METHODS: Feature selection, model training, and validation were performed using patient-level data from 12 randomized controlled trials. A gradient-boosted machine (GBM) model was trained to estimate PEP risk, and the performance of the resulting model was evaluated using the area under the receiver operating curve (AUC) with 5-fold cross-validation. A web-based clinical decision-making tool was created, and a prospective pilot study was performed using data from ERCPs performed at the Johns Hopkins Hospital over a 1-month period.
RESULTS: A total of 7389 patients were included in the GBM with an 8.6% rate of PEP. The model was trained on 20 PEP risk factors and 5 prophylactic interventions (rectal nonsteroidal anti-inflammatory drugs [NSAIDs], aggressive hydration, combined rectal NSAIDs and aggressive hydration, pancreatic duct stenting, and combined rectal NSAIDs and pancreatic duct stenting). The resulting GBM model had an AUC of 0.70 (65% specificity, 65% sensitivity, 95% negative predictive value, and 15% positive predictive value). A total of 135 patients were included in the prospective pilot study, resulting in an AUC of 0.74.
CONCLUSIONS: This study demonstrates the feasibility and utility of a novel machine learning-based PEP risk estimation tool with high negative predictive value to aid in prophylaxis selection and identify patients at low risk who may not require extended postprocedure monitoring.
Methods: This cross-sectional study, which was carried out at the Paediatric Intensive Care Unit of Hospital Universiti Sains Malaysia (HUSM) in Kelantan, Malaysia, had involved 60 neonates admitted for suspected sepsis. Sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV) and the area under receiver operating characteristics curve (AUC) for PCT were determined at initial presentation (0 h) as well as 12 h and 24 h after presentation in comparison to blood culture as the gold standard.
Results: The study consisted of 27 (45.0%) male and 33 (55.0%) female neonates with a mean (SD) age of 76.8 (48.25) h. At cut-off PCT value of > 2 ng/mL, the sensitivity, specificity, PPV and NPV were 66.7%, 66.7%, 33.3% and 88.9% at 0 h. The respective parameters were 83.3%. 56.3%, 32.3% and 93.1% at 12 h and 83.3%, 52.1%, 30.3% and 92.6% at 24 h. AUC was 71.6%, 76.6% and 71.7% at 0 h, 12 h and 24 h.
Conclusions: Diagnostic performance and discrimination values of PCT for diagnosis of neonatal sepsis varied with time of obtaining the blood samples. The PCT result at 12 h demonstrates the most optimal diagnostic performance and discrimination values.
Methods: One hundred and fifty hard-working agricultural farmers from high-prevalence area for CKDu (Madawachchiya) were screened three times for proteinuria; 66 proteinuric and 21 non-proteinuric were identified as the baseline classification. Selected individuals were analysed further for creatinine, protein and cystatin C in urine and creatinine, cystatin C in serum. Urine protein-to-creatinine ratio (UP/UC) was calculated.
Results: Based on creatinine and cystatin C cut-off levels in serum, individuals were classified as high or normal. Diagnosis of two functional markers (creatinine and cystatin C) were evaluated using receiver operating characteristic (ROC) curve and in terms of sensitivity and specificity using UP/UC as the baseline. Creatinine and cystatin C-based eGFR (estimated Glomerular filtration rate) levels were calculated, and Pearson's correlation coefficient was determined between different eGFR measurements using UP/UC. Mean (SD) UP/UC ratio, serum creatinine, and serum cystatin C levels of the proteinuric subjects were 129.0 (18.4) mg/mmol, 1.35 (0.39) mg/dL, 1.69 (0.58) mg/L. For non-proteniuric individuals, the results were found to be 14.4 (2.28), 1.22 (0.40) mg/dL, 0.82 (0.25) mg/L. The ROC analysis showed excellent accuracy in using cystatin C for identifying proteinuric patients than creatinine area under the curve (AUC): 0.9675, P < 0.001). Cut-off points were identified as 1.015 mg/dL for serum creatinine and 0.930mg/L for cystatin C. Furthermore, cystatin C based Hoek formula showed the better correlation (0.635, P < 0.001) with UP/UC compared with creatinine based modification of diet in renal disease (MDRD) formula.
Conclusion: The study showed elevated serum cystatin C in patients with persisting proteinuria compared with non-responding serum creatinine. Moreover, cystatin C-based eGFR equations were more accurate to determine the kidney function than serum creatinine in proteinuric patients who are vulnerable for CKDu in high-prevalence areas.
METHODS: The study was divided into two phases: (I) Marker discovery by miRNA microarray using paired cancer tissues (n = 30) and blood samples (CRC, n = 42; control, n = 18). (II) Marker validation by stem-loop reverse transcription real time PCR using an independent set of paired cancer tissues (n = 30) and blood samples (CRC, n = 70; control, n = 32). Correlation analysis was determined by Pearson's test. Logistic regression and receiver operating characteristics curve analyses were applied to obtain diagnostic utility of the miRNAs.
RESULTS: Seven miRNAs (miR-150, miR-193a-3p, miR-23a, miR-23b, miR-338-5p, miR-342-3p and miR-483-3p) have been found to be differentially expressed in both tissue and blood samples. Significant positive correlations were observed in the tissue and blood levels of miR-193a-3p, miR-23a and miR-338-5p. Moreover, increased expressions of these miRNAs were detected in the more advanced stages. MiR-193a-3p, miR-23a and miR-338-5p were demonstrated as a classifier for CRC detection, yielding a receiver operating characteristic curve area of 0.887 (80.0% sensitivity, 84.4% specificity and 83.3% accuracy).
