Me. Counihan et al., Human parainfluenza virus-associated hospitalizations among children less than five years of age in the United States, PEDIAT INF, 20(7), 2001, pp. 646-653
Background. Human parainfluenza viruses 1 through 3 (HPIV-1-3) are importan
t causes of respiratory tract infections in young children. This study soug
ht to provide current estimates of HPIV-1-3-associated hospitalizations amo
ng US children.
Methods, Hospitalizations for bronchiolitis, bronchitis, croup and pneumoni
a among children age <5 years were determined for the years 1979 through 19
97 using the National Hospital Discharge Survey. Average annual hospitaliza
tions during the last 4 years of the study for each of these four diseases
were multiplied by the proportions of each disease associated with HPIV-1-3
infection (as previously reported in hospital-based studies) to estimate h
ospitalizations potentially associated with HPIV-1-3 infections. Seasonal t
rends in HPIV-1-3-associated hospitalizations were compared with HPIV detec
tions in the National Respiratory and Enteric Virus Surveillance System, wh
ich prospectively monitors respiratory viral detections throughout the Unit
ed States.
Results. The proportions of hospitalizations associated with HPIV infection
for each disease varied widely in the 6 hospital-based studies we selected
. Consequently our annual estimated rates of hospitalization were broad: HP
IV-1, 0.32 to 1.59 per 1000 children; HPIV-2, 0.10 to 0.86 per 1000 childre
n; and HPIV-3, 0.48 to 2.6 per 1000 children. Based on these data HPIV-1 ma
y account for 5800 to 28 900 annual hospitalizations; HPIV-2 for 1800 to 15
600 hospitalizations; and HPIV-3 for 8700 to 52 000 hospitalizations.
Conclusions, We provide broad, serotype-specific estimates of US childhood
hospitalizations associated with HPIV infections. More precise estimates of
HPIV-associated hospitalizations would require large prospective studies o
f HPIV-associated diseases by more sensitive viral testing methods, such as
polymerase chain reaction techniques.