Table 4. Mean cost and effect differences between the SHARP and CAU group.

Analysisa			ΔC (95% CI)		ΔE (95% CI)
		euros			percentage/days/eurosd	ICER	NMBe
Total group
	CEA-incidence of RSA	1932 (-318 to 5350)			0.24 (0.03 to 0.45)	10605
	CEA-time to RSA	1358 (-945 to 4886)			55 (2.85 to 106.09)	2183
	CBA HCA only sickness absence	800 (678 to 922)			5246 (-2701 to 13192)		6046
	CBA FCA only sickness absence	800 (678 to 922)			3195 (-2214 to 8604)		3995
	CBA HCA combinedb	800 (678 to 922)			4730 (-5699 to 15158)		5530
	CBA FCA combinedb	800 (678 to 922)			3929 (-3764 to 11623)		4729
Excluding outlierc
	CEA-incidence of RSA	-133 (-1155 to 914)			0.25 (0.03 to 0.46)	-533
	CEA-time to RSA	-129 (-1266 to 964)			59 (5.95 to 111.15)	-2

CEA = cost effectiveness analysis; RSA = recurrent sickness absence; CBA = cost benefit analysis; ΔC = mean cost difference; ΔE = mean effect difference; HCA = human capital approach; FCA = friction cost approach; ICER = incremental cost effect ration; NMB = net monetary benefit

^aIn the CEA, ΔC is the mean difference in total health care costs and ΔE is the mean difference in percentage of workers that experienced recurrent sickness absence; in the CBA, ΔC is the mean difference in total occupational health care costs, including the intervention, for the employer and ΔE is the mean difference in sickness absence costs estimated by the HCA or FCA.

^bProductivity loss costs are a combination of sickness absence costs and costs due to lost productivity at work.

^cSensitivity analysis excluding one extreme outlier.

^dDifferences in CEA effects are presented in (1) percentage of workers that experienced recurrent sickness absence, (2) number of days to recurrent sickness absence; differences in CBA effects are presented as costs in euros.

^eNegative values of the NMB imply lower costs for the intervention group compared to the control group.