BACKGROUND/RATIONALE:
To develop a learning health care system (LHCS), VHA leadership must understand where quality improvement is needed via valid and actionable performance measurement and reporting. Current performance measurement and reporting systems focus exclusively on facility-level variability across diverse performance metrics. This approach assumes that facility-level variability is reliable and provides valid information for decision-making regarding performance improvement. Also, current methods of performance reporting may not be meeting the needs of VHA operational leaders for actionable information related to local variations in performance.

OBJECTIVE(S):
Our goal is to advance the science of multi-level health care performance measurement toward more actionable feedback about sources of variability in performance. To achieve this, we will (1) build an analytical model of inpatient outcomes and their predictors at multiple levels of the health system. We assert that variability in inpatient outcomes and predictors at the inpatient unit level will exceed corresponding variability at the facility level, and that unit- and facility-level service burden and resources will be predictors of clinical outcomes. We will also (2) present data in feedback reports targeted to front-line clinicians and administrators who can use the results to improve care quality.

METHODS:
For our first objective, we will build a multi-level structural equations model to simultaneously evaluate variation at the inpatient unit level and facility level. The study population includes all inpatient stays on acute care nursing units at VA acute care facilities. Inpatient outcomes include adverse events from the Inpatient Evaluation Center (IPEC) and readmissions calculated using patient data from the corporate data warehouse (CDW) and VA/CMS data from VIReC. Predictors of patient outcomes include measures of service-burden (average and total relative risk score and diagnosis-related group (DRG) index per unit per day) from CDW and Health Economics Resource Center, measures of resources (total hours worked by nurses per unit per day) calculated from payroll and staff data from Human Resources Information Services (HRIS), VA Time and Attendance System (VATAS), and Personnel and Accounting Integrated Data (PAID), and measures of nurse satisfaction from the National Center for Organization Development's (NCOD) All Employee Survey (AES). To achieve our second objective, we will conduct individual interviews with 5-6 key informants from each of five levels of VHA (national, network, facility, service line, and unit) and use qualitative data from these interviews to develop facility performance data display templates aligned with current science on feedback design. For each of the tailored set of templates, usability will be tested by 5 participants in 2-3 iterative rounds.

FINDINGS/RESULTS:
We have data for 2,524,350 subjects (patients and staff) from administrative datasets across 577 acute care nursing units in 122 facilities from FY 2015-2017. We constructed data crosswalks to support linking staff and patients to nursing units by date for AES, BCMA, and HRIS data.

IMPACT:
Consistent with the initial submission, people throughout the VHA want to provide excellent care and desire the information and resources to make excellent care possible. Our goal is to provide valid information on performance and quality of care in reports targeted to front-line clinicians and administrators who can use this information to initiate appropriate change. By providing these necessary tools to those who deliver and distribute care to Veterans, we aim to enable continuous and effective quality improvement throughout the VHA.

---

External Links for this Project

NIH Reporter

Grant Number: I01HX002034-01A2
Link: https://reporter.nih.gov/project-details/9287114

---

PUBLICATIONS:

Journal Articles

1. O'Mahen P, Mehta P, Knox MK, Yang C, Kuebeler M, Rajan SS, Hysong SJ, Petersen LA. An Alternative Method of Public Reporting of Comparative Hospital Quality and Performance Data for Transparency Initiatives. Medical care. 2021 Sep 1; 59(9):816-823. [view]
2. Massarweh NN, Chen VW, Rosen T, Dong Y, Richardson PA, Axelrod DA, Harris AHS, Wilson MA, Petersen LA. Comparative Effectiveness of Risk-adjusted Cumulative Sum and Periodic Evaluation for Monitoring Hospital Perioperative Mortality. Medical care. 2021 Jul 1; 59(7):639-645. [view]
3. Eck CS, Knox MK, Mehta PD, Petersen LA. Estimating the Relationship Between Nurse Staffing and Medication Pass Workload Using National Barcode Data. Nursing Research. 2024 Jul 17. [view]
4. Shah NR, Ahmed ST, Winchester DE, Ramsey DJ, Akeroyd JM, Wu WC, Waldo SW, Schofield RS, Ballantyne CM, Petersen LA, Virani SS. Facility-Level Variation in Stress Test Utilization in Veterans With Ischemic Heart Disease. JACC. Cardiovascular imaging. 2019 Jul 1; 12(7 Pt 1):1292-1293. [view]
5. Massarweh NN, Chen VW, Rosen T, Richardson PA, Harris AHS, Petersen LA. Relationship Between Perioperative Outcomes Used for Profiling Hospital Noncardiac Surgical Quality. The Journal of surgical research. 2021 Aug 1; 264:58-67. [view]
6. O'Mahen P, Petersen LA. Subsidizing Affordable Care Act Marketplace Insurance for Adults Below the Federal Poverty Level. JAMA health forum. 2021 Jun 4; 2(6):doi:10.1001/jamahealthforum.2021.0871. [view]

DRA: Health Systems Science
DRE: Technology Development and Assessment, TRL - Applied/Translational, TRL - Development
Keywords: Clinical Performance Measures, Efficiency
MeSH Terms: none