Read on Twitter
Just Out: Our latest #lowvaluecare paper on low-value imaging within the Veterans Health Administration in @JAMANetworkOpenThread
@PittGIM @PittGIMStrong @PittCP3 @VAequity @VAHSRD https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2770750?utm_source=twitter&utm_medium=social_jamajno&utm_term=3729475083&utm_campaign=article_alert&linkId=100221055
VA is different from other healthcare systems: 1) it’s non fee-for-service, 2) insulated from malpractice litigation, and 3) has lots of decision support when ordering health services… so there shouldn’t be much low-value care...
But earlier studies suggest that low-value care still occurs within VA, which begs the question whether low-value care in VA is more common than we might expect, how does it vary, and if so, why?
We found that low-value testing for headache, low-back pain, syncope, and sinusitis was common, affecting 5-21% of Veterans overall, and also those at greatest risk of 1-year mortality.
There was 2 to 5-fold variation in low-value testing across VA Medical Centers (VAMCs) with a uniformly small intraclass correlation coefficient for the delivery of low-value testing for each disease state…
...suggesting that unique local factors, such as those related to provider culture and variations in decision support for ordering tests at each VAMC, may drive low-value testing as opposed to simply sociodemographic and VAMC characteristics.
The correlation between use of different low-value tests within the same VAMC also suggests that the delivery of such low-value care within VA is driven by systemic factors that aren’t specific to the delivery of any one individual test.
Our study isn’t without limitations… value is inherently hard to measure using claims, this study didn’t focus on provider characteristics, and we didn’t account for dual use of VA and non-VA care (more to come on that soon…)
But together, these findings reinforce that 1) despite its unique features, low-value testing is common in VA, 2) We need to dig deeper to ID those cultural and other drivers of low-value care, and 3) effective interventions will likely need to be tailored to the VAMC-level.
A huge thanks to my mentor-extraordinaire @WalidGellad, @maizenblueheels for her collaboration to ensure this research continues, @A_Schwa for his foundational work in measuring low-value care, and the support of the @PittICRE KL-2 Award.
