of proportions of complications in patients who received endoscopic treatment
—————– 6. Meta-analysis of odds ratios of the risk of use of VPS in endoscopic vs open surgery —————————
or_endo_surg <- read.table(text = "
Study_ID number_ENDOS ENDOS_VPS number_surgical Surgical_VPS
'Apuzzo 1984' 1 0 29 7
'Kumar, 2008' 1 0 9 4
'Rangel-Castilla, 2009' 4 0 6 2
", header = TRUE, stringsAsFactors = FALSE)
print(or_endo_surg)## Study_ID number_ENDOS ENDOS_VPS number_surgical Surgical_VPS
## 1 Apuzzo 1984 1 0 29 7
## 2 Kumar, 2008 1 0 9 4
## 3 Rangel-Castilla, 2009 4 0 6 2# 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_analysis6 <- metabin(
event.e = or_endo_surg$ENDOS_VPS,
n.e = or_endo_surg$number_ENDOS,
event.c = or_endo_surg$Surgical_VPS,
n.c = or_endo_surg$number_surgical,
data = or_endo_surg,
studlab = paste(or_endo_surg$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_analysis6)## OR 95%-CI %W(random)
## Apuzzo 1984 1.0000 [0.0367; 27.2642] 34.2
## Kumar, 2008 0.4074 [0.0131; 12.6365] 31.7
## Rangel-Castilla, 2009 0.2000 [0.0073; 5.4528] 34.2
##
## Number of studies: k = 3
## Number of observations: o = 50 (o.e = 6, o.c = 44)
## Number of events: e = 13
##
## OR 95%-CI z p-value
## Random effects model 0.4342 [0.0629; 2.9985] -0.85 0.3975
##
## Quantifying heterogeneity:
## 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
## 0.46 2 0.7948
##
## 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_analysis6, layout = "JAMA")
6. Meta-analysis of odds ratios of the risk of use of VPS in endoscopic vs open surgery
—————– 7. Meta-analysis of odds ratios of the risk of use of VPS in endoscopic vs medical treatment —————————
or_endo_med <- read.table(text = "
Study_ID number_ENDOS ENDOS_VPS number_med Med_VPS
'Kumar, 2008' 1 0 1 1
'Rangel-Castilla, 2009' 4 0 1 1
", header = TRUE, stringsAsFactors = FALSE)
print(or_endo_med)## Study_ID number_ENDOS ENDOS_VPS number_med Med_VPS
## 1 Kumar, 2008 1 0 1 1
## 2 Rangel-Castilla, 2009 4 0 1 1# 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_analysis7 <- metabin(
event.e = or_endo_med$ENDOS_VPS,
n.e = or_endo_med$number_ENDOS,
event.c = or_endo_med$Med_VPS,
n.c = or_endo_med$number_med,
data = or_endo_med,
studlab = paste(or_endo_med$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_analysis7)## OR 95%-CI %W(random)
## Kumar, 2008 0.1111 [0.0012; 10.2689] 47.8
## Rangel-Castilla, 2009 0.0370 [0.0005; 2.8230] 52.2
##
## Number of studies: k = 2
## Number of observations: o = 7 (o.e = 5, o.c = 2)
## Number of events: e = 2
##
## OR 95%-CI z p-value
## Random effects model 0.0626 [0.0027; 1.4331] -1.73 0.0828
##
## Quantifying heterogeneity:
## tau^2 = 0; tau = 0; I^2 = 0.0%; H = 1.00
##
## Test of heterogeneity:
## Q d.f. p-value
## 0.12 1 0.7308
##
## 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_analysis7, layout = "JAMA")