. 2013 Mar 19;37(6):1135–1140. doi: 10.1007/s00264-013-1851-3

Table 3.

Number needed to treat with a cemented prosthesis in order to cause one fatality due to the cementation (NNH) day 0–1, according to ASA score

ASA score		Dead	Total	Risk	RD (cem−uncem)	NNH^a	95 % CI
ASA score		Dead	Total	Risk	RD (cem−uncem)	NNH^a	Lower	Upper
1 + 2	Uncemented	2	1,002	0.002
	Cemented	11	3,407	0.003	0.0012	811	218	∞
3	Uncemented	7	1,390	0.005
	Cemented	79	4,755	0.017	0.0116	86	59	157
4 + 5	Uncemented	3	144	0.021
	Cemented	26	512	0.051	0.0299	33	17	∞
1–5	Uncemented	12	2,536	0.005
	Cemented	116	8,674	0.013	0.0086	116	82	199

CI confidence interval

^aNumber of patients treated with a cemented implant in order to cause one fatality because of cementation compared to an uncemented implant if there is a direct causal effect. The NNH was calculated as an inverse value of the risk difference (RD) between the methods [1/(risk cemented − risk uncemented)]