Hi all,
My model is a mixed effects model that is testing the change of weight between different drug classes over time. I used restricted cubic splines for the time variable and the model includes an interaction term between drug class (five different categories) and time_spline -- it looks something like this:
xtmixed weight ib5.class##time_sp* var1 var2 var3
the outcome looks like this:
---------------------------------------------------------------------------------
Weight change| Coef. Std. Err. z P>|z| [95% Conf. Interval]
----------------+----------------------------------------------------------------
class |
1 | 0.62 1.50 0.414 0.679 -2.32 3.56
2 | 2.96 1.27 2.334 0.020 0.47 5.45
3 | -0.08 1.53 -0.054 0.957 -3.08 2.92
4 | 0.16 0.87 0.182 0.855 -1.55 1.86
time_sp1 | 1.50 0.20 7.501 0.000 1.11 1.90
time_sp2 | -1.76 1.51 -1.167 0.243 -4.73 1.20
time_sp3 | 1.28 2.38 0.539 0.590 -3.38 5.94
class#c.time_sp1 |
1 | -0.51 0.50 -1.035 0.301 -1.49 0.46
2 | -1.49 0.42 -3.543 0.000 -2.31 -0.67
3 | 0.30 0.56 0.531 0.596 -0.80 1.39
4 | -0.73 0.31 -2.333 0.020 -1.35 -0.12
class#c.time_sp2
1 | 5.75 3.80 1.515 0.130 -1.69 13.19
2 | 6.03 3.18 1.895 0.058 -0.21 12.26
3 | -2.23 4.17 -0.535 0.593 -10.41 5.95
4 | 3.83 2.39 1.604 0.109 -0.85 8.51
class#c.time_sp3
1 | -9.08 5.98 -1.518 0.129 -20.80 2.65
2 | -8.79 5.00 -1.756 0.079 -18.60 1.02
3 | 3.21 6.54 0.490 0.624 -9.61 16.02
4 | -5.58 3.76 -1.485 0.138 -12.94 1.78
I am trying to test whether the weight change over time is different between different drug classes. For example: Does the change in weight for patients receiving DRUG 1 any different from the change in weight for patients receiving DRUG 2 ?
Thank you for your assistance with this.
Kassem
My model is a mixed effects model that is testing the change of weight between different drug classes over time. I used restricted cubic splines for the time variable and the model includes an interaction term between drug class (five different categories) and time_spline -- it looks something like this:
xtmixed weight ib5.class##time_sp* var1 var2 var3
the outcome looks like this:
---------------------------------------------------------------------------------
Weight change| Coef. Std. Err. z P>|z| [95% Conf. Interval]
----------------+----------------------------------------------------------------
class |
1 | 0.62 1.50 0.414 0.679 -2.32 3.56
2 | 2.96 1.27 2.334 0.020 0.47 5.45
3 | -0.08 1.53 -0.054 0.957 -3.08 2.92
4 | 0.16 0.87 0.182 0.855 -1.55 1.86
time_sp1 | 1.50 0.20 7.501 0.000 1.11 1.90
time_sp2 | -1.76 1.51 -1.167 0.243 -4.73 1.20
time_sp3 | 1.28 2.38 0.539 0.590 -3.38 5.94
class#c.time_sp1 |
1 | -0.51 0.50 -1.035 0.301 -1.49 0.46
2 | -1.49 0.42 -3.543 0.000 -2.31 -0.67
3 | 0.30 0.56 0.531 0.596 -0.80 1.39
4 | -0.73 0.31 -2.333 0.020 -1.35 -0.12
class#c.time_sp2
1 | 5.75 3.80 1.515 0.130 -1.69 13.19
2 | 6.03 3.18 1.895 0.058 -0.21 12.26
3 | -2.23 4.17 -0.535 0.593 -10.41 5.95
4 | 3.83 2.39 1.604 0.109 -0.85 8.51
class#c.time_sp3
1 | -9.08 5.98 -1.518 0.129 -20.80 2.65
2 | -8.79 5.00 -1.756 0.079 -18.60 1.02
3 | 3.21 6.54 0.490 0.624 -9.61 16.02
4 | -5.58 3.76 -1.485 0.138 -12.94 1.78
I am trying to test whether the weight change over time is different between different drug classes. For example: Does the change in weight for patients receiving DRUG 1 any different from the change in weight for patients receiving DRUG 2 ?
Thank you for your assistance with this.
Kassem
Comment