Kaplan-Meier Survival Analysis: Comparing Treatment Efficacy in Migraine Management
Overview
This paper presents a Kaplan-Meier survival analysis comparing the efficacy of a new migraine treatment versus regular pain relief across 20 subjects. The analysis employs non-parametric methods to estimate survival distributions and uses log-rank testing to determine statistical significance between treatment groups, with results indicating no significant difference in survival outcomes.
What this coursework covers
- Introduction to Kaplan-Meier product-limit method and non-parametric survival estimation
- Case processing summary and survival table data for both treatment groups with cumulative proportions and standard errors
- Means and medians for survival time with 95% confidence intervals and overall comparisons using log-rank, Breslow, and Tarone-Ware tests
- Discussion of learning outcomes and challenges encountered in the course
Preview
How to Perform a Survival Analysis
Helena Lynch
Capella University
MPH5509: Principles of Biostatistics
Dr. Haley Cash
16th December, 2020
How to Perform a Survival Analysis
Kaplan-Meir is also known as the product limit method. This is a non-parametric method that is used in the estimation of survival past given certain time points. The survival distributions of two or more groups of a between-subjects factor is used to compare for equality. In this analysis, we have undertaken a Kaplan-Meir analysis using the data provided.
From the case processing summary, we can see that Total N was 20, and N of events is 15.
Case Processing Summary | ||||
Treatment | Total N | N of Events | Censored | |
N | Percent | |||
New Migraine Treatment | 10 | 8 | 2 | 20.0% |
Regular pain relief | 10 | 7 | 3 | 30.0% |
Overall | 20 | 15 | 5 | 25.0% |
Survival Table | |||||||
Treatment | Time | Status | Cumulative Proportion Surviving at the Time | N of Cumulative Events | N of Remaining Cases | ||
Estimate | Std. Error | ||||||
New Migraine Treatment | 1 | 3.000 | 1.00 | .900 | .095 | 1 | 9 |
2 | 5.000 | 1.00 | .800 | .126 | 2 | 8 | |
3 | 6.000 | 1.00 | .700 | .145 | 3 | 7 | |
4 | 7.000 | 1.00 | .600 | .155 | 4 | 6 | |
5 | 11.000 | 1.00 | .500 | .158 | 5 | 5 | |
6 | 12.000 | 1.00 | .400 | .155 | 6 | 4 | |
7 | 14.000 | 1.00 | .300 | .145 | 7 | 3 | |
8 | 17.000 | .00 | . | . | 7 | 2 | |
9 | 19.000 | .00 | . | . | 7 | 1 | |
10 | 22.000 | 1.00 | .000 | .000 | 8 | 0 | |
Regular pain relief | 1 | 7.000 | 1.00 | .900 | .095 | 1 | 9 |
2 | 11.000 | 1.00 | .800 | .126 | 2 | 8 | |
3 | 12.000 | 1.00 | .700 | .145 | 3 | 7 | |
4 | 14.000 | .00 | . | . | 3 | 6 | |
5 | 18.000 | 1.00 | .583 | .161 | 4 | 5 | |
6 | 19.000 | .00 | . | . | 4 | 4 | |
7 | 21.000 | 1.00 | .438 | .175 | 5 | 3 | |
8 | 22.000 | 1.00 | .292 | .167 | 6 | 2 | |
9 | 25.000 | 1.00 | .146 | .133 | 7 | 1 | |
10 | 26.000 | .00 | . | . | 7 | 0 | |
Means and Medians for Survival Time | ||||||||
Treatment | Meana | Median | ||||||
Estimate | Std. Error | 95% Confidence Interval | Estimate | Std. Error | 95% Confidence Interval | |||
Lower Bound | Upper Bound | Lower Bound | Upper Bound | |||||
New Migraine Treatment | 12.400 | 2.377 | 7.741 | 17.059 | 11.000 | 3.953 | 3.252 | 18.748 |
Regular pain relief | 18.808 | 2.042 | 14.806 | 22.811 | 21.000 | 3.595 | 13.955 | 28.045 |
Overall | 15.675 | 1.703 | 12.338 | 19.012 | 14.000 | 5.963 | 2.313 | 25.687 |
a. Estimation is limited to the largest survival time if it is censored. | ||||||||
Overall Comparisons | |||
Chi-Square | df | Sig. | |
Log Rank (Mantel-Cox) | 2.842 | 1 | .092 |
Breslow (Generalized Wilcoxon) | 3.251 | 1 | .071 |
Tarone-Ware | 3.118 | 1 | .077 |
Test of equality of survival distributions for the different levels of Treatment. | |||
The log rank test is the testing null hypothesis that there is no difference in the overall survival distributions between the groups.