Comparing measures of patient safety for inpatient care provided to veterans within and outside the VA system in New York. Issue 1 (1st February 2008)

• Record Type: Journal Article
• Authors: Weeks, W B; West, A N; Rosen, A K; Bagian, J P
• Is Part Of: Quality & safety in health care. Volume 17:Issue 1(2008)
• Page Start: 58
• Page End: 64
• Publication Date: 2008-02-01
• DOI: 10.1136/qshc.2006.020735
• Languages: English
• ISSNs: 1475-3898

Abstract

Abstract : Context: Previous studies have compared measures of patient safety for veterans using the VA system to the general population. Discrepancies in the results of those studies suggest that the choice of an appropriate comparison group is critical for accurate interpretation of results and for determining whether to take actions to address findings. We explored another method of providing consumer information by comparing the experiences of VA enrolled patients who received care in the VA to those who received care outside the VA system. Objective: For male veterans living in New York State and enrolled in the VA healthcare system, to determine (a) whether those who obtain care outside the VA system experience different measures of patient safety than those treated within the VA system, and (b) whether cross-system comparisons of measures of patient safety among older veterans reflect those of the full age spectrum. Design: Retrospective cohort analysis. Setting: All VA and non-VA hospitals in New York State. Patients: 353 570 male New York State residents who were enrolled in the VA system in 1998, 1999 or 2000. Main outcome measures: The Agency for Healthcare Research and Quality (AHRQ) has developed Patient Safety Indicators (PSIs) from hospital discharge data. To standardise these indicators across settings, AHRQ has provided software for risk-adjustment purposes so that the indicators can be compared across settings of care. We used the PSI software to calculate risk-adjusted PSI rates with 95% confidence intervals to compare veterans' inpatient care provided within and outside the VA system. Results: Risk-adjusted rates for nine of 15 PSIs did not differ between care provided within and outside the VA system. However, compared with care provided outside the VA system, risk-adjusted rates of decubitus ulcer, postoperative sepsis, infection due to medical care, postoperative respiratory failure and postoperative metabolic derangement occurred at lower rates within the VA system, while death in low mortality DRGs occurred at a higher rate in the VA system. Findings for patients aged 65 and older were similar to those of the entire age spectrum. Conclusions: Using AHRQ's PSI software, male veterans in New York who obtain their inpatient care within the VA received care that was comparable with or somewhat better than those who obtained their inpatient care outside the VA. The experiences of older patients reflected those of younger patients. Given that our findings are much more similar to reported comparisons between the VA and Medicare than to comparisons between the VA and the general population, we conclude that, should system comparisons be made, choice of comparison groups will be critical to accurate interpretation of findings; however, prior to such interpretation, the validity of the PSIs must be determined within VA.