Cohort studycited by the Tayebat archivefull text checked★ key study A-ABX-13 in-depth review
Azithromycin and heart deaths
Azithromycin and the risk of cardiovascular death.
Overview
Ray WA et al.. N Engl J Med 2012 · PMID 22591294 · DOI 10.1056/NEJMoa1003833
Azithromycin was linked to about 47 extra heart deaths per million courses compared with amoxicillin.
Does the study support the specific claim? Yes · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Test CV risk.
- How was the study done?
Cohort.
- How many people, animals or samples were studied?
347,795 azithromycin courses vs 1.39 million matched non-use periods and 1.35 million amoxicillin courses (Tennessee Medicaid, 1992-2006).
- What was actually found?
CV deaths during 5 days: 85.2 per million azithromycin courses vs 31.5 per million amoxicillin courses; HR 2.49 (1.38-4.50) vs amoxicillin.
- How large was the observed effect?
HR 2.49; +47/million.
- What do the results mean in everyday life?
Rare but real for high-risk patients.
- What are the study's limitations?
Observational; few deaths (29); Medicaid population; no difference from levofloxacin.
- What does it say about Elawady's specific claim?
Supports.
- Do the results apply to Germany and Europe?
Moderate.
- What can explicitly not be concluded from this study?
That antibiotics are generally dangerous.
Scientific in-depth review
1. Full reference
- Authors
- Ray WA, Murray KT, Hall K, Arbogast PG, Stein CM
- Journal
- N Engl J Med
- Year
- 2012
- Identifier
- PMID 22591294 DOI 10.1056/NEJMoa1003833 · PMC3374857
- Verification
- verified – PubMed E-utilities efetch, 2026-10-25
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
Supported by National Heart, Lung, and Blood Institute (HL081707) and AHRQ Centers for Education and Research on Therapeutics cooperative agreement (HS1-0384); sponsors had no role in conduct or reporting; disclosure forms available at NEJM.org (not in text).
2. Research question
Does azithromycin increase cardiovascular death?
3. Study design and population
- Study design
- Cohort study – Retrospective cohort (Tennessee Medicaid linked to death certificates and hospital discharges) with propensity-score-matched no-antibiotic control periods and active-comparator antibiotic groups (new-prescription course as unit of analysis)
- Population
Tennessee Medicaid enrollees aged 30-74 years without life-threatening non-cardiovascular illness (azithromycin users 77.5% women, mean age 48.6)
- Country
USA
- Period
1992-2006
- Sample size
347,795 azithromycin prescriptions; 1,391,180 matched control periods; 1,348,672 amoxicillin, 264,626 ciprofloxacin and 193,906 levofloxacin prescriptions
- Inclusion and exclusion criteria
Age 30-74; >=365 days enrolment and regular medical care; excluded life-threatening non-cardiovascular illness, drug abuse or nursing-home residence in prior year, hospitalisation in prior 30 days; person-time during/shortly after hospitalisation excluded; controls no study antibiotic in prior 30 days
4. Methods and endpoints
- Exposure / intervention
Azithromycin prescription (5-day course; 10-day window also analysed)
- Comparator
Propensity-matched periods without antibiotics (4 per course, 153 covariates); amoxicillin (primary active comparator), ciprofloxacin, levofloxacin
- Endpoints
Primary: cardiovascular death and death from any cause; secondary: sudden cardiac death (validated algorithm, PPV 88%)
- Statistical methods
Cox regression with pairwise propensity scores (including recorded indication for antibiotic comparisons), IPTW-standardised balance checks, excess risk = (HR-1) x amoxicillin cumulative incidence, by CV-risk-score decile; sensitivity analyses (repeated measures, propensity-decile stratification, other proarrhythmic drugs)
5. Actual results
- Results
Days 1-5: CV deaths azithromycin 29 (85.2 per million courses; 22 sudden cardiac deaths, 64.6/million) vs 41 (29.8/million) in no-antibiotic periods and 42 (31.5/million) with amoxicillin. Azithromycin vs no antibiotic: CV death HR 2.88 (95% CI 1.79-4.63); any-cause death HR 1.85 (1.25-2.75). Vs amoxicillin: CV death HR 2.49 (1.38-4.50); any-cause death HR 2.02 (1.24-3.30). 10-day window vs no antibiotic: CV death HR 1.86 (1.27-2.73), any-cause HR 1.27 (0.92-1.75); no excess on days 6-10. Vs ciprofloxacin: CV death HR 3.49 (1.32-9.26). Vs levofloxacin: CV death HR 1.27 (0.66-2.47), any-cause HR 1.07 (0.61-1.85). Amoxicillin itself showed no increased risk vs no antibiotic. Sensitivity analyses similar.
