The number, coverage and quality of diagnostic test accuracy studies in Nephrology

Article type
Authors
McGee R1, Nguyen B1, Mitchell R2, Craig J1, Webster A1
1Sydney School of Public Health, University of Sydney, Sydney, New South Wales, Australia
2Centre for Kidney Research, The Children’s Hospital at Westmead, Westmead, Australia
Abstract
Background: The validity of a systematic review depends on assessing the totality of evidence. It is unknown how many diagnostic test accuracy studies (DTAS) are published in Nephrology and other specialties. Objectives: To establish the number, coverage and quality of DTAS in Nephrology compared to other specialties. Methods: DTAS in 13 specialties of internal medicine were identified using MEDLINE MeSH terms and the clinical queries filter ‘diagnosis optimized’ from 1966–2008. Results were ‘standardized’ by adjusting for total citations within each specialty. Coverage within Nephrology was explored by classifying citations into eleven subspecialty topic areas. A modified QUADAS tool was used to assess risk of bias in a random sample of studies. All stages were conducted by two authors working independently. Results: Publication of DTAS increased over time across all specialties, however Nephrology (22,230 total) published fewer relative to other specialties (Mean: 59,764, SD ± 34,855). When adjusting for total citations 5.1% of citations in Nephrology were DTAS (Mean%: 5.9, SD ± 1.0, Range: Rheumatology 4%, Infectious diseases 7.6%) (Figure 1). Within Nephrology subspecialty topics, diabetic nephropathy, renal artery obstruction and urinary tract infections published a greater proportion of DTAS than the mean, and kidney transplantation, hemodialysis, nephritis, peritoneal dialysis, urolithiasis, acute kidney injury, renal tubular transport and renal failure were under-represented. Only 47 of 300 sampled (to date) citations examined accuracy of diagnostic tests in a defined population (the remaining studies examined tests in genetic techniques, microbial sensitivities or were review articles). Of these 47, quality was suboptimal (Figure 2). Conclusions: The total number and relative proportion of DTAS published in Nephrology is low compared to other medical specialties and quality of nephrology DTAS is poor. However, we did not compare quality across specialties and publications are only surrogate markers of research. Increased familiarity with DTAS methodologies and the QUADAS tool may improve research quality.