Key Points
Question
Should renin-angiotensin system inhibitors (RASIs) be discontinued after an acute decline in eGFR?
Results
This cohort study using a target trial emulation approach in 4233 patients who initiated RASI and had a greater than 15% decline in eGFR found that those who discontinued RASIs had a higher risk of end-stage kidney disease and death compared with those who continued RASIs. There was no association between RASI discontinuation and the risk of major adverse cardiovascular events or acute kidney injury.
Meaning
These findings suggest that further study is needed to understand reasons for frequent discontinuation of RASIs and devise strategies to improve their persistent use.
This cohort study examines the cardiovascular, kidney, and mortality risks associated with discontinuing vs continuing renin-angiotensin system (RAS) inhibitors after an acute decline in estimated glomerular filtration rate (eGFR).
Abstract
Importance
Although renin-angiotensin system inhibitors (RASIs) slow the progression of chronic kidney disease, these agents are frequently discontinued if acute declines in the estimated glomerular filtration rate (eGFR) occur after their initiation.
Objective
To examine the risk for cardiovascular, kidney, and mortality outcomes associated with discontinuation vs continuation of RASIs in patients with an acute decline in eGFR after RASI initiation.
Design, Setting, and Participants
This retrospective cohort study using a target trial emulation approach leveraged electronic health record data from the Manitoba Centre for Health Policy that were linked to vital statistics in the province of Manitoba, Canada. Propensity score matching was applied to identify adults (aged ≥18 years) receiving new prescriptions for RASIs between January 1, 2008, and December 31, 2021, who subsequently had an eGFR decline of more than 15% within 90 days of the prescription. Data were analyzed between January 14, 2024, and January 22, 2026.
Exposure
RASI discontinuation (vs continuation) defined based on whether a RASI prescription was refilled within 90 days after a decline in eGFR of 15% had occurred.
Main Outcomes and Measures
The primary outcomes of end-stage kidney disease (ESKD) or death, and secondary outcomes of major adverse cardiovascular events (MACEs) (including myocardial infarction, heart failure, stroke, or cardiovascular mortality) or acute kidney injury (AKI) were examined after 180 days. Cox proportional hazards models were used to compare the exposure with the outcomes in an intention-to-treat approach.
Results
A total of 4233 patients (mean [SD] age, 64.6 [16.2] years; 2697 male [51.1%]) who had a more than 15% decline in eGFR after starting a RASI were included; 1411 patients (33.3%) discontinued RASIs, and 2822 (66.6%) continued RASIs . Patients who discontinued vs continued RASIs had a higher risk of death (hazard ratio [HR], 1.23; 95% CI, 1.07-1.41) and ESKD (HR, 1.74; 95% CI, 1.28-2.38). Their risk of MACEs (HR, 1.13; 95% CI, 0.99-1.28) and AKI (HR, 1.11; 95% CI, 0.90-1.37) was not significantly higher.
Conclusions and Relevance
This cohort study found that discontinuation of RASIs after an acute decline in eGFR was associated with ESKD and death compared with continuing RASIs. These findings suggest that further study is needed to understand reasons for frequent discontinuation of RASIs and devise strategies to improve their persistent use.
Introduction
Use of angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) have both been shown to slow the progression of chronic kidney disease (CKD). Yet, these treatments are not optimally used in clinical practice. In the clinic, 15% to 20% of patients with CKD exhibit an acute decline in estimated glomerular filtration rate (eGFR) with ACEI or ARB initiation, often leading to drug discontinuation.1,2,3,4,5,6 Currently, guidelines recommend tolerating up to a 30% increase in serum creatinine after RASI initiation,1 but these agents are frequently stopped by clinicians in response to smaller declines in eGFR following therapy initiation.7,8
The objective of this study was to examine the association between discontinuation vs continuation of ACEIs or ARBs (hereafter referred to as renin-angiotensin system inhibitors [RASIs]) following an acute decline in eGFR and kidney and survival outcomes in clinical practice. Specifically, in a cohort of new users of RASIs, we examined whether discontinuation of RASIs was associated with the risk of end-stage kidney disease (ESKD) or all-cause mortality and, secondarily, major cardiovascular events (MACEs) or acute kidney injury (AKI) among those who experienced an acute decline in eGFR of more than 15% within 90 days of their RASI prescription.
