Research prioritization of men’s health and urologic diseases

ABSTRACT Objectives We sought to determine whether disease representation in the Cochrane Database of Systematic Reviews (CDSR) reflects disease burden, measured by the Global Burden of Disease (GBD) Study as disability-adjusted life-years (DALYs). Materials and Methods Two investigators performed independent assessment of ten men’s health and urologic diseases (MHUDs) in CDSR for systematic review and protocol representation, which were compared with percentage of total 2010 DALYs for the ten conditions. Data were analyzed for correlation using Spearman rank analysis. Results Nine of ten MHUDs were represented by at least one CDSR review. There was a poor and statistically insignificant positive correlation between CDSR representation and disease burden (rho = 0.42, p = 0.23). CDSR representation was aligned with disease burden for three conditions, greater than disease burden for one condition, and less than disease burden for six conditions. Conclusions These results yield high-quality estimates to inform future research prioritization for MHUDs. While prioritization processes are complex and multi-faceted, disease burden should be strongly considered. Awareness of research priority setting has the potential to minimize research disparities on a global scale.


INTRODUCTION
In order to achieve effective clinical research, scarce research funds must be distributed to appropriate diseases in order to maximize health benefits to the represented population. Systematic approaches to inform research prioritization include identifying and prioritizing research questions, recognizing existing research, and setting goals for primary research (1,2). A derivative of this approach is to value major diseases, injuries, and risk factors based on their burden to society (3). Spearheaded by the Institute for Health Metrics and Evaluation (IHME), the Global Burden of Disease (GBD) 2010 Study estimates the burden of 291 diseases and injuries across 187 countries from 1990 to 2010 (4,5). The metric of disability--adjusted life years (DALYs), in which 1 DALY is equivalent to 1 year of healthy life lost, allows for descriptive global epidemiology of a wide array of disease states. The following ten men's health and urologic diseases (MHUDs) were studied by GBD on the basis of prevalence, common case definitions, and data availability: tubulointerstitial nephritis, pyelonephritis, and urinary tract infections; kidney and other urinary organ cancers; urolithiasis; male infertility; benign prostatic hyperplasia; prostate cancer; testicular cancer; hydrocele due to lymphatic filariasis; dysuria/bladder pathology/ hydronephrosis due to schistosomiasis; and bladder cancer (Figure-1).
Systematic reviews are the cornerstone of evidence-based medicine, yet few efforts have been made to assess whether the prioritization of systematic reviews reflect global disease burden (6). The Cochrane Database of Systematic Reviews (CDSR) produces systematic reviews and protocols (published proposals for future systematic reviews) across all medical specialties as well as health systems, public health, and child development. Cochrane systematic reviews undergo exhaustive editorial processing, are more methodologically rigorous, and are updated more frequently than non-Cochrane reviews and paper-based journals (7,8). Prior studies have evaluated the association between broad categories of disease burden with randomized trials and Cochrane systematic reviews (9)(10)(11)(12). This study will assess whether the CDSR representation of ten MHUDs corresponds to GBD 2010 disability estimates.

MATERIALS AND METHODS
ICD-10 code definitions for the ten MHU-Ds have been previously published and were used to generate search terms, which were entered into the "title, abstract, keywords" CDSR search function (5,13). Systematic reviews and protocols were  considered to determine MHUD representation in CDSR, according to abstract subject content. Online publication date, the number of studies included in each systematic review, and the particular Cochrane review group that published the review or protocol were collected. Two authors (T.O and H.P.) collected data independently during February 2015. DALY metrics for each of the ten MHUDs, expressed as percentages of total DALY's of all 291 conditions measured in GBD 2010, were obtained from the GBD Compare interactive time plot, available at <http://viz.healthmetricsandevaluation.org/gbd--compare/>. Spearman rank correlation analysis was performed to assess statistical dependence between CDSR representation and disease burden. Rho, a coefficient ranging from-1 (strong negative correlation) to +1 (strong positive correlation), is interpreted with a two-tailed p-value. A line-of--best-fit was also generated between CDSR representation and % of total DALYs.
As this study did not involve human subjects, institutional review board approval was not necessary.

