Table 4.

Proportional Hazards Survival Model Predicting the Likelihood of Being Hospitalized for an Avoidable Hospitalization Among a Cohort of Infants Born in 1990 in California, Georgia, and Michigan and Enrolled in Medicaid by 1 Month of Age

FactorCalifornia*GeorgiaMichigan
HR 95% CI HR 95% CI HR 95% CI
Preventive care
 Well visits up-to-date for age0.520.50–0.55 0.540.50–0.58 0.74 0.69–0.79
 Sporadic preventive care*0.87 0.82–0.920.90 0.83–0.97 0.86 0.78–0.94
 Immunizations up-to-date for age§1.12 1.06–1.18 1.050.98–1.12 0.88 0.82–0.94
Control variables
 Average number of sick visits per week2.76 2.65–2.89 2.03 1.90–2.17 2.77 2.58–2.98
 Sick visit within the past 30 days 2.67 2.57–2.792.68 2.54–2.84 2.55 2.41–2.70
 Hospitalized during the first month of life1.48 1.25–1.751.38 1.16–1.64 1.22 0.99–1.51
 Male gender1.38 1.33–1.44 1.18 1.12–1.241.35 1.28–1.42
 Race/ethnicity#
  Black 1.14 1.07–1.21 0.850.81–0.90 1.14 1.07–1.21
  Asian/Pacific Islander0.91 0.85–0.99 0.65 0.36–1.17
  Hispanic 1.14 1.09–1.200.87 0.68–1.11 1.08 0.93–1.25
  Other 1.221.09–1.37 1.17 0.94–1.45 0.83 0.66–1.04
 Medicaid eligibility category**
  Medically needy 0.65 0.62–0.680.38 0.23–0.63 0.88 0.80–0.96
  Other poverty0.62 0.58–0.66 0.88 0.83–0.92 0.74 0.69–0.80
 Percentage of families living under the federal poverty level1.00 1.00–1.01 1.03 1.02–1.031.00 0.99–1.01
 Number of pediatricians/10 000 county residents0.97 0.93–1.01 0.74 0.72–0.760.92 0.87–0.96
 Number of hospital beds/10 000 county residents4-1641.01 0.99–1.05 1.031.02–1.04 1.03 1.00–1.05
  • * Because the size of the California data set made it impossible to run a partial likelihood model with the whole sample, a nested case-control design was used (see Methods section). The California regression consisted of 11 554 children with an avoidable hospitalization and 12 142 controls, which was 1 of 16 of the children with no avoidable hospitalization. The models used the entire Georgia and Michigan samples.

  • The child was up-to-date if he or she had 1 visit by age 3 months, 2 by age 4 months, 3 by age 6 months, 4 by age 8 months, 5 by age 11 months, 6 by age 14 months, 7 by age 17 months, and 8 by age 20 months.

  • If the child was not up-to-date by the above criteria, a sporadic well visit was defined as a recent well visit that occurred within 30 days if child <4 months of age, 60 days if 4 to 10 months, 90 days if 11 to 19 months, 120 days if 20 months, 150 days if 21 months, and 180 days if 22 months or older.

  • § This variable was based on a modification of the 1988 AAP schedule of immunizations. Children were up-to-date if they had 1 DPT by age 4 months, 2 DPTs and 1 OPV by age 6 months, 3 DPTs and 2 OPVs by 12 months, and 4 DPTs, 3 OPVs, 1 MMR, and 1 HiB by 24 months.

  • # Refers to any hospitalization not associated with birth.

    White is the reference category.

  • Female is the reference category.

  • ** AFDC eligibility category is the reference.

  • F4-164 Source: Office of Research and Planning, Bureau of Health Professions, U.S. Department of Health and Human Services, the Area Resource File, 1992 version. These data were linked to each child's record by county of birth.