Survival analysis with time dependent covariate

[Frank Furter]

Dear Colleagues:

I have a set of clinical routine data from a cohort of critically ill
patients where some laboratory measures were determined every time the
patients saw the doctor. There is, however, no fixed visit schedule, and
thus different patients have laboratory measures taken at different points
in time.

I would like to investigate the association between patient survival and
(changes in) the course of these laboratory measures over time. Is that
something that can be done using Cox regression? Or is there a more
appropriate procedure for this?

Best, Andreas



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Rich Ulrich]

I'm going to speak to the structuring of the problem, not procedures.

One thing that I found useful when looking for "final indicators"

was to index events in reverse, from the end. I had "relapse" and

I had months rather than specific dates, but the same hypotheses

seem to be in play. (We found that patients stopped taking their meds

shortly before psychotic relapse, which was no surprise.  We found that

stopping meds was as common for Placebo as for Active Med; that /was/

a surprise.  It seems, the patient quits taking when they start feeling worse.)

When you write up intriguing conclusions, you will fall back on time-spans,

like "last day" and "last three days" and "last week" and "last month."

I would try aggregating within those intervals -- if that does not show

patterns, I don't know what will. 

--

Rich Ulrich

---

From: SPSSX(r) Discussion <[hidden email]> on behalf of AndreasV <[hidden email]>
Sent: Thursday, May 28, 2020 9:26 AM
To: [hidden email] <[hidden email]>
Subject: Survival analysis with time dependent covariate

 

Dear Colleagues:

I have a set of clinical routine data from a cohort of critically ill
patients where some laboratory measures were determined every time the
patients saw the doctor. There is, however, no fixed visit schedule, and
thus different patients have laboratory measures taken at different points
in time.

I would like to investigate the association between patient survival and
(changes in) the course of these laboratory measures over time. Is that
something that can be done using Cox regression? Or is there a more
appropriate procedure for this?

Best, Andreas



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Frank Furter]

I think my research problem is different from yours. My data are from a
clinical trial  with a long-term follow-up. The trial investighated the
treatment of a chronic lung disease from which patients will die sooner or
later unless they receive a lung transplant. The treatment objective is to
slow down the deterioration of the lung function (however, damage is
irreversible). After the end of the study, patients were followed up
whenever they came to see the doctors routinely, and fom these visits we
have got the results of the lung function assessments (the spirometry
parameters). Moreover, we also know whether and when patients died during
the follow-up period.

I would like to investigate the association between lung function measures
(these are continuous outcomes changing over time) and time to death.

Best, Andreas



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Sadhana Kannan]

Hi,

You may try Joint modelling.

Sadhana

 

----- Original message -----
From: "AndreasV" <[hidden email]>
Sent by: "SPSSX(r) Discussion" <[hidden email]>
To: [hidden email]
Cc:
Subject: Re: Survival analysis with time dependent covariate
Date: Tue, Jun 2, 2020 12:17 PM
 

I think my research problem is different from yours. My data are from a
clinical trial  with a long-term follow-up. The trial investighated the
treatment of a chronic lung disease from which patients will die sooner or
later unless they receive a lung transplant. The treatment objective is to
slow down the deterioration of the lung function (however, damage is
irreversible). After the end of the study, patients were followed up
whenever they came to see the doctors routinely, and fom these visits we
have got the results of the lung function assessments (the spirometry
parameters). Moreover, we also know whether and when patients died during
the follow-up period.

I would like to investigate the association between lung function measures
(these are continuous outcomes changing over time) and time to death.

Best, Andreas



--
Sent from: http://secure-web.cisco.com/11_ZqQ0dhvGzu1T-qaYHv_P1Ek_TIVT62yNH3-PHBOyk4ltUjCJl3hhBPFISkW7X3MKxqFfw_ssPsLy-_7AYAu5puUTwvvF8nX611vO179qOWO8fCIm0opC25j40Inm--5K8QcpV2Im6xRXXbs207qJUZdt6liaHXNCn1AYys1XyhBTglk6zHbx0KzDNeJXAvhQRryxy353lqVlE4FC9hMdIVqR2SBoNBNwdP3ol31p1sA6EqoQEhefNFX4vFinUKKBO0nfbRm-lLv8NHCe7j-m4MU-FYFGi3YHEXsPOpQMZm-OC65F70mMAC0EXEGbc7O6T6E-Jf6SkxJ40ESkuvwEbza67b91jof-Jwn03YgD1RVDXEbl4fmpczQ7keLdcBPKYNTK2PF_wFTZUNR6MrIm-qLjpxklJ-ZiVivwduZUflJMx1PgojdPBCI_2j9VbB4MRMuWOA2igfSY86S2HuVw/http%3A%2F%2Fspssx-discussion.1045642.n5.nabble.com%2F

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
 

 

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Frank Furter]

Sadhana Kannan wrote
> You may try Joint modelling.

Thank you very much - I think that this is exactly what I need.

Is anyone aware of a macro or program for performing Joint Modelling in
SPSS?

Best, Andreas




--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD