Table 3.

Negative Binomial Analysis of Number of Primary Care Visits among Subjects with Any Visit in FY 2000

Parameter	Estimate
Age 40–49a	0.28	<0.0001
Age 50–64a	0.59	<0.0001
Age 65–74a	0.82	<0.0001
Age 75–84a	0.59	<0.0001
Age >84a	−0.00	0.9664
Genderb	−0.52	<0.0001
Educationc	−0.04	<0.0001
Blackd	−0.09	<0.0001
Hispanicd	0.04	0.09
Other raced	0.02	0.46
Service connection less than 50e	0.04	0.005
Service connection greater than 50e	−0.19	<0.0001
Drug use disorderf	−0.28	<0.0001
Schizophreniaf	−0.96	<0.0001
Alcohol use disorderf	−0.14	0.003
PTSDf	−0.25	0.003
Bipolar disorderf	−0.67	<0.0001
Major depressive disorderf	−0.14	0.0002
Other depressionf	0.38	<0.0001
Medical co-morbidityg	0.15	<0.0001
Physical health statush	0.00	0.006
Residence micropolitan area (population 10,000–49,000)i	0.22	<0.0001
Residence large rural towni	0.17	<0.0001
Residence isolated small rural areai	0.14	<0.0001
Small facility	−0.00	<0.0001
% of budget spent on mental health	0.69	<0.0001
% of mental health budget spent on research	−0.10	0.02

Reference groups for covariates: ^a age <40; ^b male gender; ^c High school; ^d Caucasian race; ^e no service connection; ^f patients without the specified psychiatric diagnosis; ^gpatients with no co-morbid medical diagnoses; ^hpopulation norm for the VSF-36 PCS subscale = 50; ⁱ residence in urbanized areas (population >50,000); ^j mean values for national VA medical centers