Skip to main content
SAGE
Search form
  • 00:00

    SPEAKER: Let's talk about life tables and withdrawals.In the previous video, we talked about a hypothetical studywhere participants could only leave through death.But that's not a very realistic scenario,because in reality, study participants come and leavethroughout the duration of a study.The reasons for leaving the studycan be varying from no longer being

  • 00:22

    SPEAKER [continued]: interested to moving to a different city and so forth.So we also have to learn how these withdrawals arehandled in a life table.Understanding this concept is super important alsofor later when we're going to talk about Kaplan Meier curves.So let's take another hypothetical study population.The study's duration is four years.

  • 00:44

    SPEAKER [continued]: The second column depicts the numberof participants alive and presentat the beginning of each year.The third column indicates the numberof participants who died during that year,and the fourth column indicates how many participantswithdrew during that year.So 500 people joined at the outset,223 died during the first year, and 24 withdrew.

  • 01:07

    SPEAKER [continued]: So we're left with 253 at the beginning of year twosince 500 minus 223 minus 24 equals 253.Makes sense so far?Now, in order to calculate the proportion of participantswho died in each year, we need to know the number of peoplewho died in that year and divide that by the number of people

  • 01:28

    SPEAKER [continued]: at risk in that year.We know the number who died, but how many people were at risk,let's say, during year one?We know that 500 joined at the beginning of that year,but we also know that 24 left throughout the year.So what are we to do?Well, in this situation, we make an assumption.We assume that all people left did so uniformly

  • 01:51

    SPEAKER [continued]: throughout the year.And if they left uniformly, 12 leftin the first half of the year and 12 left in the second half.So on average, half of these 24 were at riskthroughout the year.So the number at risk is the numberalive at the beginning of that year minus halfof the withdrawals.So for the first year, that will be 500 minus 24 divided by 2,

  • 02:16

    SPEAKER [continued]: or 500 minus 12, and that equals 488.Please stop the video and make the calculations for year twoto four yourself, and then come back.So here are the correct numbers.Did you get them right?Now let's calculate the proportions of peoplewho died during each year.

  • 02:36

    SPEAKER [continued]: We set the proportion of people who diedis calculated by dividing the number of peoplewho died by the number of people at risk in that year.So for year one, that would be 223 divided by 488,and that's 0.457, or 46%.Please pause the video again and make the calculationsfor years 2, 3, and 4 yourself and then come back.

  • 03:02

    SPEAKER [continued]: So these are the correct numbers.Now let's take the next step.The next step in the process is to calculatethe proportion of participants who survived in each year.And that, as you can imagine, is calculatedas 1 minus the proportion of participants who died.So for year one, this would be 1 minus all 0.457,which equals 0.543, or 53%.

  • 03:26

    SPEAKER [continued]: Again, please stop the video and calculate the proportionsfor years 2, 3, and 4 yourself, then come back.And here are the correct numbers.I hope you got them right.Now comes the final step in the process,and that's to calculate the cumulative survivalfrom the start of the study to the end of each year follow-up.

  • 03:46

    SPEAKER [continued]: You've already learned how to do thatin one of our previous videos.It's by multiplying the proportionsof the individual years.So surviving from the start of the studyto the end of year one is the same as surviving year one.So that's 0.543.Surviving from the start of the study until the end of year twois 0.543 times 0.546, and that's 0.297.

  • 04:10

    SPEAKER [continued]: So around 30% of the study's participantssurvived the first two years.Now, again, please stop the videoand calculate the cumulative proportions for each yearyourself, then come back.So these are the numbers.You see that five year survival in this study populationis 16.4%.These cumulative percentages could be nicely depicted

  • 04:32

    SPEAKER [continued]: in a survival curve if need be.[MUSIC PLAYING]

Video Info

Series Name: Expressing Disease Prognosis

Episode: 3

Publisher: Medmastery GmbH

Publication Year: 2017

Video Type:Tutorial

Methods: Clinical research, Participatory research

Keywords: clinical research; kaplan-meier estimate; life tables; participatory research; prognosis; Survival rate ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:

Keywords:

Abstract

How to adjust for withdrawals when using life tables as a technique to assess survival and express disease prognosis is explained, using a clear example.

Looks like you do not have access to this content.

Adjusting for Withdrawals in Life Tables

How to adjust for withdrawals when using life tables as a technique to assess survival and express disease prognosis is explained, using a clear example.

Copy and paste the following HTML into your website