Background In 1997, the SIPAP classification was introduced, a guide designed for MRI, to characterize pituitary adenomas with emphasis on extrasellar extensions and impact on adjacent structures. To our knowledge no previous evaluation of the inter-observer agreement of the SIPAP classification has been performed.
Purpose To evaluate the inter-observer agreement of the SIPAP classification.
Material and Methods Sixty patients operated on for growth hormone producing pituitary adenomas at Lund University Hospital 1991-2007 had an assessable preoperative MRI scan. Forty-five of them also had an assessable postoperative MRI scan. The mean time between surgery and postoperative MRI scans was 11 months. Two observers evaluated all the MRI scans independently. The outcome of the evaluation is presented as the percentage of concordance in each of the evaluated directions and the degree of discrepancy for each of the directions evaluated.
Results In 284 (79%) of 360 preoperative gradings both observers agreed completely. In 17 of the 60 preoperative MRI scans, both observers made identical assessments according to the SIPAP classification in all the six different directions of tumor extension. In 76 gradings the results differed between the observers. The difference was 1 grade (or less) in 69 cases. In 230 (85%) of 270 postoperative gradings the results were identical for both observers. In 18 of the 45 postoperative MRI scans, both observers made the same assessments according to the SIPAP classification in all the six different directions of tumor extension. In 40 gradings the results differed between the observers. The difference was 1 grade (or less) in all 40 cases.
Conclusion We found a relatively high inter-observer agreement both pre- and postoperatively, supporting the usefulness and easy applicability of the SIPAP system for grading of pituitary adenomas pre- as well as postoperatively.
2011. Vol. 52, no 7, 796-801 p.