Analysis of partial and complete protection in malaria cohort studies

Malaria Journal

Table 4 Zero-inflated negative binomial regression output for the Kintampo data

Negative binomial component			Logistic component
Incidence rate ratios (IRR)			Odds ratio (OR) for not having malaria
	IRR (95% CI)	p		OR (95% CI)	p
Rural residence	1.64 (1.21, 2.20)	0.001	Rural residence	0.25 (0.10, 0.58)	0.001
Sex (female vs. male)	0.92 (0.79, 1.07)	0.259	-
Distance from health centre
(≥5 km vs. < 5 km)	0.92 (0.78, 1.08)	0.321	-
Thatched roof	1.11 (0.93, 1.32)	0.25	Thatched roof	1.27 (0.51, 3.16)	0.612
SES			SES
Least poor	-		Least poor	-
Less poor	1.51 (1.01, 2.24)	0.044	Less poor	0.59 (0.23, 1.53)	0.276
Poor	1.71 (1.18, 2.49)	0.005	Poor	0.38 (0.14, 1.05)	0.063
More poor	1.68 (1.15, 2.46)	0.008	More poor	0.34 (0.11, 1.05)	0.062
Most poor	1.65 (1.14, 2.41)	0.009	Most poor	0.07 (0, 1.67)	0.101
Sibling antibody response category			Sibling antibody response category
Low			Low
Medium	1.03 (0.84, 1.26)	0.77	Medium	1.28 (0.57, 2.87)	0.549
High	1.13 (0.92, 1.38)	0.241	High	1.02 (0.43, 2.44)	0.964
Bed net use			Bed net use
Low			Low
Medium	1.07 (0.87, 1.32)	0.526	Medium	0.86 (0.37, 1.98)	0.723
High	1.17 (0.95, 1.45)	0.138	High	0.54 (0.19, 1.52)	0.244

Table shows output from the zero-inflated negative binomial regression model. Incidence rate ratios are from the count component and odds ratios are from the logistic component.
CI = confidence interval; SES = socio-economic status.

ISSN: 1475-2875