Methods
Study Population
This cohort study used data from the Manitoba Centre for Health Policy, which are linked to vital statistics in the province of Manitoba, Canada (eTable 1 in Supplement 1). The study was approved by the University of Manitoba Health Research Ethics Board, and the requirement for informed consent was waived due to the retrospective nature of the study. The study design emulated a target trial and adhered to the Transparent Reporting of Observational Studies Emulating a Target Trial (TARGET)9 (eTable 2 in Supplement 1) and Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
The study population included adults aged 18 years or older who started a RASI de novo (defined as a new outpatient prescription for any RASI without any active prescriptions in the 1 year prior) (eTable 3 in Supplement 1) between January 1, 2008, and December 31, 2021 (Figure 1). We excluded patients receiving dialysis or with kidney transplant at the time of RASI initiation or within the first 180 days of follow-up. We also excluded patients with potassium levels greater than 5.5 mEq/L (to convert to mmol/L, multiply by 1) prior to the start of any RASI prescription to avoid including patients who had a propensity toward hyperkalemia and may not have been eligible to continue a RASI. To be included in the study, patients were required to have at least 1 serum creatinine measurement within 1 year prior to the start of a RASI and another within 90 days of the start of the RASI prescription (Figure 1). Patients also must have experienced a more than 15% decline in eGFR within 90 days of the start of the RASI prescription based on prior work showing the potential adverse effects associated with this level of decline in eGFR.10,11 Only outpatient serum creatinine values were used for eligibility determination and analysis.
Figure 1. Schematic of the Study Design.
RASI indicates renin-angiotensin system inhibitor.
Exposure and Outcome Definitions
Continuation of RASI was defined using an intention-to-treat approach at 90 days after the initial RASI prescription (Figure 1). Patients were considered continuers of RASIs if they had an active prescription between days 90 and 180, and they were considered discontinuers of RASIs if they did not. The applicable Anatomical Therapeutic Chemical codes for the drugs were used to define RASI use (eTable 3 in Supplement 1).
Outcomes were defined as described in eTable 4 in Supplement 1 and included the 2 primary outcomes of ESKD and all-cause mortality and 2 secondary outcomes of fatal or nonfatal MACEs (including myocardial infarction, heart failure, stroke, or cardiovascular mortality) and hospitalization for AKI. Outcomes were determined using International Classification of Diseases, Ninth Revision, Clinical Modification and International Statistical Classification of Diseases, Tenth Revision, Canadian Classification codes from physician claims, hospitalizations, and the vital statistics service (eTables 1 and 4 in Supplement 1). Outcome ascertainment was completed within the province of Manitoba given the presence of its provincial public health care system.
Patients’ eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration 2009 equation. Percentage change in eGFR was calculated using the most recent serum creatinine value prior to new RASI initiation as the baseline and the lowest serum creatinine value within 90 days of RASI initiation.
Propensity Score Matching
After the exposure was defined, propensity scores were used to match RASI continuers vs discontinuers 2:1 to emulate randomization. Covariates included in the propensity score were variables known to be confounders of the association between RASI discontinuation (vs continuation) and the outcome, including age, sex, baseline eGFR, potassium, diabetes, hypertension, heart failure, coronary artery disease, stroke, and use of other antihypertensive agents that may affect potassium levels (diuretics, β-blockers, and aldosterone antagonists). Data on race and ethnicity were not available. Covariates were assessed up to 1 year before the start of the RASI prescription and defined using International Classification of Diseases, Ninth Revision, Clinical Modification or International Statistical Classification of Diseases, Tenth Revision, Clinical Modification codes (eTable 5 in Supplement 1). Propensity scores were estimated using logistic regression and used to perform nearest-neighbor matching. Standardized mean differences (SMDs) were assessed before and after matching with an absolute SMD of less than 0.1 considered the threshold for adequate covariate balance.
Statistical Analysis
Data were analyzed between January 15, 2024, and January 22, 2026. We evaluated the association between RASI discontinuation (vs continuation) among patients with an acute decline in eGFR of more than 15% and all outcomes using Cox proportional hazards regression. Unadjusted models were fitted in the prematched cohort, while propensity score matching was used to balance baseline covariates in a 2:1 matched cohort. Cox proportional hazards models were performed for all outcomes, and the models were stratified by matched sets. Time in the Cox models started at day 180 to allow for new users of RASIs to refill (or not refill) their prescription to determine exposure status (continuation vs discontinuation) following an acute decline in eGFR within 90 days of the initial RASI prescription. We tested for the proportional hazards assumption using Schoenfeld residuals.
