Showing 2 results for Likelihood Ratio
Farzad Hadaegh, Hadi Harati, Asghar Ghasemi, Maryam Tohidi, Azadeh Zabetian, Mojgan Padyab, Fereidoun Azizi,
Volume 6, Issue 1 (8-2006)
Abstract
Background: The aim of this study was to determine the level of agreement between the impaired fasting glucose (IFG) and abnormal glucose tolerance before and after application of the new IFG definition and to evaluate the impact of adding common clinical data on this agreement.
Methods: A cross sectional population based study was carried out in an Iranian urban population which enrolled 8766 men and women over 20 years. Fasting and 2-hour plasma glucose were measured in all subjects excluding those with previously diagnosed diabetes and fasting plasma glucose ≥126 mg/dl. The diagnostic parameters and kappa coefficient between the previous and revised definitions of IFG for detecting impaired glucose tolerance (IGT) and dysglycemia (IGT and diabetes) were calculated. Logistic regression and ROC curve analysis were used to determine the independent clinical risk factors and their optimal cut-points associated with IGT and dysglycemia.
Results: After using the new criteria, sensitivity of IFG for detecting IGT or dysglycemia increased but specificity and positive likelihood ratio (LR+) decreased and the κ slightly improved (0.16 to 0.29 for IGT and 0.24 to 0.35 for dysglycemia). Adding the clinical data to the revised criteria considerably improved the agreement between IFG with IGT and dysglycemia (κ increased from 0.286 to 0.470 for IGT and from 0.354 to 0.574 for dysglycemia). This also increased the LR+ from 3.86 to 14.5 and from 4.46 to 17.4 respectively for detecting IGT or dysglycemia.
Conclusion: The new IFG definition in combination with common clinical risk factors most likely predicts IGT and dysglycemia.
Farzane Saeidifard, Akbar Soltani, Fereshteh Ghadiri, Sahar Manavi, Motahareh Taleba, Moein Foroughi, Parvaneh Ansari, Mostafa Qorbani, Hamideh Moosapour,
Volume 17, Issue 4 (5-2018)
Abstract
Background: It is critical to understand how accurately physicians can estimate the importance of each clinical finding in estimating the probability of a specific diagnosis in the process of clinical decision making. This study aimed to investigate whether physicians’ estimates of the importance of various clinical findings of ascites correlated with the positive likelihood ratios of these findings in diagnosis of ascites.
Methods: One hundred and ten physicians were asked to respond to a questionnaire. In this questionnaire they were presented with a clinical scenario about a patient suspected of having ascites followed by a list of clinical findings. Participants were asked to assign a weight (between 0 and 100%) to each clinical finding based on their perception of how much the presence of that finding changed the probability of ascites for the patient. Positive likelihood ratios of those findings were extracted from current best evidence. We investigated if the weights assigned by physicians were associated with the positive likelihood ratios of those findings.
Results: Significant differences were discovered between the weights assigned by the physicians and the positive likelihood ratios for each clinical finding. Significant positive correlation was observed between the weights assigned by different groups of physicians.
Conclusion: Physicians inaccurately estimated the importance of various clinical findings in the diagnosis of ascites. Further research is needed to determine if such inaccurate estimations would lead to any adverse clinical outcomes.
