Table 3. Distribution of 14 high-risk HPV types among screened, unscreened cases and the population; impact numbers to prevent or detect one cervical cancer case by HPV type (CIs are presented in Fig 5 and S1A Table).

	% Among screened casesa	% Among unscreened casesb	% Among populationc	Estimated number of cases screenedd	Estimated number of cases unscreenede	Estimated number of women in populationf	Population impact number (number needed to screenk)		Number needed to follow-up (in test positive ones)
							To prevent one caseg	To detect one caseh	To prevent one casei	To detect one casej
HPV 16	50.61	58.39	2.67	182.7	641.7	67,815	5,527	13,885	147	371
HPV 18	18.85	13.38	0.96	68.1	147.0	24,275	32,125	37,280	307	356
HPV 31	3.47	3.06	1.58	12.5	33.6	40,194	120,247	202,547	1,905	3,208
HPV 33	2.87	3.69	0.58	10.3	40.6	14,711	83,884	245,200	486	1,421
HPV 35	0.14	0.91	0.34	0.5	10.0	8,516	267,045	5,100,308	896	17,120
HPV 39	0.70	1.28	0.43	2.5	14.0	10,852	220,334	1,008,712	942	4,314
HPV 45	6.05	7.28	1.52	21.9	80.0	38,472	43,655	116,091	662	1,760
HPV 51	0.13	0.03	1.13	0.5	0.3	28,630	∞l	5,239,990	∞l	59,133
HPV 52	1.58	1.12	1.43	5.7	12.4	36,336	380,757	445,906	5,453	6,386
HPV 56	0.81	0.75	0.82	2.9	8.2	20,705	479,556	871,466	3,913	7,112
HPV 58	0.28	1.09	0.58	1.0	12.0	14,711	230,175	2,533,751	1,334	14,692
HPV 59	0.56	0.36	0.33	2.0	3.9	8,312	1,339,680	1,252,334	4,389	4,103
HPV 66	0.10	0.30	0.50	0.4	3.3	12,779	852,629	7,065,580	4,294	35,589
HPV 68	0.04	0.31	0.18	0.1	3.4	4,453	783,010	18,162,183	1,374	31,878

^aAge-standardized proportion of HPV types among cervical cancer cases that were screened in the last 10 years.

^bAge-standardized proportion of HPV types among cervical cancer cases that were unscreened in the last 10 years.

^cAge-standardized proportion of HPV types in women population.

^dEstimated number of cases in 2011 in the pseudo-scenario that all women were screened. Calculated by number of cases per year during 2002–2011 (425) multiplied by incidence rate ratio between screened and the population (8.4/9.9 as shown in Fig 2) and multiplied by age-standardized proportion of HPV types among cervical cancer cases that were screened in the last 10 years.

^eEstimated number of cases in 2011 in the pseudo-scenario that all women were unscreened. Calculated by number of cases per year during 2002–2011 (425) multiplied by incidence rate ratio between unscreened and the population (25.6/9.9 as shown in Fig 2), multiplied by age-standardized proportion of HPV types among cervical cancer cases that were unscreened in the last 10 years.

^fCalculated as total number of women aged 23–64 years (2,536,995) multiplied by age-standardized proportion of HPV types in women population.

^gCalculated as total number of women aged 23–64 years (2,536,995) divided by the difference between number of cases in the pseudo-scenario that all women were unscreened and number of cases in the pseudo-scenario that all women were screened

^hCalculated as total number of women aged 23–64 years (2,536,995) divided by number of cases in the pseudo-scenario that all women were screened.

ⁱCalculated as estimated number of women in population for each HPV type divided by the difference between number of cases in the pseudo-scenario that all women were unscreened and number of cases in the pseudo-scenario that all women were screened.

^jCalculated as estimated number of women in population for each HPV type divided by number of cases in the pseudo-scenario that all women were screened.

^kInterpretation should be that among every X number of women in screening population, one cervical cancer case caused by a certain HPV type can be detected or prevented.

^lInfinity.

CI, confidence interval; HPV, human papillomavirus.