11 posters, 7 topics, 32 authors, 14 institutions
ePostersLive by SciGen Technologies S.A. All rights reserved.
Clinical Radiology Quality Improvement ePoster Competition 2024
22 NOVEMBER, 2024 | LONDON, UK
43
Improving parathyroid localisation in the <sup>99m</sup>Tc-Sestamibi scan
Poster Presenter
Authors
Peter Jarvis, Kat Dixon, Emma O'Shaughnessy, Jennifer Downs, Tristan Richardson, Amit Parekh
Affiliations
Part of Topic
Nuclear medicine
Introduction
•Primary hyperparathyroidism can only be cured by successful surgery
•High quality pre-operative localisation allows the surgeon to perform targeted exploration with economic and risk benefits.
Methods
•Retrospective audit of 431 patients undergoing parathyroid surgery.
•Correct localisation was determined by pathological specimen at surgery.
•5-point scale was assigned to each report based on wording and lesion description.
•Blinded to surgical outcome
•Scale used to calculate ROC curves
Initial results
•Initial audit of 194 patients
•Sensitivity 69.8
•31% failed to identify adenoma
•44% confidently identified adenoma
Interventions
Goal was to improve sensitivity of adenoma localisation as well as confidence of the radiology report.
1) Upgraded SPECT/CT scanners oSiemens Symbia T to Siemens Intevo Bold
2) Optimised image acquisition
oStep and shoot acquisition over previous continuous acquisition
oIncreased administered activity from 750Mbq to 900Mbq 99mTc-Sestamibi
oChanged SPECT reconstruction parameters (increased product of subsets and iterations)d
3) Addition of arterial phase iodinated contrast, Dropped the early phase of SPECT acquisition
Results
Pie charts showing proportions of adenomas which were successfully identified following each intervention
Baseline
Sensitivity = 69.8%
Specificity = 95.7%
AUC = 0.92
Intervention 1
Sensitivity = 77.6%
Specificity = 93.3%
AUC = 0.95 (p0.22)
Intervention 2
Sensitivity = 84.3%
Specificity = 95.9%
AUC = 0.98 (p0.004)
Intervention 3
Sensitivity = 94.7%
Specificity = 90.0%
AUC = 0.99 (p<0.001)
ROC curve following each intervention. Statistical analysis using DeLong test for 2 correlated ROC curves shown.
Conclusion
We have demonstrated significant sequential improvement in adenoma localisation and radiologist report confidence.
Improved adenoma localisation has resulted in fewer full neck exploration surgeries being performed with consequent improvement in patient safety, length of stay and reduced cost.