Optimised electronic patient records to improve clinical monitoring of HIV-positive patients in rural South Africa (MONART trial): study protocol for a cluster-randomised trial

Table 1 Numbers needed to demonstrate an improvement in viral load monitoring, for different assumptions regarding clinic size and effect sizes

Number of clinics per arm	Number of patients per clinic¹	Coefficient of variation (k)	% with VL suppressed in control	% with VL suppressed in intervention	Power (%)
5	100	0.20	45	67	71
5	100	0.20	45	70	80
5	100	0.20	50	74	71
5	100	0.20	50	77	80
5	100	0.20	55	80	70
5	100	0.20	55	84	80
5	150	0.20	45	66	70
5	150	0.20	45	69	80
5	150	0.20	50	73	70
5	150	0.20	50	77	80
5	150	0.20	55	80	70
5	150	0.20	55	84	80

ISSN: 1471-2334