Kaplan-Meier Survival Curve

Description on how the survival plot data is calculated

Background

Survival Curve Wikipedia

Kaplan-Meier Wikipedia

How to build a Kaplan-Meier tile to a dashboard

To generate a survival plot, the user must select:

  • 1 numerical field (which represents time)

  • 1 categorical field (which will be transformed into the individual's status)

    • The categorical field should use one of the following 4 terms (case-insensitive) to indicate a status of Living: “living”, “alive”, “diseasefree”, “disease-free”

Note that these fields should have a 1 to 1 relationship with one another for accurate display.

To generate the plot survival at each censor or trigger event is calculated and plotted.

Survival Percentage Calculation

To calculate survival percent at the current event the system evaluates the following formula:

ST: Survival at the current event

LT0: Number of subjects living at the start of the period or event

D: Number of subjects that died

For each time period the following values are generated:

Status

  • Each individual is considered Dead unless they qualify as Living

Number of Subjects Living at the Start (LT0)

  • For the initial value this is the total number of records returned by the backend from survival data with Living or Dead Status.

  • For followup events this is the number of subjects at the start of the previous event minus the number of subjects that died in the previous event and the subjects that dropped out or were censored in the previous event

Number of Subjects that Died (D)

  • 1 for each individual who at the event does not have a status of Living

Number of Subjects Dropped or Censored

  • 1 for each individual who at the event has a status of Living

Survival Percent at the Current Event (ST)

Cumulative Survival (S)

  • ST-1: Survival percent at the previous event

NOTE: this is the actual point drawn on the survival plot