I have a case-control study in which 21 patients with a certain clinical outcome and 20 patients without that clinical outcome were (retrospectively) selected from a larger group of patients who were followed up for the clinical outcome. Note that, to determine whether a patient had a clinical outcome or not, that patient, needed to be followed for many years. I am interested in testing whether a certain binary metric is predictive of the clinical outcome.
I was planning to use the log-rank test to compare Kaplan-Meier survival probability curves for patients grouped by the binary metric. But now I see in some literature that Kaplan-Meier survival analysis cannot be used in a case-control study. Does this mean that I cannot perform survival analysis for my study?
Can I still use Kaplan-Meier method to calculate only the survival probability curves (and use some other test like Fischer or Chi-squared to calculate p-values)?