The hazard ratio is an expression of the hazard or chance of events occurring in the

treatment arm as a ratio of the hazard of the events occurring in the control arm.

For more clarification and examples please visit this site:

http://www.medcalc.org/manual/kaplan-meier.php

Hello Karthik, confidence interval can be calculated from various sample statistics including hazard ratio or incidence rate ratio.

95% confidence interval assuming that your sample size is above 60 .. can be calculated by

incidence rate ratio(IRR) plus or minus 2 times Standard error of IRR

this has to be calculated using the natural log of the values.

Hope this is helpful..

Thank you so much for taking the time to respond. My problem is, all I have is Kaplan-Meier curve for two treatments. See the attached image that shows the curve. I need to "estimate" hazard ratio from this chart and the corresponding confidence interval. The x-axis values are (0,3,6,9,12,15) in years. Can we come up with a rough estimate of hazard ration from this curve. Thanks again for your time in advance.

I got the chart from the article (link provided). I need to find "hazard ratio for microvascular complications" from the article.

http://columbiamedicine.org/education/r/Endocrine/Outpatient%20and%20Prevention/UKPDS%2033.pdf

See the attached image.

It's never a good idea to say never but it seems to me that calculating confidence intervals from nothing more than Kaplin-Meier curves would be incredibly tedious. Without regard to other issues, the standard errors you need for each time interval are a function of both the number of incidents that have occurred prior to the interval and the number of incidents that occur during the interval. Have you considered contacting this UKPDS group... or what remains of it... to request access to the original data?

Than you so much for providing more details. I found a way to do rough estimate of hazard ratio from Kaplan-Meier curve. You can find information on the website (http://www.drtimdelivers.com/statswhi052504/statsofwhi052504.shtml). Beyond this I could not find any other information that might help me to calculate confidence interval from the curve. I will consider contacting UKPDS group. Thank you all for your response. I sincerely appreciate it.

The following article shows how to calculate the hazard ratio and its interval from a K-M curve, it includes a spreadsheet to guide the process

( Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods for incorporating summary time-to-event data into meta-analysis. Trials. 2007;8:16. Published 2007 Jun 7. doi:10.1186/1745-6215-8-16 )

Article “Practical Methods for Incorporating Summary Time-to-Event D...