Working towards a simple case definition for influenza surveillance

Area of Science:

• Epidemiology
• Infectious Diseases
• Public Health Surveillance

Background:

• Influenza surveillance relies on standardized case definitions for influenza-like illness (ILI).
• Australia has adopted a single ILI definition for national surveillance systems.
• Evaluating the predictive accuracy of clinical symptoms for influenza is crucial.

Purpose of the Study:

• To identify the most accurate combination of clinical symptoms for predicting laboratory-confirmed influenza in adult ILI patients.
• To compare the predictive value of symptom complexes against existing surveillance case definitions.

Main Methods:

• Utilized data from sentinel general practices in Western Australia and Victoria (1998-1999).
• Employed univariate analysis and stepwise logistic regression to identify independent clinical predictors of influenza.
• Calculated sensitivity, specificity, and positive predictive value (PPV) for various symptom combinations.

Main Results:

• The combination of cough, fever, and fatigue demonstrated good sensitivity (43.5-75.1%) and specificity (46.6-80.3%).
• Positive predictive values for this symptom complex ranged from 23.3% to 59.7%.
• Cough, fever, and fatigue proved more predictive of laboratory-confirmed influenza than current surveillance case definitions.

Conclusions:

• Recommended the symptom complex of cough, fever, and fatigue as a simplified case definition for ILI surveillance.
• Emphasized that accurate influenza activity monitoring still necessitates laboratory confirmation for a proportion of cases.