forest plot

[abdelkrim yannis]

Hello i m yannis, medical student

I m working on my thesis about the psychiatric disease in prison.
I made a meta analysis of proportion from prevalence and i want to make a forest plot with it, i don't practice metan and i did it with advices from good guy.

this is my command line:

clear

input study cases total

1 4 415

2 2 80

3 29 800

4 58 320

5 5 102

6 3 193

7 1 290

8 13 1008

9 9 1350

10 5 495

11 9 184

12 38 707

13 6 500

14 5 98

end

gen p = .

gen se = .

// get proportions and std errors

forv i =1(1)14 {

cii total[`i'] cases[`i']

qui replace p = r(mean) in `i'

qui replace se = r(se) in `i'

}

// get the inverse variance-weighted proportion

// use the official Stata -vwls- command

gen cons =1

vwls p cons, sd(se)

// use the user written -metan- command

// for fixed-effects meta-analysis

metan p se, nograph fixed

// for random-effects meta-analysis

metan p se, nograph random





hth,

bw


Furthermore, i think that metaprop is easier to use isn t it?

Thanx

[Madu Abuchi]

It may be helpful to provide information about the version of Stata you use. Here I present with Stata/IC 14.2.

Provide -prop- in the immediate command for confidence interval as:

cii prop total[`i'] cases[`i']

In general, to plot a forestplot, remove the -nograph- in your command line

The command line below should work

clear

input study cases total

1 4 415

2 2 80

3 29 800

4 58 320

5 5 102

6 3 193

7 1 290

8 13 1008

9 9 1350

10 5 495

11 9 184

12 38 707

13 6 500

14 5 98

end

gen p = .

gen se = .

// get proportions and std errors

forv i =1(1)14 {

cii prop total[`i'] cases[`i']

qui replace p = r(mean) in `i'

qui replace se = r(se) in `i'

}

// get the inverse variance-weighted proportion

// use the official Stata -vwls- command

gen cons =1

vwls p cons, sd(se)

// use the user written -metan- command

// for fixed-effects meta-analysis

metan p se, fixed

// for random-effects meta-analysis

metan p se, random

Regards,

[abdelkrim yannis]

Hello Thx i run with stata 14.0

it don t work, maybe i didn't explain myself enough. I want to get the forest plot graph
when i put your command, this is stata says:

. input study cases total

study cases total
1.
. 1 4 415
2.
. 2 2 80
3.
. 3 29 800
4.
. 4 58 320
5.
. 5 5 102
6.
. 6 3 193
7.
. 7 1 290
8.
. 8 13 1008
9.
. 9 9 1350
10.
. 10 5 495
11.
. 11 9 184
12.
. 12 38 707
13.
. 13 6 500
14.
. 14 5 98
15.
. end

.
. gen p = .
(14 missing values generated)

.
. gen se = .
(14 missing values generated)

.
. // get proportions and std errors
/ is not a valid command name
r(199);

.
. forv i =1(1)14 {
2.
. cii prop total[`i'] cases[`i']
3.
. qui replace p = r(mean) in `i'
4.
. qui replace se = r(se) in `i'
5.
. }
prop not found
r(111);

.
. // get the inverse variance-weighted proportion
/ is not a valid command name
r(199);

.
. // use the official Stata -vwls- command
/ is not a valid command name
r(199);

.
. gen cons =1

.
. vwls p cons, sd(se)
no observations
r(2000);

.
. // use the user written -metan- command
/ is not a valid command name
r(199);

.
. // for fixed-effects meta-analysis
/ is not a valid command name
r(199);

.
. metan p se, nograph fixed
Insufficient data to perform this meta-analysis

.
. // for random-effects meta-analysis
/ is not a valid command name
r(199);

.
. metan p se, nograph random
Insufficient data to perform this meta-analysis

.
.
.
.
.
. hth,
command hth is unrecognized
r(199);

.
. bw
command bw is unrecognized
r(199);

.
.

[abdelkrim yannis]

prop not found..

ii m not familiar with stata

Thx a lot

[abdelkrim yannis]

ok it works thank you very much just remove nograph

[abdelkrim yannis]

Sorry juste one last question
how to put the study name in the forest plot?

Thx

[Marcos Almeida]

Among the options, the study name could be a label of the variable. Check how to use - label define - command.