Skip main navigation

Angioedema Incidence in US Veterans Initiating Angiotensin-Converting Enzyme Inhibitors

Originally published 2008;51:1624–1630


Angioedema is a rare but potentially serious complication of angiotensin-converting enzyme inhibitor (ACE) use. We conducted a study to estimate incidence of ACE-related angioedema and explore its determinants in a large racially diverse patient population. We used linked medical and pharmacy records to identify all patients in the US Veterans Affairs Health Care System from April 1999 through December 2000 who received first prescriptions for antihypertensive medications. We studied 195 192 ACE initiators and 399 889 patients initiating other antihypertensive medications (OAH). New angioedema was identified by diagnosis codes using methods validated in a national sample of 869 angioedema cases with confirmation for over 95% of cases. Overall, 0.20% of ACE initiators developed angioedema while on the medication and the incidence rate was 1.97 (1.77 to 2.18) cases per 1000 person years. This compares with a rate of 0.51 (0.43 to 0.59) in OAH initiators and the adjusted relative risk estimate was 3.56 (2.82 to 4.44). Fifty five percent of cases occurred within 90 days of first ACE use but risk remained elevated with prolonged use, even beyond 1 year. We estimate that 58.3% of angioedema in patients starting antihypertensives was related to ACE. We also found that angioedema rates were nearly 4-fold higher in blacks, 50% higher in women, and 12% lower in those with diabetes. This study provides a reliable estimate of angioedema incidence associated with ACE use in a diverse nontrial patient population, confirming that the incidence is low, but finding substantial variation by race, sex, and diabetes status.

Angiotensin-converting enzyme inhibitors (ACE) are commonly prescribed medications for patients with high blood pressure, diabetes, chronic heart failure, and coronary heart disease. Their use has been associated with lower risk of myocardial infarction, delay in symptoms of heart failure, diminished progression of kidney disease, and lower mortality risk.1–4 These medicines are used by millions of patients with few side effects.

One potentially serious complication of ACE is angioedema, a subcutaneous nonpitting edema typically of sudden onset and short duration.5–9 It can involve nearly any part of the body, although the larynx, tongue, lips, and face are most often involved. When the airways are affected it can be life-threatening.

Research conducted to date suggests that the incidence of angioedema associated with ACE use is low,9–16 on the order of 1 or 2 per 1000 patients, but both its incidence and determinants are not well understood. Previous estimates have come primarily from small observational studies in select patient populations or from clinical trials of relatively healthy people. The best of these is Omapatril Cardiovascular TreAtment Versus Enalapril (OCTAVE), a randomized double-blind clinical trial of 12 557 persons with hypertension treated with enalapril,11 in which an incidence rate of 6.8 per 1000 was found over 24 weeks of follow-up. Although this is probably our best estimate, it remains uncertain how well it applies to the larger more diverse population of patients seen in clinical settings with comorbidities that may have made them ineligible for trials.

The current study was undertaken to assess the incidence of angioedema in ACE users and to explore its determinants in a large racially diverse patient population. To estimate the independent contribution of ACE use to angioedema risk, incidence of angioedema in new ACE users was compared to that among new users of other antihypertensive medications (OAH), and other factors were considered in the analysis. This study also includes detailed chart review of a large number of patients with angioedema to confirm the validity of our methods for identifying angioedema, refining the analysis, and obtaining more accurate estimates of the incidence of ACE-associated angioedema.


Data Sources

This was a national study of veterans receiving health care from the United States Veterans Affairs Health Care System (VA) from October 1, 1998 through December 31, 2000. Data sources included national VA pharmacy records obtained from the Pharmacy Benefits Management Strategic Health Group (PBM),17 computerized health encounter records from the Austin Automation Center,18,19 and Medicare claims obtained from the Centers for Medicare and Medicaid Services for VA patients receiving part of their care outside of the VA that was covered by Medicare.20,21 This research was part of the Diabetes Epidemiology Cohorts Program.22 The Institutional Review Board at the Edith Nourse Memorial VA Hospital in Bedford, Massachusetts reviewed and approved the study protocol and a waiver of informed consent. All study procedures were in accordance with institutional guidelines.

National Cohort Study Sample

We identified all VA patients who received VA prescriptions for antihypertensive medications from October 1998 through December 2000. Each patient’s prescriptions were processed to assess timing and duration of use and multiple prescriptions for a given medication were summarized into episodes of use. To study only new users of antihypertensive medications, we restricted the sample to those patients who did not receive an antihypertensive prescription in the first 6 months of the study (October 1998 through March 1999) and then subsequently received a first prescription (from April 1999 through December 2000). Prior analyses indicated that this 6-month “clean” period excludes more than 90% of past users of a class of medications. The remaining sample was split into the 2 study groups, those who first initiated ACE and those who initiated a different class of antihypertensive medication (OAH). Patients in the latter group, initiating OAH, who subsequently were prescribed ACE up through December 2000 also were excluded.

Identification of Angioedema

First episodes of angioedema were identified using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes (995.1)23 recorded during inpatient stays or outpatient visits and contained in either national VA patient records or Medicare claims. We evaluated this method in a large medical chart review study described below. The date of the first code was taken as the day of onset of the angioedema. All episodes of angioedema occurring during the study period were included and analyzed in relation to the time of antihypertensive prescriptions.

Other Patient Information

Patient records from national computerized data files also provided information to assign demographic (sex, age, race) and comorbidity status (chronic heart failure [CHF; ICD-9-CM=428], coronary artery disease [CAD; 410–414], diabetes mellitus [DM; 250], and high blood pressure [HBP; 401–405]). Deaths were identified from the VA Beneficiary Identification and Record Location file (a registry of all veterans whose families have applied for death benefits that has been shown to reasonably reliable and complete24, supplemented by searches of VA inpatient files, Medicare claims, and Social Security files.

Statistical Analysis

All analyses were conducted using Statistical Analysis Software (SAS) version 8.0.25 We counted the numbers of new angioedema cases and the total numbers of patients in each medication group in each month before, during, and after medication initiation. Counts were combined across months and case counts were divided by total patient counts to calculate unadjusted incidence rates for the time periods before the first prescription, after the first prescription, and after use had ended in each medication group. Incidence rates are presented as cases of angioedema per 1000 person-years and exact Poisson 95% confidence intervals were calculated for rates.26 Additional analyses were performed for cases with angioedema codes from inpatient stays or intensive care units and for cases dying within 90 days of their angioedema. Multivariate Poisson regression models were used to estimate adjusted relative risk of angioedema for ACE use, relative to use of OAH.27,28 Models included terms for calendar year, age, sex, race, and diagnoses of congestive heart failure, coronary artery disease, and diabetes. Unfortunately, we did not have sufficient information on other potentially important correlates of angioedema risk, including smoking and allergies.11 In separate models, we evaluated risk of angioedema associated with initiation of specific ACE medications and modification of the risk using interaction terms as a test of homogeneity across strata of covariates. Population attributable risk percentages were calculated using relative risk estimates for angioedema obtained from the regression models.29,30

Medical Chart Review Study

To evaluate methods used in the national cohort study and to inform the analysis, we conducted a national medical chart abstraction study with a subset of the angioedema cases. To maximize its applicability to our national study, we selected and recruited VA facilities from across the country. After receiving human studies approval at each hospital, we obtained medical charts for patients from 20 VA facilities throughout the United States (see acknowledgments), representing over 20% of the entire VA patient population. Complete medical charts for patients with a diagnosis code of angioedema were abstracted by trained VA health care providers (internists, residents, nurses, doctors of pharmacy) using standardized forms and methods. Confirmation of the angioedema in the medical chart was based on explicit notation of the diagnosis and description of the relevant symptoms in notes near the time of the code assignment. Additional information from earlier and later notes in the record indicating corrected or alternative diagnoses was applied to reclassify confirmation status. We assigned the likelihood of angioedema as (1) current firm (diagnosis noted with description of specific symptoms of edema of the tongue or lips, or asymmetrical facial swelling at the time of the medical visit or stay), (2) current probable (either diagnosis or symptoms at the current visit), (3) past firm (both diagnosis and symptoms but with reference made to an episode of angioedema in the recent past), (4) past probable (either diagnosis or symptoms referring to the recent past), or (5) none (neither). Confirmation assignments were evaluated by a panel of 3 VA physicians who independently reviewed 10% of the charts and assigned likelihood of angioedema to each case. In addition to confirmation of the diagnosis, medical chart data were used to confirm antihypertensive medication use and the timing of use in relation to the angioedema episode.


There were 1 715 858 veteran VA patients who were prescribed antihypertensive medications from April 1999 through December 2000 (Table 1), representing 49.4% of the entire VA population at that time. These patients were predominantly male (98%), older (64% over 65 years of age), and of white race (78%), and associated chronic medical conditions were relatively common. More than 50% of those taking antihypertensive medications (888 124) were prescribed an ACE.

Table 1. VA Patients Prescribed Antihypertensive Medications, Angioedema Cases, and Study Sample of Patients Initiating Antihypertensives—April 1999 to December 2000

Patient CharacteristicsAll Patients Prescribed AH*Angioedema CasesPatients Initiating AH
N%N%New ACENew Other AH
*AH indicates antihypertensives; ACE, angiotensin-converting enzyme inhibitor.
†Limited to those with diagnostic code from April 1, 1999 through December 31, 2000.
Total No.1 715 8585901195 192399 889
    Female53 0193.14487.650932.615 1353.8
    Male1 662 83996.9545392.4190 09997.4384 75496.2
    <4560 1983.53195.480344.121 9885.5
    45 to 54245 09414.367711.531 58116.267 75316.9
    55 to 64304 30117.775012.736 61218.870 67617.7
    65 to 74501 74629.2199933.958 48030.0110 19627.6
    75+604 51935.2215636.560 48531.0129 27632.3
    White1 234 08478.2408072.4138 40379.0276 09379.0
    Black254 36516.1129923.024 20213.853 16515.2
    Other90 1715.72574.612 6847.220 1885.8
    Unknown137 23826519 90350 443
    Chronic heart failure340 59419.8142724.239 58320.362 16815.5
    Coronary artery disease391 17122.8154726.242 56221.878 54919.6
    Diabetes mellitus599 94135.0189032.090 07146.1107 88227.0
All PrescribedFirst Prescribed
    ACE inhibitor888 12451.8218237.0195 192100.0
    Calcium channel blocker711 33741.5181430.784 28721.1
    β-Adrenergic blocker640 93137.4129521.994 02023.5
    Thiazide diuretic368 82221.593915.954 04613.5
    Antiadrenergic, peripheral466 45427.292815.766 24416.6
    Loop diuretic396 15623.194616.047 51711.9
    Other diuretic207 46412.15108.628 6297.2
    Antiadrenergic, central80 5324.72624.496542.4
    Angiotensin II receptor blocker61 9073.63586.198162.5
    α/β-Adrenergic blocker27 5891.6731.235260.9
    Vasodialator28 0001.61242.121490.5

For the follow-up study, the sample was limited to the 595 081 patients who initiated antihypertensive medications during that period, with 195 192 starting ACE and 399 889 starting another agent (Table 1). The predominant ACE was lisinopril (72%), followed by fosinopril (22%) and captopril (6%). In terms of other antihypertensive medications initiated, most were diuretics (32.6%), β adrenergic blockers (23.5%), calcium channel blockers (21.1%), and peripherally acting antiadrenergic agents (16.6%). ACE initiators were similar to other patients starting antihypertensive medications in terms of demographic characteristics but they more often had chronic heart failure and especially diabetes.

A total of 5901 VA patients were identified with new angioedema during that period (Table 1). Of these patients, 704 (11.9%) were diagnosed as inpatients, 337 (5.7%) were subsequently admitted to intensive care, and 302 (5.1%) died sometime after the episode—9 within 1 week and 33 within 1 month. The demographic and morbidity profile of these patients was similar to the overall VA patient population, except for older age (70% over 65 years) and a preponderance of blacks (25%).

With the sample restricted to those newly prescribed antihypertensive medications, there were a total of 833 new angioedema cases, 434 in ACE initiators and 399 in OAH initiators (Table 2). In the new ACE users, most (352 or 81.1%) of the cases occurred after the new prescription and while still using the drug. Overall, 0.20% of patients initiating ACE developed angioedema while on the medication and the angioedema rate was calculated to be 1.97 cases per 1000 person years with a 95% confidence interval of 1.77 to 2.18. Rates were highest in the 30 days after the new prescription (7.38 [6.12 to 8.82]). Fifty-five percent of cases occurred within 90 days of first ACE use and the rate in that period was 4.13 (3.57 to 4.75). Rates declined gradually over the following year but remained elevated, even in those with more than 1 year of ACE use (0.93 [0.64 to 1.31]). These rates compare with a rate of 0.47(0.34 to 0.63) in this group before ACE initiation and a rate of 0.51 (0.43 to 0.59) in the comparison group of OAH initiators while using the drug. Angioedema incidence was elevated after OAH initiation but the increase was slight (0.99[0.68 to 1.39]) and it dropped in the next time period with no apparent trend in rates by time. When the OAH initiation group was broken out by class of antihypertensive medication, there was little variation among the other classes: calcium channel blockers (0.52/1000), beta adrenergic blockers (0.46/1000), thiazide diuretics (0.66/1000), peripherally acting antiadrenergic agents (0.42/1000), loop diuretics (0.50/1000), other diuretics (0.60/1000), centrally acting antiadrenergic agents (0.76/1000), angiotensin II receptor blockers (0.99/1000), α- and α-/βadrenergic blockers (0.28/1000), and vasodilators (0.40/1000). Evaluation of angioedema risk associated with other specific classes of antihypertensive medications is beyond the scope of this report.

Table 2. Angioedema Rates Among Patients Initiating ACE or Other Antihypertensives (OAH)—April 1999 to December 2000

Time of AngioedemaInitiating ACE (n=195 192)Initiating OAH (n=399 889)
Person-Years nCases nPer 1000 Person-YearsPerson-Years nCases nPer 1000 Person-Years
Rate95% CIRate95% CI
All315 5274341.381.25 to 1.51738 2503990.540.49 to 0.60
Before drug use89 547420.470.34 to 0.63297 1941800.610.52 to 0.70
During drug use179 0883521.971.77 to 2.18349 8921770.510.43 to 0.59
    <+30 days16 2661207.386.12 to 8.8233 324330.990.68 to 1.39
    31 to 60 days15 367432.802.03 to 3.7730 162160.530.30 to 0.86
    61 to 90 days14 877291.951.31 to 2.8028 913120.420.21 to 0.72
    91 to 180 days37 955601.581.21 to 2.0373 920270.370.24 to 0.53
    181 to 270 days32 398401.230.88 to 1.6863 084240.380.24 to 0.57
    271 to 360 days26 760271.010.67 to 1.4752 588220.420.26 to 0.63
    >+360 days35 465330.930.64 to 1.3167 901450.660.48 to 0.89
After discontinuing use46 892400.850.61 to 1.1691 164510.560.42 to 0.74

Results from Poisson regression are presented in Table 3. The adjusted relative risk for new ACE users relative to new OAH users was 3.56 (95% confidence intervals of 2.82 to 4.44). The risk was elevated for all 3 ACE agents individually. Although the risk appeared to be somewhat lower for captopril, as has been reported from a previous study,15 the differences were not statistically significant. These models also confirm that, independent of other factors considered, the risk of angioedema was increased substantially for blacks (3.88 [2.99 to 4.95]), females (1.45 [1.15 to 1.88]), and those with chronic heart failure (1.22 [1.08 to 1.38]) or coronary artery disease (1.31 [1.16 to 1.48]), and decreased significantly for those with diabetes (0.88 [0.82 to 0.95]). We also ran an additional series of models with interaction terms to test for differences in risk of ACE-associated angioedema by demographic factors and comorbidity. None of the terms were significant in these models, and they are not presented.

Table 3. Relative Risk Estimates*for Angioedema in VA Patients Prescribed ACE or Other Antihypertensives

VariableReference CategoryRelative Risk*95% Confidence Intervals
*Estimated from multivariate Poisson regression models.
Initiating ACEother AH3.562.82 to 4.44
Blackwhite3.882.99 to 4.95
Other racewhite0.910.77 to 1.13
Femalemale1.451.15 to 1.88
Age <45 years75+ years1.170.78 to 1.77
Age 45 to 54 years75+ years0.900.80 to 1.04
Age 55 to 64 years75+ years1.170.91 to 1.51
Age 65 to 74 years75+ years1.421.15 to 1.74
Chronic heart failureNo CHF1.221.08 to 1.38
Coronary artery diseaseNo CAD1.311.16 to 1.48
Diabetes mellitusNo DM0.880.82 to 0.95
ACE initiated: in separate model
Lisinoprilother AH3.632.34 to 5.48
Fosinoprilother AH3.452.06 to 5.46
Captoprilother AH2.201.08 to 3.95

To study the more severe cases, we repeated the analysis with restriction to angioedema cases diagnosed in hospital, admitted to intensive care, or dead within 90 days of the angioedema. The results were comparable to those obtained with all cases. The incidence of inpatient angioedema was 0.136 per 1000 for new ACE users compared with 0.031 for OAH initiators; 90 day postangioedema mortality was 0.010 per 1000 for new ACE users as compared to 0.006 for new OAH users. The adjusted relative risks of severe angioedema from Poisson regression were nearly 4.0 for ACE initiation and for black race; risk estimates for other factors were not significantly different from 1.0.

Finally, we estimated the attributable risk percentages for ACE and angioedema. Among new ACE users, the proportion of angioedema related to ACE exposure was estimated at 71.9%, and the proportion of angioedema attributable to ACE among all patients newly prescribed antihypertensive medications was 58.3%.

Medical Chart Review

Medical records for 1043 of the angioedema cases were obtained, and a total of 869 were judged to be complete. Evidence confirming angioedema was found in all but 41 of the charts (95.3%), with 487 (56.0%) assigned as “current firm,” 146 (16.2%) as “current probable,” 16 (1.8%) as “past firm,” and 179 (20.6%) as “past probable.” There was consensus on the assignments for 94% of cases by the 3 reviewing clinicians.

We also used the medical chart abstraction data to verify the use and timing of medications. Out of the 384 cases indicating recent ACE use in the charts, all but 20 (5.2%) had VA prescription records for ACE. For cases with notation of OAH, only 7 of 415 were without VA prescriptions. However, OAH were infrequently recorded in the charts at the time of the angioedema (36% of those with OAH prescriptions) as compared to chart recording of ACE use for 77% of those with VA prescriptions. Verification of the dates of angioedema onset and prescription dispensation confirmed the temporal relationship between the condition and previous exposure to the ACE in all but 7 cases. Most of these exceptions were for patients who responded rapidly to first exposure and had the angioedema diagnosis on the same day as the first prescription, which we conservatively considered as being before first use. Overall, these discrepancies suggest that missed medications and date errors could have only a minimal effect on the rates and associations estimated in our analysis.

Although the relatively high confirmation rate of the diagnosis in the chart review analysis was reassuring, as a further check, we repeated the Poisson regression modeling of risk in a sample restricted to patients from facilities with chart review and to those angioedema cases with confirmed diagnosis. The results were broadly similar to those observed and presented for the national sample.


This is the largest observational study of medication-related angioedema published to date. It provides important information on the risks of this condition across the spectrum of patients seen in clinics, including many patients who would not have been included in trials because of comorbidities. It is derived from a population of about 2 million VA patients prescribed antihypertensive prescriptions, focusing on a sample of nearly 600 000 patients first initiating this treatment with 833 new angioedema cases identified over a 21-month period. Furthermore, an extensive medical record review of a 14% sample indicated confirmation for over 95% of cases. From these data, we estimate an angioedema incidence in patients newly prescribed ACE of 1.97 per 1000 person years of use with relatively narrow confidence intervals (1.71 to 2.18). We project that about 1 of every 2600 new ACE users experiences angioedema within 30 days and about one of every 1000 experiences it within a year after first use. It also is evident that risk remains elevated with continued ACE use, even more than 1 year after initiation. The majority (71.9%) of angioedema occurring in new ACE users is attributable to the drug, and more than half (58.3%) of the adverse condition occurring in the population of patients newly prescribed antihypertensive medications is attributable to ACE.

Our estimated incidence of ACE-associated angioedema agrees well with those reported from other studies,9–13 even though there are a number of reasons why some variance might be expected. Compared to other study populations,7,11,12,31,32 VA patients are older and are predominantly male, with relatively high prevalence of comorbidities such as diabetes.22,33 We found lower risk of angioedema in men and patients with diabetes, and the later findings agree with those reported from the OCTAVE Study.11 Unfortunately, we did not have information on smoking or allergies so we could not evaluate the higher angioedema risks associated with these conditions in the OCTAVE Study. In agreement with other recent studies,12,13,15,16 we found that the most important demographic factor modifying angioedema risk is race, with blacks having 4-fold higher risk of angioedema. In the VA, 15.2% of ACE users are blacks; this is less than the 26.3% from the Medicaid recipient population studies12,15 but not lower than most other populations studied, including OCTAVE,11 and it is similar to the general population of the United States.34 If we standardize our incidence rate30 by sex, age, and race to the general population reporting hypertension,34 our overall estimate of ACE-associated angioedema is 2.7 per 1000 person years of ACE use.

Errors in case ascertainment are a possible reason for inaccuracy in our estimated incidence rates. The high proportion of cases confirmed in our medical chart review was reassuring with respect to possible overcounting of angioedema.35,36 On the other hand, undercounting may have occurred if angioedema cases failed to receive a diagnosis, particularly in cases with atypical presentation or those that resolved quickly. To the extent that this was present, it would most likely affect the least severe cases. This may be the reason why a higher incidence was reported from OCTAVE (6.8 per 1000) which conducted prospective adjudication of cases and probably included many mild cases that we might have missed.11 Given that most physicians are aware of the risk of angioedema associated with ACE use and that notation of use and discontinuation after an episode were much more likely for ACE than for the other drugs in our medical chart review, a missed diagnosis was probably more likely in patients using drugs other than ACE; this may have led to greater underestimation of angioedema risk associated with antihypertensive medications other than ACE. Some episodes of angioedema also may have been missed if patients were diagnosed and treated outside of the VA system with care not covered by Medicare, but this is unlikely to represent more than a small fraction of cases.21

Errors in the medication information are another potential source of bias. We may have missed some drug exposures if patients received prescriptions from pharmacies outside of the VA. However, prior studies indicate that this is unlikely to have been common given that most patients take advantage of the low prescription copay and obtain most of their medicines from the VA.37,38 In the chart abstraction analysis, VA prescription data were found for all but a few percentages of cases with notation of ACE or OAH use. Errors in dates of medications or diagnoses also could have led to misattribution of the angioedema given the rapid onset of the condition. There was little evidence for this in the abstracted chart data with the exception of a small number of cases with diagnosis within days of first prescription but onset of symptoms several days before the diagnosis. While these cases may have had their angioedema mistakenly classified as occurring after the first prescription, it appears that this was not common enough to substantially bias the estimates given the low rates of angioedema in the group classified as “before use.”

