Clinical utility and cost effectiveness of CSF TB BACTEC MGIT cultures in the evaluation of HIV Patients with meningoencephalitis at the Kenyatta National Hospital
Background: The spread of HIV epidemic has contributed immensely to an increase in the annual incidence of Tuberculosis (TB) and especially of extra-pulmonary TB in Kenya.' The diagnosis of Tuberculous Meningitis (TBM) is difficult, usually empirical,delayed and is associated with high mortality, which can be reduced by initiation of therapy in the earlier stages of disease.Traditional solid culture methods have a sensitivity of 50 to 70% and take 4 to 8 weeks to yield results. Liquid BACTEC culture media have a sensitivity of 95% and take between 48hrs and 21 days with an average time of 7 - 14 days to give a result. Though costly, they potentially would offer an advantage of early diagnosis, treatment and positive impact on, morbidity and mortality from TBM. Objective: To determine the clinical utility and cost effectiveness of early CSF BACTEC cultures among HIV positive patients who present with features of meningoencephalitis at K.N.H. Study design: Hospital based comparative In-patient cohort study. Methodolozy: Patients admitted with a diagnosis of meningoencephalitis were screened for HIV on the morning following admission. Those who tested positive and met inclusion criteria were recruited consecutively and randomized to either BACTEC or Usual Care arms. All patients received the usual care for meningitis as deemed necessary by their primary care physicians. Those in the BACTEC arm had, in addition to their usual workup, a Mycobacterial Growth Indicator Tube (MGIT) BACTEC culture done on their CSF for TBM, on the morning immediate post admission. All patients were followed up for a period or 4 weeks in hospital stay or up to death or discharge whichever was sooner. Time to definitive diagnosis of TBM by positive culture result, time to initiation of anti- TB therapy and time to discharge were studied. Initiation of anti- TB was at the discretion of the clinical care team and was not study directed. Results: A total of 219 patients were screened. Of these, 142 were included into the study, 89 in the BACTEC arm and 53 in the usual care arm. Male patients were 70 (49.3%). The mean age was 37.2 years, range 13 and 62 years. Of those in the BACTEC arm, 9 (10%) had a positive CSF culture at 21 days and 14 (15.7%) at 8 weeks. The average time to culture positivity was 12 days, ranging between 5 and 40 days. Time to starting anti- TB therapy ranged from 1 to 22 days, with a mean of 4.99 days. The decision to start anti- TB was made empirically in all the patients, at the discretion of the clinical care team and was not influenced by the findings of the study. The average duration of hospitalization for discharged patients was 11.92 days, 10.56 and 14.48 in the BACTEC and usual care arms respectively, with a statistically significant difference in duration of hospital stay between the two groups; p=O.OOI. Total hospital mortality was 65 (45.8%). There was no statistically significant difference in mortality between the BACTEC and usual care arms. Average time to death was 7.5 days from admission, 6.93 in the BACTEC arm and 8.9 in the usual care arm, p=O.1 08. Conclusion: Among the HIV positive patients with meningoencephalitis, the prevalence of TBM by BACTEC culture was found to be 15.7%. There is no statistically significant difference in time to diagnosis, duration of hospital stay, mortality and morbidity between the BACTEC and usual care arms. It is therefore not cost effective to perform a BACTEC culture in these patients, for the purpose of early diagnosis and treatment of TBM.