RESULTS
Nine of the ten MHUD conditions studied in GBD 2010 were represented by at least one systematic review. A total of 116 systematic reviews and protocols published by nine Cochrane review groups represented the ten MHUDs (Supporting Tables 1 and 2 for included and excluded titles, respectively). The majority of reviews and protocols covered tubulointerstitial nephritis, pyelonephritis, and urinary tract infections (n=46). Hydrocele due to lymphatic filariasis had no representation in CDSR. Of the ten MHUDs, benign prostatic hyperplasia had the greatest global disease burden (0.2%) while male infertility had the lowest (0.007%) Table-1. Reviews and protocols representing the ten MHUDs were published by the following Cochrane review groups: Prostatic Diseases and Urologic Cancers Group (n=43); Renal Group (36); Incontinence Group (14); Menstrual Disorders and Subfertility Group (8); Pregnancy and Childbirth Group (7); Infectious Diseases Group (3); Pain, Palliative and Supportive Care Group (2); Gynecological Cancer Group (2); and Neonatal Group (1).
Spearman rank correlation testing between CDSR representation and DALY metrics revealed poor positive correlation that was statistically insignificant (rho=0.41, p=0.21). The majority of the MHUDs (6) were under-represented in CDSR as compared to GBD DALY (Figure-2). Most of the systematic reviews and protocols (58.6%) were published from 2011 to 2015 while 37.9% were published from 2000 to 2010; only 4 reviews were published prior to 2000. Maintaining systematic reviews up--to-date is critical to deliver consensus statements on current world literature that ultimately impact clinical decisions and patient outcomes.
Representation of tubulointerstitial nephritis, pyelonephritis, and urinary tract infections exceeded GBD disease burden. This disease category also had the greatest number of cumulative studies informing its systematic reviews (529). The one systematic review representing testicular cancer, entitled "Screening for testicular cancer," found no randomized controlled trials in the literature. Systematic reviews that find no suitable trials to address their objectives uncover areas for much-needed, high-quality research.
The World Health Organization (WHO) classifies two of the MHUDs as neglected tropical diseases: dysuria/bladder pathology/hydronephrosis due to schistosomiasis and hydrocele due to lymphatic filariasis (14). It is important to note that just as the DALY metrics reported for these two diseases include only burden due to the MHUD morbidity (dysuria, bladder pathology, hydronephrosis, hydrocele), systematic reviews were only considered representative if they included assessment of the MHUD pathology.

DISCUSSION
We acknowledge several limitations of our study. The scope of CDSR systematic reviews is subject to variability. For instance, authors may prepare one large review of multiple interventions (lumping) or several reviews of individual interventions (splitting). Therefore, treating a systematic review or protocol as one measurement unit may not be entirely accurate for every topic.       Benign prostatic hyperplasia N40 "benign prostatic hyperplasia" "median bar" "prostatic hyperplasia" "adenofibromatous hypertrophy of prostrate" "hypertrophy of prostate "prostatic obstruction" 9 (7 R, 2 P) 0.2% 35 Prostate cancer C61, D07.5, D40.0 "prostate cancer" "prostatic carcinoma in situ" "prostate neoplasm" 23 (22 R, 1 P) 0.15% 411 Kidney and other urinary organ cancers C64-C66, D41.0-D41.2 "kidney cancer" "neoplasm of the ureter" "neoplasm of the kidney" "neoplasm of the renal pelvis" 4 (4 R) 0.15% 63 Tubulointerstitial nephritis, pyelonephritis, and urinary tract infections N10-N12, N15.1-N15.9, N30, N34, N39.0 "tubulointerstitial nephritis" "pyelonephritis" "urinary tract infections" "infectious interstitial nephritis" "pyelitis" "balkan nephropathy" "renal and perinephric abscess" "cystitis" "trigonitis" "urethral abscess" "urethritis" Urolithiasis N20-N23 "urolithiasis" "urinary stones" "nephrolithiasis" "kidney stones" "ureterolithiasis" "cystolithiasis" "bladder stones"  While beyond the scope of this limited study, further exploration is warranted into potential underrepresentation of certain conditions. There remains a lack of transparency in publications and databases on the quality of data and criteria involved in prioritization decisions (15). Other important factors in priority setting include availability of research funds, knowledge gap, and impact on disadvantaged populations. Research prioritization is also inherently political and dependent on financial backing, which further demonstrates the importance of a transparent process. Attention and awareness of priority setting has the potential to minimize research disparities and, ultimately, impact populations at a global scale.

ACKNOwLEDGEMENTS:
Tyler Okland and Chante Karimkhani are contributed similarly as first authors.
Luc Coffeng MD, Department of Public Health at Erasmus MC University Medical Hydrocele due to lymphatic filariasis Cochrane system atic reviews and protocols