Advertisement

Mental Illness Drives Hospitalizations for Detained California Youth

      Abstract

      Purpose

      The purpose of the study was to describe inpatient hospitalization patterns among detained and nondetained youth in a large, total population of hospitalized adolescents in California.

      Methods

      We examined the unmasked California Office of Statewide Health Planning and Development Patient Discharge Dataset from 1997 to 2011. We considered hospitalized youth aged 11–18 years “detained” if admitted to California hospitals from detention, transferred from hospital to detention, or both. We compared discharge diagnoses and length of stay between detained youth and their nondetained counterparts in the general population.

      Results

      There were 11,367 hospitalizations for detained youth. Hospitalizations differed for detained versus nondetained youth: 63% of all detained youth had a primary diagnosis of mental health disorder (compared with 19.8% of nondetained youth). Detained girls were disproportionately affected, with 74% hospitalized for a primary mental health diagnosis. Detained youth hospitalized for mental health disorder had an increased median length of stay compared with nondetained inpatient youth with mental illness (≥6 days vs. 5 days, respectively). This group difference was heightened in the presence of minority status, public insurance, and concurrent substance abuse. Hospitalized detained youth discharged to chemical dependency treatment facilities had the longest hospital stays (≥43 days).

      Conclusions

      Detained juvenile offenders are hospitalized for very different reasons than the general adolescent population. Mental illness, often with comorbid substance abuse, requiring long inpatient stays, represents the major cause for hospitalization. These findings underscore the urgent need for effective, well-coordinated mental health services for youth before, during, and after detention.

      Keywords

      Implications and Contribution
      The vast majority of hospitalizations among youth in the juvenile justice system result from mental health conditions, often requiring prolonged inpatient stays, transfer to specialized facilities, and significant public resources.
      See Related Editorial p. 453
      Detained youth are a high-risk population with numerous unmet medical and mental health needs [
      • Joseph-DiCaprio J.
      • Farrow J.
      • Feinstein R.A.
      • et al.
      Health care for incarcerated youth. Position paper of the Society for Adolescent Medicine.
      ,
      American Academy of Pediatrics. Health care for children and adolescents in the juvenile correctional care system.
      ,
      • Feinstein R.A.
      • Lampkin A.
      • Lorish C.D.
      • et al.
      Medical status of adolescents at time of admission to a juvenile detention center.
      ]. Previous studies within the juvenile justice system demonstrated increased prevalence of a variety of medical conditions, including sexually transmitted diseases, pregnancy, asthma, and obesity [
      • Thornberry T.P.
      • Huizinga D.
      • Loeber R.
      The causes and correlates studies: Findings and policy implications.
      ,
      • Morris R.E.
      • Harrison E.A.
      • Knox G.W.
      • et al.
      Health risk behavioral survey from 39 juvenile correctional facilities in the United States.
      ,
      • Lederman C.S.
      • Dakof G.A.
      • Larrea M.A.
      • et al.
      Characteristics of adolescent females in juvenile detention.
      ,
      • Forrest C.B.
      • Tambor E.
      • Riley A.W.
      • et al.
      The health profile of incarcerated male youths.
      ]. The prevalence of psychiatric illness in detained adolescents is striking, with studies suggesting that up to two thirds of these youth have mental health disorders [
      • Teplin L.A.
      • Abram K.M.
      • McClelland G.M.
      • et al.
      Psychiatric disorders in youth in juvenile detention.
      ,
      • Domalanta D.D.
      • Risser W.L.
      • Roberts R.E.
      • et al.
      Prevalence of depression and other psychiatric disorders among incarcerated youths.
      ,
      • Cauffman E.
      A statewide screening of mental health symptoms among juvenile offenders in detention.
      ].
      Despite this increased burden of illness, little is known about the severity of these conditions among detained youth. Hospitalization, as an indicator of disease severity, can provide insight into the burden of unmet health care needs. Moreover, because many of the conditions observed in these youth are potentially amenable to coordinated outpatient care, characterization of hospitalizations for this group could help elucidate the nature and scope of their unmet health needs.
      Few previous studies have examined hospitalizations among detained youth, and none have examined the causes of hospitalization in a large, total population of hospitalized adolescents [
      • Feinstein R.A.
      • Lampkin A.
      • Lorish C.D.
      • et al.
      Medical status of adolescents at time of admission to a juvenile detention center.
      ,
      • Balla D.
      • Lewis D.O.
      • Shanok S.
      • et al.
      Subsequent psychiatric treatment and hospitalization in a delinquent population.
      ,
      • Cropsey K.L.
      • Weaver M.F.
      • Dupre M.A.
      Predictors of involvement in the juvenile justice system among psychiatric hospitalized adolescents.
      ]. Using data from the California Office of Statewide Health Planning and Development (OSHPD) Patient Discharge Database, we examined all hospital discharges among adolescents aged 11–18 years in California over a 15-year time span. Our objectives were to characterize hospitalization patterns among detained youth and to compare these patterns with those for the general, nondetained adolescent population in the large, ethnically diverse state of California utilizing a large database.

