Table 3.

Summary of the risk of bias.

Included Study	Selection Bias	Lack of Blinding	Exposure Misclassification	Outcome Misclassification	Incomplete Exposure Data	Incomplete Outcome Data	Selective Reporting of Exposures	Selective Reporting of Outcome	Differences in Numerator and Denominator	Conflict of Interest
	Any health services use
Giatti 2008 * [22]	+	+	+	+	+	+	+	+	++	+
Giatti 2011 [23]	+	+	+	+	+	+	+	+	++	+
Le 2015 [24]	-	+	-	+	+	+			++	+
Miquillan 2013 [25]	++	+	+	+			+	+	++	+
	Fatal occupational injuries
Lopez- Bonilla 2011 [26]	+	+	++	-		-	+	--	+	+
Mora 2011 [27]	+	+	+	+		--	+	-	+	+
	Non-fatal occupational injuries
Cunningham 2012 [28]	-	+	++	+	++	-	++	-	-	+
Calys-Tagoe 2017 [29]	-	+		-	++	-	++	+	-	+
Santana 2003 [30]	+	+	+	++	+	+	+	-	++	+
	Depression
Abbas 2013 [31]	+	+	++	++	++		++	+	++	+
Da Silva 2006 a [32]	-	+	-	+	+	++	+	++	+	+
Giatti 2008 * [22]	+	+	+	+	+	+	+	+	++	+
	Musculoskeletal disorders
Da Silva 2006 b [33]	-	+	-	+	++	+	+	+	+	+

RoB-SPEO [19] risk of bias ratings: definitely low (++ dark green); probably low (+ light green); probably high (- pink); definitely high (-- red); no information (yellow). * This study reported evidence on two outcomes: any health services use and depression.