Evidence Table 23Data abstraction of major harms in cohort studies

DesignTotal N
Subject recruitment
Inclusion criteriaF/U periodSettingPopulation
Age (mean (SD)
Sex (% female)
Intervention drugOther allowed HT medicationsTotal withdrawals
Withdrawals due to AEs
AE assessmentAdverse eventsAEs among subgroupsComments
 Chalmers 1992

Single-group cohort, open label, post-marketing67698

GPs report information on all their patients on captopril for HT; post-marketing surveillance
Patients on captopril for HTMedian F/U 6m
Total 39,635 pt-years
General practiceHT

Age: 60.4 (11.3) y 57% female
Captopril, dosage NRNA75% of patients completed the study

WD due to AEs: 8.2%
GPs reported AEs to a centralized agencyAngioedema: 16 pts; after median 28d (7–306)
Hypotension: 2.8/1000 (more common in >70y)
Heamatological disorders: 15 pts WD due to heam disorders; 11 leucopenia; 4 thrombocytopenia; none persisted after WD; several cases had other likely causes
Liver disease: 9 patients WD; all had other likely causes; 3 deaths from liver failure (not suspected to be related to drug)
Deaths: 1.1%; this rate was 80% of expected rate (in general populations) and 4% more than expected rate of CV deaths in general populations
Renald failure: listed as cause in 21 deaths; all had underlying disease
WD rates were higher in >70y and in females (10.4% in females >70y)
 Gonzalez-Perez 2004

Cohort with nested case-control3708 breast cancer cases; 18478 controls; total on ACE about 1000

Subjects and controls from National Practitioner Database in UK
Females 30–79 years of age; had computerized prescription at least 1y prior to entry; incident cases of breast cancer from database alsoNR (NR how long (recruitment period 1/1995–12/2001)GP officesHT

Age: NR
Captopril, enalapril, lisinopril
Dosages NR
NRNA (case-control)Incident cases of breast cancer, validated approach from UK databaseIncidence breast cancer among current users of ACE-I vs non-users:
Usage <2y: OR 1.1 (95% CI, 0.6 to 2.0)
Usage >2y: OR 0.8 (95% CI, 0.5 to 1.3)

Usage <2y: OR 0.9 (95% CI, 0.6 to 1.4)
Usage >2y: OR 0.7 (95% CI, 0.5 to 1.1)
Usage <2y: OR 0.8 (95% CI, 05 to 1.2)
Usage >2y: OR 0.7 (95% CI, 0.7 to 1.6)
NRIncidence breast cancer among users of anti-HT drugs vs non-users: OR 1.0 (95% CI, 0.9 to 1.1)
 DeBianco 1991

Single-group cohort, open label, prospective6669 with data (7658 started the trial)

Data from a trial; subjects apparently selected by participating GPs
Inclusion criteria: >18y; HF NYHA class II or III; no prior captopril; on diuretic +/− digoxin8wOfficeHF: mild-to-moderate

Captopril: start at 12.5 mg tid, titrate up to 15 mg tid

Mean dosage: 65 mg QD
NRTotal WDs: 14.8%

Of patients with AEs, 4.9% withdrew
NRAEs (1 or more) in 18.1% of patients
Total AEs: 1983 in 1386 patients)

Deaths: 3.0%, causes NR

Most common AEs: dizziness (2.4%), nausea (1.4%), cough (1.1%), hypotension (1.2%)
Postural hypotension: <1%
 Rosenthal 1996

Single-group cohort, open label, prospective, multicenter33841

Private practice physicians in Germany asked to record results of 4m of treatment in up to 5 patients
HT, not otherwise specified (33447/33841 had HT, NR what remaining patients had)mean 109dGeneral practiceHT

Age: 58.6Y (NR)

Diabetes: 5037/33841
Cilazapril: start at 1.25 mg qd, increase to 2.5 to 5 mg qd; 0.5 mg to 2.5 mg qd in elderly or with impaired renal function

Median dosage at end of observation period:
2.5 mg qd
Various; 14.4% took concomitant HT medicationsTotal WD: 6.7%

WD due to AEs: 3.7%
GPs reported serious AEs on a form, reported to central agencyOverall rate of AEs: 7.3%; 3.8% of total population considered to have drug-related AE

Severe AEs (not defined): 0.6% of total population; none felt to be related to treatment

Deaths: 44 patients (12 cardiac, 10 cerebral)

Hypotension (not defined): approximately 0.2% (graphical data)
NRRate of cough: 1.5% (most frequently reported AE)
 Messner 1995Single-group cohort, multicenter17546

Inclusion criteria: >18y; HF stabilized with diuretic and/or digitalis therapy

Exclusion criteria: on vasodilator; Cr>150 mmol/L: Na <130 meq/L; SGP <110 mm Hg; pregnant or nursing
3mGP officesHF: mild-to-moderate

Age: 70y (10.5)
Sex: 50.4% female
Race: 99.3% Caucasian
NYHA class II 67%, class III 33%
Enalapril: start 2.5 mg qd for 3d; 5 mg qd for 3d; 10 mg qd for 7d, then 20 mg qd
Mean daily dosage 16 mg

Run-in period: 1–3w; stability observed; diuretic dosage reduced by 1/2
Diuretic; digitalisTotal WDs: 3.3%

WD due to AEs: 1.4%
Patients asked to report any AE; investigator completed a case reportOverall adverse event rate: 5.6%

Hypotension: 0.34%
Postural hypotension: 0.3%
Hyperkalemia: 0.13%
Death: 127/17,546 (0.72%); none felt related to drug
Worsening HF: 95/17,546 (0.54%); none felt related to drug
MI: 0.10%
Pulmonary embolism: 0.08%
NRCough: 1.7%
Creatinine: increase 10.9 to 11.1 mg/d, P=0.0001
 Thorp 2005

Retrospective cohort18977 prescribed lisinopril; 13166 had pre and post Cr levels

Computerized database of HMO medical records
Patients taking lisinopril between 7/2000 and 6/2002; >40y6mUS HMOVarious indications
Age: reported by stratum
Sex; 49.5% female
Race: NR
DM and/or CHD: 53.8%
Lisinopril; dosage NRNANA: only subjects with pre and post Cr levels were examinedPre and post lisinopril serum creatinine levels within 6m of initial prescriptionRise in serum Cr from ≤ 1.2 mg/dl to >2.5 mg/dL: 31 patients (0.2%)
Rise in serum Cr from ≤ 1.2 mg/dl to >1.2 mg/dl: 6.8%

In N=31: possible contributors to increase in Cr: CHF (9/31), dehydration 7/31, infection 4/31
In N=31, “most patients” had decrease from rise in subsequent 6m to <2.5 md/dL

ESRD: 0 cases

Deaths: 3 patients
In N=31, 11 had no DM or CAD; 20 had one or both
 Speirs 1998

Post-marketing surveillanceo

From GP offices across France; physicians selected up to 10 patients per practice
Adults with nonaccelerated HT and DBP 95–115 mmg Hg

Exclusion: pregnancy, breast-feeding, secondary HT, history of stroke, MI, or unstable angina in last 3m;hepatic, renal, or other serious disease
12mGP officesHT

Age: 60.9 (NR)
53% women

Diabetes: 14%
Perindopril: started at 2 (>70y) to 4 mg and titrated up to 8 mgDiureticsTotal WD: 4008/47,351 (8%)

WD due to AEs: 6.3% female, 3.5% male
Case report formsOverall rate of AEs: men 14.2%, women 17.8%

Deaths: 190 (0.4%)’ 27 due to MI; 26 due to stroke
Hospital admissions: 255
Renal dysfunction: men 0.14%, women 0.17%; 3 cases of CKD referred for hemodialysis (2 had renal artery stenosis)
Angioedema: men 0.004%, women 0.02%
Serious allergic reaction: men 0.02%, women 0.01%; 3 cases were pancytopenia which started after perindopril started
Hematologic disturbance: men 0.02%, women 0.004%
Serious allergic reaction: men 0.02%, women 0.01%
Hypotension: men 0.29%, women 0.4%; 1 case related to nonfatal stroke
WD due to AEs: no difference across age and sex groups except for WD due to renal insufficiency which increased with age; rate highest in men >80yCough: 11.3% in women, 7.8% in men
 Tytus 2007