Fine-Gray models were used in sensitivity analyses to account for the competing risk of death for the outcomes of ESKD, MACEs, and AKI. In these models, matching was not accounted for given the balance in covariates between RASI discontinuers and continuers.
We then evaluated prespecified subgroups, including those with CKD (defined by eGFR <60 mL/min/1.73 m2), heart failure, or diabetes at baseline, given that these are groups for which RASI may be more strongly indicated. We also performed subgroup analyses among patients with albuminuria data available at baseline. These subgroup analyses were conducted without rematching of the cohort, leveraging subgroups of patients with each comorbidity in the primary cohort matched.
We also repeated our primary analyses using a Cox proportional hazards model to examine the negative outcome of cataract surgery using 2:1 propensity score matching. All statistical analyses were performed using R, version 4.5.0 (R Foundation for Statistical Computing). Findings were considered significant if the 95% CI did not include 1.00.
In additional sensitivity analyses, we conducted a per-protocol analysis that censored follow-up at any switches to the treatment strategy using 2:1 matching. A number of other sensitivity analyses were also conducted, including (1) defining the cohort using an acute decline in eGFR of more than 30% from baseline as the threshold for inclusion and (2) using a starting time of 150 days after the first RASI prescription (allowing only a 60-day period after the initial decline in eGFR for a refill to occur). Both analyses were conducted using 1:1 propensity score matching given fewer eligible patients meeting these inclusion criteria.
Results
We identified 5273 patients (mean [SD] age, 64.6 [16.2] years; 2576 female [48.9%] and 2697 male [51.1%]) who had an acute decline in eGFR of more than 15% following the start of a RASI and were eligible for inclusion (eFigure 1 in Supplement 1). Prior to propensity score matching, patients who discontinued RASIs compared with those who continued after an acute decline in eGFR of more than 15% were more likely to have stage G3 to G5 CKD (356 of 1411 [25.2%] vs 730 of 3862 [18.9%]) and less likely to have a history of myocardial infarction (182 [12.9%] vs 687 [17.8%]) (Table 1).
Table 1. Baseline Characteristics of the Study Population Before and After Propensity Score Matching.
| Characteristica | Patients, No. (%) | |||||
|---|---|---|---|---|---|---|
| Before propensity score matching | After 2:1 propensity score matching | |||||
| Continued RASI (n = 3862) | Discontinued RASI (n = 1411) | SMD | Continued RASI (n = 2822) | Discontinued RASI (n = 1411) | SMD | |
| Age, mean (SD), y | 64.2 (15.8) | 64.6 (16.9) | 0.02 | 64.6 (16.2) | 64.6 (16.9) | <0.01 |
| Sex | ||||||
| Female | 1863 (48.2) | 713 (50.5) | 0.05 | 1416 (50.2) | 713 (50.5) | 0.01 |
| Male | 1999 (51.8) | 698 (49.5) | 1406 (49.8) | 698 (49.5) | ||
| eGFR, mean (SD), mL/min/1.73 m2 | 79.5 (22.5) | 77.2 (25.8) | 0.09 | 77.5 (24.0) | 77.2 (25.8) | 0.01 |
| CKD stage | ||||||
| G1 | 1347 (34.9) | 465 (32.4) | 0.04 | 922 (32.7) | 465 (33.0) | 0.01 |
| G2 | 1785 (46.2) | 590 (41.8) | 0.09 | 1217 (43.1) | 590 (41.8) | 0.03 |
| G3 | 647 (16.8) | 301 (21.3) | 0.11 | 600 (21.3) | 301 (21.3) | <0.01 |
| G4 | 74 (1.92) | 49 (3.5) | 0.09 | 74 (2.6) | 49 (3.5) | 0.05 |
| G5 | 9 (0.23) | 6 (0.43) | 0.03 | 9 (0.3) | 6 (0.4) | 0.02 |
| Previous diagnosis | ||||||
| Myocardial infarction | 687 (17.8) | 182 (12.9) | 0.14 | 351 (12.4) | 182 (12.9) | 0.01 |
| Stroke | 239 (6.2) | 86 (6.1) | <0.01 | 168 (6.0) | 86 (6.1) | 0.01 |
| Congestive heart failure | 665 (17.2) | 244 (17.3) | <0.01 | 500 (17.7) | 244 (17.3) | 0.01 |
| Coronary artery disease | 949 (24.6) | 292 (20.7) | 0.09 | 561 (19.9) | 292 (20.7) | 0.02 |
| Hypertension | 1655 (42.9) | 548 (38.8) | 0.08 | 1111 (39.4) | 548 (38.8) | 0.01 |
| Diabetes | 778 (20.1) | 290 (20.6) | 0.01 | 595 (21.1) | 290 (20.6) | 0.01 |
| Potassium, mean (SD), mmol/Lb | 4.11 (0.4) | 4.12 (0.5) | 0.03 | 4.12 (0.5) | 4.12 (0.5) | 0.02 |
| Previous medication use | ||||||
| β-Blockers | 1684 (43.6) | 527 (37.3) | 0.13 | 1020 (36.1) | 527 (37.3) | 0.03 |
| Diuretics | 1105 (28.6) | 382 (27.1) | 0.03 | 794 (28.1) | 382 (27.1) | 0.02 |
| Aldosterone antagonists | 227 (5.9) | 64 (4.5) | 0.06 | 142 (5.0) | 64 (4.5) | 0.02 |
Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; RASI, renin-angiotensin system inhibitor; SMD, standardized mean difference.