      Methods

      Patients and study variables

      Data source

      We used the private, unmasked OSHPD Patient Discharge Database from 1997 to 2011. This dataset contains information (submitted biannually) on all hospital discharges from nonfederal acute care hospitals. Variables in the “unmasked” version of the dataset include all available patient identifiers (date of birth, social security number, Zone Improvement Plan (ZIP) code of residence, hospital identification number), as well as other sociodemographic information (ethnicity, race, gender), expected source of payment (health insurance), admission source, diagnosis and treatment codes, length of hospital stay, and disposition.

      Study population

      We queried the dataset for all adolescents aged 11–18 years, excluding non-California residents (n = 35,662) from the analyses. We considered patients admitted to California hospitals from detention facilities, transferred from a hospital directly to a detention facility, or both during the study period to be “detained youth.” We identified admissions from detention by Admission Source of “jail” in the OSHPD database and discharges to detention by Disposition to “jail.” Based on our clinical experience, we divided hospitalizations for detained youth into three patterns: (1) hospitalizations for youth admitted from detention and then transferred back to detention (hereafter referred to as “Detention-Hospital-Detention”); (2) hospitalizations for youth admitted from the community and then transferred to detention (hereafter referred to as “Community-Hospital-Detention”); and (3) hospitalizations for youth admitted from detention and subsequently discharged or transferred to another facility, including psychiatric hospitals or substance abuse treatment centers (hereafter referred to as “Detention-Hospital-Treatment”). Throughout the article, we use “hospitalizations” among detained youth or nondetained youth as shorthand for “discharges among detained youth” or “discharges among nondetained youth,” particularly in the tables.

      Sociodemographic, health insurance, and geographic characteristics

      Sociodemographic variables we examined included patient age, gender, and race/ethnicity (Caucasian, black, Hispanic, Asian-American, and other). We categorized health insurance as public insurance, private insurance, and other [
      • Chamberlain L.J.
      • Chan J.
      • Mahlow P.
      • et al.
      Variation in specialty care hospitalization for children with chronic conditions in California.
      ]. We used ZIP code of residence to create a dichotomous variable reflecting major metropolitan population centers versus rural areas.

      Diagnosis identification

      We determined the principal discharge diagnosis and up to 24 secondary diagnoses to identify co-occurring or comorbid conditions [
      • Chamberlain L.J.
      • Chan J.
      • Mahlow P.
      • et al.
      Variation in specialty care hospitalization for children with chronic conditions in California.
      ,
      • Pineda N.
      • Chamberlain L.J.
      • Chan J.
      • et al.
      Access to pediatric subspecialty care: A population study of pediatric rheumatology inpatients in California.
      ,
      • Wang N.E.
      • Chan J.
      • Mahlow P.
      • Wise P.H.
      Trauma center utilization for children in California 1998–2004: Trends and areas for further analysis.
      ]. We collapsed International Classification of Disease Ninth Revision, Clinical Modification (ICD-9-CM) codes into the following broad categories: (1) Mental Health; (2) Trauma; (3) Pregnancy; and (4) Other Acute and Chronic Medical conditions. The Mental Health category comprised psychiatric DSM-IV discharge diagnostic codes (ICD-9 CM codes 290xx–319xx). This category reflected the following three major groups: (1) mental disorders; (2) substance use disorders; and (3) developmental disorders. We subdivided mental disorders into clinically relevant psychiatric categories that included anxiety/stress, depressive, disruptive, and psychotic disorders [
      • Huffman L.C.
      • Wang N.E.
      • Saynina O.
      • et al.
      Predictors of hospitalization after emergency department visit for California youths with psychiatric disorders.
      ]. Finally, we identified mental health comorbidities (defined as a psychiatric diagnosis that occurred with one or more lower ranked psychiatric diagnoses) because such conditions are predictors of poorer outcomes [
      • Hawton K.
      • van Heeringen K.
      Suicide.
      ,
      • Groholt B.
      • Ekeberg O.
      • Haldorsen T.
      Adolescent suicide attempters: What predicts future suicidal acts?.
      ].
      We defined Trauma diagnoses as ICD-9 CM codes 800xx-959xx. We calculated Injury Severity Scores (ISS) for all hospitalizations with trauma diagnoses and stratified them into mild (ISS = 1–8), moderate (ISS = 9–15), and severe (ISS > 16) [
      • Wang N.E.
      • Chan J.
      • Mahlow P.
      • Wise P.H.
      Trauma center utilization for children in California 1998–2004: Trends and areas for further analysis.
      ,
      • Marcin J.P.
      • Schembri M.S.
      • He J.
      • Romano P.S.
      A population-based analysis of socioeconomic status and insurance status and their relationship with pediatric trauma hospitalization and mortality rates.
      ]. Diagnoses associated with pregnancy included 630xx-677xx. We categorized all other ICD-9 codes as “Other acute and chronic medical” diagnoses. Although we examined both primary and up to 24 secondary discharge diagnoses, we only counted each discharge once.

