Spatial patterns and secular trends in human leishmaniasis incidence in Morocco between 2003 and 2013.

BACKGROUND Few studies on spatial patterns or secular trends in human leishmanias have been conducted in Morocco. This study aimed to examine spatial patterns and trends associated with the human leishmaniasis incidence rate (HLIR) at the province/prefecture level between 2003 and 2013 in Morocco. METHODS Only the available published country data on the HLIR between 2003 and 2013, from the open access files of the Ministry of Health, were used. Secular trends were examined using Kendall's rank correlation. An exploratory spatial data analysis was also conducted to examine the spatial autocorrelation (Global Moran's I and local indicator of spatial association [LISA]), and spatial diffusion at the province/prefecture level. The influence of various covariates (poverty rate, vulnerability rate, population density, and urbanization) on the HLIR was tested via spatial regression (ordinary least squares regression). RESULTS At the country level, no secular variation was observed. Poisson annual incidence rate estimates were 13 per 100 000 population (95 % CI = 12.9-13.1) for cutaneous leishmaniasis (CL) and 0.4 per 100 000 population (95 % CI = 0.4-0.5) for visceral leishmaniasis (VL). The available data on HLIR were based on combined CL and VL cases, however, as the CL cases totally outnumbered the VL ones, HLIR may be considered as CL incidence rate. At the provincial level, a secular increase in the incidence rate was observed in Al Hoceima (P = 0.008), Taounate (P = 0.04), Larache (P = 0.002), Tétouan (P = 0.0003), Khenifra (P = 0.008), Meknes (P = 0.03), and El Kelaa (P = 0.0007), whereas a secular decrease was observed only in the Chichaoua province (P = 0.006). Even though increased or decreased rate was evident in these provinces, none of them showed clustering of leishmaniasis incidence. Significant spatial clusters of high leishmaniasis incidence were located in the northeastern part of Morocco, while spatial clusters of low leishmaniasis incidence were seen in some northwestern and southern parts of Morocco; there was spatial randomness in the remaining parts of the country. Significant clustering was seen from 2005 to 2013, during which time the Errachidia province was a permanent 'hot spot'. Global Moran's I increased from 0.2844 (P = 0.006) in 2005 to 0.5886 (P = 0.001) in 2011, and decreased to 0.2491 (P = 0.004) in 2013. It was found that only poverty had an effect on the HLIR (P = 0.0003), contributing only 23 % to this (Adjusted R-squared = 0.226). CONCLUSION Localities showing either secular increase in human leishmaniasis or significant clustering have been identified, which may guide decision-making as to where to appropriately allocate funding and implement control measures. Researchers are also urged to undertake further studies focusing on these localities.