Table 3.

Assessment of risk of bias of the included studies.

	Aimetti et al. 2019	Sanchez-Martos et al. 2020	Tenore et al. 2020
Random sequence generation (selection bias)	Low risk Reported in the article “A balanced randomly permuted block was used to prepare the randomization table”	Low risk Reported in the article “using a randomized system based on stratified blocks”	Low risk Reported in the article “patients were randomly allocated from a computer-generated list of random numbers”
Allocation concealment (selection bias)	Low risk Reported in the article “To conceal assignment, forms with the treatment modality were put into identical and opaque envelopes”	Low risk Reported in the article “The allocation concealment was carried out through the use of sealed opaque envelopes”	Low risk Reported in the article “Allocation concealment was achieved through the provision, by professionals not involved in patient enrolment, of a numbered sequence of opaque and sealed envelopes”
Blinding of outcome assessment (detection bias)	Low risk Reported in the article “Two examiners, who were blinded to the group assignment, performed all measurements of clinical assessment”	High risk Reported in the article “The interventions assigned to each group were performed by a calibrated and trained examiner not blind to the group assignment”	High risk No information in the article
Incomplete outcome data (attrition bias)	Low risk All data presented	Low risk All data presented	Low risk Number and reasons for withdrawals were reported. It does not seem that the lost data had affected the results
Selective reporting (reporting bias)	Low risk All outcomes appear to be detected	Low risk All outcomes appear to be detected	Low risk All outcomes appear to be detected
Other bias	None detected	None detected	None detected