Sample size and power calculations for detecting changes in malaria transmission using antibody seroconversion rate

Table 6 Expected bias and precision (in percentage) of estimates for current SCR using the sample sizes shown in Table 5

				Assumption on the true change point
				Known			Unknown
Estimation	SCR before^a	SCR after^a	True change point^b	80 %	90 %	95 %	80 %	90 %	95 %
Bias^c	0.0969	0.0324	3	−3.62	−2.18	−1.44	−17.15	10.16	−5.06
			5	−1.30	0.18	0.10	−23.62	−16.29	−10.27
			10	0.34	0.14	0.08	−5.20	−3.24	−2.39
		0.0108	3	−12.86	−4.85	−1.13	−13.95	−11.42	−7.91
			5	−3.20	−0.78	−0.04	−28.33	−23.17	−18.07
			10	0.23	−0.64	−0.88	−28.42	−23.02	−17.75
	0.0324	0.0108	3	−0.35	1.14	1.89	−19.30	−17.64	−16.67
			5	−2.10	−0.69	0.16	−22.70	−14.89	−9.85
			10	−0.25	−0.22	−0.34	−6.00	−3.96	−2.91
		0.0036	3	0.39	−4.13	−6.48	−0.37	−3.35	−4.96
			5	−2.82	−0.53	1.02	−27.64	−21.20	−16.04
			10	0.86	0.53	0.17	−17.20	−15.22	−13.42
Precision^d	0.0969	0.0324	3	1.96	1.50	1.27	2.01	1.78	1.66
			5	1.48	1.21	1.06	1.65	1.65	1.46
			10	0.80	0.70	0.64	1.02	0.85	0.66
		0.0108	3	7.56	3.97	3.42	4.75	4.68	4.23
			5	3.62	3.22	3.10	3.13	3.13	3.05
			10	2.95	2.55	2.23	2.73	2.13	2.13
	0.0324	0.0108	3	2.04	1.74	1.55	2.10	1.84	1.68
			5	1.66	1.38	1.23	1.79	1.66	1.60
			10	1.01	0.83	0.73	1.40	1.04	0.80
		0.0036	3	5.23	4.43	3.97	4.82	4.14	3.73
			5	4.30	3.60	3.19	3.65	3.37	3.10
			10	2.70	2.70	2.58	2.96	2.95	2.86

^aExpected number of events per year
^bYears before sampling
^cAs percentage in relation to the true parameter value
^dDifference between 2.5 and 97.5 % quantiles of the distribution of current−SCR estimates divided by the true parameter value that generated the corresponding data

ISSN: 1475-2875