|
BACKGROUND/RATIONALE:
Patient safety initiatives have rapidly escalated in the last few years due to increased pressure on hospitals to reduce medical errors. The Agency for Healthcare Research and Quality (AHRQ) has recently developed a methodology to screen for potential patient safety events using administrative data. As safety measures, the Patient Safety Indicators (PSIs) are attractive because they are based on administrative data, which is readily available; they are standardized and therefore may be useful for benchmarking; they are risk-adjusted; and they are easy to implement. Although AHRQ originally developed the PSIs for quality improvement (QI) initiatives, they are currently being applied in ways that are inconsistent with their intended use.6 At present, over 100 organizations are using the PSIs for national, state, and regional public reporting; pay-for-performance; and hospital profiling.6-8 Increasing interest in these activities makes it essential that the PSIs accurately reflect hospital safety performance. There is an urgent need to validate these indicators before the PSIs are recommended as measures for assessing hospital safety and for rewarding better performance.
OBJECTIVE(S):
We propose to validate the AHRQ PSIs relevant to the VA, modify them to improve their validity and clinical meaningfulness, and test their utility for both identifying adverse events and measuring comparative patient safety performance. Our specific objectives include: 1) develop collaborations with key stakeholders to guide us in selecting and validating the PSIs; 2) investigate the validity of the PSIs by review of the VA's electronic medical record (EMR); 3) identify explicit processes and structures of care associated with individual PSIs; 4) revise and improve the PSIs using multiple data sources and settings of care; and 5) assess the utility of the PSIs for QI and performance measurement.
METHODS:
To accomplish Objective 1, we will work with a steering committee composed of key VA and non-VA senior managers and leaders who will assist in prioritizing the PSIs for validation and evaluating the usefulness of the measures. For Objective 2, we will compare clinical data abstracted from EMRs with that obtained from administrative data-based information to determine the criterion validity of the PSIs (i.e., false positives and false negatives). In Objective 3 we will conduct site visits to six selected facilities to assess the attributional validity of the PSIs (i.e., is some of the variation in risk-adjusted PSI rates explained by structures and processes of care) in Objective 3. In Objective 4 we will explore a number possible methods for revising and improving the PSIs, such as adding present on admission data to the PSIs, adding limited clinical data to PSI algorithms, and linking the index hospitalization with preceding inpatient and outpatient data as well as post-discharge data. Finally, we will evaluate the utility of the PSIs through focus groups with potential end-users, and evaluate the ability of selected PSIs to improve care through a modified Breakthrough Series utilizing quality improvement techniques.
FINDINGS/RESULTS:
We have results from the chart abstraction of PSI PE/DVT cases. We found a PPV of 44%. Out of 112 cases abstracted, we found the following:
49 cases (44%) of true post-operative PE/DVT.
24 cases (21%) of coding-related inaccurate diagnosis.
10 cases (9%) of remote history of PE/DVT.
13 cases (12%) of pre-procedure diagnosis.
16 cases (14%) of present of admission.
IMPACT:
Upon completion of the study, the potential impact associated with this project is as follows:
This study will provide critical information regarding which PSIs are currently useful and which are not, which PSIs may be useful for QI, quality monitoring, or case-finding, and which may be appropriate for public reporting and performance measurement. It will also identify deficiencies in process of care and examine whether other data sources, such as the VA's electronic medical record (EMR), may help improve the validity and utility of current measures.
PUBLICATIONS:
None at this time.
DRA:
Health Services and Systems
DRE:
Technology Development and Assessment, Communication and Decision Making, Resource Use and Cost
Keywords:
Risk factors, Safety, Implementation
MeSH Terms:
none
|
