Premium
VA Rapidly Reduces Adverse Clinical Events Following Randomized Clinical Trial Showing Harm
Author(s) -
Campbell H.,
Murata A.,
Conner T.
Publication year - 2020
Publication title -
health services research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.706
H-Index - 121
eISSN - 1475-6773
pISSN - 0017-9124
DOI - 10.1111/1475-6773.13446
Subject(s) - medicine , hyperkalemia , kidney disease , population , albuminuria , randomized controlled trial , emergency medicine , clinical trial , acute kidney injury , intensive care medicine , diabetes mellitus , environmental health , endocrinology
The VA has policies directing communication between Research, Pharmacy Benefits Management (PBM) and health care providers to decrease time between research findings and clinical practice implementation. A randomized double‐blind trial of type 2 diabetes mellitus patients found increased acute kidney injury (AKI) and hyperkalemia from combination therapy with an angiotensin‐converting enzyme inhibitor and an angiotensin receptor blocker compared to monotherapy. Specific to this trial, we sought to assess the apparent clinical impact of these policies, which create a multicomponent intervention comprising of four dissemination strategies. We conducted a retrospective longitudinal database analysis with data from the VA Corporate Data Warehouse using pooled cross‐sectional data to conduct an interrupted time series analysis (ITSA) using separate models for AKI and hyperkalemia. As the intervention occurred February 2013 to December 2014, we assessed the rate of AKI and hyperkalemia between July 2008 and November 2017. Our denominator was comprised of the population of patients utilizing VA health care services at any VA Health Care System who were diagnosed with type 2 diabetes mellitus, chronic kidney disease stages 1‐3, and moderately or severely increased albuminuria. Although we report our results as the number of events, our numerator was the patients who used combination therapy and who experienced an AKI or hyperkalemia. We also stratified the population due to differences in renal risk. Among all patients, a mean of 66.07 AKI events per month occurred in those using combination therapy in the preintervention period while a mean of 11.32 per month was experienced postintervention. Before the PBM communication, the number of AKI events per 100 000 patients was decreasing at 0.54 per month (95% CI: −0.96, −0.13, P = .011). After the PBM communication, there was a significant change in level: there were 17.97 fewer AKI events per 100 000 patients (95% CI: −28.19, −7.75, P = .001) in the first month; compared to the base trend, there were 0.19 more AKI events per 100 000 patients (95% CI: −0.23, 0.62, P = .37) per month. This translates into the PBM communication being associated with a 53.02%, 55.01%, and 57.51% relative decrease in AKI events experienced per 100 000 patients in the sixth, twelfth, and eighteenth month, respectively, compared to the base trend. For hyperkalemia, among all patients, we observed a mean of 9.49 hyperkalemia events per month in the preintervention period and a mean of 1.77 events post‐intervention. Although ITSA indicated a nonsignificant ten‐fold decrease in events the month immediately following the PBM communication, it found some groups at higher renal risk had a significant decrease in events. We cannot discern whether the change in observed data seems to be solely due to the PBM communication or more than one component of the intervention. Although ITSA cannot determine causality, we estimate 431 AKI events appear to have been averted in the first 18 months if the intervention did not occur. Compared to the seventeen‐year benchmark, the apparent speed and impact of VA’s dissemination strategies of research results between VA Research, PBM, and health care providers are a model for other health care organizations. Department of Veterans Affairs.