Measurements for albuminuria were missing in 4053 patients (76.8%) in the full cohort and 3199 (75.6%) in the matched cohort and, hence, are not shown.
To convert to mEq/L, divide by 1.
After propensity score matching in a 2:1 ratio of continuers vs discontinuers, the cohort size included for analysis was 4233. A total of 1039 patients (24.5%) had stage G3 to G5 CKD at baseline, 744 (17.6%) had congestive heart failure, and 885 (20.9%) had diabetes. All characteristics were balanced after propensity score matching, with small SMDs (Table 1; eFigure 2 in Supplement 1).
During a mean (SD) follow-up of 4.7 (3.4) years, 1310 patients (30.9%) died, 256 (6.0%) experienced an ESKD event, 1486 (35.1%) had a MACE, and 592 (14.0%) had an AKI event in the propensity score–matched cohort. The incidence rate for each event by whether RASI had been discontinued vs continued is shown in Table 2. The causes of death in this cohort are shown in eTable 6 in Supplement 1, with the most common related to the circulatory system, followed by neoplasms.
Table 2. Association Between RASI Discontinuation vs Continuation and Clinical Outcomes.
| Model | Event rate after matching, No. (%) | Incidence rate per 1000 person-years after matching | HR (95% CI) | |
|---|---|---|---|---|
| Unadjusted | RASI discontinuation after 2:1 propensity score matching in ITTa | |||
| All-cause mortality | ||||
| All | 1310 (30.9) | 67.1 | NA | NA |
| RASI discontinuation | 480 (34.0) | 75.1 | 1.29 (1.16-1.43) | 1.23 (1.07-1.41) |
| RASI continuation | 830 (29.4) | 63.2 | 1 [Reference] | 1 [Reference] |
| ESKD | ||||
| All | 256 (6.0) | 13.7 | NA | NA |
| RASI discontinuation | 104 (7.3) | 17.2 | 1.69 (1.33-2.15) | 1.74 (1.28-2.38) |
| RASI continuation | 152 (5.3) | 12.0 | 1 [Reference] | 1 [Reference] |
| MACEs | ||||
| All | 1486 (35.1) | 93.2 | NA | NA |
| RASI discontinuation | 517 (36.6) | 99.9 | 1.16 (1.05-1.29) | 1.13 (0.99-1.28) |
| RASI continuation | 969 (34.3) | 90.0 | 1 [Reference] | 1 [Reference] |
| AKI | ||||
| All | 592 (14.0) | 32.2 | NA | NA |
| RASI discontinuation | 201 (14.2) | 33.5 | 1.14 (0.97-1.35) | 1.11 (0.90-1.37) |
| RASI continuation | 391 (13.9) | 31.7 | 1 [Reference] | 1 [Reference] |
Abbreviations: AKI, acute kidney injury; ESKD, end-stage kidney disease; HR, hazard ratio; ITT, intention to treat; MACE, major adverse cardiovascular event; NA, not applicable; RASI, renin-angiotensin system inhibitor.
