Physician Prognostic Accuracy for In-Hospital Mortality in Percutaneous Coronary Intervention. Michael E. Matheny, MD Medical Informatics Fellow Decision Systems Group Brigham & Women’s Hospital Boston, MA. Specific Aims. Primary Hypothesis - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
A Free sample background from www.powerpointbackgrounds.com
Slide 2
Specific AimsSpecific Aims
Primary HypothesisPrimary Hypothesis
– Accuracy of subjective physician estimations of Accuracy of subjective physician estimations of in-hospital mortality will be similar or improved in-hospital mortality will be similar or improved when compared with accepted objective risk when compared with accepted objective risk assessment methods for percutaneous coronary assessment methods for percutaneous coronary intervention (PCI)intervention (PCI)
A Free sample background from www.powerpointbackgrounds.com
Slide 3
Specific AimsSpecific Aims
Secondary HypothesesSecondary Hypotheses
– Accuracy of subjective physician estimations of in-Accuracy of subjective physician estimations of in-hospital major adverse cardiac events (MACE) will be hospital major adverse cardiac events (MACE) will be similar or improved when compared with accepted similar or improved when compared with accepted objective risk assessment methods for PCIobjective risk assessment methods for PCI
– Qualitative collection of risk factors could identify Qualitative collection of risk factors could identify additional important risk factors currently not included in additional important risk factors currently not included in the objective risk modelsthe objective risk models
– Incorporating subjective physician estimates into an Incorporating subjective physician estimates into an objective risk model will outperform either separatelyobjective risk model will outperform either separately
A Free sample background from www.powerpointbackgrounds.com
Slide 4
BackgroundBackground
DiscriminationDiscrimination– Ability to predict an outcome on a population levelAbility to predict an outcome on a population level– Area under the Receiver Operating Characteristic Curve Area under the Receiver Operating Characteristic Curve
A Free sample background from www.powerpointbackgrounds.com
Slide 5
BackgroundBackground
CalibrationCalibration– Ability to predict an outcome on a case/small group levelAbility to predict an outcome on a case/small group level– Hosmer-Lemeshow Goodness-of-Fit Test (HL-GF)Hosmer-Lemeshow Goodness-of-Fit Test (HL-GF)– Brier ScoreBrier Score
A Free sample background from www.powerpointbackgrounds.com
Slide 7
BackgroundBackground
Subjective vs LR Acute CHF 90 day and 1 Subjective vs LR Acute CHF 90 day and 1 year Mortality year Mortality 3 43 4
– Discrimination: Discrimination: No DifferenceNo Difference– Calibration:Calibration: No DifferenceNo Difference– All estimations poorAll estimations poor
Subjective vs SNAP Neonatal ICU Mortality Subjective vs SNAP Neonatal ICU Mortality 55
– Discrimination:Discrimination: No DifferenceNo Difference– Calibration:Calibration: No DifferenceNo Difference
A Free sample background from www.powerpointbackgrounds.com
Slide 11
BackgroundBackground
No work has been done evaluating No work has been done evaluating subjective physician estimates for in-hospital subjective physician estimates for in-hospital mortality in percutaneous coronary mortality in percutaneous coronary interventions.interventions.
– Population referral biasPopulation referral bias• Temporal BiasTemporal Bias
– Medical Care StandardsMedical Care Standards– Data DocumentationData Documentation
• Data NoiseData Noise– Heterogeneous Data StandardsHeterogeneous Data Standards– Variation in Data Element CollectionVariation in Data Element Collection– Data Entry Quality VariationsData Entry Quality Variations
A Free sample background from www.powerpointbackgrounds.com
Slide 18
StudyStudyData CollectionData Collection
Paper SurveyPaper Survey– Administration refused to allow survey to be part Administration refused to allow survey to be part
of medical recordof medical record
Subjective mortality assessment (0-100%) Subjective mortality assessment (0-100%) before and after procedurebefore and after procedure– AttendingsAttendings– FellowsFellows– Scrub NurseScrub Nurse
Qualitative additional risk factors from Qualitative additional risk factors from AttendingsAttendings
A Free sample background from www.powerpointbackgrounds.com
Slide 19
ExposuresExposures
Percutaneous Coronary Transluminal Percutaneous Coronary Transluminal Angiography with or without Coronary Angiography with or without Coronary StentingStenting
A Free sample background from www.powerpointbackgrounds.com
Slide 22
Analysis PlanAnalysis Plan
Measure Discrimination & Calibration on local data Measure Discrimination & Calibration on local data for:for:– Objective MACE & Mortality ModelsObjective MACE & Mortality Models
Pair-wise Comparison of Objective and Subjective Pair-wise Comparison of Objective and Subjective models for statistical differencesmodels for statistical differences
Develop LR model with subjective data as a Develop LR model with subjective data as a covariate, and perform pair-wise comparisons with covariate, and perform pair-wise comparisons with objective and subjective models to determine if new objective and subjective models to determine if new model shows improvementmodel shows improvement