Supplementary Table 1: Quality of the evidence assessment for studies reporting on the impact of bed rest immediately after IUI/ET on ART outcomes

Quality assessment / Quality
№ of studies / Study design / Risk of bias / Inconsistency / Indirectness / Imprecision / Other considerations
Positive pregnancy test - ET
3 / randomised trials / serious a / not serious b / serious c / serious d / none / ⨁◯◯◯
VERY LOW
Clinical pregnancy - ET
4 e / randomised trials / serious f / not serious b / serious g / serious h / none / ⨁◯◯◯
VERY LOW
Live birth - ET
1 / randomised trials / serious i / serious j / not serious k / serious l / none / ⨁◯◯◯
VERY LOW
Positive pregnancy test - IUI
1 / randomised trials / serious m / serious j / not serious n / not serious o / none / ⨁⨁◯◯
LOW
Clinical pregnancy - IUI
3 / randomised trials / serious p / not serious q / not serious r / not serious s / none / ⨁⨁⨁◯
MODERATE
Live birth - IUI
1 / randomised trials / serious i / serious j / not serious k / serious l / none / ⨁◯◯◯
VERY LOW
Positive pregnancy test - ET
1 / Observation-al studies / serious t / not serious u / serious c / serious l / none / ⨁◯◯◯
VERY LOW
Clinical pregnancy - ET
1 / Observation-al studies / serious v / serious w / serious g / serious l / none / ⨁◯◯◯
VERY LOW
Explanations

a. Unclear allocation concealment for one study; high risk of other sources of bias for all studies (no documentation of bed rest periods after ART procedure when back home)

b. Two studies reported OR (95% CI) showing overlap of CI

c. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (pregnancy test); indirect comparison (some studies assessed the effect of two different duration of bed rest. Immediate mobilization was not the comparator).

d. Two studies reported 95% CI that crosses the statistical threshold; one study reported no difference (%) in the rate of positive pregnancy test between the groups (95% CI will cross the statistical threshold).

e. One non-randomized controlled trial and 3 RCTs

f. No blinding and concealment for one study; unclear randomization procedures for one study; high risk of other sources of bias for all studies (no documentation of bed rest periods after ART procedure when back home)

g. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (clinical pregnancy, surrogates=fetal heart rate at the ultrasonographic screening); indirect comparison (some studies assessed the effect of two different duration of bed rest. Immediate mobilization was not the comparator).

h. Two studies reported 95% CI that crosses the statistical threshold; two studies reported no difference (%) in the rate of positive pregnancy test between the groups (95% CI will cross the statistical threshold).

i. Unclear allocation concealment; high risk of other sources of bias (no documentation of bed rest periods between ART procedure and confirmation of pregnancy)

j. n=1. Impossible to assess inconsistency

k. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (live birth from medical record); direct comparison

l. n=1. Impossible to assess imprecision.

m. Unclear randomization sequence generation; high risk of other sources of bias (no documentation of bed rest periods between ART procedure and confirmation of pregnancy)

n. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (pregnancy test); direct comparison

o. Only one RCT reported statistically positive effect of bed rest after IUI on the rate of positive pregnancy test (95% CI will not cross the statistical threshold). However, 3 RCT reported statistically positive effect of bed rest on clinical pregnancy rate after IUI. Is it very likely that if the rate of positive pregnancy test would have been assessed, similar effect would have been observed.

p. Unclear randomization procedures for two studies; high risk of other sources of bias for all studies (no documentation of bed rest periods after ART when back home)

q. Three RCT reported % difference in clinical pregnancy rate between bed rest and mobilization and found a significant difference of 9-15% (point estimates do not vary widely across studies)

r. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (clinical pregnancy, surrogates=fetal heart rate at the ultrasonographic screening); direct comparison

s. One study reported 95% CI, with CI not crossing the statistical threshold. Two studies reported % differences between groups that were statistically significant (95% CI will not cross statistical threshold)

t. Other sources of risk of bias: significant difference in the number of embryos transferred between the pregnant and non-pregnant group.

u. No other observational study to compare consistency of the results but results are consistent with those of 3 RCTs

v. High risk of bias for exposure measurement in the historical cohort ("Most in vitro fertilization units in the UK practiced bed rest to some extent following embryo transfer during the studied period". No way to make sure wether the patient bed-rest or not and for how long); high risk of other sources of bias (No information about the characteristics of the historical cohort. Unable to know whether they were similar or not. No definition of clinical pregnancy for the historical cohort).

w. No other observational study to compare consistency of the results but results are inconsistent with those of 3 RCTs.

Supplementary Table 2: Quality of the evidence assessment for studies reporting on the impact of physical activity after ET on pregnancy outcomes.

Quality assessment / Quality
№ of studies / Study design / Risk of bias / Inconsistency / Indirectness / Imprecision / Other considerations
Positive pregnancy test - ET
1 / Observation-al studies / not serious a / serious b / not serious c / serious d / none / ⨁◯◯◯
VERY LOW
Clinical pregnanct - ET
2 / Observation-al studies / not serious a / serious e / not serious f / serious g / none / ⨁◯◯◯
VERY LOW
Live birth - ET
2 / Observation-al studies / not serious a / serious e / not serious h / serious i / none / ⨁◯◯◯
VERY LOW
Explanations

a. No risk of bias detected

b. n=1. Impossible to assess inconsistency.

c. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (pregnancy test); direct comparison

d. n=1. Impossible to assess imprecision

e. Conflicting results are reported by the two studies

f. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (clinical pregnancy, surrogates=fetal heart rate at the ultra-sonographic screening); direct comparison

g. One study reported OR (95% CI), with CI crossing the statistical threshold; one study report a significant differences in clinical pregnancy rate between the groups (95% CI will not cross the statistical threshold)

h. Direct evidence: infertile women undergoing ART; applicability of intervention; common outcome measurement (live birth from medical records); direct comparison

i. One study reported OR (95% CI), with CI crossing the statistical threshold; one study report a significant differences in live birth rate between the groups (95% CI will not cross the statistical threshold)