      Statistical analysis

      We compared discharge diagnoses between detained adolescents (including stratification by gender, race/ethnicity and by the three hospitalization patterns) and their nondetained counterparts using chi-square tests. We compared length of stay (LOS) between detained and nondetained youth within diagnosis categories (i.e., Mental Health, Trauma, Pregnancy, Other Medical). To account for skewness in LOS, we used nonparametric tests (Wilcoxon and Kruskal–Wallis) to compare sociodemographics, insurance status, disposition subgroup, diagnosis categories, and hospital type among detained and nondetained youth. We designated p values less than .05 to indicate statistical significance, and we report 95% confidence intervals. We used SAS 9.3 (SAS Institute, Cary, NC) for all analyses.

      Human subjects

      The institutional review board at Stanford University and the State of California Committee for the Protection of Human Subjects reviewed and approved this study.

      Results

      Of the 3,562,644 pediatric discharges during the study period, 1,936,513 involved adolescent California residents. Six tenths of one percent (.6%) of adolescent discharges (n = 11,367) either originally came to the hospital directly from a juvenile detention facility, were discharged to such a facility, or both (Figure 1).
      Figure thumbnail gr1
      Figure 1Flowchart of study population from California OSHPD hospital discharge dataset, 1997–2011.
      Table 1 shows the significant differences in general characteristics and distribution of hospital discharge diagnoses between the detained and nondetained adolescent populations. Detained youth were more likely to be older (mean age 16.1 vs. 15.5 years, p < .001), male (63% vs. 32%, p < .001), and publicly insured (71.8% vs. 51.8%, p < .001) compared with the nondetained, inpatient adolescent population. In addition, hospitalized detained youth were disproportionately black and from larger metropolitan counties: Los Angeles (24.9%), Alameda (13.7%), San Diego (7.8%), Orange (5.9%), Contra Costa (4.9%), and Santa Clara (4%; data not shown). Sixty-two percent of all hospitalizations among detained youth versus 53.6% (p < .001) of hospitalizations among nondetained adolescents in the general population came from these six, mostly urban counties.
      Table 1Comparison of demographic characteristics among detained and nondetained youth hospitalized in California
      Detained, hospitalized, N = 11,367Nondetained, hospitalized, N = 1,925,146p
      Mean age, years16.115.5<.001
      Sex, n (%)
       Male7,772 (63)718,441 (32)<.001
       Female3,595 (37)1,206,705 (68)
      Race/ethnicity, n (%)
       Black2,550 (22.8)193,847 (10.2)<.001
       Caucasian3,834 (34.3)718,273 (37.7)
       Hispanic4,092 (36.6)845,041 (44.4)
       Asian303 (2.7)85,471 (4.5)
       Other401 (3.6)62,500 (3.3)
      Payer, n (%)
      “Public” includes Medicare, Medicaid, county indigent programs, State Children's Health Insurance program, and Title V-supported California Children's Services (CCS); “Other” includes worker's compensation, self-pay, and other payer.
       Private2,285 (20.1)843,501 (43.9)<.001
       Public8,156 (71.8)997,268 (51.8)
       Other918 (8.1)83,479 (4.3)
      County
       Major metropolitan
      Larger metropolitan counties include Los Angeles, Alameda, San Diego, Orange, Contra Costa, and Santa Clara counties.
      6,959 (61.2)1,031,108 (53.6)<.001
      a “Public” includes Medicare, Medicaid, county indigent programs, State Children's Health Insurance program, and Title V-supported California Children's Services (CCS); “Other” includes worker's compensation, self-pay, and other payer.
      b Larger metropolitan counties include Los Angeles, Alameda, San Diego, Orange, Contra Costa, and Santa Clara counties.
