Table 1

Effectiveness of training strategies on the practices of professional healthcare providers

Strategies tested*		Outcome scale	No of study comparisons (risk of bias: low, moderate, high, very high)	Median MES†, in %-points (IQR; range)
Intervention arm	Reference arm	Outcome scale		Median MES†, in %-points (IQR; range)
Group in-service training
Group in-service training	Controls	Percentage	60 (9, 19, 17, 15)	7.3‡ (3.6–17.4; –21.3 to 68.1)
Group in-service training	Controls	Continuous	16 (3, 5, 2, 6)	17.4§ (–2.3 to 28.5; –25.0 to 81.4)
Group in-service training plus other strategy components	Other strategy components	Percentage	13 (6, 3, 4, 0)	3.7¶ (–0.1 to 5.8; –2.7 to 23.6)
Group in-service training plus other strategy components	Other strategy components	Continuous	4 (1, 1, 2, 0)	–7.3 (–20.6 to 3.6; –25.8 to 6.4)
Educational outreach visits
Educational outreach visits	Controls	Percentage	8 (0, 2, 3, 3)	9.9 (4.3–20.6; 2.8–30.9)
Educational outreach visits plus other strategy components	Other strategy components	Percentage	3 (2, 1, 0, 0)	21.5 (NA; 5.4–30.7)
Group in-service training versus (or combined with) educational outreach visits
Group in-service training	Educational outreach visits	Percentage	1 (0, 0, 1, 0)	0.8 (NA; NA)
Group in-service training plus other strategy components	Educational outreach visits plus other strategy components	Percentage	2 (2, 0, 0, 0)	–6.4 (NA; –5.8 to –7.0) (ie, both studies favoured educational outreach visits)
Group in-service training plus educational outreach visits	Controls	Percentage	1 (0, 0, 0, 1)	–2.5 (NA; NA)
Group in-service training versus (or combined with) self-study in-service training
Group in-service training	Self-study in-service training	Percentage	2 (1, 1, 0, 0)	9.3 (NA; 4.6–14.0)
Group in-service training plus other strategy components	Self-study in-service training plus other strategy components	Percentage	2 (0, 0, 1, 1)	2.0 (NA; –1.0 to 5.0)
Group in-service training plus self-study in-service training	Controls	Percentage	1 (0, 0, 1, 0)	24.0 (NA; NA)
Group preservice training
Group preservice training	Controls	Percentage	3 (1, 1, 1, 0)	16.9 (NA; 15.0–46.7)
Peer-to-peer training
Peer-to-peer training	Controls	Percentage	1 (0, 0, 0, 1)	4.0 (NA; NA)
Peer-to-peer training plus group in-service training	Controls	Percentage	3 (0, 0, 0, 3)	8.4 (NA; 1.8–66.2)
Peer-to-peer training plus group in-service training plus other strategy components	Other strategy components	Percentage	1 (0, 0, 1, 0)	25.0 (NA; NA)

*See boxes 1 and 2 for descriptions of the strategies and the comparisons, respectively.
†Effect sizes calculated as the intervention arm improvement minus reference arm improvement.
‡Among studies with a low or moderate risk of bias, median MES=5.1 %-points (IQR: 2.5–14.0; range: –3.0 to 42.8); median MES for high or very high risk of bias studies=9.7 (IQR: 5.1–19.8; range: –21.3 to 68.1).
§Among studies with a low or moderate risk of bias, median MES=15.1 %-points (IQR: –3.8 to 21.2; range: –25.0 to 81.4); median MES for high or very high risk of bias studies=20.3 (IQR: 1.9–41.4; range: –19.2 to 57.3).
¶Among studies with a low or moderate risk of bias, median MES=4.5 %-points (IQR: 2.1–5.8; range: –2.0 to 23.6); median MES for high or very high risk of bias studies=1.4 (IQR: –1.8 to 5.4; range: –2.7 to 7.0).
MES, median effect size; NA, not applicable; %-points, percentage-points.