Rufener Stephanie L, Patel Smita, Kazerooni Ella A, Schipper Matthew, Kelly Aine M
Department of Radiology, University of Michigan Health System, Ann Arbor, MI 48109, USA.
Acad Radiol. 2008 Jan;15(1):71-6. doi: 10.1016/j.acra.2007.06.030.
On-call radiology residents frequently interpret computed tomography (CT) pulmonary angiography and CT venography studies outside of routine working hours. The purpose of this study was to compare resident and faculty interpretation concordance rates and to see if concordance rates differed depending on the number of CT detectors used.
The study population included 122 consecutive CT pulmonary angiography (CTPA) and CT venography (CTV) examinations performed on a four-row multidetector CT (MDCT) and 125 consecutive CTPA examinations performed using a 16-row MDCT scanner with CTV performed in 124 patients. Preliminary resident reports and final faculty reports were compared. Discrepant cases were independently reviewed by three cardiothoracic radiologists who were unaware of the initial interpretations. Interpretation concordance rates were calculated for both 4- and 16- row MDCT studies and compared using Fisher's exact test.
Resident and faculty CTPA and CTV interpretations were concordant in 80% of the 4-row cases and 94% of the 16-row cases. When comparing resident interpretation to the final expert reference standard, the corrected resident error rate was 11% and 2% for 4-row CTPA and CTV, respectively and 4% and 2% for 16-row CTPA and CTV, respectively. Overall CTPA and CTV concordance was significantly lower for 4-row MDCT (80% versus 94%, P < .001 [two-sided] by Fisher's exact test).
Radiology resident interpretation of CTPA and CTV studies demonstrates a high level of agreement with radiology faculty interpretation. Concordance rates are significantly higher for 16-row MDCT than 4-row MDCT which may be due to improved image quality.
值班放射科住院医师经常在非工作时间解读计算机断层扫描（CT）肺动脉造影和CT静脉造影研究。本研究的目的是比较住院医师和教员的解读一致率，并观察一致率是否因使用的CT探测器数量而异。
研究人群包括在四排多层螺旋CT（MDCT）上进行的122例连续CT肺动脉造影（CTPA）和CT静脉造影（CTV）检查，以及在16排MDCT扫描仪上进行的125例连续CTPA检查，其中124例患者进行了CTV检查。比较住院医师的初步报告和教员的最终报告。由三位不知道初始解读结果的心胸放射科医生独立审查有差异的病例。计算4排和16排MDCT研究的解读一致率，并使用Fisher精确检验进行比较。
在4排病例中，住院医师和教员对CTPA和CTV的解读一致率为80%，在16排病例中为94%。将住院医师的解读与最终专家参考标准进行比较时，4排CTPA和CTV的校正住院医师错误率分别为11%和2%，16排CTPA和CTV的校正住院医师错误率分别为4%和2%。4排MDCT的总体CTPA和CTV一致性显著低于16排MDCT（80%对94%，Fisher精确检验的双侧P <.001）。
放射科住院医师对CTPA和CTV研究的解读与放射科教员的解读高度一致。16排MDCT的一致率显著高于4排MDCT，这可能是由于图像质量提高所致。