Allowed baseline hazards to vary among matched pairs. Hazard ratios were estimated using Cox proportional hazards models fitted in a 2:1 propensity score–matched cohort, with stratification by matched sets. Propensity scores were estimated using logistic regression that included the following covariates: sex, age, potassium level, baseline estimated glomerular filtration rate, chronic kidney disease, myocardial infarction, congestive heart failure, stroke, coronary artery disease, hypertension, diabetes, diuretics, β-blockers, and aldosterone antagonists.
The hazard ratios (HRs) for all-cause mortality and ESKD were 1.23 (95% CI, 1.07-1.41) and 1.74 (95% CI, 1.28-2.38), respectively, among patients who discontinued RASIs (vs those who continued) after an acute eGFR decline had occurred (Table 2; Figure 2). The findings were consistent for the model that treated death as a competing risk for the outcome of ESKD (subdistribution HR [SHR], 1.37; 95% CI, 1.07-1.76).
Figure 2. Kaplan-Meier Curves of the Cumulative Probability of the Primary Outcomes in the Propensity Score–Matched Cohort for All-Cause Mortality and End-Stage Kidney Disease.
RASI indicates renin-angiotensin system inhibitor.
In terms of secondary outcomes, MACEs were not significantly greater for RASI discontinuation compared with continuation (HR, 1.13; 95% CI, 0.99-1.28) (Table 2; eFigure 3 in Supplement 1). The effect size was further attenuated after accounting for the competing risk of death in the sensitivity analysis (SHR, 1.08; 95% CI, 0.97-1.20).
Discontinuation of RASIs (vs continuation) was not associated with AKI (HR, 1.11; 95% CI, 0.90-1.37) (Table 2; eFigure 3 in Supplement 1). Findings were similar in the competing-risk analysis for the outcome of AKI (SHR, 1.02; 95% CI, 0.86-1.21).
Among patients with CKD, heart failure, or diabetes at baseline, characteristics generally remained balanced in these subgroups, though SMDs were larger than for the primary analysis (eTables 7-9 in Supplement 1). Overall, results were similar for our primary outcomes of all-cause mortality and ESKD, though findings were not always statistically significant for outcomes for which a fewer number of events occurred in subgroups (Table 3). However, there was a large effect size between RASI discontinuation and risk of ESKD in patients with a history of diabetes or CKD. Findings were also similar among the subset with albuminuria measures available (n = 1034), although the results were only significant for ESKD (HR, 2.44; 95% CI, 1.23-4.87) (eTables 10 and 11 in Supplement 1). RASI discontinuation vs continuation was not associated with the negative control outcome of cataract surgery (HR, 1.10; 95% CI, 0.87-1.38) (eTable 12 in Supplement 1).
Table 3. Subgroup Analysis Among Patients With CKD, Diabetes, or Heart Failure.
| Subgroup and model | Event rate after matching, No. (%) | Incidence rate per 1000 person-years after matching | RASI discontinuation, HR (95% CI) | |
|---|---|---|---|---|
| Unadjusted | Adjusteda | |||
| CKD | ||||
| All-cause mortality | ||||
| All | 527 (50.5) | 126.7 | NA | NA |
| RASI discontinuation | 200 (56.2) | 157.5 | 1.39 (1.17-1.65) | 1.28 (0.99-1.65) |
| RASI continuation | 327 (0.48) | 113.5 | 1 [Reference] | 1 [Reference] |
| ESKD | ||||
| All | 141 (13.0) | 38.2 | NA | NA |
| RASI discontinuation | 56 (15.7) | 51.9 | 1.63 (1.17-2.28) | 2.08 (1.28-3.39) |
| RASI continuation | 85 (12.1) | 32.6 | 1 [Reference] | 1 [Reference] |
| Diabetes | ||||
| All-cause mortality | ||||
| All | 323 (36.5) | 75.7 | NA | NA |
| RASI discontinuation | 123 (42.4) | 92.9 | 1.41 (1.14-1.75) | 1.43 (0.84-2.42) |
| RASI continuation | 200 (33.6) | 68.0 | 1 [Reference] | 1 [Reference] |
| ESKD | ||||
| All | 121 (13.7) | 31.2 | NA | NA |
| RASI discontinuation | 52 (17.9) | 45.8 | 2.09 (1.47-2.96) | 2.88 (1.21-6.88) |
| RASI continuation | 69 (11.5) | 25.2 | 1 [Reference] | 1 [Reference] |
| Heart failure | ||||
| All-cause mortality | ||||
| All | 414 (55.6) | 151.5 | NA | NA |
| RASI discontinuation | 149 (61.0) | 180.3 | 1.41 (1.16-1.71) | 1.41 (0.92-2.17) |
| RASI continuation | 265 (53.0) | 139.0 | 1 [Reference] | 1 [Reference] |
| ESKD | ||||
| All | 34 (4.6) | 12.8 | NA | NA |
| RASI discontinuation | 13 (5.3) | 16.5 | 1.74 (0.89-3.43) | Not availableb |
| RASI continuation | 21 (4.2) | 11.3 | 1 [Reference] | 1 [Reference] |
Abbreviations: CKD, chronic kidney disease; ESKD, end-stage kidney disease; HR, hazard ratio; NA, not applicable; RASI, renin-angiotensin system inhibitor.
