UMHS LOGOUniversity of Michigan
Department of Pediatrics

Evidence-Based Pediatrics Web Site

All Negative DFA Tests Should Be Followed By Culture For the Diagnosis of B. pertussis


  • A 2 year old boy with recent onset cough, rhinnorrhea, questionable apneic episode and sick exposure, presents to my clinic. A direct flourescent antibody (DFA) test is sent off and comes back negative. Consideration is given as to whether or not to send off a culture specimen.

Clinical Bottom Lines

  1. DFA is sufficiently less sensitive (sensitivity = 68.5%) than culture in the diagnosis of B. pertussis.  Therefore a negative DFA test should always be followed up with a culture specimen.
  2. A positive DFA result should prompt one to treat.
  3. Interpretation of culture results should be approached with caution as well - oftentimes your clinical suspicion is your best diagnostic tool.
  4. Treatment should not be delayed, and the decision to treat should rely upon diagnostic criteria and clinical relevance.

Summary of Key Evidence

  1. Nasopharyngeal secretions were collected from 223 children (regardless of age)  who presented to St. Christopher's Hospital for Children with symptoms suggestive of B. pertussis infection. Symptoms included cough- sometimes accompanied by cyanosis, vomiting, whooping, and/or choking- rhinnorrhea and congestion.1
  2. Secretions from each individual patient were prepared by culturing on Regan-Lowe agar and by direct fluorescent antibody testing. The cultures were examined daily for 7 days for the presence of small, round glistening colonies. The  DFA tests were analyzed and considered positive when short rods with rims of  fluorescence were visualized with respect to antiglobulin and not visualized with respect to goat antirabbit globulin.
  3. B.pertussis was detected in 38 of the 223 patients entered into the study.  Twelve patients were culture positive and DFA negative. No patient was DFA positive and culture negative. 178 patients were culture negative and DFA negative.
  4. According to this study- the sensitivity of DFA when compared to culture  is  68.5% -The specificity of DFA is 100%.
  5. The prevalence of pertussis infection in this population was approximately 18%, thereby giving a positive DFA result post-test probability approaching 100%, and a negative DFA result post-test probability of approximately 7-10%.

Additional Comments

  • The role of culture as the goal standard in diagnosing pertussis has come into question as other tests such as serology and polymerase chain reaction have proven to be both more specific and sensitive.2
  • Specimen collection from the nasopharynx appears to be more appropriate than from the throat.3
  • Lymphocytosis > 50 % was observed in all patients. Lymphocytosis > 70% was observed in 47% of 30 patients studied with documented pertussis infection.  Thus, this finding remains important clinically.
  • 95% of patients had a cough - only 16% of patients had the characteristic 'whoop' associated with the cough. It usually occurres late in the clinical course, making this less useful in deciding whom to test.
  •  63% of patients with documented pertussis infections were hospitalized. Out of the 38 patients, complications included 6 cases of OM, 3 cases of pneumonia and 1 case of apnea requiring mechanical ventilation.


  1. Gilligan PH. Fisher MC.  Importance of Culture in Laboratory diagnosis of Bordetella pertussis infections.  Journal of Clin Microbiology.  20(5):891-93, 1984.
  2. vanderZee A. Agterberg C. Peeters M.  A clinical validation of B. pertussis and B. parapertussis polymerase chain reaction: comparison with culture and serology using samples from patients with suspected whooping cough from a highly immunized population. Journal of Infectious Diseases. 174(1):89-96, 1996.
  3. Marcon, MJ. Hamoudi AC.  Comparisons of throat and nasopharyngeal swab specimens for culture diagnosis of Bordetella pertussis infection.  Journal of  Clinical Microbiology. 25(6):1109-1110, 1987.
  4.  Waggoner-Fountain L. Hayden GF.  Pertussis in primary care practice.  Primary Care Clinics in Office Practice   vol 23, number 4 1996.

CAT Author: Rudo Benjamin, MD

CAT Appraisers: <Reviewers>, MD

Date appraised: August 31, 1998

Last updated June 15, 2003
Department of Pediatrics and Communicable Diseases
© 1998-2002 University of Michigan Health System