Survival Analysis
This chapter reviews some of the fundamental concepts and basic methods in survival analysis. Frequently, event rates such as mortality or occurrence of nonfatal myocardial infarction are selected as primary response variables. The analysis of such event
- PDF / 375,939 Bytes
- 23 Pages / 439.37 x 666.142 pts Page_size
- 33 Downloads / 192 Views
Survival Analysis
This chapter reviews some of the fundamental concepts and basic methods in survival analysis. Frequently, event rates such as mortality or occurrence of nonfatal myocardial infarction are selected as primary response variables. The analysis of such event rates in two groups could employ the chi-square statistic or the equivalent normal statistic for the comparison of two proportions. However, when the length of observation is different for each participant, estimating an event rate is more complicated. Furthermore, simple comparison of event rates between two groups is not necessarily the most informative type of analysis. For example, the 5-year survival for two groups may be nearly identical, but the survival rates may be quite different at various times during the 5 years. This is illustrated by the survival curves in Fig. 15.1. This figure shows survival probability on the vertical axis and time on the horizontal axis. For Group A, the survival rate (or one minus the mortality rate) declines steadily over the 5 years of observation. For Group B, however, the decline in the survival rate is rapid during the first year and then levels off. Obviously, the survival experience of the two groups is not the same, although the mortality rate at 5 years is nearly the same. If only the 5-year survival rate is considered, Group A and Group B appear equivalent. Curves such as these might reasonably be expected in a trial of surgical versus medical intervention, where surgery might carry a high initial operative mortality.
Fundamental Point Survival analysis methods are important in trials where participants are entered over a period of time and have various lengths of follow-up. These methods permit the comparison of the entire survival experience during the follow-up and may be used for the analysis of time to any dichotomous response variable such as a nonfatal event or an adverse event. © Springer International Publishing Switzerland 2015 L.M. Friedman et al., Fundamentals of Clinical Trials, DOI 10.1007/978-3-319-18539-2_15
319
320
15
Survival Analysis
1 Group A
Group B
Survival Probability
0.75
0.5
0.25
0 0
1
2 3 Follow-up Time (Years)
4
5
Fig. 15.1 Survival experience for two groups (A and B)
A review of the basic techniques of survival analysis can be found in elementary statistical textbooks [1–6] as well as in overview papers [7]. A more complete and technical review is in other texts [8–11]. Many methodological advances in the field have occurred and this book will not be able to cover all developments. The following discussion will concern two basic aspects: first, estimation of the survival experience or survival curve for a group of participants in a clinical trial and second, comparison of two survival curves to test whether the survival experience is significantly different. Although the term survival analysis is used, the methods can be applied to any dichotomous response variable when the time from enrollment to the time of the event, not just the fact of its occurrenc
Data Loading...
