Microsoft word - harris1020508.doc

Secondary data analysis Participant selection Ads Meta-analysis: Objective: Understand what a meta-analysis is, how to interpret, and where to go for further guidance Evidence-based medicine: Clinical practice should follow the best supported information on outcomes. Why do we do meta-analysis? Presumption– no one definitive study as any study is unlikely to address all known and unknown sources of bias. Meta-analysis is a systematic review and statistical analysis of data from studies relevant to the question. Two major types: 1. Studies themselves are “units” of an analysis 2. Subjects within studies are pooled Should be as carefully planned as any other research project with a detailed, written protocol in advance and a priori definitions of eligibility for studies Chart showing “What is a Systematic Review?” For more information: Egger, Smith, Phillips. Meta-analysis: Principles and procedures. BMJ 1997:315;1533-1537 (series) Simple issues: 1. Inclusion criteria 3. Precision does not = truth if there is a Recent past: Cox-2 inhibitors and risk of myocardial infarction Next up: Rosiglitazone and cardiovascular events Type 2 diabetes mellitus-insulin resistance and beta-cell dysfunction Many metabolic abnormalities What are thiazolidinediones and how do they work? Rosiglitazone and pioglitazone are potent inhibitors of peroxisome-proliferator activator-receptor γ – improve whole body insulin sensitivity with actions on adipose tissue and liver. Graph showing “The Cardiovascular Role of the TZDs in Type 2 Diabetes J Diabetes Complications. 2007 Sep Oct;21(5):326-34 Article: Glycemic Durability of Rosiglitazone, Metformin, or Gluburide Monotherapy Table 4: showing Effect of Rosiglitazone on the Risk of Myocardial Infarction and Death from Cardiovascular Causes Table 5: showing Rosiglitazone on the Risk of Myocardial Infartction and Death from Cardiovascular Causes Press release: GlaxoSmithKline Responds to NEJM Article on Avandia Rosiglitazone Evaluated for Cardiovascular Outcomes – An Interim Analysis Table 2: Hospitalization or Death from Cardiovascular Causes Summary: Rosiglitazone for type 2 diabetes mellitus (Review) Summary: Rosiglitazone for type 2 diabetes mellitus (Review) contd. Press Release: GSK Responds to Outline Review of Rosligliazone by the Cochrane Collaboration Philadelphia, PA – July 17, 2007. Flow Chart of Congestive Heart Failure and cardiovascular death in patients with pre-diabetes and type 2 diabetes given thiazolidinediones” a meta-analysis of randomized clinical trials. Interpretation of clinical trials from previous table Overall risk for congestive heart failure: status, weight, risk ratio at 95% ci numbers and figures Overall risk for cardiovascular death: status, weight, risk ratio at 95% ci numbers and figures Glycemic Durability of Rosiglitazone, Metformin, or Glyburide Monotherapy. Results of study with note added: While this article was in production, further examination of data on adverse events identified a higher rate of fractures in the group receiving rosiglitazone. This was an unexpected event that was not part of the pre-specified analysis plan. Graph showing risk of fractures in women in ADOPT at 5 years. Rosiglitazone and Cardiovascular Risk Rosiglitazone – Continued Uncertainty about Safety Rosiglizasone and Cardiotoxicity – Weighing the Evidence Cardiovascular Risk and the Thiazolidinedoines Déjà Vu All Over Again? Thiazolidinediones, deadly sins, surrogates, and elephants Secondary data analysis Participant selection Ads New horizon for meta-analysis: GENOME-WIDE ASSOCIATION STUDIES How to Interpret a Genome-wide Association Study Thomas A. Pearson; Teri A. Manolio JAMA. 2008;299(11);1335-1344(doi:10.1001/jama.299.11.1225) Data Chart: Association of three genetic loci with Eric acid concentration and risk of gout: a genome-wide association study Table1: Characteristics of participants to the studies Data: Association of three genetic loci with Eric acid concentration and risk of gout: a genome-wide association study Graph: Association of three genetic loci with Eric acid concentration and risk of gout: a genome-wide association study Overview: Meta-analysis Secondary data analysis C-reactive protein Participant selection Ads Secondary data analysis: Objectives: Open up possibilities for obtaining preliminary data Consider the range of secondary data analysis in addition to meta-analysis Benefits: Data often available, therefore study should be cheap to perform. Good way to work through the problems of the study design including case definition, controls, potential biases and develop statistical techniques. Preliminary data for applications Networking and collaborations Asking for data: Sharing and collaborating, not appropriating. Most large studies have data resources available or have standard procedures for collaborations. Creative add-ons to existing studies—nested studies—use the original sample to answer a different but related questions. May involve using laboratory specimens. Don’t be shy! Sources of data: Published statistics Federal or local survey data (geocoding) National Center for Health Statistics Computerized medical records Industrial records dbGAP Published studies Observation studies – case/control Trials – pre and post Chart: Collected data – new hypothesis- CRP and obesity in children National Center for Health Statistics - CDC Chart: Novel marker – measure changes before and after intervention Table: Ridker et al. Nested case-control WHI C-reactive