There are a number of isolated reports of angioedema after OAH use in the literature,39–42 but we found little evidence for increased risk associated with antihypertensives other than ACE. There were only small increases in incidence after new OAH use, and trends in the rates were not consistent. Some of the OAH, such as calcium channel blockers, are commonly used to treat patients with congestive heart failure or coronary artery disease, and these patients may be at increased risk of angioedema from the disease or from other medications that they are taking for these diseases (eg, aspirin, beta adrenergic blockers). Nevertheless, the estimates did not change with adjustment for these conditions or for other antihypertensive medications. Finally, for this analysis, we did not have sufficient numbers of users of angiotensin II receptor blockers so evaluation of angioedema risks associated with use of these agents is reserved for future studies with more recent data.


This study represents the largest observational study to date examining the association between antihypertensive drug therapy and incidence of angioedema. It is remarkably consistent with clinical trials in showing a similarly modest increase in incidence associated with use of angiotensin-converting enzyme inhibitors when compared with other antihypertensive medications. This increased risk extends beyond the initial period of ACE use and does not appear to be shared with other antihypertensive medications. The strength of this study is that it demonstrates that this increased risk extends across a broad spectrum of patients, including the elderly and those with multiple comorbidites, who typically may not be included in clinical trials. Yet, specific subgroups exist with different risks of ACE-associated angioedema including increased risk in blacks and decreased risk in patients with diabetes. The challenge for future research will be to better identify those subgroups at the highest risk for angioedema in whom the risks of ACE therapy may outweigh the benefits. This information is essential in making appropriate clinical decisions about medical therapy for hypertensive patients and patients with congestive heart failure.

We are grateful for the assistance of the principal investigators at the collaborating VA sites: Joe Englehart, Samuel S. Stratton VA Medical Center, Albany, NY; Douglas D. Bradham, VA Maryland Health Care System, Baltimore, Md; Alfredo Selim, VA Boston Healthcare System, Boston, Mass; Christine Clark, Jesse Brown VA Medical Center, Chicago, Ill; David Aron, Louis Stokes VA Medical Center, Cleveland, Ohio; Hayden B. Bosworth, Durham VA Medical Center, Durham, NC; Monika Safford, VA New Jersey Health Care System, East Orange, NJ; Maude Rittman, Gainesville VA Medical Center, Gainesville, Fla; Francesca Cunningham, Edward Hines Jr, VA Hospital, Hines, Ill; Anita Deswal, Michael E. DeBakey VA Medical Center, Houston, Tex; Hanna A. Bloomfield, Minneapolis VA Medical Center, Minneapolis, Minn; Jan Knoefel, New Mexico VA Hospital, Albuquerque, NM; Mary K. Goldstein, VA Palo Alto Health Care System, Palo Alto, Calif; Elizabeth Stephens, Portland VA Medical Center, Portland, Ore; Scott Sherman, VA Greater Los Angeles Healthcare System, Sepulveda, Calif; Britta Neugaard, James A. Haley VAMC, Tampa, Fla; Nancy Small, Central Texas Veterans Health Care System, Temple, Tex. We are also grateful for the assistance of Elaine Czarnowski, Mariscelle Sales, Elaine Hickey, Ann Borzecki, and Qing Shao.

Sources of Funding

This project was supported in part by Department of Veterans Affairs, Boston University School of Public Health, and Bristol-Myers Squibb in association with Galt Associates. The funding sources had no role in the design, conduct, and reporting of the study or in the decision to submit the manuscript for publication. The views expressed are solely the responsibility of the authors.


D.R.B. has served as a consultant for Bristol-Myers Squibb, a manufacturer of antihypertensive medications. The remaining authors report no conflicts.


Correspondence to Donald R. Miller, Center for Health Quality, Outcomes, and Economic Research, VAMC, Bldg 70 (152), 200 Springs Rd, Bedford, MA 01730. E-mail


