Rabies Postexposure Prophylaxis Noncompletion After Dog Bites: Estimating the Unseen to Meet the Needs of the Underserved

International audiencePostexposure prophylaxis (PEP) prevents human rabies and is accessible in Cambodia principally in Phnom Penh, the capital. Timely, affordable access to PEP is a challenge for the mainly rural population. We aimed to identify districts independently associated with PEP noncompletion to position frontline vaccination centers. We analyzed the 2009–2013 database at the Rabies Prevention Center at the Institut Pasteur du Cambodge, Phnom Penh. Logistic regressions identified nongeographic determinants of PEP noncompletion as well as the districts that were independently associated with noncompletion after adjustment for these determinants. The influence of distance by road was estimated using a boosted regression-trees model. We computed a population attributable fraction (rabies index (RI)) for each district and developed a map of this RI distribution. A cartographic analysis based on the statistic developed by Getis and Ord identified clusters of high-RI districts. Factors independently associated with noncompletion were pa-tients' district of residence, male sex, age 15–49 years, initial visit during rice harvest, the dog's status (culled or disappeared), and a prescribed PEP protocol requiring more than 3 PEP sessions (4 or 5). Four clusters of high-RI districts were identified using this analytical strategy, which is applicable to many vaccination or other health services. Positioning frontline PEP centers in these districts could significantly widen access to timely and adequate PEP