Prabda Prapasiri. Epidemiology and etiology of radiographycally confirmed community acquired pneumonia in Nakhon Phanom province Thailand; population based active surveillance. Doctoral Degree(Epidemiology). Mahidol University. : Mahidol University, 2007.
Epidemiology and etiology of radiographycally confirmed community acquired pneumonia in Nakhon Phanom province Thailand; population based active surveillance
Abstract:
Community-acquired pneumonia (CAP) is a major cause of morbidity and
mortality but little is known about its incidence and etiology in rural Thailand. Better
understanding of factors associated with an increased risk of mortality could help guide
clinician and public health decisions. An active, population-based pneumonia surveillance
system captures all hospitalized patients in Nakhon Phanom Province with evidence of
acute infection and respiratory signs and symptoms. Local clinicians order and interpret
chest radiographs within 48 hours of admission as well as routine hospital laboratory
tests. Blood is cultured on a Bac T/Alert 3D automated system and pathogens are
confirmed by standard methods. Trained surveillance staffs collect data using a standard
form and data were analyzed from a period of 3 years 6 months, August 2003-December
2006. Incidence, etiology, mortality and risk factors of hospitalized CAP patients with
chest radiograph confirmation are described.
Of 22,144 cases of clinical CAP, 10,706 (48.3%) had a chest radiograph, and
5,604 (52.3%) were radiographically-confirmed. Incidence of radiographically CAP was
179-269 per 100,000 persons per year; 186 (3.3%) patients died. Mortality was 4.9-8.5
per 100,000 persons and high in children aged 0-1 years (20-47/100,000) and adults aged
>65 (37.2-63.8/100,000). High fever and tachypnea were the predominant vital signs that
brought patients to the hospitals. Bacterial pathogens identified were B. pseudomallei in
6.3%-13.6%, E. coli in 7.7%-12.1%, S. pneumoniae in 4.9%-9.8%, K. pneumoniae in
1.2%-4.9% and S. aureus in 2.3%-2.8% and viral pathogens identified were RSV in 351,
46.9%, Flu A in 16.2%, HPIV3 in 10%, adenovirus in 6%, HMPV in 5.6% and Flu B in
38, 5.1%. In multiple logistic regression models, eight factors significantly associated
with death were admission to an ICU (RR=18.28, 95%CI=8.94-37.41), leukopenia
(RR=14.54, 95%CI=3.07-68.83), hemoptysis (RR=3.21, 95%CI=1.11-9.26), cavitations
(RR=6.53, 95%CI=2.3-18.47), dyspnea (RR=6.53, 95%CI=2.31-18.47) and bacteremia
(RR=2.59, 95%CI=1.09-6.16). Patients aged 5-19 years were 8.53 times (95% CI=1.98-
36.76), aged 20-39 were 8.42 times (95%CI=2.32-30.51), aged 40-65 were 5.65 times
(95%CI=2.07-15.38), and aged >65 were 7.93 times (95%CI=2.82-19.43), more likely to
die than patients aged 0-4 years. CAP incidence highly presented in young children and
the elderly. The evidence showed high mortality particulary in the elderly. B.
pseudomomallei was the dominant bacterial causing CAP in the hospitalized patients and
frequently found in the northeastern part of Thailand.