Allowed baseline hazards to vary among matched pairs. Hazard ratios were estimated using Cox proportional hazards models fitted in a 2:1 propensity score–matched cohort, with stratification by matched sets. Propensity scores were estimated using logistic regression that included the following covariates: sex, age, potassium level, baseline estimated glomerular filtration rate, chronic kidney disease, myocardial infarction, congestive heart failure, stroke, coronary artery disease, hypertension, diabetes, diuretics, β-blockers, and aldosterone antagonists.
Insufficient number of events to provide a reliable HR due to an unstable model.
In sensitivity analyses, results were consistent when we defined the cohort based on the presence of an acute decline in eGFR of greater than 30% after RASI initiation (n = 1558) (eTables 13 and 14 in Supplement 1) for the primary outcomes. Findings were also consistent when we changed the start time to 150 days after the initial RASI prescription (eTables 15 and 16 in Supplement 1). Finally, findings were more robust with a per-protocol analysis approach, with larger HRs showing RASI discontinuation to be associated with all outcomes except MACEs (eTable 17 in Supplement 1).
Discussion
This retrospective cohort study of patients treated within a closed health care system in Manitoba, Canada, found that overall, discontinuation of RASIs (compared with continuation of RASIs) in new users with acute declines in eGFR was associated with a higher risk of ESKD, including among those with a history of heart failure, diabetes, or CKD. Discontinuation of RASIs was also associated a higher risk of all-cause mortality in the propensity score–matched cohort. We did not find a difference in the risk of MACEs and AKI when RASIs were discontinued after an acute decline in eGFR. Taken together, these findings highlight the potential risks of discontinuing RASI therapy when small declines in kidney function following their initiation occur.
Ku et al11 previously showed that acute declines in eGFR of 15% to 20% over a 1-month period were needed for the kidney benefit of starting RASIs (vs placebo or comparator antihypertensive agents) to be attenuated among patients with CKD in clinical trial settings. Here, we found that discontinuation of RASIs was associated with more harm in terms of progression to ESKD. While we do not know definitively whether the acute decline in eGFR triggered clinicians to stop the RASI, our findings suggest that there is benefit to continuing these agents and no signal of harm, even when declines in eGFR were more than 30%. Of note, a larger proportion of patients with CKD at baseline had their RASI stopped. We speculate that clinicians may have been more concerned about the loss of eGFR with RASI initiation in patients who already had a low eGFR.12
Overall, our data are generally consistent with studies examining acute declines in eGFR following the start of sodium-glucose cotransporter 2 inhibitors, which are also thought to be benign, and guidelines have recommended not reacting to such acute changes in the setting of both RASI and sodium-glucose cotransporter 2 inhibitor use.13,14 Although there is great excitement in the nephrology community about the availability of newer agents for slowing CKD progression and increasing cardiovascular benefit,15 discontinuation of RASIs may be associated with worse kidney and survival outcomes and, potentially, a higher tendency toward MACEs.
Strengths and Limitations
The strengths of our study included its large cohort size and use of data from a closed health care system linked to comprehensive vital statistics and prescription fill data to examine the study question using a trial emulation design. In addition, our results were consistent across multiple sensitivity and subgroup analyses.
Limitations to our study included the lack of information on the exact reason for discontinuation of RASIs and the potential for selection bias. Patients who discontinued RASIs may have had greater severity of illness and, therefore, were more prone to cardiovascular or kidney outcomes, though there was no association between our exposure and the control outcome of cataract surgery. We also did not have data on adherence to medications and socioeconomic factors, and residual confounding may have been present. Although we included a large cohort of patients using clinical data, findings may reflect the practices of clinicians in Manitoba and not be generalizable to other practice settings, especially settings without publicly funded health care systems. Other limitations included potential misclassification of outcomes based on International Classification of Diseases diagnostic codes and the lack of albuminuria ascertainment in all patients, though findings were generally consistent in the subcohort with available measurements.
