Chronic cough in adults is a common clinical problem that accounts for about 30 million visits to primary care physicians in the United States each year. However, several published studies have suggested that infectious causes may be overlooked. One organism known to produce cough in children with acute infection is Bordetella pertussis. Diagnosis of infection with B. pertussis is problematic in adults because the infection is often not suspected and, even when it is, serologic diagnosis is time consuming and expensive. Ironically, it is adults who often infect susceptible infants, who then develop the condition known as “whooping cough.” Birkebaek and colleagues performed a study to evaluate the frequency of the presence of B. pertussis as a cause of chronic cough in adults.

Patients 16 years of age or older who presented to a pulmonary medicine clinic were evaluated for the study. Inclusion criteria were a cough of two to 12 weeks' duration, a normal chest radiograph, normal spirometry results and no history of chronic cardiopulmonary disease. Eligible patients were compared with healthy adult control subjects who did not have a history of cough during the previous 12 weeks.

At study entry, the following samples were obtained: a nasopharyngeal aspirate for B. pertussis polymerase chain reaction (PCR) testing, a calcium alginate nasopharyngeal swab for B. pertussis culture and a blood specimen for B. pertussis antibodies (acute serum). Four weeks later, all patients (including the control subjects) had a second serum sample obtained for B. pertussis antibodies (convalescent serum).

A total of 201 patients who had a persistent cough for a mean of six and one-half weeks were enrolled in the study. Four (2 percent) of the patients had a nasopharyngeal culture positive for B. pertussis, and 11 (5.5 percent) patients had PCR testing positive for the organism. Of these 11 patients, four had positive cultures. Serologic evidence of infection was found in 33 patients. The mean titers of IgG antibody to pertussis toxin were significantly higher in the patients with positive PCR tests compared with the control subjects. Of interest, mean IgG antibody titers were significantly higher in the acute phase serum samples compared with convalescent sera. Moreover, none of the patients with B. pertussis had a fourfold increase in IgG titers, which is typically used as the diagnostic standard for many infectious diseases. A survey of family members of patients who tested positive for pertussis revealed a greater incidence of cough among these persons compared with those who had negative PCR or serologic tests.

In their discussion, the authors conclude that, as noted in several other studies, B. pertussis infection is a fairly common cause of chronic cough in adults. The best way to diagnose acute infection appears to be with PCR testing of a nasopharyngeal specimen. Serology is still valuable but is dependent on the cutoff values that are used in quantifying the IgG titers to determine if the infection is recent or from past exposure. The study participants did not have a significant rise in convalescent antibody titers; however, they had been symptomatic for an average of six weeks, and peak titers may have actually been the acute serum values. By the same reasoning, a larger number than the 5.5 percent who had positive PCR testing might have been diagnosed if they had been seen at an earlier stage of their illness. Physicians should consider infection with B. pertussis in an adult patient who presents with acute cough. Diagnostic testing may be warranted if the patient may be exposed to susceptible infants. Antibiotic treatment was not addressed in this study.