Lead/Presenter: Jacob Kurlander, COIN - Ann Arbor
All Authors: Kurlander JE (VA Ann Arbor)
Kolbe M (University of Michigan)
Scheiman JM (University of Michigan)
Weissman A (American College of Physicians)
Piette J (VA Ann Arbor)
Rubenstein JH (VA Ann Arbor)
Saini SD (VA Ann Arbor)
Objectives:
Proton pump inhibitors (PPIs), which have been linked with multiple serious adverse effects, are commonly overused in VA. We sought to evaluate provider attitudes towards stopping PPIs in case vignettes of patients with varying degrees of upper gastrointestinal bleeding (UGIB) risk.
Methods:
Online survey of a representative sample of the American College of Physicians in 2013. Participants were presented at random with 2 of 3 patient vignettes, all featuring a 70-year-old woman taking omeprazole 20 mg daily and recently diagnosed with osteopenia, increasing her risk of bone fracture, a PPI adverse effect. PPI indication and UGIB risk varied in the vignettes: (1) gastroesophageal reflux disease (GERD) well controlled for years with normal upper endoscopy (low risk); (2) prevention of UGIB while on low-dose aspirin (LDASA) and warfarin, asymptomatic (moderate risk); or (3) prevention of UGIB while on LDASA with remote NSAID-related peptic ulcer disease (PUD), asymptomatic (high risk). Participants were asked how likely they would be to stop PPI (1-7 scale). Responses were dichotomized as > 5 (stop PPI) or < 5 (continue PPI). We used Poisson regression with robust error variance to estimate the impact of PPI indication on stopping PPI, controlling for physician age, gender, training status, and concern about PPI adverse effects, measured on 1-7 scale.
Results:
487 invited participants completed the survey for a response rate of 53%. 21.8% were trainees, 57.9% practiced general medicine, 2.3% gastroenterology, and 18.1% another subspecialty. In multivariable analysis, respondents shown the moderate-risk vignette (LDASA and warfarin) were 1.47 times (95% CI: 1.21-1.78, p < 0.001) more likely to stop PPI relative to the low-risk (GERD) vignette, while respondents shown the high-risk vignette (LDASA and prior PUD) were 1.93 (95% CI: 1.62-2.31, p < 0.001) times more likely to stop PPI relative to the low-risk (GERD) vignette. Concern about PPI adverse effects (RR 1.11 for one-point change, 95% CI: 1.05-1.17, p < 0.001) was positively associated with stopping PPI. Older physician age (RR 0.92 for each decade of life, 95% CI: 0.86-0.98, p = 0.007) and being a current trainee (RR 0.80, 95% CI: 0.66-0.98, p = 0.03) were both independently associated with a lower probability of stopping PPI. Gender had no effect.
Implications:
In this national sample, decisions about whether to stop or continue PPIs are highly discordant with bleeding risk.
Impacts:
Without better guidance, physicians' well-intended efforts to reduce PPI use, and possible adverse effects, may have serious unintended consequences.