Operational evaluation of the deployment of Malaria/CRP Duo and Dengue Duo rapid diagnostic tests for the management of febrile illness by village malaria workers in rural Cambodia

Table 2 Interrupted time series analysis: VMW consultations before and after RDT deployment using Prais-Winsten regression

	Pre-intervention trend ^a (consultations per month), (95% CI); p value	Step change at RDT introduction (consultations per month), 95% CI); p value	Post-intervention trend (consultations per month), 95% CI); p value
Study VMWs (combo-RDTs) \(\:n\)= 84	-0.18 (-0.40 to 0.04); p = 0.11	4.4 (1.5 to 7.4); p < 0.01	-0.02 (-0.43 to 0.39); p = 0.92
All VMWs \(\overline n\)= 152	-0.12 (-0.24 to 0.01); p = 0.07	N/A	N/A

^a Pre-intervention trend in the study area was derived from VMW data collected in the 18 months prior to deployment.
Pre-intervention trend for “All VMWs” was derived from routine national VMW data collected in the 3 ½ years prior deployment
p-values refer to a null effect, i.e., trend: slope = 0 and step-change = 0, \(\overline n\) is the mean number of VMWs over the 3 ½ years
Study VMWs: ρ = 0.38, Durbin-Watson = 1.88
Adjusting for seasonality with a categorical month indicator and a categorical wet season indicator did not improve the model
VMWs– Village malaria workers, CI– confidence intervals, N/A– Not applicable

ISSN: 1471-2334