Dr LAU Chun Wing Arthur
Associate Consultant, Department of Intensive Care; Honorary Associate Consultant, Division of Respiratory and Critical Care Medicine, Department of Medicine, Pamela Youde Nethersole Eastern Hospital
Presented on 30 March 2008 at the Annual Scientific Meeting of the Hong Kong Thoracic Society
Tuberculosis (TB) is mostly a chronic wasting disease for which intensive care unit (ICU) admission is uncommon. We have reviewed 50 (mean age 62, 84% male) microbiologically and/or histologically proven pulmonary tuberculosis patients admitted to our Medical High Dependency Unit or ICU from 1996 – 2007, and performed a literature review.
Chief admission reasons were acute respiratory failure (78%) and shock (12%). Oriental series showed less chronic alcoholism and HIV positivity. Diagnosis mainly depended on AFB smear in respiratory specimens. Management issues include: renal and hepatic impairment, gastrointestinal dysfunction, inability to monitor vision in sedated patients, clotting dysfunction precluding invasive diagnosis and intramuscular injection, drug interactions, and paradoxical reaction being more common in disseminated TB. Clear benefits of corticosteroid in “sick” TB patients are not established, except in pre-existing or TB-related adrenal failure receiving rifampicin. For respiratory failure, benefits of corticosteroid in acute respiratory distress syndrome in general is not substantiated by the 2006 NHLBI ARDS Clinical Trials Network Study, in which corticosteroid in persistent ARDS after onset for 2 weeks actually increased the risk of death. For shock, the recent 2008 CORTICUS study of hydrocortisone in septic shock showed that corticosteroid did not improve survival, though hastened shock reversal. Cut-off values for intervention in both absolute (low total cortisol level due to hypoalbuminaemia but actually normal free cortisol response in stressed condition) and relative adrenal insufficiency have not been clearly defined. The APACHE II Standardized Mortality Ratios (SMR) of TB-ICU patients ranged from 0.49 – 2.64 (1.37 in the present series), which were higher than the 0.6 – 0.8 in Hong Kong ICU patients overall. Mortality in TB-associated vs ARDS overall were 41.4 – 88.9% vs 30 – 60% respectively. We look forward to greater collaboration between TB experts and critical care physicians in clinical management and research.