Evaluation of post-exposure prophylaxis practices to improve the cost-effectiveness of rabies control in human cases potentially exposed to rabies in southern Bhutan

Table 6 Final multivariate logistic regression model of factors associated with clinicians making an accurate rabies exposure risk assessment, defined by agreement of clinicians’ risk assessment with an independent assessment based on criteria listed in the NRMG

Variable	Coefficient Estimate	SE	Multiplier^1,3	P-value
Gender
Male
Female	−0.77	0.55	0.5	0.16
Designation
Medical Doctor
Clinical Officer	−0.70	0.63	0.5	0.26
Health Assistant	1.10	0.49	3.0	0.02
Health Centre Type
Basic Health Unit
District Hospital	2.05	0.73	7.8	0.00
Regional Hospital	2.87	1.16	17.6	0.01
Female*Health Assistants²	−1.70	0.84	0.2	0.04

¹ Multiplicative effect on the odds of the outcome, due to being in this stratum compared to the reference stratum. This corresponds to the odds ratio only for variables for which there were no interactions in the model
²The interaction between Gender and Designation; there were no female Clinical Officers in the study
³The Odds Ratio for varying levels of gender and designations can be obtained by combining the corresponding coefficients, e.g. OR of correct risk assessment for male health assistant compared to female health assistants is exp.(1.10)/exp.(− 0.77 + 1.10–1.70) = 12.

ISSN: 1471-2334