Gravar-mail: Proportional hazards and threshold regression: their theoretical and practical connections