Comparing two KM estimates outputs without raw data

[Mustafa Acar]

Some raw data has gone missing, but I have the outputs of two KM curves - Both their survival estimates and people at risk.

Is there any way for me to use just the estimates to test the difference in the curves and generate a hazard ratio + CI's?

[Carlo Lazzaro]

Mustafa:
welcome to this forum.
The usual approach to estimate the difference in survival estimates after KM is -logrank-.
However, the recently passed away and greatly missed Doug Altman in his https://www.crcpress.com/Practical-S.../9780412276309 proposes a method to calculate HR (and estimated CI) using -logrank- outcome (see pages 375-376).