1 figure omitted
Pertussis (i.e., whooping cough) is a prolonged cough illness caused
by the bacteria Bordetella pertussis and associated
typically with an inspiratory "whoop," paroxysmal cough, and posttussive vomiting. B. pertussis can cause severe illness or death, especially
in infants who have not completed their pertussis vaccination series. Adolescents
(i.e., persons aged 13-17 years), adults, and recently vaccinated persons
often report atypical symptoms, resulting in delay of recognition and creation
of infectious reservoirs for further transmission. In 2003, the Minnesota
Department of Health (MDH) investigated a fatal case of unsuspected B. pertussis infection in an elderly adult. This report
summarizes the case investigation, which documents the rare isolation of B. pertussis from blood and underscores the need for clinicians
to consider pertussis infection in adolescents and adults who have a prolonged
In February 2003, a woman aged 82 years on immunosuppressive medications
for multiple myeloma was admitted to a local hospital for control of pain
from vertebral compression fractures. A chest radiograph revealed a nodular
infiltrate, which was thought to be a residual finding from pneumonia diagnosed
in early January. Two days after hospitalization, the patient had a cough;
rales were observed on lung examination. She remained afebrile and was discharged
to a nursing care facility. The patient's cough worsened; she had a fever
of 102.2°F (39°C) and labored respirations that required rehospitalization
and mechanical ventilation. On readmission, a chest radiograph revealed diffuse
left-lung infiltrates. She was placed on multiple nonmacrolide broad spectrum
antibiotics but had hypotension and a sepsis-like syndrome. Respiratory support
was withdrawn, and the patient died.
During her rehospitalization, tests for Legionella spp. and influenza A and B were negative. Gram stain of tracheal secretions
showed gram-negative bacilli and white blood cells; the culture was negative.
Blood cultures revealed gram-negative bacilli, but further identification
of the pathogen by routine culture media was unsuccessful. B. pertussis was identified after a special culture medium containing
charcoal and sheep blood was used. Because infection with B. pertussis was not suspected originally, no nasopharyngeal (NP) specimen
was obtained for B. pertussis isolation.
The patient had lived with her daughter, a high-school nurse, who reported
having an intermittent, nonproductive cough for approximately 1 month preceding
her mother's illness. Her cough changed to a pertussis-like, paroxysmal cough
approximately the same time as her mother's cough onset. By the time the patient's B. pertussis results were known, the daughter was asymptomatic
and therefore was not tested. She reported no contact with students with pertussis-like
A contact investigation identified 47 persons who were exposed to the
index patient. Exposure was defined as ≥10 hours per week of close (i.e.,
<3 feet or "arms length") contact with the patient while she was symptomatic
or direct face-to-face contact during an episode of coughing or sneezing,
regardless of duration. NP specimens were obtained from exposed persons if
they reported having a cough illness ≤20 days after their last exposure
to the patient. Two (12%) of 17 exposed family members reported cough illness
and had NP specimens tested; one tested positive for B.
pertussis by culture and polymerase chain reaction (PCR). Ten (34%)
of 29 exposed nursing home staff reported cough illness and were tested; one
(10%) person tested positive by culture and PCR. Hospital personnel were asked
if they had exposure consistent with the definition; one physician reported
mild upper-respiratory symptoms and was tested, but his NP specimen was lost.
The epidemiologic link between the patient and both infected contacts was
confirmed by pulsed-field gel electrophoresis (PFGE) analysis of isolates
at MDH and CDC.
C Kenyon, MPH, C Miller, MS, K Ehresmann, MPH, Immunization, Tuberculosis,
and International Health Section; D Boxrud, MS, Public Health Laboratory,
Minnesota Dept of Health. P Cassiday, MS, GN Sanden, PhD, Div of Bacterial
and Mycotic Diseases, National Center for Infectious Diseases; KM Bisgard,
DVM, National Immunization Program; K Kiang, MD, EIS Officer, CDC.
The case described in this report is an example of a fatal case of unsuspected B. pertussis infection in an adult and the rare occurrence
of B. pertussis isolated from blood. This is the
second reported case of B. pertussis isolated from
blood; this organism does not generally invade the bloodstream. B. pertussis also is difficult to recover through routine culture including
NP specimens because of growth inhibitors encountered in standard culture
medium. A special culture medium that contains activated charcoal (e.g., Regan-Lowe
agar) or potato-derived starch (i.e., Bordet-Gengou agar) and defibrinated
horse or sheep blood to neutralize the inhibitory substances must be used
to isolate B. pertussis.1 Testing is not
performed for B. pertussis unless a specific request
Molecular epidemiologic techniques (e.g., PFGE) have enhanced surveillance
for pertussis by helping to identify infection, track transmission in outbreaks,
and describe geographic and temporal trends. In this investigation, the epidemiologic
link between the index patient and contacts was confirmed by PFGE analysis
of bacterial isolates.
Adolescents and adults account for a substantial proportion of pertussis
cases.2-4 In Minnesota during 1997-2000, adolescents accounted
for 15% of cases reported annually (18.3 per 100,000 population), and adults
accounted for 23% of cases (3.1 per 100,000 population) (MDH, unpublished
data, 1997-2000). The incidence of pertussis reported in adolescents and adults
has increased markedly in Minnesota and throughout the United States. This
increase might be attributable in part to heightened awareness and improved
detection of pertussis, with the introduction of PCR as a diagnostic tool.
Adolescents and adults who have pertussis are potential sources of infection
for infants, who are most at risk for severe illness and death.2,5-7 During 1997-2001, three pertussis-associated deaths were reported in
Minnesota; all occurred among infants aged <2 months. When a source of
infection for infants is identified, household contacts are the most common
source.2,7 In Minnesota during 1998-2001, the probable source of
exposure to pertussis was determined in 50% of cases in infants aged <1
year; 67% of the source-patients were either adolescents (6%) or adults (61%).
Because B. pertussis infection is a common cause
of cough illness among adolescents and adults, heightened clinical suspicion
for pertussis and appropriate testing of these persons is warranted.
This report is based on data contributed by V Miller, C Nassif, Mayo
Clinic, Rochester; L Rahn, Chatfield, Minnesota.
References: 7 available
Fatal Case of Unsuspected Pertussis Diagnosed From a Blood Culture—Minnesota, 2003. JAMA. 2004;291(13):1557–1558. doi:10.1001/jama.291.13.1557