Related Experiment Video
Updated: May 22, 2026
An Experimental Model to Study Tuberculosis-Malaria Coinfection upon Natural Transmission of Mycobacterium tuberculosis and Plasmodium berghei
Published on: February 17, 2014
Distinguishing malaria and influenza: early clinical features in controlled human experimental infection studies
Patrick J Lillie1, Christopher J A Duncan, Susanne H Sheehy
1Centre for Clinical Vaccinology and Tropical Medicine, Churchill Hospital, Oxford, UK. patricklillie@doctors.net.uk
Abstract:
During the H1N1 influenza pandemic (pH1N1/09) diagnostic algorithms were developed to guide antiviral provision. However febrile illnesses are notoriously difficult to distinguish clinically. Recent evidence highlights the importance of incorporating travel history into diagnostic algorithms to prevent the catastrophic misdiagnosis of life-threatening infections such as malaria. We applied retrospectively the UK pH1N1/09 case definition to a unique cohort of healthy adult volunteers exposed to Plasmodium falciparum malaria or influenza to assess the predictive value of this case definition, and to explore the distinguishing clinical features of early phase infection with these pathogens under experimental conditions. For influenza exposure the positive predictive value of the pH1N1/09 case definition was only 0.38 (95% CI: 0.06-0.60), with a negative predictive value of 0.27 (95% CI: 0.02-0.51). Interestingly, 8/11 symptomatic malaria-infected adults would have been inappropriately classified with influenza by the pH1N1/09 case definition, while 5/8 symptomatic influenza-exposed volunteers would have been classified without influenza (P = 0.18 Fisher's exact). Cough (P = 0.005) and nasal symptoms (P = 0.001) were the only clinical features that distinguished influenza-exposed from malaria-exposed volunteers. An open mind regarding the clinical cause of undifferentiated febrile illness, particularly in the absence of upper respiratory tract symptoms, remains important even during influenza pandemic settings. These data support incorporating travel history into pandemic algorithms.
Insights
During the H1N1 influenza pandemic, diagnostic algorithms struggled to differentiate febrile illnesses. Incorporating travel history is crucial for accurate diagnosis, preventing misdiagnosis of infections like malaria.
Area of Science:
- Infectious Diseases
- Epidemiology
- Clinical Diagnostics
Background:
- Diagnostic algorithms for pandemic influenza (pH1N1/09) were developed to guide antiviral treatment.
- Distinguishing febrile illnesses clinically is challenging.
- Misdiagnosis of life-threatening infections like malaria can occur.
Purpose of the Study:
- To retrospectively assess the UK pH1N1/09 case definition's predictive value.
- To explore distinguishing clinical features of early malaria and influenza infections under experimental conditions.
- To evaluate the utility of travel history in diagnostic algorithms.
Main Methods:
- Retrospective application of the UK pH1N1/09 case definition.
- Analysis of a cohort of healthy adult volunteers experimentally exposed to Plasmodium falciparum malaria or influenza.
- Comparison of clinical features between influenza-exposed and malaria-exposed groups.
Main Results:
- The pH1N1/09 case definition had a low positive predictive value (0.38) for influenza.
- A significant proportion of malaria cases (8/11) were misclassified as influenza.
- Cough and nasal symptoms were key differentiators between influenza and malaria.
- Influenza-exposed volunteers often lacked influenza symptoms (5/8).
Conclusions:
- The pH1N1/09 case definition showed poor predictive value in distinguishing influenza from malaria.
- Clinical presentation alone is insufficient for diagnosing febrile illnesses during pandemics.
- Travel history is essential for accurate diagnosis of undifferentiated febrile illnesses, especially when upper respiratory symptoms are absent.