Single-group cohort, open label, prospective, multicenterEnrolled: 2096
Completed 14-w titration period: 1683

Recruited from randomly-selected primary care practices; NR how provider selected patients
Stage 1 to 2 HT, no prior HT treatment or HT uncontrolled on current monotherapy with a diuretic or calcium channel blocker

Exclusion criteria: on steroids, secondary HT, clinically significant CVD
26wPrimary care clinics in CanadaHT (newly treated or uncontrolled on current first-line medications)

Age: 56.6(12.6) (years)
Trandolapril: 1 mg qd, titrated up to 4 mg qdVerapamil 240 mg qd, diuretic

Excluded: beta-blockers, another ACE-I
Total WD during 14-w titration period 413/2096 (19.7%)

WD during remaining 12w: 33/1683 (2%)

WD due to serious AE: 19 (0.9%)
WD due to nonserious AE: 169 (8.1%) (cough, nausea, headache)
Treating physician asked about AE at each visit and determined if AE causally related to drugTotal of 343 AEs attributed to study drugs in 252 patients (15.3%)

Serious AEs: pregnancy, cerebral aneurysm, diabetic crisis, TIA, carcinoma, others (rates NR)
None attributed to trandolapril
 Bramlage 2004

Single-group cohort, open label, prospective, multicenter17,284; 16,600 “could be used in the analysis”

GPs selected patients for treatment with irbesartan; study period 10/2002 to 6/2003
>=18y, with HT and type 2 diabetes3mGerman GP officesHT and DM2

Age: 62.2 (10.7)
Irbesartan 300 mg qd (Aprovel 300) or combined with HCTZ 12.5mg qd (CoAprovel 300)HCTZ allowed, other HT drugs as neededData available on 16,600/17,284 (96.0%); no other detailsCollected by GPs; no other details62 AEs noted in 48 patients (0.3% of total); 2 serious AEs: terminal renal insufficiency “not related to study medication” and tremor “likely related”

No deaths during study
Included subjects had DM2; no other subpopulations examined
 Schrader 2007

Post-marketing surveillance; prospective, multi-center14200

Physicians collected data on patients they elected to treat with irbesartan
Adults with uncontrolled HTUp to 9mGeneral practiceUncontrolled HT, with or without metabolic syndrome

Age: 62 (10.8) y

Diabetes: 31.1%
Metabolic syndrome: 65.4%
Irbesartan 75 to 300 mg daily or irbesartan/HCTZ 150/12.5 or 300/12.5 mg qdAs needed; no restrictionsTotal WD: NR

WD due to AEs: NR
GPs reported serious AEs on a form, reported to central agencyOverall AE rate: 0.62% (141 events in 88 patients)

Number of patients (n=14,200)
Serious AEs (not defined): 34 patients (0.24%) (not all were listed in table or described)
Deaths: 16 over 9-m F/U
Cardiogenic shock: 1
Cerebral infarction: 1
Gastrointestinal hemorrhage: 1
MI: 2
 Mann 1999

Post-marketing surveillance14522

Data from prescription event monitoring database; forms sent to physicians who prescribed the drug
Physician completed drug AE form for patients’ first prescriptionPatients on drug at least 6mGP officesHT or HF

Age: 63.5 (12.1) y

Sex: 59.3% female
LosartanNASurvey response rate 60%; additional 7.8% had no event data; useful information obtained on 14,522 subjects

Total WD: 17.5% after 6m
WD due to AEs: 5.1%
Prescription event monitoring system303 adverse drug reactions (defined as attributed to the drug by the GP) (including dizziness, headache, malaise, nausea, cough, etc)

Incidence density per 1000 patient-months: month 1; month 2–5
Cardiac failure: 53; 115
Renal dialysis: 13; 2

Number of cases:
Angioedema: 8
Renal failure and electrolyte abnormalities: researchers unable to differentiate from pre-existing disease
Death: 363; none attributed to losartan (Table 5 lists causes)
Incidence density higher for>76y vs <76 (P<0.05) for cough, dizziness, edema, nausea/vomiting
 Schmidt 2008