protein and HRT-Synergistic effect? Novel marker in cases and controls--?? Risk Chart: Do statins reduce both lipids and CRP? Figure 3: Mean change in CRP levels over time according to observed changed in LDL cholesterol. Data are shown for those allocated to prevastatin (solid bars) or to placebo (open bars). Chart: Result from clinical trial- lowering LDL cholesterol and CRP is synergistic Downside: Hypothesis-generating, power, compromises in design and problems of data-dredging Easy to make errors because you didn’t design study-need to learn as much as you can before starting to use data Statistical help Still need to assess the validity of the results Does the literature support your observation? Is the result biologically plausible? Overview: Meta-analysis Secondary data analysis Participant selection Estrogen risk story Ads Overview: Your research question: What are the health effects of estrogen therapy for postmenopausal women? Meta-analysis: Estrogen use and CHD risk, 1991 Figure2: Summary relative risks and 95% interval estimates for studies of estrogen use and risk of coronary disease, by study design. There was significant (p<0.001) heterogeneity by study design. Meta-analysis: Estrogen and Breast Cancer, 1991 Table 3: Effects of Estrogen Replacement Therapy on Breast Cancer Risk in Women Data From 15 Case-Controlled Studies Stratified and Pooled by High, Moderate, and Low Quality Scores, 1976-1989. Can a meta-analysis reach the wrong conclusion?? Even if biologically plausible, may be wrong Biases may overestimate benefit, underestimate risk Healthy use selection bias Continue to use=better health=better outcomes Choosing subjects to address potential bias THE HERS STUDY Heart and Estrogen/progestin Replacement Study guidelines Exclusions from HERS: CHS event within 6 months of randomization Serum triglycerides >300 mg/dL Use of hormones within 3 months of screening History of DVT or pulmonary embolism History of breast cancer or suggestive mammogram History of endometrial cancer, abnormal uterine bleeding, endometrial thickness of greater than 5mm on screening Abnormal PAP test Serum aspartate aminotransferase level > 1.2 times normal Planning to move within 4 years Disease other than CHD deemed likely to be fatal within 4 years NYHA Class IIII or IV congestive failure Alcoholism Uncontrolled hypertension, diabetes Participation in another clinical study Less than 80% compliance with placebo run0in prior to randomization History of intolerance to hormone therapy Participant selection: On exposure: Say HRT…. “Are you currently using HRT?” Criteria: What type of HRT? [Is it really an HRT?] Which formulation? [Response may vary by type] How long has it been taken? Taken continuously or intermittently? [Years taken may affect the effect] Has she taken the same type for the whole time? Participant selection: On “case” status: Defining a case also identifies your controls You want your controls from the same reference population, but to truly differ from your cases in terms of the underlying feature you are studying Mixing of “cases” in your “control” group pushes toward a null result! HOW MANY WAYS TO IDENTIFY CHD? Large, simple: Told of MI by MD Cholesterol Coagulation, inflammation markers Adhesion molecules Table 1: Baseline Characteristics of HERS Participants (n=2763) by Treatment Group* Hers indicates Heart and Estrogen Replace Study; CHD coronary heart disease LDL, low-density lipoprotein; and HDL high density-lipoprotein. Pvalues are for difference between treatment groups by t test or x2. Table 3 – Outcome by Treatment Group and Year Since Randomization Cartoon of man pushing large question mark up a steep hill. Graph: WHI Results for CHD and Breast Cancer Graph: WHI Results for colorectal cancer and hip fracture Updated meta-analysis of estrogen risks, 2003 Table4: Hormone Replacement Therapy Use in 10,000 women: Benefits and Harms per Year Table 4: Results of Randomized Trials of the Effect of Hormone Replacement Therapy on Atherosclerosis Progression Measured by Coronary Angiography Health Risks After Stopping Estrogen and Progretin Figure 2: Risk of Death from all Causes and Global Risk by Randomized Assignment to Conjugated Equine Estrogen Plus Medroxyprogesterone Acetate or Placebo Before and After Termination of the Intervention in the Women’s Health Initiative Estrogen Plus Progestin Trial. Table: Effects of raloxifene on cardiovascular events and breast cancer in postmenopausal women. Estrogen Therapy and Coronary-Artery Calcification Conclusions: Among women 50 – 59 years old at enrollment the calcified-plaque burden in the coronary arteries after trial completion was lower in women assigned to estrogen than in those assigned to placebo. However, estrogen has complex biologic effects and may influence the risk of cardiovascular events and other outcomes through multiple pathways. (ClinicalTrials.gov number, NCT00000611.) Data related to previously mentioned study. HRT and the Young at Heart The translation of basis research to the bedside and to public guidelines a collaborative and interactive process conducted with patience and persistence. Just such an interactive process has enabled our emerging appreciation for the potential cardiovascular benefits of hormone-replacement therapy in younger women who have recently undergone menopause. Review Prevention of cardiovascular events in early menopause: A possible role for hormone replacement therapy.

Source: http://ippcr.nihtraining.com/handouts/2009/Harris1020508.pdf