Etanercept

Preterm (< 37 weeks)

R analysis
id Study   Lib. in paper Exposition period    Study type 
 Control type 
 
Tags OR 95%CI x1/n1 x0/n0 no cases no exposed ROB Ref.
S4894
R12921		 Broms (Controls exposed to other treatments), 2020 Preterm birth (<37 weeks of gestation) 3 months (or more) before pregnancy or during pregnancy population based cohort retrospective exposed to other treatment, sick excluded Adjustment: No 1.30 [0.97;1.74] C
excluded (control group) 		54/510   786/9,393 840 510
ref
S4895
R12926		 Broms (Controls unexposed, disease free), 2020 Preterm birth (<37 weeks of gestation) 3 months (or more) before pregnancy or during pregnancy excluded population based cohort retrospective unexposed, disease free Adjustment: No 2.45 [1.84;3.24] C
excluded (exposition period) 		54/510   74,981/1,623,489 75,035 510
ref
S4961
R13015		 De Lorenzo (Controls unexposed, disease free), 2020 Preterm birth at least 1st trimester retrospective cohort unexposed, disease free excluded Adjustment: No 17.00 [1.99;145.01] C
excluded (control group) 		3/6   2/36 5 6
ref
S4960
R13009		 De Lorenzo (Controls unexposed, sick), 2020 Preterm birth at least 1st trimester retrospective cohort unexposed, sick Adjustment: No 9.67 [1.32;70.97] C 3/6   3/32 6 6
ref
S4994
R13071		 Drechsel, 2020 Preterm delivery (<37 0) 1st trimester prospective cohort unexposed, sick Adjustment: No 1.05 [0.11;9.72] C 1/9   7/66 8 9
ref
S3011
R8691		 Carman (Control unexposed, disease free), 2017 Birth before 37 wk during pregnancy (anytime or not specified) retrospective cohort (claims database) unexposed, disease free excluded Adjustment: No 1.70 [1.15;2.51] C
excluded (control group) 		40/217   127/1,081 167 217
ref
S3010
R8690		 Carman (Control unexposed, sick), 2017 Birth before 37 wk during pregnancy (anytime or not specified) retrospective cohort (claims database) unexposed, sick Adjustment: No 1.40 [0.97;2.03] C 40/217   269/1,940 309 217
ref
S3004
R4284		 Hoxha, 2017 Premature birth (<37 weeks of gestation) 1st trimester prospective cohort unexposed, sick Adjustment: No 0.64 [0.01;34.59] C 0/17   0/11 0 17
ref
S3137
R4914		 Chambers, 2015 Preterm delivery (NOS) at least 1st trimester prospective cohort unexposed, sick Adjustment: Yes 1.60 [0.86;2.98] -/-   -/- - -
ref
Total 5 studies 1.50 [1.10;2.04] 323 249
x1: number of endpoints among exposed, n1: number of exposed; x0: number of endpoints among non exposed, n0: number of non exposed; C: calculated odds ratio from numbers of events and effectives

Forest plot

StudyTE95% CIn casesn exposedweightROBABCDEF De Lorenzo (Controls unexposed, sick), 2020De Lorenzo, 2020 1 9.67[1.32; 70.97]662%ROB confusion: NAROB selection: NAROB classification: NAROB missing: NAROB mesure: NAROB reporting: NA Drechsel, 2020Drechsel, 2020 1.05[0.11; 9.72]892%ROB confusion: NAROB selection: NAROB classification: NAROB missing: NAROB mesure: NAROB reporting: NA Carman (Control unexposed, sick), 2017Carman, 2017 2 1.40[0.97; 2.03]30921771%ROB confusion: NAROB selection: NAROB classification: NAROB missing: NAROB mesure: NAROB reporting: NA Hoxha, 2017Hoxha, 2017 0.64[0.01; 34.59]0171%ROB confusion: NAROB selection: NAROB classification: NAROB missing: NAROB mesure: NAROB reporting: NA Chambers, 2015Chambers, 2015 1.60[0.86; 2.98]--25%ROB confusion: NAROB selection: NAROB classification: NAROB missing: NAROB mesure: NAROB reporting: NA Total (5 studies) I2 = 0% 1.50[1.10; 2.04]3232490.220.01.0ROB: A: confusion, B: selection, C: classification, D: missing, E: measurement, F: reportinglow,moderate,serious,critical,unclear,

1: Controls unexposed, sick; 2: Control unexposed, sick;

Sensitivity analysis

SubsetTE95% CIn casesn exposedkI2 Type of studies cohort studiescohort studies 1.50[1.10; 2.04]3232490%NADe Lorenzo (Controls unexposed, sick), 2020 Drechsel, 2020 Carman (Control unexposed, sick), 2017 Hoxha, 2017 Chambers, 2015 5 case control studiescase control studies 0 Type of controls unexposed, sickunexposed, sick 1.50[1.10; 2.04]3232490%NADe Lorenzo (Controls unexposed, sick), 2020 Drechsel, 2020 Carman (Control unexposed, sick), 2017 Hoxha, 2017 Chambers, 2015 5 Tags Adjustment   - No  - No 1.71[0.77; 3.80]32324920%NADe Lorenzo (Controls unexposed, sick), 2020 Drechsel, 2020 Carman (Control unexposed, sick), 2017 Hoxha, 2017 4   - Yes  - Yes 1.60[0.86; 2.98]-- -NAChambers, 2015 1 All studiesAll studies 1.50[1.10; 2.04]3232490%NADe Lorenzo (Controls unexposed, sick), 2020 Drechsel, 2020 Carman (Control unexposed, sick), 2017 Hoxha, 2017 Chambers, 2015 50.220.01.0

Publication bias and p-hacking diagnosis

funnel plot
0.0-4.85.22.4460.000De Lorenzo (Controls unexposed, sick), 2020Drechsel, 2020Carman (Control unexposed, sick), 2017Hoxha, 2017Chambers, 2015

Asymetry test p-value = 0.5872 (by Egger's regression)

slope=0.2879 (0.2579); intercept=0.4451 (0.7342); t=0.6062; p=0.5872

p values plot
0.01.00.01.0

Sub-groups analysis using all included studies

excluded 3011, 4894, 4961

Sub-groupsTE95% CIn casesn exposedkI2ROB type of controls unexposed controls (disease free or unspecified)unexposed controls (disease free or unspecified) 4.18[0.46; 37.86]17222377%NADe Lorenzo (Controls unexposed, disease free), 2020 Carman (Control unexposed, disease free), 2017 2 unexposed, sick controlsunexposed, sick controls 1.50[1.10; 2.04]3232490%NADe Lorenzo (Controls unexposed, sick), 2020 Drechsel, 2020 Carman (Control unexposed, sick), 2017 Hoxha, 2017 Chambers, 2015 50.510.01.0