981. Tuberculosis and Brucellosis Meningoencephalitis: the Debate Continues Beyond Scoring Systems
Session: Poster Abstract Session: CNS Infection
Friday, October 9, 2015
Room: Poster Hall
  • P 981.pdf (61.7 kB)
  • Background: Tuberculous (TME) and Brucella meningoencephalitis (BME) display a great challenge because of similar clinical features and difficulty in microbiologic techniques to confirm diagnosis. The Lancet and Thwaites scoring systems may contribute to strengthen the diagnosis of TME. We aim to define how useful are those scores in BME.

    Methods: Forty seven patients hospitalized in the department of Infectious Diseases between 1995 and 2013 were retrospectively enrolled and divided into 2 groups: G1 (15 patients with BME) and G2 (32 patients with TME). We calculated the Lancet and Thwaites scores. When the score of Lancet is ≥ 12, the TME diagnosis is assigned as ‘probable’ and when between 6 and 11 as ‘possible’. If Lancet score is less than 5, TME probability is ‘negative’. When Thwaites score is ≤ 4, the diagnosis was TME and if it is greater than 4, it was labeled as bacterial meningitis. Confirmed TME with isolation of Mycobacterium tuberculosis (MT) in cerebrospinal fluid (CSF) was excluded.  

    Results: The mean age was 38.2±17 years in G1 and 36.9±16 years in G2 (p=0.8). Culture and PCR technique from a site other than CSF was revealed MT in 15.6% and 3.1% respectively. The Thwaites score was -3.13±2.4 in G1 and -0.94±3 in G2 (p=0.02). The Lancet score was significantly greater in G2 (5.27±1.38 vs. 12.48±2.8; p<0.0001). The Thwaites score was less than 4 in 14 cases among G1 (93.4%) and in 24 cases among G2 (75%). According to the lancet scoring system, TME was classified into ‘probable’ category in 20 cases (62.5%) which was more frequently than G1 (0%; p<0.0001). Negative category included 7 cases from G1 (46.7%) and no one from G2 (p<0.0001). BME was listed more frequently in ‘possible’ category in 8 cases (53.3%) against 12 cases with TME (37.5%; p<0.0001). The area under receiver operating characteristic curve (AUROC) for predicting BME was 0.98 with the Lancet score. When Lancet score was less than 8.5, we found a sensitivity of 100% and specificity of 91% to predict BME. The same threshold was 90% sensitive and 100% specific to predict TME.

    Conclusion: The Thwaites and Lancet scoring system are insufficient to discriminate between TME and BME. This latter disease should be kept in mind when dealing with neurological infections especially in the brucellosis endemic countries.

    Houda Ben Ayed, MD1, Makram Koubaa, MD1, Sahar Ben Kahla, MD1, Chakib Marrakchi, MD1, Hela Fourati, MD2, Imed Maaloul, MD1, Zeinab Mnif, MD2 and Mounir Ben Jemaa, MD1, (1)Department of Infectious Diseases, Hedi Chaker University Hospital, Sfax, Tunisia, (2)Department of Radiology, Hedi Chaker University Hospital, Sfax, Tunisia


    H. Ben Ayed, None

    M. Koubaa, None

    S. Ben Kahla, None

    C. Marrakchi, None

    H. Fourati, None

    I. Maaloul, None

    Z. Mnif, None

    M. Ben Jemaa, None

    See more of: CNS Infection
    See more of: Poster Abstract Session

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 7th with the exception of research findings presented at the IDWeek press conferences.