How do I perform a Cox proportional hazard analysis?

[#1]

In Sweden I seem to be the only person who uses the excellent statistical software JMP. I have used it for several years and now have version 14.2.0. The majority of the research colleagues that I work with use SPSS but I think JMP is much better. However, since I am a surgeon my statistical knowledge is somewhat limited. I know JMP fairly well but I mostly do univariate and multivariate analyses of nominal data. However, now I got a question from a journal regarding the Cox proportional hazard analysis of the data. Since I do not know how to do that in JMP I would appreciate if the community could help me. The dependent (y) variabel is continuous and is time to event (in bold, see a small sample below). To the left (in italic and book antiqua) are the x variables as nominal values and to the right of the y variabel are the x variables converted to numerical (1 or 0) values where 1 is a "positive event" and 0 is the reference.

I wonder how I do in JMP to get the results of these sample data (these are just examples and not the real data) and I would like to be able to present it like below. I would prefer an instruction of how to do it in an interactive mode. However, if possible also in a scripting answer. My knowledge of how to use JMP regarding programming is somewhat better than my statistical knowledge.

	Univariable analysis	Multivariable analysis
	Hazard ratio (95% confidence interval)	p	Hazard ratio (95% confidence interval)	p
Hilar stricture (Reference non-hilar)	0.00 (0.00-0.00)		0.00 (0.00-0.00)
Male (Reference Female)	0.00 (0.00-0.00)		0.00 (0.00-0.00)
Age ≥ 75 years (Reference <75 years)	0.00 (0.00-0.00)		0.00 (0.00-0.00)
ASA III-IV (reference ASA I-II)	0.00 (0.00-0.00)		0.00 (0.00-0.00)

Below is the example file:

Hilar stenosis	Gender	Age ≥ 75 years	ASA III-IV	Time to event	Hilar structure (GS) Yes=1 No=0	Men (GS) Male=1 Female=0	Age ≥ 75 years =1 else 0	ASA III-IV=1 else 0
No	Male	No	No	115	0	1	0	0
No	Female	Yes	No	157	0	0	1	0
No	Female	No	Yes	101	0	0	0	1
No	Female	Yes	Yes	138	0	0	1	1
No	Female	Yes	No	114	0	0	1	0
No	Female	No	Yes	177	0	0	0	1
No	Female	No	No	133	0	0	0	0
No	Female	Yes	No	39	0	0	1	0
Yes	Male	Yes	Yes	29	1	1	1	1
Yes	Female	Yes	Yes	7	1	0	1	1
No	Female	Yes	No	156	0	0	1	0
Yes	Female	Yes	Yes	169	1	0	1	1
No	Female	No	No	894	0	0	0	0
Yes	Female	Yes	No	303	1	0	1	0
Yes	Male	No	Yes	79	1	1	0	1
No	Female	Yes	Yes	32	0	0	1	1
No	Female	No	No	211	0	0	0	0
Yes	Male	No	Yes	248	1	1	0	1
No	Female	Yes	Yes	252	0	0	1	1
No	Male	No	Yes	456	0	1	0	1

Of JSL code I have no knowledge.

Sincerely yours

Lars Enochsson

Professor of Surgery

Umeå University 

Sweden

[#2 · Accepted solution]

Cox Proportional Hazards Model is available under the Fit Model Platform.  Here is a link to the JMP Help on it.