Skip to main content


Fig. 2 | Malaria Journal

Fig. 2

From: Using ante-natal clinic prevalence data to monitor temporal changes in malaria incidence in a humanitarian setting in the Democratic Republic of Congo

Fig. 2

ac The concept of a crossbasis function in this context, in a the explanatory metric has corresponding effect on the response metric, the function that explains this relationship is the transmission effect basis. In b for a given value of the explanatory metric, this may have delayed effects on the response metric—in this plot for 3 months afterwards. This relationship is characterized by the temporal lag basis. In c, these two basis functions are combined into a bi-dimensional plot, the shape of the crossbasis function is restricted by the choice of functions in a and b. The precise shape of the crossbasis is determined during the fitting of the DLNM model. df How subsequent changes in one metric (Metric 1) can cause unpredictable patterns in another metric (Metric 2). d The different changes in Metric 1 differentiated by colour (yellow for the change in month 4, green for the change in month 5 and brown for month 6). e Each of these changes in Metric 1 have lagged effect that may differ with the size of the observation in Metric 1 and start at different times. These lagged effects are then observed as changes in Metric (2) over multiple months (f) with the lagged effects of three different changes in Metric 1 stacking up to create complex patterns in Metric 2. This is illustrated in this example where month 4 saw the greatest increase in Metric 1 whilst Metric 2 peaked in month 6

Back to article page