      Among all nondetained adolescents in California, the main causes of hospitalization by diagnosis category included: (1) Other Acute and Chronic Medical conditions (43%; most commonly appendicitis, pulmonary/gastroenterologic/genitourinary infections, and hematologic conditions); (2) Pregnancy-related conditions (28.2%; most commonly normal deliveries, gestational diabetes, pre-eclampsia, and pre-delivery and post-delivery infections); (3) Mental Health (19.7%; most commonly depression and substance use disorders); and (4) Trauma (9.1%; most commonly extremity fractures, abdominal/thorax/head injuries, and “other trauma”) (Table 2; data in parenthesis not in table). Most primary trauma diagnoses were “unintentional” (77.5%), and the vast majority was mild-to-moderate in severity (86.5%; not in table). Of note, for nondetained youth, the number of discharges associated with co-occurring conditions across diagnostic categories (e.g., Mental Health and Trauma) was quite small.
      Table 2Primary hospital discharge diagnoses and length of stay for detained and nondetained youth in California, stratified by mental health diagnostic categories, race/ethnicity, and insurance type, 1997–2011
      Primary discharge diagnoses (by race/ethnicity and payer type)Detained, N = 11,367Nondetained, N = 1,925,146p value
      n (%)Median LOS (IQR)n (%)Median LOS (IQR)LOS
      Mental health
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      7,158 (63)6 (3–14)380,183 (20)5 (3–8)<.001
       Race/ethnicity
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Black1,630 (23)7 (3–14)45,935 (12)6 (3–9)<.001
      Caucasian2,805 (39)6 (2–12)195,836 (52)5 (3–8).001
      Asian204 (3)6 (3–12.5)13,757 (4)5 (3–8)<.005
      Hispanic2,125 (30)7 (3–15)105,260 (28)5 (3–8)<.001
      Other254 (4)6 (3–17)12,949 (3)5 (3–8).038
       Insurance
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Public5,225 (73)7 (3–16)176,454 (46)5 (3–9)<.001
      Private1,632 (23)5 (3–10)190,364 (50)5 (3–8).0001
      Other298 (4)4 (2–9)12,895 (3)4 (2–7).458
       Mental Health subcategories
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Depressive
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      2,621 (37)6 (3–11)199,846 (53)5 (3–9)<.001
      Anxiety/stress
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      1,900 (27)6 (2–18)77,515 (20)5 (3–8)<.001
      Disruptive
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      2,616 (37)7 (3–19)86,409 (23)6 (3–9)<.001
      Psychotic
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      1,377 (19)8 (4–15)46,024 (12)7 (4–12)<.001
      Substance abuse
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      3,015 (42)6 (3–17)12,949 (3)5 (3–8).038
      Trauma
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      1,314 (12)2 (1–4)175,359 (9)2 (1–4)<.001
       Race/ethnicity
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Black292 (22)2 (1–5)15,764 (9)2 (1–4).008
      Caucasian232 (18)2 (1–4)74,305 (42)2 (1–3).110
      Asian35 (3)2 (1–6)8,191 (5)2 (1–4).201
      Hispanic577 (44)2 (1–5)58,983 (34)2 (1–4).339
      Other48 (4)2 (1–4.5)6,685 (4)2 (1–4).979
       Insurance
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Public905 (69)2 (1–5)64,986 (37)2 (1–4).009
      Private218 (17)2 (1–3)95,233 (54)2 (1–3).476
      Other189 (14)2 (1–3)15,072 (9)2 (1–3).702
      Pregnancy
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      363 (3)2 (2–3)542,585 (28)2 (2–3).027
       Race/ethnicity
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Black89 (25)2 (2–3)48,638 (9)2 (1–3).507
      Caucasian96 (26)2 (1–3)109,359 (20)2 (1–3).738
      Asian10 (3)3 (2–3)16,089 (3)2 (1–2).022
      Hispanic150 (41)2 (2–3)352,278 (65)2 (2–3).238
      Other13 (4)3 (2–3)12,329 (2)2 (2–3).223
       Insurance
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Public261 (72)2 (2–3)397,376 (73)2 (2–3).111
      Private51 (14)2 (2–3)123,898 (23)2 (1–3).469
      Other50 (14)2 (2–3)21,203 (4)2 (1–3).035
      Other acute/chronic medical
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      2,532 (22)2 (1–4)827,020 (43)2 (1–4).766
       Race/ethnicity
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Black539 (21)2 (1–5)83,510 (10)3 (1–5).352
