New Merged Facility: Data Management and Analysis Center (DMAC)

As of July 1, 2018, the Data Management Center (DMC) and the Biostatistical Consulting Unit (BCU) merged into one Shared Facility: the Data Management and Analysis Center (DMAC). This new core remains part of the Department of Pediatrics, Division of Biostatistics and Epidemiology (DBE). Dr. Mekibib Altaye, PhD, Professor, is now the Director for DMAC. Rachel Akers, Cyndie Baker, and Matthew Fenchel will continue to lead the management of operations for the combined unit.

While the DMC and BCU have always worked closely together, being under “one roof” further promotes the idea that researchers are served best when study design, data management, programming, and statistical analysis are well-coordinated – from the beginning of a study to the end. This new organization will promote more efficient processes, leading to higher quality and lower costs.

The new DMAC can also function as a unified, in-house data coordinating center (DCC) for multi-site studies, and DMAC leadership can help research faculty compete for multi-site grants like the Bench to Bassinet (B2B) study led by Drs. Eileen King, PhD, and Pete White, PhD, or the International Childhood Cardiovascular Cohort (i3C) Consortium study led by Jessica Woo, PhD; Elaine Urbina, MD; and John Morrison, MD.

Scientific Rigor and Reproducibility

National Institutes of Health (NIH) requires applicants to address scientific rigor and transparency in their grant and progress report submissions. The expertise of the DMAC personnel (as well as faculty in DBE) can greatly assist researchers with meeting these NIH requirements. This includes, but is not limited to, the following areas:

• Formulating clear, testable hypotheses, and design valid measures (exposure, process, outcome)
• Designing robust studies to minimize various forms of bias
• Design data collection processes that optimally support hypothesis testing and measurement goals
• Developing Case report forms (CRFs) to minimize errors and maximize efficiency
• Designing databases with appropriate quality checks and query management
• Programming for clear and accurate reports to meet regulatory requirements and analytic goals
• Selecting the best analytic methodology to accurately detect “signals” in the data, while accounting for other variables of importance

Decentralized Geomarker Assessment for Multi-Site Studies (DeGAUSS)

Drs. Cole Brokamp, PhD, and Patrick Ryan, PhD, MD, along with Mr. Chris Wolfe, have recently published an approach and accompanying software package that allows for decentralized and reproducible geocoding and characterization of community and environmental exposures for multisite studies in the Journal of the American Medical Informatics Association. Their article details the free and open source software package with a proof-of-concept application in the Electronic Medical Records and Genomics (eMERGE) Network: Since its initial release less than a year ago, DeGAUSS has been used by:

•Projects with principal investigators from Columbia, Wright State, Vanderbilt, Kaiser Permanente Washington Health Research Institute, Henry Ford Health System, Marshfield, Mayo Clinic, University of Wisconsin, Harvard, University of Arizona, Northwestern, Cornell

•NIH-funded Environmental Children’s Health Outcomes (ECHO) study

•University of Cincinnati Biomedical Informatics Research Services and Cincinnati Children's Hospital Medical Center Biomedical Informatics Research Information Technology

•Cystic Fibrosis Foundation Registry Network

•Transplantation and Immunology Research Network, American Society of Transplantation

Geospatial Research Accelerator for Precision Population Health (GRAPPH)

Drs. Brokamp and Ryan (along with Dr. Andrew Beck, MD, MPH, in the Division of General & Community Pediatrics) began this year to establish the Geospatial Research Accelerator for Precision Population Health (GRAPPH), a shared research facility and hub that works to democratize the use of geospatial techniques and data. Termed, “geomarkers”, these data include community- and individual-level exposures and characteristics, such as substandard housing, green space, air pollution, poverty, and crime. Substantial infrastructure investment and methodological development in conjunction with meaningful community engagement and partnership is taking place in order to harness the full potential of these data to predict and prevent adverse outcomes. GRAPPH currently provides support through both consulting and as a budgeted component of extramural funding across Cincinnati Children's for research, clinical, and quality improvement projects that utilize geospatial data and methods. To increase GRAPPH capacity and make this resource available to the entire institution, its organization and services will become part of the DMAC, whose successful business model will secure long-term sustainability.

Collaboration on Quality Improvement with Anderson Center Learning Networks

Under the leadership of Dr. Maurizio Macaluso, MD, DPH, the Data Management and Analysis Center (DMAC) collaborates extensively with the Division of James M. Anderson Center for Health Systems Excellence and provides comprehensive data management and statistical support for quality improvement initiatives within the learning networks. These include Improve Care Now (ICN), National Pediatric Cardiology Quality Improvement Collaborative (NPC-QIC), Improving Renal Outcomes Collaborative (IROC), Ohio Perinatal Quality Collaborative (OPQC) and the Autism Treatment Network (ATN). Division of Biostatistics and Epidemiology and DMAC staff provide expertise on study design, database development, case report form (CRF) design, data quality and query management, programming for data merges and reporting, measure development, and statistical analysis. Because traditional statistical process control (SPC) methods do not easily accommodate complex data issues encountered by the learning networks, such as repeated measures and clustering of observations (each network has multiple sites involved), Dr. Macaluso and DMAC statisticians are working at integrating advanced statistical modeling techniques (time series analysis, generalized mixed linear models) to more accurately meet the needs of the growing learning network data needs.

Pulmonary and GI Research

In FY 2018, the Division of Biostatistics and Epidemiology faculty and staff collaborated with investigators in the Divisions of Pulmonary Medicine, and Gastroenterology, Hepatology and Nutrition. We are proud of our partnership with these divisions, which ranked #2 and #1 in U.S. News & World Report - 2018 Pediatric Hospital Ranking by Specialty. Pediatric Pulmonology published a study by Dr. Nanhua Zhang, PhD, and colleagues in the Division of Pulmonary Medicine, documenting a significant relationship between glucose and insulin values obtained by OGTT with EC among non-diabetic patients with preserved lung function. In the same journal, Dr. Md Monir Hossain, PhD, MSc, and colleagues in the Division of Pulmonary Medicine documented that opacities, lucencies, and overall heterogeneity of lungs detected via quantitative CT can distinguish bronchopulmonary dysplasia (BPD) patients from healthy controls and that these abnormalities decrease with age across BPD patients. The Science -Translational Medicine Journal published a study by Dr. Lin Fei, PhD, and a colleague in the Division of Gastroenterology, Hepatology and Nutrition identifying serum MMP-7 (alone or in combination with GGT) as a diagnostic biomarker for biliary atresia that may serve as a therapeutic target. PLoS ONE published another study co-authored by Lin Fei and members of Division of Gastroenterology, Hepatology and Nutrition that found that the exposure of Cxcr2-/- mice to SMZ/TMP suppressed cholestasis, decreased the incidence of biliary obstruction, and improved survival.