Hepler Staci A, Kaufeld Kimberly A, Kline David, Greene Andrew, Gorris Morgan E
Department of Statistical Sciences, Wake Forest University, Winston-Salem, NC 27109, United States.
Statistical Sciences Group, Los Alamos National Laboratory, Los Alamos, NM 87545, United States.
Am J Epidemiol. 2025 Jan 8;194(1):56-63. doi: 10.1093/aje/kwae199.
Coccidioidomycosis, or Valley fever, is an infectious disease caused by inhaling Coccidioides fungal spores. Incidence has risen in recent years, and it is believed the endemic region for Coccidioides is expanding in response to climate change. While Valley fever case data can help us understand trends in disease risk, using case data as a proxy for Coccidioides endemicity is not ideal because case data suffer from imperfect detection, including false positives (eg, travel-related cases reported outside of endemic area) and false negatives (eg, misdiagnosis or underreporting). We proposed a Bayesian, spatio-temporal occupancy model to relate monthly, county-level presence/absence data on Valley fever cases to latent endemicity of Coccidioides, accounting for imperfect detection. We used our model to estimate endemicity in the western United States. We estimated high probability of endemicity in southern California, Arizona, and New Mexico, but also in regions without mandated reporting, including western Texas, eastern Colorado, and southeastern Washington. We also quantified spatio-temporal variability in detectability of Valley fever, given an area is endemic to Coccidioides. We estimated an inverse relationship between lagged 3- and 9-month precipitation and case detection, and a positive association with agriculture. This work can help inform public health surveillance needs and identify areas that would benefit from mandatory case reporting. This article is part of a Special Collection on Environmental Epidemiology.
球孢子菌病，即山谷热，是一种因吸入球孢子菌真菌孢子而引发的传染病。近年来其发病率有所上升，据信球孢子菌的流行区域正因气候变化而不断扩大。虽然山谷热病例数据有助于我们了解疾病风险趋势，但将病例数据用作球孢子菌流行程度的替代指标并不理想，因为病例数据存在检测不完美的问题，包括假阳性（例如在流行区域之外报告的与旅行相关的病例）和假阴性（例如误诊或报告不足）。我们提出了一种贝叶斯时空占用模型，将山谷热病例的月度、县级存在/不存在数据与球孢子菌的潜在流行程度联系起来，同时考虑到检测不完美的情况。我们使用该模型估计了美国西部的流行程度。我们估计在南加州、亚利桑那州和新墨西哥州，以及包括得克萨斯州西部、科罗拉多州东部和华盛顿州东南部等没有强制报告要求的地区，球孢子菌流行的可能性很高。我们还量化了在球孢子菌流行的地区，山谷热可检测性的时空变异性。我们估计滞后3个月和9个月的降水量与病例检测之间呈负相关，与农业呈正相关。这项工作有助于为公共卫生监测需求提供信息，并确定将从强制病例报告中受益的地区。本文是环境流行病学特刊的一部分。
