Nuclear medicine image interpretation by radiographers: findings of an accredited postgraduate module

Introduction:
The study aimed to analyse the results of radiographer’s image interpretation of nuclear medicine (NM) examinations following a nine-month postgraduate module.

Methods:
Twenty participants completed 60 summative image commentaries each at the end of the module from prospective NM worklists in England. Each submitted a mixed selection of examinations in bone, lung, renal, and thyroid scans. Prevalence of abnormalities was 51% incorporating acute and chronic pathology, normal variants and incidental findings. Every commentary was marked against reference standard radiologist definitive reports. Statistical analysis included Kappa (k), intraclass correlation coefficient (ICC) and Spearman’s rank correlation coefficient (RS).

Results:
Bone scan sensitivity and specificity was 93% (95% CI 91.3-95.6) and 88% (95% CI 84.3-90.9) respectively, accuracy at 91.5% (95% CI 88.6-93.7), with k=0.82, ICC= 0.904, RS=0.826. Lung scans demonstrated a sensitivity of 92.6% (95% CI 85.7-96.8), specificity was 92.1% (95% CI 88.7-94.1), accuracy 92.3% (95% CI 87.7-95.0), k=0.83, ICC=0.910, RS=0.835. Renal scan sensitivity was 95% (95% CI 91.0-97.3), with 95.2% specificity (95% CI 91.8-97.3), accuracy were 95% (95% CI 91.4-97.3), k=0.90, ICC=0.948, RS=0.907. Thyroid scans sensitivity was 88% (95% CI 83.1-91.4), with 93% specificity (95% CI 85.9-96.8), accuracy were 90.2% (95% CI 84.3-93.8), k=0.80, ICC=0.897, RS=0.813.

Conclusion:
In this small pilot study, the image interpretation ability in assessing prospective NM examinations in a clinical environment displayed encouraging results. Further work is recommended to evaluate a larger sample and case selection.