TABLE 1

Thirty-Day MH Readmission Rates by Patient Characteristics

CharacteristicsIndex Admissions No. (%) (n = 47 397)a30-d Readmissions
No.Rate, %Bivariate Odds Ratio (95% CI)bMultivariate Odds Ratio (95% CI)c
Sex
 Boy20 157 (42.5)16888.41.08 (1.01–1.15)1.02 (0.95–1.09)
 Girl27 240 (57.5)20957.7ReferenceReference
Age, y
 5–82540 (5.4)2118.31.22 (1.05–1.42)1.22 (1.04–1.43)
 9–128115 (17.1)81610.11.49 (1.36–1.63)1.48 (1.35–1.63)
 13–1412 816 (27.0)10768.41.23 (1.13–1.33)1.24 (1.14–1.34)
 15–1723 926 (50.5)16807.0ReferenceReference
No. non-MH CCIsd
 ≥24733 (10.0)4228.91.25 (1.12–1.40)1.22 (1.09–1.36)
 112 450 (26.3)10378.31.14 (1.05–1.23)1.11 (1.03–1.20)
 None30 214 (63.7)23247.7ReferenceReference
Primary diagnosise
 Depression28 437 (60.0)21877.7ReferenceReference
 Bipolar disorders5169 (10.9)5139.91.34 (1.21–1.50)1.33 (1.20–1.49)
 Externalizing disorders3074 (6.5)2648.61.11 (0.96–1.27)1.00 (0.87–1.15)
 Reaction disorders2740 (5.8)1425.20.58 (0.48–0.70)0.58 (0.48–0.70)
 Psychosis2324 (4.9)26911.61.56 (1.36–1.79)1.60 (1.39–1.84)
 Anxiety disorders2136 (4.5)1406.60.86 (0.72–1.03)0.82 (0.69–0.99)
 ADHD1506 (3.2)1389.21.15 (0.96–1.39)1.01 (0.83–1.23)
 Autism spectrum disorder484 (1.0)449.11.25 (0.91–1.72)1.14 (0.83–1.56)
 Eating disorders474 (1.0)398.21.12 (0.80–1.58)1.09 (0.77–1.54)
 Substance-related disorders446 (0.9)122.70.33 (0.19–0.60)0.38 (0.21–0.68)
 Other MH diagnoses607 (1.3)355.80.79 (0.56–1.12)0.73 (0.51–1.03)
Insurancef
 Public26 264 (55.5)22818.71.26 (1.17–1.36)1.20 (1.11–1.29)
 None1102 (2.3)615.50.75 (0.58–0.98)0.78 (0.60–1.02)
 Other2307 (4.9)1948.41.12 (0.94–1.32)1.11 (0.94–1.31)
 Private17 653 (37.3)12427.0ReferenceReference
  • a MH was the primary diagnosis of 47 397 (18.7%) of 253 309 index admissions.

  • b To assess whether readmission rates varied by each patient characteristic, the P value from multiple degrees of freedom block test on all categories of each characteristic was calculated from a hierarchical logistic regression model with fixed effect for the characteristic and random effect for hospital. P = .03 for block test of sex, and P < .001 for block tests of all other characteristics in bivariate analysis.

  • c aOR from hierarchical logistic regression model with random effect for hospital and fixed effects for all characteristics that were significant in bivariate analysis with P < .20. P = .60 for block test of sex, and P < .001 for block tests of all other characteristics in multivariate analysis.

  • d The CCIs, developed by the AHRQ, categorize ∼14 000 ICD-9-CM diagnosis codes as chronic or not chronic and assign codes into 1 of 18 mutually exclusive body system groups.

  • e Other MH diagnoses include personality, reactive attachment, motor, elimination, developmental, and sexual and/or gender identity disorders, each accounting for <1% of MH admissions.

  • f Index admissions with missing insurance type (0.1%, n = 71) were excluded in the respective bivariate analysis and in the multivariate model.