Selecting patients for randomized trials: a systematic approach based on risk group

Table 3 Sensitivity analysis. Net benefit when relative risk and NNT_t are varied.

		Relative Risk
NNT _t	Cut-off for risk of recurrence	0.75	0.8	0.85	0.9
10	10%	-0.00710	-0.0162	-0.02530	-0.03440
	25%	0.00880	0.00160	-0.00560	-0.01280
	50%	0.01090	0.00620	0.00150	-0.00320
15	10%	0.01043	0.00133	-0.00777	-0.01687
	25%	0.01787	0.01067	0.00347	-0.00373
	50%	0.01510	0.01040	0.00570	0.00100
20	10%	0.01920	0.01010	0.00100	-0.00810
	25%	0.02240	0.01520	0.00800	0.00080
	50%	0.01720	0.01250	0.00780	0.00310
25	10%	0.02446	0.01536	0.00626	-0.00284
	25%	0.02512	0.01792	0.01072	0.00352
	50%	0.01846	0.01376	0.00906	0.00436
33	10%	0.02956	0.02046	0.01136	0.00226
	25%	0.02776	0.02056	0.01336	0.00616
	50%	0.01968	0.01498	0.01028	0.00558

The relative risk of the intervention is 0.75, the NNT_t is 33 and the event rate in the absence of intervention is 20%. The decrease in event rate is the event rate in the population in the absence of intervention minus the event rate in the population when individuals meeting the high-risk definition receive intervention.

ISSN: 1745-6215