1. Critical Appraisal DR Joshna Rajbaran

2. CARDIAC TROPONIN and OUTCOME in ACUTE HEART FAILURE NEJM 358;20 MAY 15,2008

3. THE AIM: To describe the association between elevated cardiac troponin levels and adverse events in hospitalized patients with ACUTE DECOMPENSATED HEART FAILURE

4. WHY?? Because an objective risk- stratification process for the evaluation of acute decompensated heart failure is lacking.

5. The value of measuring serum cardiac troponin when a patient presents with acute decompensated heart failure remains uncertain.

6. NB: Troponins Trop T & Trop I are regulatory proteins with a very high specificity for cardiac injury. They are released early ( 2-4 hrs) & can persist for up to 7 days. Troponin testing is primarily used as a tool in diagnosing myocardial infarctions. Elevated levels suggest myocardial or some form of cardiac damage. Insignificant if used in the absence of S&S of cardiac disease!!

7. THE KEY DIFFERENCES LARGE STUDY SHORT TERM OUTCOMES IN HOSPITALIZED PATIENTS WITH ACUTE DECOMPENSATED HEART FAILURE.

8. METHOD Registry data: –ADHERE( Acute Decompensated Heart Failure National Registry) –Observational registry –274 hospitals –TIME FRAME :October 2001  January 2004

9.  Inclusion criteria: Hospitalization & documentation of the measurement of trop I or trop T at “INITIAL” evaluation

10.  Exclusion criteria: serum creatinine level>2.0mg/dl or 176.8umol/l  Ischemic heart failure defined as cause if : hx coronary artery disease OR hx myocardial infarction Not as exclusion criteria!!!

11. METHOD Troponin measurement:  Trop T & trop I were interchangeable levels considered positive, with cut-off based on expert consensus!!  Trop T≥0.1µg/l & Trop I ≥1.0µg/l

12. Method Statistical analysis:  Primary out-come all causes  Secondary out-come differences in medical mx / procedures / length of stay between +ve & -ve cohorts  All outcomes were specified before the data were examined

13. Statistical analysis ( cont)  Associations between therapy & mortality  Controls used in this regard  Mortality was adjusted for relevant prognostic factors

14.  Logistic regression adjusted for: age / blood urea nitrogen / SBP / DBP / serum creatinine / serum sodium / HR /dyspnea at rest  1.2% records excluded due to missing values

15.  SAS software  Study designed by all authors  ADHERE statisticians

16. METHOD Source Time period Inclusion criteria Exclusion criteria IHD/Race / Gender troponin measurements justified Statistical analysis explained Tools and teams stated

17. RESULTS Troponin levels & characterists of the patients

18. 105,388 84,872 ( 80.5% ) Hospitalized Trop tested Cr < 2mg/dl 67,924 Positive Negative 4240 (6,2%) 63,684

19. There were small but significant differences between the two cohorts!!!

20. Troponin- positive patients on admission:  Lower SBP  Lower EF  Less likely AF  Summary of characteristics given +ve vs –ve Trop  No comparison made for the two proteins as only 2% had both tested!!

21. REVISION OF TERMINOLOGY Odds ratio : provides a more useful way of presenting diagnostic data & can be applied to individual patients in a way that specificity & sensitivity cannot. It is a number btw 0 to infinity IF > 1 indicates that the information increases the likelihood of the suspected diagnoses. IF <1 it decreases the likelihood of the suspected diagnoses!!

22. SPECIFICITY: the proportion of patients WITHOUT the disease who are correctly identified by the test. SENSITIVITY: the proportion of patients WITH the disease who are correctly identified by the test.

23. RESULTS In-hospital mortality  Trop Positive (8.0%) > Trop Negative (2.7%) patients.......... (P<0.001)  Actuarial analysis  Trop as a continuous variable  Adjusted odds ratio for death (P<0.001)

24.  IHF was not a useful discriminator of Troponin status, nor was it predictive of mortality.  IHF Trop +ve 53% Trop –ve 52%  Trop +ve mortality 8,4% IHF 7,4% non-IHF  Trop –ve mortality 2,8% IHF 2,6% non-IHF

25. RESULTS Treatment, Troponin status & Mortality  Diuretics  +ve more likely to receive: nitroglycerine, inotropes & vasodilators  Resource utilization and mortality  No interaction between treatment & Troponin status with respect to mortality

26. RESULTS Sample size large but justified Basic data adequately described Variables taken into account Missing data accounted for Numbers add up High risk cohort established Statistical significance assessed

27. Main findings and their value: Prognostic value / cost Early assessment of risk/ triage & management Add to existing risk-stratification data for predicting the short term risk of death among patients with acute decompensated heart failure... Blood urea>15.4mmol/l SBP < 115mm Hg Cr >243.1µmol/l More aggressive therapeutic approach justified

28. Value of findings from Trop negative cohort Identifying low risk patients/ planning Rx Other studies  the impact of early risk stratification has been supported BASEL TRIAL EFFECT STUDY SMALLER STUDIES-98 CONSECUTIVE PTS -159 PTS -RITZ-4 STUDY

29. Studies correlating Troponin with physiological variables Impact on guidelines : National-ACS Trop & brain natriuretic peptide or N- terminal pro-brain peptide. Current for Heart Failure Trop NOT mentioned & brain nitriuretic peptide only if dx uncertain!!!

30. Suggested guideline!!! Measurement of Troponin levels in patients who present with heart failure provides independent prognostic information regarding in hospital death & other clinical outcomes & can be useful for risk stratification of such patients!!!!

31. LIMITATIONS Retrospective analysis ADHERE large data set : investigator discretion, diagnosis not objectively ascertained, cause of death not consistently recorded Troponin tests  Introduction of variability/ bias  Measurement only at admission  Interaction with other biomarkers Under represented adverse outcomes

32. Critical appraisal INFORMATIVE STUDY AIM/METHOD/FINDINGS SIGNIFICANCE STRENGTHS & LIMITATIONS WITH SUGGESTIONS OFFERED I FOUND NO REASON TO QUESTION THE STATISTICAL APPROACH SUGGESTIONS FOR FUTURE STUDIES OTHER RELEVANT STUDIES DOCUMENTED

33. With relevance to SA South African statistics :10 473 mortalities per annum d/t Heart Failure vs. US 55,704 Further evaluation of other biomarkers vs Trop T required Cost factors need to be examined Ischaemic heart disease is the commonest cause for acute heart failure in America.

34. HOWEVER, in Sub- Saharan Africa the causes in Africans are largely ( > 90%) NON-ISCHAEMIC viz.: HPT / cardiomyopathy / Rheumatic heart disease / chronic lung disease / pericardial disease Coronary artery disease and it’s complications remain uncommon in Africa but the situation is changing!!

35. I found the journal article rather transparent in it’s limitations However, there was one limitation that seemed to stand out: that some patients with both heart failure and ACS may have been included!!!! I think that with urbanization,varying risk profiles amongst race groups, risk prone behaviour & diet, that the findings are worthy of consideration in our setting.

36. Finally, EARLY RISK STRATIFICATION may help identify patients who are likely to receive the greatest benefit from intensive therapy.....that in itself highlights it’s relevance to emergency medicine!!!!