Evidence based testing and outcomes in transplantation

Abstract

The use of diagnostic tests is central to the practice of modern medicine, but knowing which test to use, and when, can be problematic. To make evidence-based diagnoses, clinicians need efficient ways of accessing diagnostic studies, interpreting the results of several studies, and checking the applicability of studies to their own setting. The aim of this thesis was to explore solutions to these problems by addressing a specific clinical question; What is the best screening test for latent tuberculosis in patients undergoing transplantation? In a study of diagnostic filter performance in MEDLINE, we found that current ‘specific’ clinical queries limit for diagnosis (used in PubMed and Ovid SP) missed up to 80% of studies in nephrology journals. Other filters (Deville 2000 Broad, Deville 2000 Balanced, Haynes 2004 Balanced, and Vincent 2003 Narrow) had similar specificity to the ‘specific’ clinical queries limit, but identified a greater proportion of the total evidence. Using systematic review methodology, we found that current available data was inadequate to determine whether interferon gamma release assays performed better, worse or the same as the tuberculin skin test for diagnosing tuberculosis in candidates for solid organ transplant. Current international guidelines recommend using either the tuberculin skin test or an interferon gamma release assay, or both in combination. Our findings support these guidelines. We conducted a cross-sectional descriptive of candidates for kidney transplantation and found that despite a high prevalence of risk factors among the group, less than a quarter of candidates were screened for latent tuberculosis before transplant, and only 36% of the 101 patients with risk factors for tuberculosis were tested. This study demonstrates that candidates for kidney transplant are at increased risk of tuberculosis and highlights the need for a nation-wide tuberculosis screening protocol in work-up for transplant.