RSNA 2003 Scientific Papers > Application of Standards for Reporting of Diagnostic ...
  Scientific Papers
  SESSION: Health Services, Policy and Research (Issues in Research Methodology)

Application of Standards for Reporting of Diagnostic Accuracy (STARD) to the MR/US Literature on Benign Uterine Abnormalities: A Systematic Review

  DATE: Wednesday, December 03 2003
  START TIME: 11:20 AM
  END TIME: 11:27 AM
  CODE: K16-1001

Richard Sharpe
Washington , DC
Susan Ascher MD
Evan Myers MD, MPH

Radiology and radiologists, research
Uterus, MR
Uterus, US

Purpose: Repeated studies indicate the quality of reporting in radiology trials falls below optimal levels. However, complete and accurate reporting is essential to determine the potential for bias in a study (internal validity) and to judge the generalisability of its results (external validity). Standards for Reporting of Diagnostic Accuracy (STARD) have been recently published in ten major journals (Radiology 2003; 226:24-8) to improve the quality of reporting. This study applied the STARD criteria to retrospectively evaluate the quality of reporting in a series of radiological studies involving premenopausal patients presenting with signs and symptoms (meno/metrorrhagia, pelvic pain) of benign uterine conditions (fibroids, adenomyosis, polyps, endometriosis). Accurate diagnosis is important because these conditions can present similarly but may have widely varying treatments (myomectomy vs hysterectomy).

Methods and Materials: English language manuscripts published from 1986 to present were identified by a pertinent Medline search and an additional hand search was performed. Diagnostic accuracy studies of premenopausal patients presenting with the above signs and symptoms of benign uterine conditions (fibroids, adenomyosis, polyps, endometriosis) undergoing MR and/or US evaluation were included. Each article was independently evaluated by two of the authors using the STARD checklist as to whether it met each recommendation (yes/no evaluation); disagreements were resolved by consensus.

Results: Of 46 evaluated manuscripts, no study met all of the STARD criteria for adequately evaluating the quality of methodological design and reporting. Title, Abstract, and Keywords sections met the STARD criteria in 18 (39.13%) of the articles and the Introduction in 46 (100.00%). Methods, Results and Discussion sections met STARD guidelines in 28 (60.87%), 21 (45.65%), and 37 (80.43%) manuscripts, respectively.

Conclusion: The series of articles reviewed is deficient in reporting items to assess their internal and external validity. The large resources spent yearly treating patients with benign uterine conditions may be misguided, as current patient management is based on the evaluated publications. Therefore, it is important for researchers to begin to apply the STARD recommendations to improve the accuracy and completeness of reporting of studies of diagnostic accuracy. Furthermore, the inadequacy of these articles may not allow them to be useful for metanalysis.




Questions about this event email: