Intention-to-treat analysis may be more conservative than per protocol analysis in antibiotic non-inferiority trials: a systematic review

Table 2 Risk of bias assessment

	All comparisons within studies (N = 164)
Randomization
High risk	3 (1.8%)
Low risk	110 (67.1%)
Unclear	51 (31.1%)
Allocation concealment
High risk	3 (1.8%)
Low risk	74 (45.1%)
Unclear	87 (53.1%)
Performance bias
High risk	75 (45.7%)
Low risk	84 (51.2%)
Unclear	5 (3.1%)
Detection bias
High risk	58 (35.4%)
Low risk	100 (61.0%)
Unclear	6 (3.7%)
Attrition bias
High risk	51 (31.1%)
Low risk	108 (65.9%)
Unclear	5 (3.1%)
Reporting bias
High risk	28 (17.1%)
Low risk	136 (82.9%)
Unclear	0 (0%)

ISSN: 1471-2288