Correct use and interpretation for Surveyreg

jcsimmo — Sun, 16 Dec 2018 02:11:03 GMT

Good evening everyone,

I am still relatively new to SAS and I wanted to make sure I am correctly implementing and interpreting my code.

I am looking at the age (in months) that patients (<=2yo) with cleft palates undergo cleft palate repair and seeing how race, gender, location, income level, and hospital type affects the timing of that repair.

This is my code to determine the estimate age of repair by each of those variables (while controlling for one-another).

proc surveyreg data=kid.KID_complete;
class race female hosp_region zipinc hosp_locteach;
weight discwt;
where age<=2 and cleftpalateonly=1 and cleftpalaterepair=1;
model agemonth = race female hosp_region zipinc hosp_locteach/ solution;
domain race female hosp_region zipinc hosp_locteach;
Format race racef.;
Format female femalef.;
Format hosp_region regionf.;
Format zipinc zipincf.;
Format hosp_locteach hosp_locteachf.;
ods output
DataSummary=output
DomainSummary=domain
ParameterEstimates = MyParmEst;
run;

In the results, for each domain analysis, I get something like this (this is just for the domain race = white children)

The SURVEYREG Procedure

I:Race=White

Domain Regression Analysis for Variable AGEMONTH

Domain Summary
Number of Observations	2459
Number of Observations in Domain	1480
Number of Observations Not in Domain	979
Sum of Weights in Domain	2639.9
Weighted Mean of AGEMONTH	12.56934
Weighted Sum of AGEMONTH	33182.4

Fit Statistics
R-Square	0.0295
Root MSE	5.0756
Denominator DF	2458

Tests of Model Effects
Effect	Num DF	F Value	Pr > F
Model	9	3.24	0.0006
Intercept	1	787.25	<.0001
RACE	0	.	.
female	1	0.99	0.3207
HOSP_REGION	3	4.73	0.0027
zipinc	3	3.29	0.0197
HOSP_LOCTEACH	2	0.19	0.8286

Note: The denominator degrees of freedom for the F tests is 2458.


Estimated Regression Coefficients
Parameter	Estimate	Standard Error	t Value	Pr > \|t\|
Intercept	11.718497	0.369853	31.68	<.0001
RACE Asian	0	0	.	.
RACE Black	0	0	.	.
RACE Hispanic	0	0	.	.
RACE Native American	0	0	.	.
RACE Other	0	0	.	.
RACE White	0	0	.	.
female Female	0.3051803	0.30724	0.99	0.3207
female Male	0	0	.	.
HOSP_REGION Midwest	1.1362171	0.460781	2.47	0.0137
HOSP_REGION Northeast	-0.2683534	0.365375	-0.73	0.4627
HOSP_REGION South	-0.5941962	0.42009	-1.41	0.1574
HOSP_REGION West	0	0	.	.
zipinc 25-50th	1.2050844	0.404636	2.98	0.0029
zipinc 51-75th	0.386581	0.373655	1.03	0.301
zipinc <25th	0.7636029	0.409625	1.86	0.0624
zipinc >75th	0	0	.	.
HOSP_LOCTEACH Rural	-0.2140308	1.106108	-0.19	0.8466
HOSP_LOCTEACH Urban non-teaching	0.4966877	0.852428	0.58	0.5602
HOSP_LOCTEACH Urban teaching	0	0	.	.

Would I correct in interpreting this as:

In white children, when controlling for age, gender, location, income, and hospital type, the average age of repair is 11.7months (p<0.0001)?

Re: Correct use and interpretation for Surveyreg

PGStats — Fri, 14 Dec 2018 05:05:49 GMT

Those results are unreadable. Could you please edit your message above and paste the results using the {i} icon in the editor.

Re: Correct use and interpretation for Surveyreg

jcsimmo — Fri, 21 Dec 2018 05:28:01 GMT

Fixed

topic Re: Correct use and interpretation for Surveyreg in SAS Programming

Correct use and interpretation for Surveyreg

Re: Correct use and interpretation for Surveyreg

Re: Correct use and interpretation for Surveyreg