• Projects
  • Active Flow Control
  • Adaptive Reconfigurable Control
  • Diagnostics and Prognostics
  • Intelligent Guidance and Trajectory Reshaping
  • Real-Time Modeling and Prediction
  • Tools for Healthcare Assessment
    • Personal Health and Fitness Wizard
    • Quality of Care Assessment
    • Nurses

Nurses

Quality improvement principles are being used increasingly to improve productivity and efficiency of health care delivery and to help contain costs. Structural changes are often made to the health care system without benefit of clinical research to evaluate appropriateness. These changes have the potential to impact safety and quality of patient care, in part because the changes often lead to a reduction in the number of registered nurses (RNs) having direct patient contact and substitution of unlicensed personnel for RNs. Under such circumstances, there is a need for reliable information that can be used to evaluate the medical consequences of changes in nurse staffing numbers and skill mix. Such information is essential to the design of workable practice guidelines and sound therapeutic and financial decisions.

The purpose of this project was to develop, implement, and demonstrate a methodology for costing the American Nurses Association (ANA) Nursing Quality Indicators. In particular, focus was on investigating the relationships between nurse staffing levels and patterns, patient outcomes (e.g., nosocomial infections), and in-hospital patient charges. As part of the effort, Barron Associates performed an outcomes and costing analysis of patient discharge data from two states for each of two years, representing a total of 12 million patients. Our analytical work identified several important reasons why conventional modeling approaches often fail to uncover relationships that exist among data variables. Data mining results demonstrated that there are varying sensitivities, as a function of patient illness (e.g., Diagnosis Related Group category), in the relationships between percentage of registered nurse (RN) hours devoted to patient care and corresponding patient total charges. Patient-level modeling performed by Barron Associates yielded first evidence that increasing RN hours as a percentage of total nursing hours reduces patient total charges in a significant number (approximately 10%) of identifiable (and often related) diagnostic categories.

As part of the effort, an on-line Web site was launched and hosted by Barron Associates, wherein all data, analysis and synthesis tools (including GNOSIS and GNOSIS-generated "what-if" models), and findings were made available to authorized users.