Gene Xpert Diagnosis of TB Using Tracheal Aspirates From Intensive Care Unit (ICU) Patients

• Rates of TB treatment initiation [ Time Frame: 48 hours ] [ Designated as safety issue: No ]
  The proportion of patients on TB treatment in each arm will be compared longitudinally
  
  

• Mortality in ICU [ Time Frame: up to 2 months ] [ Designated as safety issue: No ]
  The time-specific proportion of patients who have died prior to discharge will be compared across arms
  
  
• Duration of mechanical ventilation [ Time Frame: Up to 2 weeks ] [ Designated as safety issue: No ]
  The median Duration of mechanical ventilation will be compared between arms prior to discharge up to 2 weeks
  
  
• ICU length of stay [ Time Frame: up to 3 months ] [ Designated as safety issue: No ]
  The median ICU length of stay will be compared up to 3 months
  
  

Background Tuberculosis (TB) is a serious global public health emergency and there is a great need for the improvement of current diagnostic tools and the evaluation of newer and novel technologies for the rapid diagnosis of TB. The most common testing method for TB is sputum smear microscopy, which has remained largely unchanged in its sophistication and sensitivity. Smear microscopy detects around 50% of all active TB cases and is not capable of identifying drug resistance. Furthermore, even though culture is more accurate results are only available in several weeks. Treating patients empirically for TB exposes patients to the toxic side-effects of drugs, and if there is no response to treatment, raises the question of MDR-TB. These considerations apply to clinical decision-making in the ICU where TB is not infrequently encountered. A diagnosis of TB in this setting has important implications for drug-drug interactions, route of administration, and drug toxicity.

The Gene Xpert® system's MTB/RIF Assay, an on-demand molecular test for simultaneous detection of Mycobacterium tuberculosis (M.tb) and rifampicin (RIF) resistance, has recently been released. This is the only system that combines on-board sample preparation with real-time PCR [1]. Helb and colleagues developed and performed the first analysis of the Cepheid Gene Xpert System's MTB/RIF assay. They evaluated the diagnostic utility of the system using sputum samples and showed that the system was highly sensitive and simple-to-use [2]. A recent paper (NEJM, 2010, in press) that enrolled almost 2000 patients found Xpert to have a sensitivity of almost 90% in smear-negative TB. However, how the test will perform when using tracheal aspirates is unknown. It is also unclear whether the test will have any impact in the ICU where the decision to treat is often made clinically and empiric treatment is initiated. To address these questions we aim to obtain tracheal aspirates from about 120 mechanically ventilated patients with suspected TB from the Groote Schuur Hospital Intensive Care Unit. Tests will be conducted at the Lung Infection and Immunity Unit laboratory (Groote Schuur Hospital, Cape Town) to evaluate the diagnostic utility of the Xpert® MTB/RIF Assay in this group of patients. The expected study period will be 12 months.