Two outcome events (death and death or hospitalization) on ONE Kaplan Meier Curve ?

[jafary]

Hi Folks

As always, this newsgroup is a godsend for someone like me.

I'm analysing a dataset in which I'm looking at two outcome events -
death
death or hospitalization for recurrent heart failure
(combined outcome)

There are no comparator groups - just one cohort and it's survival
experience.

My question is : can I somehow place the survival curves of both
outcomes on ONE graph instead of two seperate graphs? And if so, can
someone kindly guide me as to how to do it in SPSS ?

Many thanks

Fahim H. Jafary, MD
Aga Khan University Hospital
Karachi, Pakistan.

[Haris]

I don't think there are simple ways of doing this. Most likely you'll
have to run the analysis two times: first treating hospitalizations as
censored, next deaths as censored. After that you can simply
superimpose the two graphs on each other as long as you keep your X and
Y axes the same for both analyses. I export EMF files and edit them in
Power Point.

You can also create two datasets with survival times for deaths and
hospitalizations in each, merge them together and plot off of that
merged dateset. It should not be very hard.

[jafary]

Haris

That was a great suggestion - did as you suggested, got two EMF files.
Then (after ungrouping the death survival curve) copied the death curve
onto the death+hospitalization one. I did all this in Powerpoint as
you suggested. Then I saved the whole slide as a .png file.

However, I do note one thing - after creating the "new" curve, the
image quality is not as crisp as the original images. It's not bad,
don't get me wrong, but I'm wondering if the image will be good enough
for print. Any ideas.

Once again, thanks for taking the time out

Fahim Jafary

[Haris]

I am not familiar with PNG format, but there should be no quality loss
in PowerPoint. I believe, once you ungroup your EMF files, they use
vector graphics for editing. If you save the product in a loseless
format such as TIFF or low compression JPEG your quality will be as
good as it gets. I use JPEG and never had a problem with image
quality. Journals frequently accept PowerPoint files without any
conversion and we had several articles published in Circulation and
JACC without any issues.

[jafary]

Haris

Unfortunately there clearly is a loss in quality once I convert the
Powerpoint file to either PNG or JPEG (or TIFF). I'll be happy to send
you pre and post images. You're correct that journals do accept
Powerpoint files but as a generic rule, I'm a bit concerned about
losing image quality. I am sure there's a way around it but haven't
been able to nail it down. I've been told that a "vector graphics
editor" should enable one to preserve quality. Now to search for a
"vector graphics editor" !

Thanks for your help. Appreciate any further thoughts

[jafary]

I figured it out (I think).

After editing the image in Powerpoint I saved the image as an .emf
file. Then I opened it in Microsoft Digital Image Starter Edition (came
preinstalled on my laptop) and saved it as a PNG or JPEG image.
Excellent quality with no degredation. I suspect (someone may correct
me if I'm wrong) the engine in Powerpoint that converts the image to
PNG/JPG/TIFF is not as good as it is in dedicated graphics packages - I
may be wrong of course. As not all journals accept Powerpoint files
for images, I this is helpful.

Once again, thanks for the fab. tip

Fahim Jafary

[JKPeck]

KM lets you save the survival and cum values as variables. If you can
reconstruct the survival plot from these, then it is easy to draw the
combined plot using Ggraph as a multiple line plot by including both
series. You would need to transform the saved series to get exactly
the form that KM produces, but it should be possible.

-Jon Peck