Search

BACKGROUND: Sputum smear microscopy is the most useful tool used for the screening of presumptive tuberculosis cases. Conventionally Ziehl-Neelsen (ZN) based staining is used to microscopically screen sputum from the presumptive Tuberculosis Cases. However, Fluorescent microscopy (FM) is now gradually replacing the conventional ZN microscopy because of its higher efficiency and efficacy in detecting Mycobacterium tuberculosis. In patients, who do not produce sufficiently high number of bacilli, and are known as Smear-negative, efficacy of both types of microscopy remains questionable as no Mycobacteria are detected in the sputum of these patients and they are falsely labelled as smear-negative. OBJECTIVE: This retrospective study was conducted on the patient samples received from different parts of Sindh to study the proportions of false-negative tuberculosis using fluorescent microscopy. MATERIALS AND METHODS: A retrospective cross-sectional study was conducted from June 2014 to June 2016 at the Tuberculosis Laboratories of Dow University of Health Sciences. Data analyzed for 1439 samples received during 24-month period, which had subsequently been confirmed by culture. All laboratory work was performed as per guidelines of the National Program. Statistical analysis was performed on MedCalc® and Open Epi®. RESULTS: During 24-month period, among the 1439 cases, 832(57.81%) were from males and 499 (34.67%) were smear negative. Culture showed that 72 (14.42%) were false-negative among all-negatives. Overall sensitivity observed was 92.21% [95%CI 90.29%-93.85%] and specificity 82.91% [95%CI 79.38%-86.06%]. The proportion of false negativity among all negatives in females (16.11%) was significantly higher (p=0.002) as compared to males (13.47%). Conclusions: Even with the introduction of Fluorescent Microscopy, smear microscopy cannot be solely relied upon for the diagnosis of tuberculosis, as considerable number of patients producing insufficient number of bacilli to be detected by microscopy. [ABSTRACT FROM AUTHOR]

Entropy is a measure of uncertainty and often used in information theory to determine the precise testimonials about unclear situations. Different entropy measures available in the literature are based on the exact form of the observations and lacks in dealing with the interval-valued data. The interval-valued data often arises from the situations having ambiguity, imprecise, unclear, indefinite, or vague states of the experiment and is called neutrosophic data. In this research modified forms of different entropy measures for normal probability distribution have been proposed by considering the neutrosophic form data. The performance of the proposed neutrosophic entropies for normal distribution has been assessed via a simulation study. Moreover, the proposed measures are also applied to two real data sets for their wide applicability. The results of the study suggested the use of the proposed methods in the presence of fuzzy, interval-valued, or neutrosophic data. [ABSTRACT FROM AUTHOR]