Public Health Informatics

The overarching goal of this work is to research multi-disciplinary approaches to the analysis of significant changes in time series measurements, relevant for occupational health scenarios, by incorporating contemporary methods from the fields of sensor development and data analytics.  Significant changes in environmental (e.g., particulate matter, chemical hazards) and physiological measurements (e.g., heart rate, core body temperature) can be viewed retrospectively to assess the significance of an intervention, or prospectively which can lead to technologies for real-time feedback mechanisms or alert systems to prompt implementation of controls.  This interdisciplinary and translational project also aims to increase the diffusion of innovation of methodologies for analysis of dependent data into the practice of occupational health.   This project is funded by the National Institute for Occupational Safety and Health. 

This environmental epidemiology project is a mentored research award (K99/R00) from the National Institute of Environmental Health Sciences (NIEHS). The project's goals are to 1) create models of atmospheric dispersions of agricultural organophosphorus and pyrethroid pesticide applications over a 20 year period in Arizona, 2) to construct a case control study of ADHD in Arizona by applying a validated algorithm to Medicaid (AHCCCS) claim records, and 3) to examine associations between the modeled ambient pesticide concentrations during fetal/early life and ADHD. The project also employs the use of novel mixture methods, primarily Bayesian Kernel Machine Regression, to assess the potential for mixture effects of different pesticide exposures. Beate Ritz is a Co-Mentor for this project with Paloma Beamer. Avellino Arellano is a collaborator for this project as well. 

Colorectal cancer is one of the most common malignancies in the United States. There is an increasing number of studies using recurrent colorectal adenomas to evaluate the prevention effect for some promising agents. The number of recurrent colorectal adenomas is often measured by performing colonoscopy, which is known to miss a small percentage of existing adenomas and result in misclassification on recurrence status. In addition, some participants might not comply with the schedule of follow-up colonoscopy, which is scheduled to be performed once at the end of the study and, therefore, have variable follow-up lengths compared to the compliant participants. The reasons that a participant cannot comply with the schedule of follow-up colonoscopy could be informative of the risk of recurrence and then bias the results derived from statistical methods that do not adjust for noncompliance. Conventional statistical methods for colorectal adenoma prevention trials cannot simultaneously incorporate misclassification and variable follow-up into analysis and cannot adjust for informative non-compliance without strong assumptions and, furthermore, may incorrectly produce equivocal results for some promising nutritional or chemopreventive agents. The purpose of this application was to develop sophisticated and appropriate statistical models to describe the relationship between the preventive agents and recurrence of colorectal adenomas. We used a latent variable recurrence model, which assumed a portion of non-recurrent participants were misclassified due missing existing adenomas at follow-up colonoscopy, to handle misclassification and a weight function to incorporate the length of follow-up into analysis. The prognostic factors for risk of recurrence was incorporated into the weight function to adjust for potential informative non-compliance. The hope was that a  better understanding of the relationship between preventive agents and recurrence of colorectal adenomas would allow clinical investigators to identify an agent that truly reduces recurrence of colorectal adenomas. This project was funded through a grant from the National Cancer Institute. 

ADHS16-114281 Derksen (PI) 2016-2020 CDC award to Arizona Department of Health Services, ADHS-AzCRH Interagency Services Agreement (ISA) - provides free online continuing medical education via the UA Virtual Lecture Hall. Find more information at: Prescription Drug Misuse/Abuse Reduction Initiative.

This project was 100% FTE from July 2011 to June 2012.  Project research involved data cleaning/scrubbing, data manipulation, statistical analysis, data reporting. This project responded to a data request regarding health care workforce information. The training involved with this project included training an MIS graduate student in public health informatics. This student was a research assistant hired with a fund to setup a data warehouse. Joe Tabor was a contributor on this project but has since left the University of Arizona.

Development of a multi-organ transplant information system

The International Pancreas Transplant Registry (IPTR) was founded in 1980. In the beginning it collected information for pancreas and islet transplants but split in 1989 into the IPTR and the ‘International Islet Transplant Registry (ITR). Despite the split, the members of the IPTR are actively involved in the Collaborative Islet Transplant Registry (CITR) to promote its mission. This is especially important since physicians treating diabetic patients still underestimate the beneficial effect of ß-cell replacement therapy.  The IPTR has collected core information on over 50,000 pancreas transplants done worldwide; performed analyses on outcomes according to multiple variables; and communicated the information by publications and presentations at scientific meetings, making it freely available to the health care community as well as the general public to advance the field of ß-cell replacement therapy for diabetes mellitus.    To minimize transplant center effort, maximize data completeness, and work effectively, the IPTR cooperates with other registries such as the United Network for Organ Sharing (UNOS) and Eurotransplant. These core data from all centers is supported by more comprehensive information from the University of Minnesota’s and University of Arizona's multi-organ databases. We collect in-depth information about pre-transplant comorbidity, surgical techniques, donor pancreas quality, post-transplant complications, immunosuppressive protocols, and long-term follow-up. This enables us to do more comprehensive multivariate analyses for pancreas as well as for islet transplants.    Over the last decade we saw a substantial improvement in the short- and long-term outcome of pancreas transplants. We estimated that the risk-adjusted half-life of patients which received a simultaneous pancreas/kidney transplant in 1998/99 reached 148 months. Patients which received a solitary pancreas transplant are estimated to have a half-life of 74-79 month. The success can be attributed to the refinement in surgical technique, better less diabetogenic immunosuppressive drugs, and better pre- and post-operative prophylaxes. Since especially the long-term outcome of solitary pancreas transplants needs improvement more strategies and better immunosuppressive protocols have to be developed to reach this goal.     Pancreas transplantation is an invasive procedure with the risk of surgery. Islet transplantation in contrast contains a much lower risk. We plan to translate the lessons we learnt to improve pancreas transplantation into islet transplantation.   Models have to be refined to find the best kind of ß-cell replacement therapy for a patient with a specific comorbidity profile. Important impact factors on outcome are the status of secondary diabetic complications of a potential recipient.    There are also criteria to be developed for the allocation of a deceased donor pancreas. At the moment mostly older overweight donor pancreata are used for islet transplantation, while the better quality donors are used for whole organ transplants. More comprehensive models have to be developed to show the impact of those donor factors on the outcome of pancreas and islet transplants.    Answering such questions will allow us to improve the outcomes of pancreas transplants and also gain important information for the advancement of islet transplants. 

The goal of this project was to develop and implement an electronic decision rule within the health system’s electronic medical record to detect hospitalized patients at risk of having severe sepsis. To do this, the sensitivity, specificity, and predictive value of the rule needed to be determined. This project was part of data generation and manipulation, such as statistics, modeling, results generation, intervention effectiveness, and program evaluation. Project funding came from the University of Alabama at Birmingham.