Can magnetic resonance imaging scan replace or triage the use of laparoscopy in establishing a diagnosis among women with chronic pelvic pain? A comparative test accuracy study.

ABSTRACTBackgroundChronic Pelvic Pain (CPP) in women is common, debilitating condition with a high economic burden to healthcare systems. A lack of accurate tools to efficiently diagnose and direct cases of patients, leading to a wide variation in clinical practice concerning diagnosis of CPP. Diagnostic laparoscopy, which is an invasive, expensive and potentially risky procedure, is frequently used in investigation. Magnetic resonance imaging (MRI) may be an easily accessible, non-invasive alternative, but it is currently not recommended in guidelines nor used in routine practice.ObjectivesThis technology assessment investigated if MRI could reduce the need for a diagnostic laparoscopy and triage cases to better plan therapy. We estimated the accuracy of post-MRI diagnoses and post-laparoscopy diagnoses for (a) the absence of any pathology, i.e. idiopathic CPP, and (b) the main pathological causes of CPP, using panel consensus as reference standard.MethodsDesignWe performed a comparative test accuracy study with panel consensus for determining reference standard and a health economic evaluation (ISRCTN registry No. ISRCTN13028601).Participants and setting: We recruited women aged 16 and over, who had been referred to a gynaecologist with CPP of at least 6 month’s duration and in whom there was an indication for diagnostic laparoscopy, from 26 UK hospitals. Index tests:Diagnostic laparoscopy and MRI scans were the index tests. Reference standard: There were many target conditions. Idiopathic CPP was the primary condition of interest. Other conditions included structural gynaecological causes, such as endometriosis, adhesions, adenomyosis and non-gynaecological causes, e.g. gastrointestinal or urinary causes. A consensus panel whose members were experts in CPP and were not involved in the study recruitment as investigators determined the reference diagnosis. Sample size: A sample size of 340 women was chosen to address the primary research question of how many women could have avoided laparoscopy if MRI was routinely used in practice for diagnosing women with CPP. The study had over 90% power (at p=0.05) to detect a reduction of 10% in the number of laparoscopies needed (i.e. from 100% down to 90%). Analysis:By comparing the index test with the reference standard, the result of the index test can be categorised as a true positive, false positive, true negative or false negative.Economic evaluation:Utilising primary data collected in this study along with secondary data sources, a model based economic evaluation was conducted from the health-care provider perspective (NHS, UK) in a secondary care (hospital) setting used a decision-tree model in TreeAge Pro Healthcare 2001 (TreeAge Software, Inc., Williamstown, MA, USA). The primary outcome measure was the quality adjusted life year over a time horizon of 6 months. The analysis compared diagnostic laparoscopy versus MRI amongst women presenting with chronic pelvic pain.ResultsWhen setting the test cut-off for the sensitivity and specificity so maximize cost-effectiveness it was found that that diagnostic laparoscopy has an ICER value of £16,038 (cost/QALY) compared to MRI. This is below the NICE acceptance threshold of £20,000-£30,000 for the QALY meaning that diagnostic laparoscopy would be acceptable to the UK health care provider. In contrast, if the test cut-off is set to maximize patient outcomes (QALYs) then diagnostic laparoscopy has an ICER of £80,261 compared to MRI, which would not be acceptable to UK health care provider.ConclusionsThe health economic analysis has shown that diagnostic laparoscopy is more cost-effective than MRI, although the importance of the cut-off for the sensitivity and specificity of the tests has been demonstrated.