Temporal variation in confirmed diagnosis of fever-related malarial cases among children under-5 years by community health workers and in health facilities between years 2013 and 2015 in Siaya County, Kenya

Table 5 Negative binomial regression univariable analyses (P ≤ 0.1) for cases diagnosed in health facilities

Variable	Variable category	Coefficient	95% confidence interval	P > \|z\|	Likelihood ratio test P*	α	Likelihood-ratio test of α = 0
Year	2014	− 0.16	[− 0.49, 0.17]	0.337	0.399	0.17	0.000^╩
Year	2015	− 0.22	[− 0.55, 0.11]	0.188
Month	February	− 0.13	[− 0.46, 0.18]	0.403	0.000	0.04	0.000
	March	− 0.10	[− 0.42, 0.22]	0.536
	April	0.11	[− 0.21, 0.43]	0.505
	May	0.50	[0.17, 0.82]	0.003
	June	0.56	[0.23, 0.88]	0.001
	July	0.59	[0.26, 0.91]	0.000
	August	− 0.06	[− 0.39, 0.26]	0.698
	September	− 0.33	[− 0.66, − 0.01]	0.044
	October	− 0.42	[− 0.75, − 0.09]	0.012
	November	− 0.33	[− 0.66, − 0.01]	0.043
	December	− 0.58	[− 0.91, − 0.25]	0.000

* P used to test the statistical significance (P ≤ 0.1) of the contribution of the variable to the univariable model; ^╩ P used to test the statistical significance of α (the overdispersion parameter). When the statistical significance of α is significant (P ≤ 0.05), it suggests that the variance in the data is higher than would be expected for a Poisson regression

ISSN: 1475-2875