      Caucasian701 (28)2 (1–4)338,773 (41)2 (1–4).820
      Asian54 (2)3 (1–4)47,434 (6)3 (1–5).345
      Hispanic1,240 (49)2 (1–4)328,520 (40)2 (1–4).875
      Other86 (3)2 (0–4)30,537 (4)2 (1–5).005
       Insurance
      Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Public1,765 (70)2 (1–4)358,452 (43)3 (1–5).002
      Private384 (15)2 (1–4)434,006 (52)2 (1–4).136
      Other381 (15)2 (1–4)34,309 (4)2 (1–3).083
      IQR = interquartile range; LOS = length of stay.
      a Distributions across discharge diagnoses, race/ethnicity, insurance status, and mental health subcategories were ALL significantly different for detained versus nondetained hospitalized adolescents, with p values of <.001. (p values for LOS are listed separately for each comparison in the table).
      Table 2 demonstrates that for detained youth, the distribution of hospital discharge diagnoses was strikingly different: Mental Health was the most common diagnostic category (63.0%), followed by Other Acute and Chronic Medical conditions (22.3%), Trauma (11.6%), and Pregnancy-related conditions (3.2%). The vast majority of hospitalizations for detained youth in all categories were covered by public insurance, and racial/ethnic distributions for detained versus nondetained youth were also significantly different. Detained blacks and Hispanics were disproportionately hospitalized in every diagnostic category except pregnancy (nondetained Hispanics were more likely to have a pregnancy-related discharge diagnosis than detained Hispanics).
      Mental Health diagnoses included a wide range of overlapping psychiatric illness, including substance use (42.1%), depressive (36.6%), and disruptive disorders (36.6%). Of interest (but not shown in table), the diagnostic distributions within the categories of Other Acute and Chronic Medical conditions and Pregnancy-related conditions were quite similar to those described for the general adolescent population. Trauma diagnoses for detained youth were overrepresented by intentional assault-related injuries (35.6% vs. 14.6% in nondetained youth; most commonly facial fractures, head traumas/concussions, and open chest wounds) and suicide attempts (5.0% vs. .8% in nondetained adolescents). Detained youth had fewer motor vehicle accidents leading to hospitalization (22.6% vs. 27.7%) and more injuries classified as “mild” (67% vs. 60%) than nondetained youth.
      The proportion of hospitalizations for detained boys with a primary diagnosis in the Mental Health category was 57.8% versus 74.1% for detained girls (not shown in table). However, hospitalizations for detained boys were >5 times more likely to be associated with a primary diagnosis in the Trauma category than those for detained girls (15.6% vs. 2.8%). Finally, approximately one in 15 hospitalizations among all detained youth was associated with co-occurring diagnoses in the Mental Health and Trauma categories (e.g., concurrent principal mental health diagnosis with a secondary trauma diagnoses, or a principal trauma diagnosis with a secondary mental health diagnosis).
      Median LOS for hospitalizations in Pregnancy, Trauma, and Other Acute and Chronic Medical categories was 2 days for both detained and nondetained youth, as Table 2 demonstrates. However, for hospitalizations of youth with a primary diagnosis in Mental Health category, median LOS was increased for those who were detained compared with nondetained (6 and 5 days respectively; p < .001) and in particular for blacks (7 vs. 6 days, p < .001), Hispanics (7 vs. 5 days, p < .001), and those with public health insurance (7 vs. 5 days, respectively; p < .001). Discharge diagnoses of all subcategories of Mental Health were associated with increased LOS for detained youth versus nondetained youth.