Single-group cohort, open label, multicenter, prospective4252

Physicians collected data on patients they elected to treat with irbesartan
Adults with HT treated in GP offices6w; mean 44.1d (SD 21.7)General practiceMild-to-moderate HT

Age: 62.5 (11.9)

Diabetes: 20.9%
CHD: 16.4%
HF: 9.9%
Renal failure: 3.4%
Olmesartan 10 to 40 mg qd; mean dosage 19.9 (7.1) mgAs needed; no restrictionsTotal WD: NR

WD due to AEs: NR
GPs reported serious AEs on a form, reported to central agencyOverall AE rate: 0.66%

Serious AEs (not defined): 2 patients: circulatory collapse and aortic bypass surgery
NRDizziness most common AE (0.19%)

Text resembles Schrader 2007
 Schumacher 2008

Retrospective cohort derived from the Micardis project database, which includes results from 30 double-blind and 20 open-label clinical studies5013 for telmisartan monotherapy in RCTs; 5907 in open-label studies

NR, likely varied across included studies
Adults with HT; varied somewhat across studiesVaried across studies: 7d to 2y; mean duration in double-blind studies 67dNRHT

Age: double-blind studies: 55.9 (11.2)
open-label studies:56.3 (11.3)

Race: double-blind: 90.0% non-black, 10.0% black
Telmisartan 20–160 mg +/− HCTZ 6.25 to 25 mg qd
or placebo
Other HT medications allowed in most studiesTreatment discontinuations due to AEs in double-blind studies:
0.33 PPY (4.4%) with placebo, 0.14 PPY (2.6%) with monotherapy; in open-label studies 0.07 PPY (4.0%)
AEs were spontaneously reported by the patient or detected by the investigatorAEs PPY in double-blind (open label) studies: monotherapy 2.03 (37.4%) (0.65, 49.6%), placebo 2.73(36.1%)

Serious AEs: monotherapy 0.07 (1.2%) (0.07, 4.4%)), placebo 0.09 (1.2%); NSD between active treatment groups in double-blind studies

Open-label studies, events PPY with monotherapy MIs: 0.004, (0.3%)

Deaths: overall 0.004 PPY with monotherapy

Hepatobiliary laboratory abnormalities: <0.05% with monotherapy
The incidences of all-cause AEs PPY were lower in patients >65y than <65y; serious AEs were higher in older groupAEs occurring at >1% in double-blind studies: headache, dizziness, fatigue; cough, peripheral edema, erectile dysfunction occurred at 0.3% or less in double-blind studies, 0.7% or less in open-label studies
 Biswas 2002

Single-group cohort, open label, prospective12881

GPs completed questionnaire on dispensed prescriptions
Patients with prescriptions dispensed between 12/96 and 11/98At least 6m after start of drugGP officesHT (assumed as was indication)

Age: men 61.1(12.1);
women 65.4 (12.5)

59% females
Valsartan: dosage NRNAReturn rate on questionnaires: 55%

WD at 6m F/U: 19.9%
GPs completed questionnaire on dispensed prescriptions; adverse drug reactions were reviewed in detail and additional questionnaire sent to the GPTotal AEs: 295 events in 209 (1.5%) of patients
Most common reasons for WD due to AEs: malaise (0.3%), dizziness (0.1%)
Deaths: 1.5% (78/85 due to CVD or cancer)

Angioedema: 0.03%
Abnormal liver function tests: 0.2% (1 case of jaundice and 1 of hepatitis improved after stopping the drug)
Hyperkalemia: 0.13%
Hyponatremia: 0.12%
Spontaneous bleeding: hematuria, hemoptysis, ect: 59 cases; unclear if related to drug

From: Evidence Tables

Cover of Drug Class Review: Direct Renin Inhibitors, Angiotensin Converting Enzyme Inhibitors, and Angiotensin II Receptor Blockers
Drug Class Review: Direct Renin Inhibitors, Angiotensin Converting Enzyme Inhibitors, and Angiotensin II Receptor Blockers: Final Report [Internet].
Norris S, Weinstein J, Peterson K, et al.
Portland (OR): Oregon Health & Science University; 2010 Jan.
Copyright © 2009, Oregon Health & Science University, Portland, Oregon.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.