Table 2.

Association of new receipt and time-updated exposure to ACE-Is or ARBs with recurrent AKI

VariableRecurrent AKI, NFollow-Up Person-YrCrude Rate of Recurrent AKI per 100 Person-Yr (95% Confidence Interval)aAdjusted Odds Ratio (95% Confidence Interval)a
Off ACE-I/ARB, n=10,242192531,4496.1 (5.9 to 6.4)Reference=1
On ACE-I/ARB, n=185316635175.7 (4.9 to 6.5)0.71 (0.45 to 1.12)
  • We used variables listed in Table 1 as candidate covariates to estimate the separate inverse probability weights for ACE-I/ARB exposure and for right-censoring events, and we specified acute heart failure, eGFR, proteinuria, and systolic BP levels as the relevant potential time-dependent confounders (see Supplemental Figure 1 for directed acyclic graph). ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker.

  • a Crude rates and marginal odds ratios of recurrent AKI among eligible adults with no prior heart failure or prior exposure to ACE-Is or ARBs and who experienced an episode of AKI while hospitalized between January 1, 2006 and December 31, 2013. Rates are presented stratified by periods receiving or not receiving ACE-I or ARB therapy during follow-up. Marginal odds ratios weighted by the inverse probability weights for ACE-I/ARB exposure, right-censoring due to death, end of membership, and occurrence of ESKD.