02469nas a2200229   4500000000100000000000100001008004100002260007000043653000900113653001300122653001700135653001800152653004300170100001200213700001700225245010200242250000800344300001200352490000800364520185300372022001402225       2017                            d  c03/2017bSchlütersche Verlagsgesellschaft mbH & Co. KGaHannover10aCCHF10aPakistan10arisk factors10azero inflated10anormalized difference vegetation index1 aT Abbas1 aM Tahir Riaz00aZero-inflated model to identify risk factors of human Crimean-Congo hemorrhagic fever in Pakistan  a3/4  a165-1690 v1303 aCrimean-Congo-hemorrhagic-fever (CCHF) is considered an emerging tick-borne viral zoonotic disease in Pakistan. CCHF viremic animals are often asymptomatic. Understanding the distribution and quantification of risk factors for CCHF infection in humans can help to identify places, times and sub-populations of livestock which should be given priority in veterinary surveillance. In this study, we quantified the effect of the previous history of outbreaks (from 2003 to 2008) and the food security index on number of human cases of CCHF reported during the year 2013. In addition, we ascertained the temporally smoothed normalized difference vegetation index (NDVI) as a predictor of structural or true zeros. The data about human cases reported for the year 2013 contained excess zeros (89%) and overdispersion (variance to mean ratio: 12.3). A zero inflated Poisson model was therefore selected. For the logistic portion of the model, the absence of CCHF cases in the disease-free districts was significantly associated with the mean NDVI (OR = 1.01, 96% CI: 1.02–1.11), meaning that the higher the NDVI, the higher the probability of not having outbreaks resulting in true zeros. For the Poisson part of the model, the presence of previous CCHF outbreaks in affected districts increased the expected number of cases by 26.89 times (96% CI: 7.61–88.89). Compared to the reference category i. e. extremely food insecure areas, the expected count of cases was relatively higher in food insecure (OR: 4.82, 96% CI: 1.03–22.68, borderline (OR: 33.41, 96%CI: 7.44–160.49), and secure areas (OR: 9.71, 96% CI: 2.07–46.61). Based on the outcome of our analysis, we propose that Islamabad, Rawalpindi and Quetta districts should be focus for surveillance, research, and one health approach to prevent transmission at tick-host-virus interface.   a0006-9366