| Saline infusion sonography (SIS) and hysteroscopy are equally good for detecting submucous fibroids and better than ultrasound (TVS). Hysteroscopy and SIS are however, associated with higher rates of failure and discomfort than TVS. |
Citation/s:
Farquhar C, Ekeroma A, Furness S, Arroll B. A systematic review of transvaginal ultrasonography, sonohysterography and hysteroscopy for the investigation of abnormal uterine bleeding in premenopausal women. Acta Obstet Gynecol Scand 2003; 82: 493-504
Lead author's name and fax: Farquhar C
Three-part Clinical Question: Which is the most accurate test (ultrasound or hysteroscopy) for the diagnosis of benign endometrial pathology in premenopausal women with abnormal uterine bleeding?
Search Terms: MesH terms (hysteroscopy or ultrasound) and ((sensitivity and specificity) or predictive). Limited to reviews
The Review:
Data Sources: : Medline, Embase, hand search,
Study Selection: Population: Premenopausal women with abnormal bleeding.
Intervention and Reference standard : Comparative studies of diagnostic test (TVS, SIS, Hysteroscopy) and reference standard (histology from hysteroscopically directed biopsy or hysterectomy specimen)
Outcomes: 'Intrauterine pathology', Submucous fibroids, Endometrial hyperplasia
Data Extraction: Not independant, ? done in duplicate by 2 reviewers (unclear)
Data extracted to populate 2x2 tables for accuracy. Also data regarding failure rates, side-effects, patient discomfort and resource utilisation
| Multiple independent reviews of individual reports? | can't tell |
| Tested for heterogeneity? | yes |
The Evidence:
| Outcome | Time to Outcome | Typical CER | Typical OR | RRR | NNT | p Value |
| 95% Confidence Intervals: | to | |||||
Comments:
Focussed and clinically relevant question. Comprehensiveness of study identification limited as restricted to English language and 'grey' literature not sought. Study selection and data extraction not done independently or in duplicate potentially introducing error. Quality assessment performed aiding later inferences, but limited (and questionable) criteria used limiting its discriminative power. Heterogeneity assessed and pooling limited to homogeneous studies (+ subgroup analyses based on quality) facilitating cautious data interpretation. Overall well conducted study with appropriate population and reference standard, but clinical inferences limited due to poor quality of studies and marked inconsistencies between findings which probably reflect methodological weaknesses of primary studies as well as populations. Inter-rater variability not reported which may cause some of this variation. Findings from studies of hysteroscopy most homogeneous although this reflects the few included studies. Future research required includes comparison of combination strategies (e.g. TVS/SIS vs. hysteroscopy), utility of test (i.e. influence of diagnosis on patient management if any) and effectiveness (RCTs or decision analyses with clinical AND economic endpoints).
Appraised by: TJ Clark; Monday, September 08, 2003
Email: t.j.clark@bham.ac.uk
Kill or Update By: 2006 or unless further systematic reviews with different methodology published