You are here

Cohort Builder for Healthcare Quality Assurance and Comparative Health Effectiveness Research


OBJECTIVE: Define and prototype architectural alternatives resulting in an easy-to use cohort builder for clinicians, nurses, and QA personnel. The cohort builder would be used for conducting quality assurance and comparative health effectiveness research; to recapitulate findings in the literature; and to remediate patient care issues for chronically ill patients. DESCRIPTION: The mantra of the Military Health System is to deliver available, accessible, acceptable, high quality, continuous, and cost effective healthcare to 9.6M beneficiaries. Each budget cycle the Military Health System is called upon to deliver continuously improving healthcare, yet has ever-increasing constraints placed on its resources. In order to do more with the same or less level of resources, and without adversely impacting health outcomes and patient safety, the Military Health System should engage in comparative health research studies to ascertain which treatments provide the best outcomes at equal or less cost than other therapies. Additional studies should be conducted to determine why one clinician may get better outcomes at equal or less cost than his/her peers, considering case mix (i.e. clinician profiling). Yet the Military Health System does not currently have the clinical analytical tools to conduct such analyses. Although the Military Health System (MHS) has a number of Automated Information Systems (AIS) in its inventory, including AHLTA, Clinicomp Essentris, CHCS, ICDB, MDR, M2, CDM, HSDW, AFCHIPS, Air Force Population Health Portal, Medical Home Clinical Data Mart, and others, there is no system that allows clinicians to build cohorts for quality assurance studies or comparative health effectiveness research, and compare those cohorts as to differences in health outcomes. In the current state, clinical quality assurance and comparative health effectiveness research studies cannot be easily conducted. Executing such studies requires extensive expertise in the form of data priests and statisticians, to identify appropriate data, pull it from the right sources, consolidate it, and analyze it. Conducting such studies can take months. Issues impacting the care of chronic care patients at the point of care cannot be easily identified or remediated at the point of care. Studies in the literature cannot be re-capitulated in a clinician"s panel in a matter of minutes. Of additional concern is the current rapid promulgation of domain-specific data marts and registries in the new Centers of Excellence, with no robust toolsets to conduct health outcomes on those registries. In the envisioned state, access to information would be democratized, with no need for data priests or statisticians. Clinicians, nurses, and QA personnel would access an easy-to-use cohort builder on their desks, that allows one to quickly build cohorts on any combination of parameters collected in base transaction systems; which supports different time events for different parameters; and provides for risk windows and blackout periods; for the primary purpose of remediating issues identified in chronic care patients. In addition, the cohort builder would be self-documenting in terms of identifying the basis for the cohorts, and produce analysis suitable for publication in research journals. The tool would incorporate a statistical service to make automated, on the-fly comparisons between cohorts as to differences in morbidity and mortality. All parameter lists, cohorts, and outcomes could be saved as objects and reused. The output of one study could become the input for another study. PHASE I: In Phase I, awardees will outline strategic, operational, and technical architecture alternatives for a prototype cohort builder to support health outcomes studies. The alternatives would include an analysis of existing GOTS and COTS products and their ability to be integrated into current clinical intelligence frameworks. The output of Phase I is a report that would provide details on the cohort builder would integrate with existing and planned MHS transactional, data marts, registries, and data warehouses, and which would serve as a complete design document to actually build a prototype cohort builder. PHASE II: In Phase II, the SBIR recipient(s) will build or buy a cohort builder, integrate it with various MHS data sources, and pilot the tool with approximately 20 clinicians at the Walter Reed Army Medical Center, USU, and Centers of Excellence. During the prototype, the Phase II recipient(s) would collect data on the usability of the cohort builder in conducting health outcomes studies, and conduct measurements on how the cohort builder improves the effectiveness and efficiency of outcomes studies over previously conducted manual efforts. Any particular studies conducted would require IRB approval in advance. PHASE III: Clinical intelligence and/or electronic health record vendors that support the analysis of healthcare delivery in traditional inpatient and outpatient settings could commercialize this research for use in new markets, including the life sciences. Those conducting comparative effectiveness of commercial products might even adopt the work outside the healthcare of life sciences arena.
US Flag An Official Website of the United States Government