      Table 3 summarizes the analyses of the three clinically distinct hospitalization discharge patterns (based on the admission source and disposition of detained youth) and their differences in diagnoses. In the “Detention-Hospital-Detention” pattern (n = 3,024), 47% of hospitalizations for detained youth were associated with primary diagnoses within Mental Health category (vs. 20% in the general population; p < .0001); of those, almost 40% were for substance use disorders (vs. 25% in the general population; p < .0001; not shown in table). Next, in the “Community-Hospital-Detention” pattern (n = 4,924), 57% of hospitalizations for detained youth were associated with primary diagnoses within Mental Health category (vs. 20% in the general population; p < .0001), with approximately 35% for substance use disorders (vs. 26% in general population; p < .0001; not shown in table). Eighteen percent of this discharge pattern were associated with primary diagnoses within the Trauma category (vs. 9% in nondetained youth; p < .0001). Finally, in the “Detention-Hospital-Treatment” pattern (n = 3,419), 85% of hospitalizations for detained youth were associated with primary diagnoses in the Mental Health category (vs. 20% in the nondetained population; p < .0001), with about 44% for substance use disorders (vs. 26% in general population; p < .0001; not shown in table).
      Table 3LOS for mental health (by disposition facility type) and trauma among detained and nondetained youth in California by discharge pattern, 1997–2011
      DetainedNot detained
      (1) Detention→ Hospital→ Detention (n = 3,024, 26.6%)(2) Community →Hospital→ Detention (n = 4,924, 43.3%)(3) Detention→ Hospital→ Treatment (n = 3,419, 30%)Hospitalized Adolescents (n = 1,925,146)
      n%Median LOS (IQR)n%Median LOS (IQR)n%Median LOS (IQR)n%Median LOS (IQR)
      Mental Health1,41847
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      7 (3–14)2,83057
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      6 (3–11)2,91085
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      6 (2–20)380,182205 (3–8)
       “Chemical Dependency”
      “Chemical Dependency” inpatient facility represents the term the Office of Statewide Health Planning and Development dataset uses for inpatient substance abuse facilities.
      inpatient facility
      13471 (26–145)4943 (14–94)68552 (12–205)2,91028 (11–45)
       Psychiatric facility6567 (4–11)1,5196 (3–10)1,7374 (2–9)271,3905 (3–8)
      Trauma33211
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–4)86218
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–5)1204
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–4)175,35992 (1–4)
      Other
      In this instance, “Other” includes pregnancy-related diagnoses, infectious processes, appendicitis, and other medical diagnoses.
      1,27442
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–4)1,23225
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–4)38911
      p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      2 (1–4)1,369,605712 (1–4)
      IQR = interquartile range; LOS = length of stay.
      a “Chemical Dependency” inpatient facility represents the term the Office of Statewide Health Planning and Development dataset uses for inpatient substance abuse facilities.
      b In this instance, “Other” includes pregnancy-related diagnoses, infectious processes, appendicitis, and other medical diagnoses.
      c p < .0005 for comparison of each column value to the reference Not Detained Hospitalized Adolescents (last column).
      Median LOS was increased for hospitalizations associated with diagnoses in Mental Health category (Detention-Hospital-Detention 7 days; Community-Hospital-Detention and Detention-Hospital-Treatment both 6 days; vs. 5 days for nondetained adolescents). For all youth with primary diagnoses in the Mental Health category, detained or nondetained, the discharge disposition to substance abuse-related facilities (referred to as “Chemical Dependency” facilities in the OSHPD dataset) was associated with longest stays (median LOS of 71, 43, and 52 days for Detention-Hospital-Detention, Community-Hospital-Detention, and Detention-Hospital-Treatment, respectively). Nondetained youth transferred to similar facilities for substance abuse/“chemical dependency” treatment had a median LOS of 28 days.