CONCLUSION: Dysregulations in circulating blood miRNAs are reflective of those in colorectal tissues. The triple miRNA classifier of miR-193a-3p, miR-23a and miR-338-5p appears to be a potential blood biomarker for early detection of CRC.
METHODS: In this single-centre retrospective study, comparative analysis on clinical presentations and laboratory findings was performed between confirmed leptospirosis versus non-leptospirosis cases.
RESULTS: In multivariate logistic regression evidenced by a Hosmer-Lemeshow significance value of 0.979 and Nagelkerke R square of 0.426, the predictors of a leptospirosis case are hypocalcemia (calcium <2.10mmol/L), hypochloremia (chloride <98mmol/L), and eosinopenia (absolute eosinophil count <0.040×109/L). The proposed diagnostic scoring model has a discriminatory power with area under the curve (AUC) 0.761 (p<0.001). A score value of 6 reflected a sensitivity of 0.762, specificity of 0.655, a positive predictive value of 0.38, negative predictive value of 0.91, a positive likelihood ratios of 2.21, and a negative likelihood ratios of 0.36.
CONCLUSION: With further validation in clinical settings, implementation of this diagnostic scoring model is helpful to manage presumed leptospirosis especially in the absence of leptospirosis confirmatory tests.
OBJECTIVE: The aim of this study was to assess the diagnostic characteristics of inferior turbinate tissue biopsy sIgE in asymptomatic and rhinitic patients.
METHODS: A diagnostic cross-sectional study was undertaken, involving patients who underwent inferior turbinate surgery with or without other surgical interventions. Inferior turbinate tissue biopsy was performed during surgery and was assessed for allergen sIgE (dust mite, grass [temperate or subtropical], and animal epithelium) using an automated immunoassay. Tissue sIgE was assessed among asymptomatic patients and those with nasal symptoms. Data were presented as median (interquartile range). A receiver operating curve was used to predict the diagnostic utility of turbinate tissue sIgE in determining allergic rhinitis.
RESULTS: A total of 160 patients (41.89 ± 14.65 years, 36.9% females) were included. The median tissue sIgE concentration among the asymptomatic nonatopic group of patients was 0.09 (0.08-0.10) kUA/L and tissue sIgE > 0.10 kUA/L was determined as a positive threshold. Inferior turbinate tissue sIgE was shown to be a predictive test for allergic rhinitis (area under curve: 0.87, 95% confidence interval: 0.84-0.90) with 90% sensitivity and 89% negative predictive value.
CONCLUSION: Inferior turbinate tissue biopsy sIgE is a sensitive tool to predict allergic rhinitis. The threshold value of 0.1 kUA/L corresponded well with the asymptomatic nonatopic group of patients. This method detects sIgE in the nasal mucosa and may be a useful test for allergic rhinitis in future research.
METHODS: A cross-sectional study was conducted among 134 geriatric patients with a mean age of 68.9 ± 8.4 who stayed at acute care wards in Hospital Tuanku Ampuan Rahimah, Klang from July 2017 to August 2017. The SGA, MNA, and GNRI were administered through face-to-face interviews with all the participants who gave their consent. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the GNRI and MNA were analyzed against the SGA. Receiver-operating characteristic (ROC) curve analysis was used to obtain the area under the curve (AUC) and suitable optimal cutoff values for both the GNRI and MNA.
RESULTS: According to the SGA, MNA, and GNRI, 26.9%, 42.5%, and 44.0% of the participants were malnourished, respectively. The sensitivity, specificity, PPV, and NPV for the GNRI were 0.622, 0.977, 0.982, and 0.558, respectively, while those for the MNA were 0.611, 0.909, 0.932, and 0.533, respectively. The AUC of the GNRI was comparable to that of the MNA (0.831 and 0.898, respectively). Moreover, the optimal malnutrition cutoff value for the GNRI was 94.95.
CONCLUSIONS: The prevalence of malnutrition remains high among hospitalized elderly patients. Validity of the GNRI is comparable to that of the MNA, and use of the GNRI to assess the nutritional status of this group is proposed with the new suggested cutoff value (GNRI ≤ 94.95), as it is simpler and more efficient. Underdiagnosis of malnutrition can be prevented, possibly reducing the prevalence of malnourished hospitalized elderly patients and improving the quality of the nutritional care process practiced in Malaysia.
METHOD: A total of 3825 trauma patients from 2011 to 2016 were extracted from the Hospital Sultanah Aminah Trauma Surgery Registry. Patients were split into a development sample (n = 2683) and a validation sample (n = 1142). Univariate analysis is applied to identify significant anatomic predictors. These predictors were further analyzed using multivariable logistic regression to develop the new score and compared to existing score systems. The quality of prediction was determined regarding discrimination using sensitivity, specificity and receiver operating characteristic [ROC] curve.
RESULTS: Existing simplified score systems (GAP & mGAP) revealed areas under the ROC curve of 0.825 and 0.806. The newly developed HeCLLiP (Head, cervical spine, lung, liver, pelvic fracture) score combines only five anatomic components: injury involving head, cervical spine, lung, liver and pelvic bone. The probabilities of mortality can be estimated by charting the total score points onto a graph chart or using the cut-off value of (>2) with a sensitivity of 79.2 and specificity of 70.6% on the validation dataset. The HeCLLiP score achieved comparable values of 0.802 for the area under the ROC curve in validation samples.
CONCLUSION: HeCLLiP Score is a simplified anatomic score suited to the local Malaysian population with a good predictive ability for trauma mortality.