Almost one-third of community-dwelling older adults fall every year and some of these falls result in major injuries and costly medical care. Evidence-based clinical guidelines recommend asking older adults about falls annually, and implementing interventions to reduce fall risk among patients in need. Since 2006, the VA Office of Quality and Performance has been monitoring these two quality indicators. For Fiscal Year 2007, data from a national random sample of VA outpatients showed that 33% of adults age 75 were asked about falls in the past year, and 36% of those with two falls, or one fall with injury, received a basic fall evaluation. These VA performance rates are consistent with those seen in a non-VA managed care sample and demonstrate significant room for improvement.
We hypothesized that performance gaps across VA facilities are the result of both clinician factors (e.g., knowledge, attitudes and beliefs) and organizational factors (e.g., availability of resources such as a falls clinic) that affect primary care clinicians' performance on these process indicators. The objectives of this study were to 1) develop a survey to identify factors that impede or facilitate implementation of falls QI and guidelines; 2) pilot test and refine the survey in a sample of primary care clinicians in two VISNs; and 3) recommend directions for QI implementation strategies.
An iterative mixed method approach was used to develop a survey instrument to describe potential causes of the variation in performance. The fall survey was developed using input from: 1) Key respondent interviews with primary care providers working in a medical center, Community-Based Outpatient Clinic (CBOC) and free-standing ambulatory care center, and 2) Technical advice from an expert panel. The draft survey was pre-tested and refined using intensive cognitive interview techniques (n=10). The resulting pilot survey was disseminated to a random sample of 88 primary care providers in 2 VISNs. Descriptive and non-parametric analyses examined variation in responses between responder characteristics and conceptual constructs.
Cognitive interviews revealed that survey items that had been designed for community physicians were not well-received by VA staff. General concepts could be retained but most items were re-written to improve comprehension, relevance to midlevel clinicians as well as physicians, or reflect VA processes of care. The pilot survey had a 28% response rate and most (74%) respondents were mid-level professionals. There was no difference in response rates across VISNs or clinical settings (medical centers or CBOCs), but statistical power to detect differences was limited. Among these primary care respondents, confidence in managing falls was lower than the average confidence level in managing other medical conditions (p = .002). Future roll-out of the survey to VA primary care providers will require a variety of incentives or administrative leadership buy-in to gain higher levels of participation by physicians.
We recommend using the refined survey in a broader sample of VA clinicians in order to guide the development of QUERI Step 4 educational initiatives and related implementation strategies to improve patient outcomes by implementation falls guidelines and quality performance indicators.
External Links for this Project
None at this time.