Integrated Report Generation for Myocardial Perfusion SPECT: Efficiency Comparison to Report Dictation C Bui, EP Ficaro, JN Kritzman, G Wu, JR Corbett University of Michigan Health System, Ann Arbor, MI 49 th SNM Annual Meeting, Los Angeles, 2002
Disclosure The University of Michigan receives royalities from the sales of the cardiac quantification software application 4D-MSPECT. 49 th SNM Annual Meeting, Los Angeles, 2002
Objective Evaluate the efficiency of Integrated Report Generation (IRG) within cardiac SPECT quantification and display software compared to Conventional Report Dictation (CRD). 49 th SNM Annual Meeting, Los Angeles, 2002
Patient Population 76 consecutive patients referred for Stress-Rest myocardial SPECT 44 patients had abnormal SPECT reports 32 patients had normal SPECT reports 49 th SNM Annual Meeting, Los Angeles, 2002
Methods: Time requirements for CRD and IRG reporting methods were analyzed and compared. 49 th SNM Annual Meeting, Los Angeles, 2002
Methods - IRG Semi-automatic report generation software was developed and integrated into 4D-MSPECT Quantitative results are exported from the quantification and display screens of the applications directly into reports Image interpretations may be generated using either a full-screen or more recently floating on-screen reporting dialogue 49 th SNM Annual Meeting, Los Angeles, 2002
Data Entry: Methods: Time Requirements IRG Technologists: Demographic data, radiotracer doses. Exercise physiologist / Nurse: Clinical indications, hemodynamic response & EKG findings. Interpretation & Clinical Scan Significance: Physicians: Two observers 49 th SNM Annual Meeting, Los Angeles, 2002
Methods CRD Time Requirements Written Preliminary Results: resident or attending Dictation of Stress Test and EKG Reports: attending Proofing of Typed EKG Reports: attending Dictation of Scan Results: resident Proofing of Typed Scan Reports: attending 49 th SNM Annual Meeting, Los Angeles, 2002
Clinical Data & EKG Input Dialogue
Dosage Data Entry Dialogue
Function Interpretation Floating Dialogue
Perfusion Interpretation Floating Dialogue
Sample HTML Report
Results Conventional vs IRG Seconds 600 500 552 475 400 Conventional 368 300 IRG 200 185 217 142 100 0 Combined Normal Abnormal
Results All Scans Combined Impression Entry 500 37 Clinical/EKG 450 400 350 300 99 60 Dosage Attending EKG Proof Attending Scan Proof 250 EKG Dictation 200 226 39 Scan Dictation 150 100 106 Preliminary W riting 50 53 40 0 Conventional IRG
Results Normal Scans Impression Entry 400 35 Clinical/EKG 350 300 76 Dosage Attending EKG Proof 250 200 57 11 Attending Scan Proof EKG Dictation 150 165 Scan Dictation 100 100 Preliminary W riting 50 34 31 0 Conventional IRG
Results Abnormal Scans Impression Entry 600 38 Clinical/EKG 500 115 Dosage Attending EKG Proof 400 63 Attending Scan Proof 300 EKG Dictation 200 270 60 Scan Dictation 110 Preliminary W riting 100 67 47 0 Conventional IRG
Results Seconds "Full-screen" vs "Floating" Impression Input 160 140 120 "Full-screen" impression 100 "Full-screen" & prelim writing 80 60 p < 0.01 79 p < 0.01 60 "Floating impression" 40 20 0 39 146 52 p < 0.01 106 16 51 11 Combined Normal Abnormal "Full-screen" impression 52 16 79 "Full-screen" & prelim writing 106 51 146 "Floating impression" 39 11 60
Results Conventional total time vs IRG Impression Seconds 600 500 552 475 400 300 368 Conventional total 200 Impression Entry 100 39 11 60 0 All Scans Normal Abnormal
Results Scan & EKG dictation vs Impression input Seconds 300 250 270 200 226 Scan & EKG dictation 150 165 Impression Entry 100 50 39 11 60 0 All Scans Normal Abnormal
Conclusion Reporting of SPECT myocardial perfusion and function studies using integrated report generation (IRG) within SPECT quantification and display software results in highly significant time saving compared to conventional reports dictated (CRD). 49 th SNM Annual Meeting, Los Angeles, 2002
Sample Text Report
Closing Dialogue