TABLE 2

Adjusted Likelihood Meeting Follow-up After Psychiatric Hospitalization Measure at 7 and 30 Days (N = 1072) for Enrollee Characteristics

Likelihood of Follow-up Within 7 d (95% CI)PLikelihood of Follow-up Within 30 d (95% CI)P
Age 6–10 yReferenceReference
Age 11–15 y−0.13 (−0.24 to −0.01).03−0.10 (−0.21 to 0.01).07
Age 16–19 y−0.09 (−0.21 to 0.03).13−0.07 (−0.19 to 0.04).22
Female sex0.01 (−0.05 to 0.07).670.03 (−0.03 to 0.10).27
WhiteReferenceReference
African American−0.05 (−0.12 to 0.02).19−0.04 (−0.12 to 0.04).31
Other race0.09 (0.01 to 0.17).020.08 (−0.01 to 0.16).06
Cost-sharing group
 Low feeReferenceReference
 Fee−0.05 (−0.13 to 0.02).160.02 (−0.06 to 0.09).66
 Expansion−0.08 (−0.16 to 0.01).06−0.01 (−0.09 to 0.08).95
 Exempt−0.25 (−0.46 to −0.03).03−0.18 (−0.46 to 0.10).21
Rural-urban commuting code
 UrbanReferenceReference
 Large rural−0.01 (−0.10 to 0.09).990.03 (−0.08 to 0.14).63
 Small rural0.09 (−0.01 to 0.20).090.10 (−0.01 to 0.21).07
 Isolated0.09 (−0.05 to 0.23).190.04 (−0.10 to 0.18).56
Mental health providers within 25 miles
 0–42ReferenceReference
 43–1080.05 (−0.04 to 0.13).270.06 (−0.04 to 0.15).23
 109–2320.04 (−0.06 to 0.14).430.07 (−0.04 to 0.17).22
 233 or more−0.01 (−0.11 to 0.09).790.07 (−0.04 to 0.19).22
Primary diagnosis group
 DepressionReferenceReference
 Mood disorder−0.09 (−0.17 to −0.01).02−0.07 (−0.16 to 0.01).08
 Behavioral disorder−0.09 (−0.19 to −0.01).04−0.07 (−0.16 to 0.02).14
 Bipolar disorder−0.11 (−0.23 to 0.01).07−0.08 (−0.19 to 0.04).20
 Adjustment disorder−0.03 (−0.18 to 0.12).68−0.02 (−0.17 to 0.14).84
 Psychosis0.07 (−0.09 to 0.22).390.05 (−0.11 to 0.21).56
 Other−0.06 (−0.20 to 0.08).43−0.07 (−0.23 to 0.09).37
Hospital length of stay, d
 0–3ReferenceReference
 4–7−0.06 (−0.14 to 0.02).15−0.03 (−0.11 to 0.05).50
 8 or more−0.18 (−0.26 to −0.10)<.001−0.26 (−0.35 to −0.17)<.001
2013ReferenceReference
20140.09 (0.01 to 0.17).030.03 (−0.05 to 0.12).48
20150.02 (−0.06 to 0.10).69−0.02 (−0.11 to 0.07).70
20160.04 (−0.05 to 0.12).37−0.04 (−0.13 to 0.06).43
  • Estimates are from logit regression, the model also contains hospital fixed effects, and SEs are clustered to account for repeated observations among enrollees. —, not applicable.