Z Ahmadinejad, Sh Phyroosbakhsh, Z.n Hatmy, B Bagherian, H Sabery, M Bahador, M Nikzad, M Jamali Zavare, A Hadady, M Hajiabdolbaghi, M Mohraz, M. Rasolinejad, A Soudbakhsh, A Yalda,
Volume 64, Issue 2 (30 2006)
Abstract
Background and Aim: Tuberculous pleural effusion occurs in 30% of patients with tuberculosis (TB). Rapid diagnosis of a tuberculous pleural effusion would greatly facilitate the management of many patients. The purpose of this study was to determine sensitivity, specificity, and predictive values of clinical, laboratory, radiographic findings in patients with tuberculous pleural effusion.
Materials and Methods: The cross sectional study was performed between august 2002 and March 2004 at a referral teaching hospital. Major clinical, laboratory, and radiographic findings were evaluated in 88 cases of pleural effusion, 33 with confirmed TB pleural effusion (TBPE) and 55 with a diagnosis other than TB (NTBPE).
Results: The sensitivity of culture of pleural effusion and tissue were 3% and 9.1% respectively. The mean of adenosine deaminase (ADA) values in TBPE was 36.7 U/L (±18.72), and the mean in the NTBPE was 28.2 U/L (±17.0). Both the sensitivity and specificity of ADA estimation in diagnosing tuberculosis were 55%. The sensitivity of PCR was 3% with specificity of 12.7% (positive predictive value, 50% negative predictive value, 70%). Younger age (p<0.024), positive history of exposure to TB patient (p<0.02), and the combination of fever, weight loss and sweating (p<0.01), were associated with tuberculous pleural effusion. There were also significant association between Positive sputum smear (p<0.001), positive sputum culture (p<0.006), positive pleural biopsy (p<0.001), pleural LDH>200 (p<0.005), pleural lymphocytes>50% (p<0.015) and TBPE.
Conclusions: In our region with a high incidence of tuberculosis, the most frequent cause of exudative pleural effusion is tuberculosis. We suggest that the diagnostic planning of pleural effusion should be determined in each region with a view to the adoption of regionally optimized diagnostic and therapeutic facilities.
Hadadi A, Rasoulinejad M, Maleki Z, Mojtahedzadeh M, Younesian M, Ahmadi S.a, Bagherian H,
Volume 65, Issue 4 (3 2007)
Abstract
Background: The object of this study was to investigate the antimicrobial resistance pattern among common nosocomial Gram-negative bacilli isolated from patients with nosocomial infections.
Methods: From June 2004 to December 2005, 380 isolates of common Gram-negative bacilli (Klebsiella, Pseudomonas, Acinetobacter and E. coli) from 270 patients with nosocomial infections in Sina and Imam Hospitals, Tehran, Iran, were evaluated for susceptibility to Imipenem, Cefepime, Ciprofloxacine, Ceftriaxone and Ceftazidime by Disc diffusion and E-test methods.
Results: The most frequent pathogens isolated were Klebsiella spp. (40%), followed by Pseudomonas (28%), Acinetobacter spp. (20%) and E. coli (12%). The most active antibiotic was imipenem (84%). 26% of all isolates were sensitive to Cefepime, 26% to Ciprofloxacin, 20% to Ceftazidime and 10% to Ceftrixone. The susceptibility rates of Klebsiella to Imipenem, cefepime, ciprofloxacin, Ceftazidime and Ceftriaxone were 91, 25, 21, 13 and 7 percent, respectively and 91, 19, 17, 21 and 21 percent, respectively, for E. coli. Among Acineto- bacter spp., the susceptibility rate was 77% for Imipenem and 21% for Ciprofloxacin. Among Pseudomonas spp., 75% of isolates were susceptible to Imipenem and 39% to Ciprofloxacin. The comparison of the resistance status of microorganisms by both Disc diffusion and E-test methods showed a clinically noticeable agreement between these two tests.
Conclusions: Since antibiotic resistance among Gram-negative bacilli has increased, enforcement of policy regarding proper antibiotic use is urgently needed in order to delay the development of resistance. Although it is widely accepted that E-test is more accurate in determining the resistance of microorganisms, our study showed that the Disc diffusion test will give the same results in most occasions and is therefore still considered useful in clinical practice.
Hossein Bagherian, Shaghayegh Haghjooy Javanmard, Mehran Sharifi, Mohammad Sattari,
Volume 79, Issue 3 (june 2021)
Abstract
This review was conducted between December 2018 and March 2019 at Isfahan University of Medical Sciences. A review of various studies revealed what data mining techniques to predict the probability of survival, what risk factors for these predictions, what criteria for evaluating data mining techniques, and finally what data sources for it have been used to predict the survival of breast cancer patients. This review is based on the Prism statement consisting of published studies in the field of predicting the survival of breast cancer patients using data mining techniques from 2005 to 2018 in databases such as Medline, Science Direct, Web of Science, Embase data and Scopus. After searching in these databases, 527 articles were retrieved. After removing duplicates and evaluating the articles, 21 articles were used. The three techniques of logistic regression, decision tree, and support vector machine have been most used in articles. Age, tumor grade, tumor stage, and tumor size are used more than other risk factors. Among the criteria, the accuracy criterion was used in more studies. Most of the studies used the Surveillance, Epidemiology, and End Results Program (SEER) dataset. Typically, in the field of survival probability prediction, data mining techniques in the field of classification are given more attention due to their adaptation to this field. Accordingly, data mining techniques such as decision tree techniques, logistic regression, and support vector machine were used in more studies than other techniques. The use of these techniques can provide a good basis for clinicians to evaluate the effectiveness of different treatments and the impact of each of these methods on patients' longevity and survival. If the output of these techniques is used to provide the data input required by a decision support system, clinicians can provide risk factors related to the patient, the patient's age, and the patient's physical condition when providing services to breast cancer patients. Through the outputs provided by the decision support system, they provided the most optimal decision to choose the best treatment method and consequently increase patient survival.