  • 1 Brown N, Vaughan D. Angiotensin converting enzyme inhibitors. Circulation. 1998; 97: 1411–1420.CrossrefMedlineGoogle Scholar
  • 2 Heart Outcomes Prevention Evaluation Study Investigators. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. N Engl J Med. 2000; 342: 145–153.CrossrefMedlineGoogle Scholar
  • 3 Strippoli GF, Craig M, Deeks JJ, Schena FP, Craig JC. Effects of angiotensin converting enzyme inhibitors and angiotensin II receptor antagonists on mortality and renal outcomes in diabetic nephropathy: systematic review. BMJ. 2004; 329: 828.CrossrefMedlineGoogle Scholar
  • 4 Kjeldsen SE. Julius S. Hypertension mega-trials with cardiovascular end points: effect of angiotensin-converting enzyme inhibitors and angiotensin receptor blockers. Amer Heart J. 2004; 148: 747–754.CrossrefMedlineGoogle Scholar
  • 5 Israili ZH, Hall WD. Cough and angioneurotic edema associated with angiotensin-converting enzyme inhibitor therapy. Ann Intern Med. 1992; 117: 234–242.CrossrefMedlineGoogle Scholar
  • 6 Jett GK. Captopril-induced angioedema. Ann Emerg Med. 1984; 13: 489–490.CrossrefMedlineGoogle Scholar
  • 7 Slater EE, Merrill DD, Guess HA, Roylance PJ, Cooper WD, Inman WH, Ewan PW. Clinical profile of angioedema associated with angiotensin-converting enzyme inhibition. JAMA. 1988; 260: 967–970.CrossrefMedlineGoogle Scholar
  • 8 Agah R, Bandi V, Guntupalli KK. Angioedema: the role of ACE inhibitors and factors associated with poor clinical outcome. Intensive Care Med. 1997; 23: 793–796.CrossrefMedlineGoogle Scholar
  • 9 Vleeming W, van Amsterdam JG, Stricker BH, de Wildt DJ. ACE inhibitor-induced angioedema. Incidence, prevention and management. Drug Safety. 1998; 18: 171–188.CrossrefMedlineGoogle Scholar
  • 10 Leenen FH, Nwachuku CE, Black HR, Cushman WC, Davis BR, Simpson LM, Alderman MH, Atlas SA, Basile JN, Cuyjet AB, Dart R, Felicetta JV, Grimm RH, Haywood LJ, Jafri SZ, Proschan MA, Thadani U, Whelton PK, Wright JT, Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial Collaborative Research Group. Clinical events in high-risk hypertensive patients randomly assigned to calcium channel blocker versus angiotensin-converting enzyme inhibitor in the antihypertensive and lipid-lowering treatment to prevent heart attack trial. Hypertension. 2006; 48: 374–384.LinkGoogle Scholar
  • 11 Kostis JB, Kim HJ, Rusnak J, Casale T, Kaplan A, Corren J, Levy E. Incidence and characteristics of angioedema associated with enalapril. Arch Int Med. 2005; 165: 1637–1642.CrossrefMedlineGoogle Scholar
  • 12 Burkhart G, Brown N, Griffin M, Wayne R, Weiss SR. Angiotensin converting enzyme inhibitor-associated angioedema: higher risk in blacks than whites. Pharmacoepidemiology Drug Safety. 1996; 5: 149–154.CrossrefMedlineGoogle Scholar
  • 13 Gibbs CR, Lip GY, Beevers DG. Angioedema due to ACE inhibitors: increased risk in patients of African origin. Br J Clin Pharmacol. 1999; 48: 861–865.CrossrefMedlineGoogle Scholar
  • 14 Huston DP, Bressler RB. Urticaria and angioedema. Med Clin North Am. 1992; 76: 805–837.CrossrefMedlineGoogle Scholar
  • 15 Brown NJ, Ray WA, Snowden M, Griffin MR. Black Americans have an increased rate of angiotensin-converting enzyme inhibitor-associated angioedema. Clin Pharmacol Ther. 1996; 60: 8–13.CrossrefMedlineGoogle Scholar
  • 16 Morimoto T, Gandhi TK, Fiskio JM, Seger AC, So JW, Cook ejectionfactor (EF), Fukui T, Bates DW. An evaluation of risk factors for adverse drug events associated with angiotensin-converting enzyme inhibitors. J Eval Clin Prac. 2004; 10: 499–509.CrossrefMedlineGoogle Scholar
  • 17 Smith MW, Joseph GJ. Pharmacy data in the VA health care system. Medical Care Research & Review. 2003; 60 (3 Suppl): 92S–123S.CrossrefMedlineGoogle Scholar
  • 18 Cowper DC, Hynes DM, Kubal JD, Murphy PA. Using administrative databases for outcomes research: select examples from VA Health Services Research and Development. J Med Syst. 1999; 23: 249–259.CrossrefMedlineGoogle Scholar
  • 19 Boyko EJ, Koepsell TD, Gaziano JM, Horner RD, Feussner JR. US Department of Veterans Affairs medical care system as a resource to epidemiologists. Am J Epidemiol. 2000; 151: 307–314.CrossrefMedlineGoogle Scholar