- Absolute risks
Per 1 million 5-day courses vs amoxicillin: 47 additional CV deaths overall; 245 additional in the highest CV-risk decile (which accounted for 59% of CV deaths during azithromycin therapy). Crude CV death rates: 85.2 (azithromycin) vs 31.5 (amoxicillin) vs 29.8 (no antibiotic) per million.
- Authors' conclusion
During 5 days of azithromycin therapy there was a small absolute increase in cardiovascular deaths, most pronounced among patients with high baseline cardiovascular risk; the data cannot establish a specific causal mechanism.
6. Strengths
- Very large linked database with validated death outcomes
- Two control strategies: propensity-matched non-users and active comparator with similar indications (amoxicillin)
- Pre-specified hypothesis based on proarrhythmia reports
- Risk confined to treatment days 1-5, consistent with acute pharmacological effect
- Multiple sensitivity analyses; absolute risks reported
7. Weaknesses
- Observational; residual confounding (smoking, BMI, infection severity) possible as acknowledged
- Small numbers of events (29 CV deaths on azithromycin)
- Medicaid population (largely women, low income), 1992-2006
- Cause-of-death misclassification (up to 25% for sudden death definition)
- No difference vs levofloxacin; mechanism not established
8. Risk of bias
- Orienting tool
- ROBINS-I
- Type of assessment
- structured assessment based on the full text by a single reviewer (no formal double review)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Single-reviewer structured ROBINS-I-oriented judgement based on the full text ('some_concerns' used as equivalent of ROBINS-I 'moderate'); not formal (needs two independent reviewers and a protocol).
| Domain | Judgement | Rationale |
|---|---|---|
| Confounding | some concerns | Propensity scores from 153 covariates plus active comparator with similar indications; unmeasured lifestyle factors and infection severity remain. |
| Selection of participants | low | New prescriptions with eligibility assessed at fill date; exclusions apply equally to all groups. |
| Classification of interventions | low | Pharmacy dispensing records; adherence unknown but non-differential. |
| Deviations from intended interventions | not applicable | Intention-to-treat style analysis of prescription courses. |
| Missing data | low | Administrative linkage with complete death ascertainment. |
| Measurement of outcomes | some concerns | Death certificates and validated algorithm; cause misclassification possible but all-cause death also elevated. |
| Selection of reported result | low | Pre-specified primary endpoints and comparators; sensitivity analyses reported. |
Full-text review (25 October 2026): Added period 1992-2006, eligibility details, crude event counts and rates (29 CV deaths; 85.2 vs 31.5 vs 29.8 per million), CIs for all comparisons including all-cause HR 2.02 (1.24-3.30) and levofloxacin/ciprofloxacin, 10-day results, funding; full ROBINS-I domains (overall unchanged).
9. Assessment of the statistical interpretation
Relative risk roughly 2.5-fold but absolute excess small (47 per million courses, 245 per million in the highest-risk decile); CIs wide because of few events; no significant difference from levofloxacin.
10. Relevance to the specific claim
11. Independent contrary findings
12. Transferability to Germany and Europe
Moderate.
13. Overall methodological judgement
High-quality pharmacoepidemiological study; the archive reports its numbers accurately and with absolute risks. It shows a rare acute cardiac risk of one antibiotic in high-risk adults, not that antibiotics in general are dangerous.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | moderate |
| Directness | direct |
| Reproducibility | partly replicated |
| Consistency with the overall evidence | mixed |
| Regional transferability | moderate |
| Adequacy of the interpretation (by the archive) | correctly reported |
14. Open questions
Replication in other populations; risk with other macrolides and fluoroquinolones in direct comparisons.
15. Linked fact checks
Use by the Tayebat archive
Note: according to the archive, the studies were compiled by the archive operators. Whether Dr. Elawady himself cited this study is not established.
- What the archive says about the study
Five days of azithromycin carried a 2.49-fold higher cardiovascular death risk versus amoxicillin (47 / 245 extra deaths per million courses).
- A. What was actually studied?
Azithromycin CV risk.
- B. What did the authors find?
As stated.
- C. What conclusions do the authors themselves draw?
Small absolute increase.
- D. What does the archive derive from it?
Reported with absolute numbers.
- E. Is this derivation scientifically justified?
Correct.
- Adequacy of the interpretation (by the archive)
- correctly reported