Simulated repeat measurement and time-to-event data

A simulated dataset which is a combination of longitudinal (repeat measurement) data and time-to-event data. This dataset contains simulated data from 3000 patients.

The longitudinal (repeat measurement) data is formed using an LME model, whose parameters were based on CVD risk assessments recorded at primary care practices in New Zealand.

A LME model was fitted to this dataset and values of sbp_stnd and tchdl_stnd were estimated at landmark age 60. These values (along with the other baseline covariates) were used to simulate time-to-event data from a cause specific model with parameters based on CVD events of patients at primary care practices in New Zealand.

data_repeat_outcomes

Format

A dataset with 9031 rows and 12 columns:

id

Patient ID

smoking

Smoking status, 0 indicates the patient has never smoked, 1 indicates the patient has quit smoking, and 2 indicates the patient is a current smoker

diabetes

Diabetes status, 0 indicates the patient is not diagnosed with diabetes, and 1 indicates the patient is diagnosed with diabetes

ethnicity

Ethnicity, one of five ethnicities

dob

Date of birth

index

An index indicating assessment number for a patient

sbp_stnd

Standardised systolic blood pressure

tchdl_stnd

Standardised total cholesterol to HDL ratio

response_time_tchdl_stnd

Age that total cholesterol to HDL ratio was recorded

response_time_sbp_stnd

Age that systolic blood pressure was recorded, this is the same as the date that the fixed measures were recorded

event_status

Event status, 0 indicates censoring, 1 indicates CVD event, and 2 indicates death from other causes

event_time

Event time