The overall goal of this project is to assess the extent to which nurse staffing in VA settings is related to patient care outcomes. A VHA nursing outcomes database of clinically relevant, comparable quality indicators is intended to provide strategic decision support through timely internal and external benchmarking to VA, expedite testing and evaluation of interventions to define evidence-based nursing practices, and to impact health care policy decisions.
To develop a standardized methodology to collect data related to nurse-sensitive quality indicators within VHA facilities and integrate these into a national database. The purpose of this effort is for benchmarking and comparing local, VISN, and national nursing and patient quality outcomes.
For the pilot project, VA Medical Centers (VAMCs) were recruited in three stages: (1) a stratified random sample of twelve VAMCs with acute care beds; subsequently, additional groups of VAMCs will be included: (2) an additional 39 VAMCs in 8 VISNs from the Clinical Nurse Leadership Project; and (3) all remaining VAMCs with acute care inpatient facilities. The first two stages of the recruitment were completed in December 2005. Using Class III software developed for this program and installed in VistA at local VAMCs, participating facilities extracted and submitted data to the VANOD database on a monthly basis for four indicators: nursing hours per patient day (HPPD); skill mix; staff musculoskeletal injuries; and patient falls. Pressure ulcer prevalence is collected using a manual, inpatient census-wide survey twice a year at each VAMC. Data on patient satisfaction with nursing care were obtained from the Office of Quality and Performance for the 12 pilot project sites. Registered nurse (RN) job satisfaction was collected by web-based surveys, adapted from Lake’s (2002) Practice Environment Scale of the Nursing Work Index. Reports with benchmarking information were presented as ProClarity web reports. To evaluate reliability of data collected, we compared data collected using the Class III VistA application with (1) those previously collected using Excel spreadsheets for the first four indicators and (2) nurse staffing data reported through the VA DSS databases.
Cross-validation was completed for 10 facilities recruited in stage 1. Results suggest that HPPD and skill mix calculated based on nursing hours in VistA NURS AMIS 1106 ManHours match well with data reported in Excel spreadsheets. However, nurse staffing data reported in DSS databases may be 10-15% higher than those collected using Excel spreadsheets. Patient falls from VistA QA Incident Report and staff musculoskeletal injuries from VistA ASISTS database match reasonably well with data collected using Excel spreadsheets. However, patient falls based on VistA fall templates and nursing hours based on VistA Time & Attendance present three major data issues in using VistA as the principal data source: (1) lack of standardization within and across facilities in certain fields/files adds difficulty in interpreting the data submitted from sites; (2) missing or out-dated data in extracted files makes it difficult to achieve accurate benchmarking across time and facilities; and (3) unavailability of important data elements in the system compromises the accuracy and granularity of the data reported. Approximately 70% of respondents to the inpatient satisfaction survey stayed on only one unit during their admission, making it possible to attribute satisfaction to a single nursing unit for these patients. More than 90% of patients were satisfied with their overall hospitalization experience. Approximately 87%, 85%, and 72% of patients reported satisfaction with nursing care, discharge planning, and pain management, respectively. Of the five aspects measured by the Practice Environment Scale for RN job satisfaction, RNs were most satisfied with nurse-physician collegial relations and least satisfied with staffing and resource adequacy.
The VHA nursing outcomes database of quality indicators is intended ultimately to improve care for veterans. Standardizing data entry practice and processes are critical for building a valid and reliable database to enable ongoing examination of processes and relationships between nurse-sensitive patient outcomes and structural/organizational elements.
External Links for this Project
- Sales A. Lessons from the VA EHR: Applications to nursing care. Paper presented at: AcademyHealth Annual Research Meeting; 2006 Jun 27; Seattle, WA. [view]
- Li Y, Rick C, Charns M, Buffum M, Meterko M, Lowy E, Collins B, Alt-White A. Value and Challenge in Measuring and Managing Staffing and Other Nursing Factors in Acute Care [workshop]. Paper presented at: VA HSR&D National Meeting; 2006 Feb 15; Arlington, VA. [view]