Table 2.

Methodological features of included studies

First author & publication year	Country, design & funding	Outcome definition	Inclusion (1) & exclusion (2) criteria	Study size	Sample size calculation and power	Comparable at baseline	Adequate random sequence generation	Adequate allocation concealment	Adequate blinding	Adequate follow-up	Attrition, % (range)	ITT data
Wilkes, 2013 [41]	USA, cluster RCT, non-profit	Primary & Secondary	1	N ≥ 200	yes	yes	unclear	yes	HCP: yes Participants: yes OA: unclear	noa	≤20% (0–20%)	yesb
Landrey, 2013 [42]	USA, parallel RCT, non-profit	Primary & Secondary	1, 2	N ≥ 200	nr	yes	unclear	nr	HCP: nr Participants: nr OA: yes	yes	≥20% (6.60–51.16%)	no
Krist, 2007 [43, 44](Woolf, 2005)	USA, parallel RCT, non-profit	Primary & Secondary	1, 2	N ≥ 200	nr	Partialc	yes	yes	HCP: no Participants: nr OA: unclear	yes	< 20% (0–13.29%)d	yesb
Gatellari, 2003 [45]	Australia, parallel RCT, non-profit	Primary	1	N ≥ 200	ire	yes	yes	yes	HCP: yes Participants: yes OA: unclear	yes	≥20% (13.71–27.82%)	no

ir incomplete reporting, nr not reported, HCP Healthcare Professionals, OA Outcome Assessors, ITT Intention To Treat

^aFollow-up was driven by the timing of the standardised patient visit and varied from 6 to 16 weeks depending on the study arm

^bWilkes 2013: ITT for physician-reported screening behaviour and role in decision-making, doctors’ recommendations towards screening, physician-reported outcomes. Krist 2007: ITT for decisional conflict, PSA tests ordered by physicians or self-reported by patients, but unclear for other outcomes

^cMore physicians from the website and brochure groups reported to know the group patients were in, with a ratio of 1:3:3 between groups. This was intentionally done to be free of other potential biases

^dUnclear for two outcomes

^eReported on power only