Conclusions
This retrospective cohort study of clinical data found that discontinuation of RASIs after an acute decline in eGFR was associated with an increased risk of ESKD and mortality but not MACEs or AKI. Further study is needed to understand reasons for the frequent discontinuation of these agents after their initiation and to devise strategies to improve their persistent use, including in patients who have strong indications for RASIs, such as CKD, heart failure, and diabetes.
eFigure 1. Flow Diagram
eFigure 2. Propensity Score Distribution Before and After 2:1 Propensity Matching
eFigure 3. Kaplan-Meier Curves for Secondary Outcomes
eTable 1. Databases Accessed From the Population Research Data Repository at the Manitoba Centre for Health Policy
eTable 2. Target Trial Emulation Framework
eTable 3. Applicable Anatomical Therapeutic Chemical (ATC) Codes for Drugs Used to Define Exposure to RASI
eTable 4. Primary Outcome Definitions
eTable 5. Definitions of Covariates
eTable 6. Causes of Death in the Full Cohort Based on WHO ICD-10 Codes
eTable 7. Baseline Characteristics of Patients With CKD Within RASI Continuation and Discontinuation Groups
eTable 8. Baseline Characteristics of Patients With Diabetes Mellitus Within RASI Continuation and Discontinuation Groups
eTable 9. Baseline Characteristics of Patients With Heart Failure Within RASI Continuation and Discontinuation Groups
eTable 10. Baseline Characteristics of the Population With Albuminuria Data Available
eTable 11. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had Albuminuria Data Available
eTable 12. Negative Control Analysis for the Outcome of Cataract Surgery to Assess Residual Confounding
eTable 13. Baseline Characteristics of Patients With a 30% Decline in eGFR Within RASI Continuation and Discontinuation Groups
eTable 14. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had an Acute Decline of 30%
eTable 15. Baseline Characteristics of the Population Who Had a Refill of RASI Within 60 Days of RASI Initiation
eTable 16. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had a Refill of RASI Within 60 Days of First RASI Prescription
eTable 17. Adjusted Association Between Assigned RASI Strategy in the Per-Protocol Analysis
Data Sharing Statement
References
- 1.Bakris GL, Weir MR. Angiotensin-converting enzyme inhibitor-associated elevations in serum creatinine: is this a cause for concern? Arch Intern Med. 2000;160(5):685-693. doi: 10.1001/archinte.160.5.685 [DOI] [PubMed] [Google Scholar]
- 2.Clase CM, Barzilay J, Gao P, et al. Acute change in glomerular filtration rate with inhibition of the renin-angiotensin system does not predict subsequent renal and cardiovascular outcomes. Kidney Int. 2017;91(3):683-690. doi: 10.1016/j.kint.2016.09.038 [DOI] [PubMed] [Google Scholar]
- 3.McCallum W, Tighiouart H, Ku E, Salem D, Sarnak MJ. Acute declines in estimated glomerular filtration rate on enalapril and mortality and cardiovascular outcomes in patients with heart failure with reduced ejection fraction. Kidney Int. 2019;96(5):1185-1194. doi: 10.1016/j.kint.2019.05.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Weir MR. Acute changes in glomerular filtration rate with renin-angiotensin system (RAS) inhibition: clinical implications. Kidney Int. 2017;91(3):529-531. doi: 10.1016/j.kint.2016.11.019 [DOI] [PubMed] [Google Scholar]
- 5.Mansfield KE, Nitsch D, Smeeth L, Bhaskaran K, Tomlinson LA. Prescription of renin-angiotensin system blockers and risk of acute kidney injury: a population-based cohort study. BMJ Open. 2016;6(12):e012690. doi: 10.1136/bmjopen-2016-012690 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ohkuma T, Jun M, Rodgers A, et al. ; ADVANCE Collaborative Group . Acute increases in serum creatinine after starting angiotensin-converting enzyme inhibitor-based therapy and effects of its continuation on major clinical outcomes in type 2 diabetes mellitus. Hypertension. 2019;73(1):84-91. doi: 10.1161/HYPERTENSIONAHA.118.12060 [DOI] [PubMed] [Google Scholar]
