The outcome of hospitalized patients with community-acquired pneumonia in Hospital Universiti Sains Malaysia
Loading...
Date
2006
Authors
Abdullah, Shaharudin
Journal Title
Journal ISSN
Volume Title
Publisher
Pusat Pengajian Sains Perubatan, Universiti Sains Malaysia
Abstract
Background:
Community-acquired pneumonia (CAP) remains an important cause of hospital
admission. Studies have shown that the mortality of patients admitted with
community-acquired pneumonia is high. The outcomes have been shown to be
influenced by various clinical variables at presentation. Applications of these
variables as predictor of severity have been shown to improve patients
management outcome. The outcome of patients admitted due to CAP has not
been studied in Hospital University Science Malaysia (HUSM). The main purpose
of this study was to evaluate outcome in patients who required admission due to
community acquired pneumonia in HUSM and to determine factors that
influenced their poor outcome.
Methodology:
This was a retrospective cohort study between January 2004 to December 2004.
Records of patients with community-acquired pneumonia admitted to HUSM
were screened. This study included all patients aged more than 12 years old who
met the inclusion criteria. The following information; demographic data, initial
clinical findings, laboratory investigations and type of antibiotics regime given
were recorded into customized data collection sheet. Variables obtained were
examined for association with mortality. Severity prediction criteria were
formulated from identified variables that showed significant association with
mortality.
Results:
Records of 155 patients' that met the inclusion criteria were evaluated. The mean
age at presentation was 62 ± 17 years. The mortality rate was 19.4%. Variables
that significantly influenced the mortality on multivariate analysis at presentation
were presence of important co-morbid illnesses (OR 11.13; p = 0.001 ), confusion
(OR 18.72; p = 0.001) and hypoxaemia (OR 10.62; p = 0.002). Other factors
identified were low diastolic blood pressure and random blood sugar greater than
13 mmol/1 with odds ratio of 1.08 (p=0.002) and 6.37 (p=0.007) respectively.
The presence of any three of following variables on admission; presence of comorbid
illness, confusion, low diastolic blood pressure of s 60 mmHg, low oxygen
saturation and random blood sugar equal or greater than 13 mmol/1 was
associated with a 46.3 fold increase in death.
The suggested predictive severity rule identified 21 of the 27 patients who died
as having severe community acquired pneumonia. The sensitivity of the
suggested severity model for predicting death was 0. 70 and specificity of 0.95.
The rule had a negative predictive value of 0.93.
Conclusion:
The mortality from community-acquired pneumonia requiring hospitalization in
our centre is high compared to previous studies. We found that certain factors
that influenced the outcome of our patients were almost similar with other
previous studies. We found that the presences of three of the five variables (comorbid
illness, confusion, low diastolic blood pressure, hypoxaemia and
hyperglycaemia) would allow us to detect patients who at risk of poor outcome.
Description
Keywords
Pneumonia