SAS Output

Example of running a series of univariate models and multivariate models with the logistic macro

1
Model	Variable	Value	# Outcome = 1/ # subgroup (% Outcome = 1)	Odds Ratio	Lower 95% CI	Upper 95% CI	P-Value	Quadratic P-Value	Type 3 P-Value
1	age			1.10	1.08	1.12	<.0001	0.0001
	age (per 10 units)			2.5	2.2	3.0
	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	477.125
	H-L Test	10 groups					0.0028

2	age	20-35	18/129 (14.0%)	1.00
	age	36-51	22/125 (17.6%)	1.32	0.67	2.6	0.43		<.0001
	age	52-66	80/128 (62.5%)	10.3	5.6	19.0	<.0001
	age	67-80	105/118 (89.0%)	50	23	107	<.0001

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	471.802
	H-L Test	4 groups					1.00
Pink cells indicate effects with p<0.05 Red cells indicate evidence for violations of model assumptions.

Example of running a series of univariate models and multivariate models with the logistic macro

2
Model	Variable	Value	# Outcome = 1/ # subgroup (% Outcome = 1)	Odds Ratio	Lower 95% CI	Upper 95% CI	P-Value	Quadratic P-Value	Type 3 P-Value

3	female	No	108/235 (46.0%)	1.00
	female	Yes	117/265 (44.2%)	0.93	0.65	1.32	0.69

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	687.975
	H-L Test	< 3 groups

4	treatment	Placebo	118/250 (47.2%)	1.00
	treatment	Active	107/250 (42.8%)	0.84	0.59	1.19	0.32

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	687.161
	H-L Test	< 3 groups

Pink cells indicate effects with p<0.05 Red cells indicate evidence for violations of model assumptions.

Example of running a series of univariate models and multivariate models with the logistic macro

3
Model	Variable	Value	# Outcome = 1/ # subgroup (% Outcome = 1)	Odds Ratio	Lower 95% CI	Upper 95% CI	P-Value	Quadratic P-Value	Type 3 P-Value
5	bmi	16.58-24.24	28/125 (22.4%)	1.00
	bmi	24.26-27.06	54/126 (42.9%)	2.6	1.50	4.5	0.0007		<.0001
	bmi	27.08-29.89	60/124 (48.4%)	3.2	1.88	5.6	<.0001
	bmi	29.9-37.06	83/125 (66.4%)	6.8	3.9	12.0	<.0001

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	636.432
	H-L Test	4 groups					1.00

6	obese	No	144/379 (38.0%)	1.00
	obese	Yes	81/121 (66.9%)	3.3	2.1	5.1	<.0001

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	656.911
	H-L Test	< 3 groups
Pink cells indicate effects with p<0.05 Red cells indicate evidence for violations of model assumptions.

Example of running a series of univariate models and multivariate models with the logistic macro

4
Model	Variable	Value	# Outcome = 1/ # subgroup (% Outcome = 1)	Odds Ratio	Lower 95% CI	Upper 95% CI	P-Value	Quadratic P-Value	Type 3 P-Value

7	overweight	No	162/368 (44.0%)	1.00
	overweight	Yes	63/132 (47.7%)	1.16	0.78	1.73	0.46

	Outcome	(+)Died Yes=225
	Outcome	(-)Died No =275
	-2 Log L	687.601
	H-L Test	< 3 groups

8	age	20-35	18/129 (14.0%)	1.00
	age	36-51	22/125 (17.6%)	1.27	0.64	2.5	0.49		<.0001
	age	52-66	80/128 (62.5%)	9.6	5.1	18.1	<.0001
	age	67-80	105/118 (89.0%)	43	19.8	95	<.0001

	treatment	Placebo	118/250 (47.2%)	1.00
	treatment	Active	107/250 (42.8%)	0.97	0.62	1.53	0.90

Pink cells indicate effects with p<0.05 Red cells indicate evidence for violations of model assumptions.

Example of running a series of univariate models and multivariate models with the logistic macro

5
Model	Variable	Value	# Outcome = 1/ # subgroup (% Outcome = 1)	Odds Ratio	Lower 95% CI	Upper 95% CI	P-Value	Quadratic P-Value	Type 3 P-Value
	female	No	108/235 (46.0%)	1.00
	female	Yes	117/265 (44.2%)	0.91	0.57	1.45	0.69

	weightcat	Not over	81/247 (32.8%)	1.00
	weightcat	Overweight	63/132 (47.7%)	0.99	0.57	1.72	0.96		0.19
	weightcat	Obese	81/121 (66.9%)	1.65	0.91	3.0	0.096

	Outcome	(+)died Yes=225
	Outcome	(-)died No =275
	-2 Log L	468.149
	H-L Test	10 groups					0.58

Pink cells indicate effects with p<0.05 Red cells indicate evidence for violations of model assumptions.