- 7.Mahmoudpour SH, Asselbergs FW, Souverein PC, de Boer A, Maitland-van der Zee AH. Prescription patterns of angiotensin-converting enzyme inhibitors for various indications: a UK population-based study. Br J Clin Pharmacol. 2018;84(10):2365-2372. doi: 10.1111/bcp.13692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Qiao Y, Shin JI, Sang Y, et al. Discontinuation of angiotensin converting enzyme inhibitors and angiotensin receptor blockers in chronic kidney disease. Mayo Clin Proc. 2019;94(11):2220-2229. doi: 10.1016/j.mayocp.2019.05.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cashin AG, Hansford HJ, Hernán MA, et al. Transparent Reporting of Observational Studies Emulating a Target Trial-the TARGET statement. JAMA. 2025;334(12):1084-1093. doi: 10.1001/jama.2025.13350 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ku E, Jamerson K, Copeland TP, McCulloch CE, Tighiouart H, Sarnak MJ. Acute declines in estimated glomerular filtration rate in patients treated with benazepril and hydrochlorothiazide versus amlodipine and risk of cardiovascular outcomes. J Am Heart Assoc. 2024;13(15):e035177. doi: 10.1161/JAHA.124.035177 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ku E, Tighiouart H, McCulloch CE, et al. Association between acute declines in eGFR during renin-angiotensin system inhibition and risk of adverse outcomes. J Am Soc Nephrol. 2024;35(10):1402-1411. doi: 10.1681/ASN.0000000000000426 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bhandari S, Mehta S, Khwaja A, et al. ; STOP ACEi Trial Investigators . Renin-angiotensin system inhibition in advanced chronic kidney disease. N Engl J Med. 2022;387(22):2021-2032. doi: 10.1056/NEJMoa2210639 [DOI] [PubMed] [Google Scholar]
- 13.Mc Causland FR, Claggett BL, Vaduganathan M, et al. Decline in estimated glomerular filtration rate after dapagliflozin in heart failure with mildly reduced or preserved ejection fraction: a prespecified secondary analysis of the DELIVER randomized clinical trial. JAMA Cardiol. 2024;9(2):144-152. doi: 10.1001/jamacardio.2023.4664 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group . KDIGO 2024 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int. 2024;105(4S):S117-S314. doi: 10.1016/j.kint.2023.10.018 [DOI] [PubMed] [Google Scholar]
- 15.Lafayette RA. Stop sitting on our RAASes: get moving with the most effective care for our patients. J Am Soc Nephrol. 2024;35(11):1610-1612. doi: 10.1681/ASN.0000000000000473 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. Flow Diagram
eFigure 2. Propensity Score Distribution Before and After 2:1 Propensity Matching
eFigure 3. Kaplan-Meier Curves for Secondary Outcomes
eTable 1. Databases Accessed From the Population Research Data Repository at the Manitoba Centre for Health Policy
eTable 2. Target Trial Emulation Framework
eTable 3. Applicable Anatomical Therapeutic Chemical (ATC) Codes for Drugs Used to Define Exposure to RASI
eTable 4. Primary Outcome Definitions
eTable 5. Definitions of Covariates
eTable 6. Causes of Death in the Full Cohort Based on WHO ICD-10 Codes
eTable 7. Baseline Characteristics of Patients With CKD Within RASI Continuation and Discontinuation Groups
eTable 8. Baseline Characteristics of Patients With Diabetes Mellitus Within RASI Continuation and Discontinuation Groups
eTable 9. Baseline Characteristics of Patients With Heart Failure Within RASI Continuation and Discontinuation Groups
eTable 10. Baseline Characteristics of the Population With Albuminuria Data Available
eTable 11. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had Albuminuria Data Available
eTable 12. Negative Control Analysis for the Outcome of Cataract Surgery to Assess Residual Confounding
eTable 13. Baseline Characteristics of Patients With a 30% Decline in eGFR Within RASI Continuation and Discontinuation Groups
eTable 14. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had an Acute Decline of 30%
eTable 15. Baseline Characteristics of the Population Who Had a Refill of RASI Within 60 Days of RASI Initiation
eTable 16. Association Between RASI Use (Discontinuation vs Continuation) and Clinical Outcomes in Patients Who Had a Refill of RASI Within 60 Days of First RASI Prescription
eTable 17. Adjusted Association Between Assigned RASI Strategy in the Per-Protocol Analysis
Data Sharing Statement
