Influenza

Essential Evidence

Last Updated on 2022-06-24 © 2022 John Wiley & Sons, Inc.

Printer Friendly

Authors:
Mark H. Ebell, MD, MS, Professor, College of Public Health, University of Georgia

Editors:
Randall Forsch, MD, MPH, Assistant Professor, University of Michigan
Mark H. Ebell, MD, MS, Professor, College of Public Health, University of Georgia
Kenny Lin, MD, MPH, Professor of Family Medicine, Georgetown University

Overall Bottom Line

  • Yearly influenza vaccination is recommended by ACIP (US) for persons over the age of 6 months. A
  • Patients classically present with rapid onset of fever, cough, chills or rigors, sore throat, and myalgias. A clinical decision rule can be used to identify patients at low, moderate or high risk of influenza. B
  • Rapid antigen tests are most useful during the beginning and end of the flu season. C
  • Treatment is largely supportive; antiviral medications are recommended only if the likelihood of influenza is high and the patient presents within 24 hours of symptom onset. There is no evidence that they reduce serious complications or hospitalization. A
  • Predictors of adverse outcome in the elderly include increased age, previous hospitalization, comorbidities, gender, and number of outpatient visits in the previous year (Table 1). A

Diagnosis

Bottom Line

  • Patients classically present with rapid onset of fever, cough, chills or rigors, sore throat, and myalgias. B 6
  • Overall, physician judgment is useful with early but not late onset of symptoms. B
  • Rapid antigen tests are most useful during the shoulder season (beginning and end of the flu season), when the pretest probability of influenza is 5% to 10%. Their accuracy varies, with older tests having approximately 55% sensitivity compared to 80% to 95% for newer molecular tests. 77B 44

Differential Diagnosis

DiagnosisFeatures
PneumoniaEvidence of consolidation on physical examination and CXR
Pulmonary embolismHypoxia, tachycardia, and pleuritic chest pain common; fever uncommon
Other viral lower respiratory tract infectionsDifficult to distinguish without laboratory investigation
Dengue, chikungunya, and Zika virusProminent myalgias and joint pains; conjunctivitis and rash with Zika

Using the History and Physical

Selecting Diagnostic Tests

  • A systematic review identified 162 studies (130 of rapid influenza diagnostic tests [RIDTs], 19 of digital immunoassays [DIAs], and 13 of rapid nucleic acid amplification tests [NAATs]). The older RIDTs had lower sensitivity (55%) than the newer DIAs (80%) and NAATs (92%); all tests had excellent specificity (98% or higher). This corresponds to very high positive likelihood ratios to rule-in influenza when any test is positive (LR+ 27-46), but varying negative likelihood ratios to rule-out flu when negative (0.46 for RIDT, 0.20 for DIA, and 0.08 for NAAT). 77
  • Rapid antigen tests are most useful during the shoulder season (beginning and end of the flu season) when the pretest probability of influenza is 5% to 10%. During flu season, patients with a negative rapid antigen test still have a significant likelihood of flu, so the test is less useful (see Figure 1). 44
  • Rapid antigen tests for H1N1 pandemic influenza has been shown to have reasonable sensitivity and specificity. 50
  • Confirmatory testing with PCR or culture is generally done only for disease surveillance.
  • A procalcitonin value <0.25 ng/ml (especially if <0.1) makes bacterial infection unlikely and during an influenza epidemic makes influenza correspondingly more likely. 24
  • Order CXR to rule out pneumonia in patients with lower respiratory tract findings or significant tachypnea or hypoxia. Two studies with approximately 2000 patients found no evidence that CXR improves clinical outcomes in patients with respiratory infection, though. 15
  • Consider pulmonary embolism in any patient with dyspnea, although fever makes it less likely.
  • Consider Dengue fever in areas where the disease is endemic (including southern Florida and the Keys), especially in patients with leukopenia and thrombocytopenia. 67 Also consider Chikungunya in Puerto Rico and Central and South America, and Zika Virus infection.

Clinical Decision Rules

  • A "FluScore" decision rule has been developed and internally validated to classify patients at low, moderate, or high risk of influenza. It uses four patient reported symptoms: Fever + cough (2 points), Myalgias (2 points), Chills or sweats (1 point) or Sudden onset (1 point). The risk of influenza was 8% for 0 to 2 point, 30% for 3 points, and 59% for 4 or more points. Two-thirds of patients fell into the low or high risk groups, and required no further testing. 51
  • A decision rule has been developed and validated to predict the need for hospitalization and mortality in patients presenting with acute respiratory infection. While not validated in an epidemic, it could be useful in such a setting. 2

Approach to the Patient

History and Physical Tests

  • Influenza suspected -> influenza
  • Influenza suspected (pediatric) -> influenza
  • Diagnostic Tests

  • Influenza suspected (all ages) -> influenza
  • Influenza suspected (child) -> influenza
  • Influenza suspected (adult) -> influenza