C:\_TEACHING\EPI-820\HMW5_Start_Here.doc

Department of EpidemiologyPage 110/21/2018

EPI-820 Evidence-Based Medicine

Homework # 5: META-ANALYSIS

Dr. Mat Reeves BVSc, PhD

Dr. Mark Ebell MD, MS

In this homework you will use an Excel spreadsheet to illustrate some general principles of assessing the quality of randomized clinical trials and in the use of Mantel-Haenszel (fixed effect) analysis methods. Download the file (HWK5_metaanal_lab.xls) from the EPI-820 web site and on a separate page complete the answers for the two exercises (again, for the most part, I am looking for just a few short words on each question).

This homework requires you to read and evaluate the following three RCT’s, which all deal with the use of thrombolytic treatment for acute ischemic stroke (these articles will be handed out in class):

The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue Plasminogen Activator for Acute Ischaemic Stroke. The New England Journal of Medicine 1995;333:1581-1587.

European Cooperative Acute Stroke Study (ECASS). Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke. The Journal of the American Medical Association 1995;274:1017-1025.

Mori E, Yoneda Y, Tabuchi M, et al. Intravenous recombinant tissue plasminogen activator in acute carotid artery territory stroke. Neurology 1992;42:976-982.

Also, I have included (as a reference) the following article, which describes the assessment tool for RCTs:

Chalmers TC, Smith H, Blackburn B, at al. A method for assessing the quality of a randomized control trial. Controlled Clinical Trials 1981; 2: 31-49.

Exercise 1 – Quality Assessment

Directions:

Open Excel for Windows (you should be able to use any version 5.0 or higher)

Open Excel file ‘HWK5_metaanal_.xls’ file.

Click on first worksheet or page labeled ‘Quality assessment (blank)'

This worksheet shows 30 items used to evaluate the quality of RCT’s. Each item had a short description and a column (labeled “Max”) showing the total number of points available (for further details refer to the Chalmers article).

Each of the three studies in question have been evaluated and their scores evaluated as shown. However, for the ECASS study and the NINDS study, the scores for item numbers 2, 8, 11, 14, 18, 27 and 30 have been left blank.

1. Read and evaluate the ECASS and NINDS studies and complete the missing item scores. For each score you give, provide a brief (meaning 1 sentence!) justification [14 points].

2. Based on this scoring system, what do you think are the three most important criteria in determining RCT quality? Does this make sense? [3 points].

3. What are the total quality scores for each trial? What criteria end up having the greatest influence on determining the variability in the overall quality scores of these three trials? [4 points].

4. How could these quality scores be used in a meta-analysis of thrombolysis and acute stroke? [4 points].

Exercise 2 – Mantel-Haenszel (MH) Chi-square

Click on second worksheet or page labeled ‘M-H chi-square (blank)'

This worksheet shows the calculations for the MH fixed effects method, including the M-H homogeneity test (Q), and the summary MH estimate of effect with 95% confidence intervals.

The spreadsheet shows the 2x2 table results for each of the three studies. Right now, the tables include “dummy” values in each of the 4 cells (denoted in blue) to illustrate that the spreadsheet works.

1. After examining each article, complete each 2x2 table using values selected from each publication. Where more than one set of values could be selected, justify your choice [10 points].

2. Interpret the Q statistic [2 points].

3. Interpret the summary MH estimate and confidence intervals. Is a summary estimate justified given the results of Q? [3 points].