Frequency and costs of low-value preoperative tests for patients undergoing low-risk procedures in the veterans health administration

High-value, patient-centered health care includes selectively ordering preoperative tests that may inform clinical management or improve patient outcomes. Tests unlikely to meet these criteria may cause avoidable harm, inconvenience, and waste of resources that could be used for higher-value services. For years, government healthcare agencies and professional organizations have been recommending the avoidance of routine preoperative testing for low-risk procedures (Balk et al. 2014; National Institute for Health and Care Excellence 2016; Fleisher et al. 2015). However, we found that almost half of OSS 1 procedures received by VA patients in FY19 were preceded by at least one potentially low-value preoperative test. Using Medicare Average Costs, we estimated that the 321,917 preoperative screening tests prior to OSS 1 procedures may represent up to $11,505,170 in low-value care.

As noted, some portion of this testing may represent high-quality care (e.g., opportunities for overdue screening) or tests that are unrelated to the OSS1 procedures. If the distribution of testing was uniformly modest (e.g., 5 or 10%), the case for investments in quality improvement might be easier to dismiss. However, we can see in Figs. 1 and 2 that there exist many VA facilities that routinely test the vast majority of patients prior to low-risk procedures, and 27 facilities with estimated associated costs over $200,000. For these facilities, these results highlight a significant opportunity to improve quality by providing less unnecessary care.

Beyond describing the magnitude and distribution of opportunities for quality improvement, another purpose of this study was to identify the patient characteristics associated with preoperative testing. We found older age, being female, not married, black, and having comorbidities were all associated with higher odds of testing. The large ICC in the model predicting receipt of at least one test (0.189) suggests that much of the variance in preoperative testing is at the facility-level. In other words, the likelihood of getting tested is as much a function of where you are treated as your specific medical profile.

Although excluded from our primary analyses, general anesthesia was received for 16.5% of OSS1 procedures in FY19, of which 82.2% underwent preoperative testing. Anesthetic risk, not just procedural risk, needs to be factored into the decision to order preoperative tests. We previously found that much of the variance in using general anesthesia for a low-risk procedure (carpal tunnel release) is driven by clinician or facility factors rather than patient characteristics or preferences (Harris et al. 2020). Therefore, although preoperative testing may be justified for patients undergoing general anesthesia, there may be quality improvement opportunities in facilities that commonly or routinely use general anesthesia even for low-risk procedures.

Several limitations are worth noting. First, there is no way to be sure that tests in the 30 days prior to OSS1 procedures were ordered for preoperative screening purposes. Some of the tests we identified may be justified by factors independent of the upcoming low-risk procedure. However, we have no reason to expect that such justifications for the tests differ systematically between facilities. In our previous work on cataract and carpal tunnel release surgery, we excluded 10% of preoperative tests because they were not preceded by a ‘plausible ordering visit, such as ophthalmology or anesthesia consult. As we could not implement this methodology in this study due to the diversity of procedures, it is possible that our estimates of low-value testing are 10% too high due to this limitation. Also, it is unknown to what extent these results might generalize outside of the VA system.

留言 (0)

沒有登入
gif