Title: | Competing Risks Combined |
---|---|
Description: | Use two independently estimated risks of death from other causes and death from disease to get a corrected estimate of risk of death from disease. For example, given a patient's risk of death from cancer based on their oncologic burden and their risk of death from other causes based on their comorbidities, we calculate an adjusted risk of death from cancer. To illustrate the importance of this correction, image two patients with identical disease burden, and therefore, identical predicted risk of death from disease. Now imagine, that one of these patients also has sever heart disease and it likely we pass away from hear disease before they would succumb to the cancer. Our correction makes the appropriate update to the risk of death from disease incorporating the risk of death from other causes. |
Authors: | Daniel D. Sjoberg [aut, cre] |
Maintainer: | Daniel D. Sjoberg <[email protected]> |
License: | MIT + file LICENSE |
Version: | 0.0.0.9000 |
Built: | 2025-01-17 05:53:41 UTC |
Source: | https://github.com/ddsjoberg/crc |
Function takes a vector of probability estimates and a time frame,
e.g. probability of death within 10 years, and converts the probability
into an exponential distribution lambda parameter. The lambda parameter
is the rate parameter, e.g. f(x) = lambda * exp(-lambda * x)
prob_to_exp_lambda(p, time = 1)
prob_to_exp_lambda(p, time = 1)
p |
numeric vector of probability estimates |
time |
a single numeric time that is the time frame for the probability
in |
numeric lambda greater than zero
prob_to_exp_lambda(0.5)
prob_to_exp_lambda(0.5)
Provided a risk estimate of death from cancer and an estimate for death from other causes, this function will combine the two risks assuming an exponential distribution for both.
update_cancer_death_risk(risk_cancer_death, risk_other_cause, time)
update_cancer_death_risk(risk_cancer_death, risk_other_cause, time)
risk_cancer_death |
risk of death from cancer within time |
risk_other_cause |
risk of death from other causes within time |
time |
time frame associated with the risk predictions above. |
numeric probability between zero and one
update_cancer_death_risk(0.5, 0.5, time = 10)
update_cancer_death_risk(0.5, 0.5, time = 10)