  • 20 Agency for Health Care Policy and Research. Report to Congress: The feasibility of linking research-related data bases to Federal and non-Federal medical administrative databases. Agency for Health Care Policy and Research Pub. No. 91-003. Rockville, MD: Agency for Health Care Policy and Research 1991.Google Scholar
  • 21 Shen Y, Hendricks A, Zhang S, Kazis LE. VHA enrollees’ health care coverage and use of care. Med Care Res Rev. 2003; 60: 253–267.CrossrefMedlineGoogle Scholar
  • 22 Miller DR, Safford MM, Pogach LM. Shao Q, Wang F. Who has diabetes? Best estimates of diabetes prevalence based on computerized patient data. Diabetes Care. 2004; 27: B10–B21.CrossrefMedlineGoogle Scholar
  • 23 Medicode. International Classification of Diseases, 9th Revision, Clinical Modification, V Edition 1996.Google Scholar
  • 24 Fisher SG, Weber L, Goldberg J, Davis F. Mortality ascertainment in the veteran population: alternatives to the National Death Index. Am J Epidemiol. 1995; 141: 242–250.CrossrefMedlineGoogle Scholar
  • 25 Statistical Analysis Software I. SAS OnlineDoc Version Eight. Cary, North Carolina: The SAS Institute 2002.Google Scholar
  • 26 Ulm K. A simple method to calculate the confidence interval of a standardized mortality ratio (SMR) Amer J Epidemiol. 1990; 131: 373–375.CrossrefMedlineGoogle Scholar
  • 27 Selvin S. Statistical Analysis of Epidemiological Data. New York: Oxford University Press 1996.Google Scholar
  • 28 McCullogh P, Nelder JA. Generalized Linear Models. II ed. New York: Chapman and Hall; 1989: 193–208.Google Scholar
  • 29 Hennekens CH, Buring JE. Epidemiology in Medicine. Boston: Little Brown and Company; 1987.Google Scholar
  • 30 Rothman KJ. Modern Epidemiology. Boston, MA: Little Brown and Co; 1986.Google Scholar
  • 31 Irvin JD, Viau JM. Safety profiles of the angiotensin converting enzyme inhibitors captopril and enalapril. Am J Med. 1986; 81: 46–50.CrossrefMedlineGoogle Scholar
  • 32 Singer DR, Macgregor GA. Angioneurotic edema associated with two angiotensin-converting enzyme inhibitors. BMJ. 1986; 293: 1243.CrossrefGoogle Scholar
  • 33 Kazis LE, Ren XS, Lee A, Skinner K, Rogers W, Clark J, Miller DR. Health status in VA patients: results from the Veterans Health Study. Am J Med Quality. 1999; 14: 28–38.CrossrefMedlineGoogle Scholar
  • 34 U. S. Census (2000) National Population Estimates: Selected Years from 1990 to 2000. U.S. Census Bureau, Population Division, Population Distribution Branch (www/census/gov).Google Scholar
  • 35 Kashner MT. Agreement between administrative files and written medical records. Medical Care. 1998; 36: 1324–1336.CrossrefMedlineGoogle Scholar
  • 36 Brown NJ, Snowden M, Griffin MR. Recurrent angiotensin-converting enzyme inhibitor-associated angioedema. JAMA. 1997; 278: 232–233.CrossrefMedlineGoogle Scholar
  • 37 Miller DR, Cunningham F. Evaluation of VA pharmacy data for pharmaco-epidemiologic research. Pharmacoepidemiol Drug Safety. 2002; 11: S81.Google Scholar
  • 38 Hoff RA, Rosenheck RA. Cross-system service use among psychiatric patients: data from the Department of Veterans Affairs. J Behav Health Serv Res. 2000; 27: 98–106.CrossrefMedlineGoogle Scholar
  • 39 Bernstein RK. Angioedematous urticaria in a diabetic patient successfully treated with nifedipine. Diabetes Care. 1985; 8: 197–198.CrossrefMedlineGoogle Scholar
  • 40 Hedner T, Samuelsson O, Lindholm L, Andren L, Wiholm BE. Precipitation of angioedema by antihypertensive drugs. J Hyperten. 1991; 9: 360–361.Google Scholar
  • 41 Romano A, Pietrantonio F, Garcovich A, Rumi C, Bellocci F, Caradonna P, Barone C. Delayed hypersensitivity to diltiazem in two patients. Ann Allergy. 1992; 69: 31–32.MedlineGoogle Scholar
  • 42 Ruscin JM, Page RL II, Scott J. Hydrochlorothiazide-induced angioedema in a patient allergic to sulfonamide antibiotics: evidence from a case report and a review of the literature. Amer J Ger Pharm. 2006; 4: 325–329.Google Scholar


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.