Hi everyone!

I’m trying to estimate the difference in blood pressure between hypertensive and non-hypertensive individuals using a complex survey, but doing so by age-group and gender.

I’m having trouble with the matrix code for outputting the results of my svy command. Here is what I am writing currently:

svyset psu [pweight=samplewt_bp], strata(stratum)
svy , subpop(HTN1) : mean sbp_final, over(sex age5)
Survey: Mean estimation
Number of strata = 2 Number of obs = 5,091
Number of PSUs = 258 Population size = 2,439,648
Subpop. no. obs = 1,645
Subpop. size = 703,978.23
Design df = 256
Linearized
Mean Std. Err. [95% Conf. Interval]
c.sbp_final@sex#age5
1 0 136.3088 5.042258 126.3792 146.2384
1 30 139.3294 1.698257 135.9851 142.6738
1 35 147.6446 2.197222 143.3177 151.9716
1 40 144.5797 1.80772 141.0198 148.1395
1 45 154.1876 4.811669 144.7121 163.6631
1 50 157.141 4.226026 148.8188 165.4632
1 55 155.97 2.916671 150.2263 161.7137
1 60 154.9218 3.047789 148.9199 160.9238
1 65 158.1695 2.623057 153.004 163.335
2 0 136.8597 2.010842 132.8998 140.8196
2 30 138.7637 2.321715 134.1916 143.3358
2 35 145.7804 2.492297 140.8723 150.6884
2 40 147.0154 2.922702 141.2598 152.771
2 45 157.0043 3.423124 150.2632 163.7454
2 50 159.3642 5.456736 148.6184 170.11
2 55 154.3976 2.330003 149.8092 158.9861
2 60 158.9386 4.417497 150.2393 167.6378
2 65 158.8922 3.301328 152.3909 165.3934
Now, trying to extract these estimates (which I will later merge by country and hypertension status) is proving to be a challenge.
matrix M = e(b)

mat li M
M[1,18]
M[1,18]
c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@
1.sex# 1.sex# 1.sex# 1.sex# 1.sex# 1.sex# 1.sex# 1.sex#
0.age5 30.age5 35.age5 40.age5 45.age5 50.age5 55.age5 60.age5
y1 136.3088 139.32943 147.64465 144.57965 154.18764 157.14103 155.96998 154.92182
c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@ c.sbp_final@
1.sex# 2.sex# 2.sex# 2.sex# 2.sex# 2.sex# 2.sex# 2.sex#
65.age5 0.age5 30.age5 35.age5 40.age5 45.age5 50.age5 55.age5
y1 158.16949 136.85968 138.76374 145.78036 147.0154 157.00432 159.36418 154.39765
c.sbp_final@ c.sbp_final@
2.sex# 2.sex#
60.age5 65.age5
y1 158.93856 158.89217

I’ve tried it many different ways but to no avail. What I would really like is a matrix with the one variable containing the strata specification and one variable containing the means. If anyone could help me figure out how to code this I would greatly appreciate it!