Summary of findings 1. Ultrasound guidance versus landmark method for peripheral venous cannulation in adults classed as difficult.

Patient or population: adults undergoing peripheral venous cannulation classed as difficulta Settings: emergency department, ICU, operating room Intervention: USG Comparison: LM
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with LM	Risk with USG
First‐pass success of cannulation Follow‐up: immediately after the procedure	358 per 1000	537 per 1000 (421 to 698)	RR 1.50  (1.15 to 1.95)	815 (10 RCTs)	⊕⊕⊝⊝ Lowb
Overall success of cannulation Follow‐up: immediately after the procedure	575 per 1000	806 per 1000 (633 to 1000)	RR 1.40  (1.10 to 1.77)	670 (10 RCTs)	⊕⊝⊝⊝ Very lowc
Pain NRS: from 0 to 10, where 10 is maximum pain Follow‐up: immediately after the procedure	The mean pain score was 3.97	MD 0.20 lower (1.13 lower to 0.72 higher)	‐	323 (4 RCTs)	⊕⊝⊝⊝ Very lowd
Procedure time for first‐pass cannulation (seconds) Follow‐up: immediately after the procedure	The mean procedure time for first‐pass cannulation was 130.5 seconds	MD 119.9 seconds longer (88.6 longer to 151.1 longer)	‐	219 (2 RCTs)	⊕⊕⊝⊝ Lowe
Number of cannulation attempts Follow‐up: immediately after the procedure	The mean number of cannulation attempts was 2.15	MD 0.33 lower (0.64 lower to 0.02 lower)	‐	568 (9 RCTs)	⊕⊝⊝⊝ Very lowf
Patient satisfaction NRS from 0 to 10 or 4‐step Likert scale The higher the score, the higher the level of satisfaction Follow‐up: immediately after the procedure	The mean patient satisfaction score was 5.61	SMD 0.49 higher (0.07 higher to 0.92 higher)	‐	333 (5 RCTs)	⊕⊝⊝⊝ Very lowg
Overall complications Follow‐up: immediately after the procedure	121 per 1000	78 per 1000 (45 to 133)	RR 0.64 (0.37 to 1.10)	431 (5 RCTs)	⊕⊕⊝⊝ Lowh
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; ICU: intensive care unit; LM: landmark method; MD: mean difference; NRS: numeric rating scale; RCTs: randomised controlled trials; RR: risk ratio; SMD: standardised mean difference; USG: ultrasound guidance
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aParticipants were classified according to the original studies' definitions.
^bWe downgraded by a total of two levels to low certainty due to risk of bias concerns (quasi‐randomised trials, lack of blinding of the outcome assessors) and substantial inconsistency. There was minimal risk of publication bias (one small study was asymmetrical but would have had little impact).
^cWe downgraded by a total of three levels to very low certainty due to risk of bias concerns (quasi‐randomised trials, lack of blinding of the outcome assessors) and serious inconsistency due to the lack of a standardised definition of failure. There was minimal risk of publication bias (two quasi‐randomised trials were asymmetrical).
^dWe downgraded by a total of three levels to very low certainty due to risk of bias concerns (a quasi‐randomised trial, lack of blinding of the outcome assessors, and incomplete outcome data), substantial inconsistency, and imprecision.
^eWe downgraded by a total of two levels to low certainty due to risk of bias concerns (a quasi‐randomised trial, lack of blinding of the outcome assessors), and imprecision.
^fWe downgraded by a total of three levels to very low certainty due to risk of bias concerns (quasi‐randomised trials, lack of blinding of the outcome assessors, and incomplete outcome data) and inconsistency due to heterogeneity and the lack of a standardised definition of failure.
^gWe downgraded by a total of three levels to very low certainty due to risk of bias concerns (a quasi‐randomised trial, lack of blinding of the outcome assessors, and incomplete outcome data), substantial inconsistency, and imprecision.
^hWe downgraded by a total of two levels to low certainty due to risk of bias concerns (quasi‐randomised trials, lack of blinding of the outcome assessors), and imprecision.