Department of Population Health Sciences, School of Medicine and Public Health, University of Wisconsin at Madison, 610 Walnut Street, WARF 703, Madison, WI, 53726-2397, USA.
Center for Biomedical Research, Universidad Autónoma de Bucaramanga, Bucaramanga, Colombia.
BMC Public Health. 2018 May 24;18(1):656. doi: 10.1186/s12889-018-5566-7.
We evaluated whether outbreaks of Zika virus (ZIKV) infection, newborn microcephaly, and Guillain-Barré syndrome (GBS) in Latin America may be detected through current surveillance systems, and how cases detected through surveillance may increase health care burden.
We estimated the sensitivity and specificity of surveillance case definitions using published data. We assumed a 10% ZIKV infection risk during a non-outbreak period and hypothetical increases in risk during an outbreak period. We used sensitivity and specificity estimates to correct for non-differential misclassification, and calculated a misclassification-corrected relative risk comparing both periods. To identify the smallest hypothetical increase in risk resulting in a detectable outbreak we compared the misclassification-corrected relative risk to the relative risk corresponding to the upper limit of the endemic channel (mean + 2 SD). We also estimated the proportion of false positive cases detected during the outbreak. We followed the same approach for microcephaly and GBS, but assumed the risk of ZIKV infection doubled during the outbreak, and ZIKV infection increased the risk of both diseases.
ZIKV infection outbreaks were not detectable through non-serological surveillance. Outbreaks were detectable through serologic surveillance if infection risk increased by at least 10%, but more than 50% of all cases were false positive. Outbreaks of severe microcephaly were detected if ZIKV infection increased prevalence of this condition by at least 24.0 times. When ZIKV infection did not increase the prevalence of severe microcephaly, 34.7 to 82.5% of all cases were false positive, depending on diagnostic accuracy. GBS outbreaks were detected if ZIKV infection increased the GBS risk by at least seven times. For optimal GBS diagnosis accuracy, the proportion of false positive cases ranged from 29 to 54% and from 45 to 56% depending on the incidence of GBS mimics.
Current surveillance systems have a low probability of detecting outbreaks of ZIKV infection, severe microcephaly, and GBS, and could result in significant increases in health care burden, due to the detection of large numbers of false positive cases. In view of these limitations, Latin American countries should consider alternative options for surveillance.
本研究旨在评估拉丁美洲现有的寨卡病毒（Zika virus，ZIKV）感染、新生儿小头畸形和格林-巴利综合征（Guillain-Barré syndrome，GBS）暴发监测系统是否能够发现 ZIKV 感染暴发，以及通过监测发现的病例可能会如何增加卫生保健负担。
我们使用已发表的数据来估计监测病例定义的敏感性和特异性。我们假设在非暴发期间，ZIKV 感染的风险为 10%，而在暴发期间的风险假设增加。我们使用敏感性和特异性估计值来校正非差异错误分类，并计算两个时期的校正错误分类的相对风险。为了确定导致可检测到的暴发的最小假设风险增加，我们将校正错误分类的相对风险与地方性通道上限（均值+2 个标准差）对应的相对风险进行了比较。我们还估计了暴发期间检测到的假阳性病例的比例。对于小头畸形和 GBS，我们采用了相同的方法，但假设暴发期间 ZIKV 感染风险增加一倍，并且 ZIKV 感染增加了这两种疾病的风险。
非血清学监测无法发现 ZIKV 感染暴发。如果感染风险增加至少 10%，则可以通过血清学监测发现暴发，但超过 50%的病例都是假阳性。如果 ZIKV 感染使小头畸形的流行率增加至少 24.0 倍，则可以发现严重小头畸形的暴发。如果 ZIKV 感染没有增加严重小头畸形的流行率，则取决于诊断准确性，所有病例中有 34.7%至 82.5%是假阳性。如果 ZIKV 感染使 GBS 的风险增加至少 7 倍，则可以发现 GBS 暴发。为了达到最佳的 GBS 诊断准确性，假阳性病例的比例范围为 29%至 54%，取决于 GBS 模拟病例的发病率。
现有的监测系统发现 ZIKV 感染、严重小头畸形和 GBS 暴发的可能性较低，并且由于检测到大量的假阳性病例，可能会导致卫生保健负担显著增加。鉴于这些局限性，拉丁美洲国家应考虑替代监测选项。
