Lubell Yoel, Blacksell Stuart D, Dunachie Susanna, Tanganuchitcharnchai Ampai, Althaus Thomas, Watthanaworawit Wanitda, Paris Daniel H, Mayxay Mayfong, Peto Thomas J, Dondorp Arjen M, White Nicholas J, Day Nicholas P J, Nosten François, Newton Paul N, Turner Paul
Mahidol-Oxford Tropical Medicine Research Unit (MORU), Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand.
Centre for Tropical Medicine and Global Health, Nuffield Department of Medicine, University of Oxford, Oxford, UK.
BMC Infect Dis. 2015 Nov 11;15:511. doi: 10.1186/s12879-015-1272-6.
Poor targeting of antimicrobial drugs contributes to the millions of deaths each year from malaria, pneumonia, and other tropical infectious diseases. While malaria rapid diagnostic tests have improved use of antimalarial drugs, there are no similar tests to guide the use of antibiotics in undifferentiated fevers. In this study we estimate the diagnostic accuracy of two well established biomarkers of bacterial infection, procalcitonin and C-reactive protein (CRP) in discriminating between common viral and bacterial infections in malaria endemic settings of Southeast Asia.
Serum procalcitonin and CRP levels were measured in stored serum samples from febrile patients enrolled in three prospective studies conducted in Cambodia, Laos and, Thailand. Of the 1372 patients with a microbiologically confirmed diagnosis, 1105 had a single viral, bacterial or malarial infection. Procalcitonin and CRP levels were compared amongst these aetiological groups and their sensitivity and specificity in distinguishing bacterial infections and bacteraemias from viral infections were estimated using standard thresholds.
Serum concentrations of both biomarkers were significantly higher in bacterial infections and malaria than in viral infections. The AUROC for CRP in discriminating between bacterial and viral infections was 0.83 (0.81-0.86) compared with 0.74 (0.71-0.77) for procalcitonin (p < 0.0001). This relative advantage was evident in all sites and when stratifying patients by age and admission status. For CRP at a threshold of 10 mg/L, the sensitivity of detecting bacterial infections was 95% with a specificity of 49%. At a threshold of 20 mg/L sensitivity was 86% with a specificity of 67%. For procalcitonin at a low threshold of 0.1 ng/mL the sensitivity was 90% with a specificity of 39%. At a higher threshold of 0.5 ng/ul sensitivity was 60% with a specificity of 76%.
In samples from febrile patients with mono-infections from rural settings in Southeast Asia, CRP was a highly sensitive and moderately specific biomarker for discriminating between viral and bacterial infections. Use of a CRP rapid test in peripheral health settings could potentially be a simple and affordable measure to better identify patients in need of antibacterial treatment and part of a global strategy to combat the emergence of antibiotic resistance.
抗菌药物使用不当导致每年数百万疟疾、肺炎及其他热带传染病患者死亡。虽然疟疾快速诊断检测改善了抗疟药物的使用,但尚无类似检测来指导未分化发热患者使用抗生素。在本研究中,我们评估了两种公认的细菌感染生物标志物——降钙素原和C反应蛋白(CRP)在东南亚疟疾流行地区鉴别常见病毒感染和细菌感染方面的诊断准确性。
在柬埔寨、老挝和泰国进行的三项前瞻性研究中,对发热患者的储存血清样本检测血清降钙素原和CRP水平。在1372例经微生物学确诊的患者中,1105例患有单一病毒、细菌或疟疾感染。比较这些病因组的降钙素原和CRP水平,并使用标准阈值估计它们在区分细菌感染和菌血症与病毒感染方面的敏感性和特异性。
两种生物标志物的血清浓度在细菌感染和疟疾中均显著高于病毒感染。CRP鉴别细菌感染和病毒感染的曲线下面积(AUROC)为0.83(0.81 - 0.86),而降钙素原为0.74(0.71 - 0.77)(p < 0.0001)。这种相对优势在所有地点以及按年龄和入院状态对患者进行分层时均很明显。对于CRP,阈值为10mg/L时,检测细菌感染的敏感性为95%,特异性为49%。阈值为20mg/L时,敏感性为86%,特异性为67%。对于降钙素原,低阈值为0.1ng/mL时,敏感性为90%,特异性为39%。高阈值为0.5ng/μL时,敏感性为60%,特异性为76%。
在东南亚农村地区单一感染的发热患者样本中,CRP是鉴别病毒感染和细菌感染的高敏感性和中等特异性生物标志物。在基层医疗卫生机构使用CRP快速检测可能是一种简单且经济实惠的措施,可更好地识别需要抗菌治疗的患者,也是全球应对抗生素耐药性出现战略的一部分。