      Discussion

      This study provides a first look at hospitalization patterns among a large population of adolescents detained within the juvenile justice system. Hospitalizations among these youth occur for reasons that are different than for adolescents in the general population in California. Mental health-related diagnoses account for a far larger proportion of hospital discharges among detained youth in California than those among nondetained adolescents in the state. Furthermore, compared with nondetained youth, detained youth who were hospitalized for mental health-related diagnoses had an increased LOS, especially among racial/ethnic minorities, those with public insurance and those with co-occurring substance abuse disorders requiring specialized inpatient chemical dependency care. These findings underscore the frequency and severity of mental health conditions among detained youth, as well as the special role the juvenile justice system can play in assuring health-related services for youth with severe, comorbid psychiatric illness in California.

      Mental health

      The finding that mental illness was the principal cause of hospitalization for detained youth supports the evidence that these youth have significant unmet mental health needs. The literature suggests youth in the justice system suffer disproportionately from mental illness and have poor access to needed care in the community setting [
      • Thornberry T.P.
      • Huizinga D.
      • Loeber R.
      The causes and correlates studies: Findings and policy implications.
      ,
      • Teplin L.A.
      • Abram K.M.
      • McClelland G.M.
      • et al.
      Psychiatric disorders in youth in juvenile detention.
      ,
      • Shelton D.
      Failure of mental health policy: Incarcerated children and adolescents.
      ,
      • Ståhlberg O.
      • Anckarsäter H.
      • Nilsson T.
      Mental health problems in youths committed to juvenile institutions: Prevalences and treatment needs.
      ,
      • Teplin L.A.
      • Welty L.J.
      • Abram K.M.
      • et al.
      Prevalence and persistence of psychiatric disorders in youth after detention: A prospective longitudinal study.
      ,
      • Schubert C.A.
      • Mulvey E.P.
      • Glasheen C.
      Influence of mental health and substance use problems and criminogenic risk on outcomes in serious juvenile offenders.
      ]. Although studies have documented that youth previously involved in juvenile court proceedings were more likely to be hospitalized for psychiatric illness later in life, none have systematically compared hospitalizations among detained youth with their nondetained counterparts [
      • Balla D.
      • Lewis D.O.
      • Shanok S.
      • et al.
      Subsequent psychiatric treatment and hospitalization in a delinquent population.
      ,
      • Cropsey K.L.
      • Weaver M.F.
      • Dupre M.A.
      Predictors of involvement in the juvenile justice system among psychiatric hospitalized adolescents.
      ,
      • Kjelsberg E.
      Adolescence-limited versus life-course-persistent criminal behaviour in adolescent psychiatric inpatients.
      ]. Further highlighting the disproportionate impact of mental illness for detained youth, a recent national study examining pediatric inpatient care found that approximately 10% of all hospitalizations in the age group <21 years were for mental health-related conditions, a frequency that is still almost six times lower than demonstrated among detained youth in this study [
      • Bardach N.S.
      • Coker T.R.
      • Zima B.T.
      • et al.
      Common and costly hospitalizations for pediatric mental health disorders.
      ].

      Gender differences

      Although detained girls' physical and mental health needs are significant and may have lifelong consequences [
      • Lederman C.S.
      • Dakof G.A.
      • Larrea M.A.
      • et al.
      Characteristics of adolescent females in juvenile detention.
      ,
      • Odgers C.
      • Robins S.
      • Russell M.
      Morbidity and mortality risk among the “Forgotten Few”: Why are girls in the justice system in such poor health?.
      ,
      • Abram K.M.
      • Teplin L.A.
      • McClelland G.M.
      • Dulcan M.K.
      Comorbid psychiatric disorders in youth in juvenile detention.
      ,
      • Fazel S.
      • Doll H.
      • Långström N.
      Mental disorders among adolescents in juvenile detention and correctional facilities: A systematic review and metaregression analysis of 25 surveys.
      ,
      • Zoccolillo M.
      • Rogers K.
      Characteristics and outcome of hospitalized adolescent girls with conduct disorder.
      ,
      • Pajer K.A.
      What happens to “bad” girls? A review of the adult outcomes of antisocial adolescent girls.
      ], few studies have examined the severity of their health needs or documented the services required to address them in large populations. The finding that almost three quarters of hospital discharges for detained female youth were related to a primary mental health diagnosis underscores the critical need for mental health services for this group. Detained female youth were less likely to be hospitalized for pregnancy-related conditions than nondetained female adolescents; a finding that stands in contrast to previous studies documenting their higher risk sexual behaviors, decreased use of contraception and condoms, and high rates of pregnancy [
      • Morris R.E.
      • Harrison E.A.
      • Knox G.W.
      • et al.
      Health risk behavioral survey from 39 juvenile correctional facilities in the United States.
      ,
      • Williams R.A.
      • Hollis H.M.
      Health beliefs and reported symptoms among a sample of incarcerated adolescent females.
      ,
      • Breuner C.C.
      • Farrow J.A.
      Pregnant teens in prison. Prevalence, management, and consequences.
      ,
      • Mertz K.J.
      • Voight R.A.
      • Hutchins K.
      • et al.
      Findings from STD screening of adolescents and adults entering corrections facilities: Implications for STD control strategies.
      ]. However, clinical experience suggests that few pregnant girls remain in detention for the entirety of their pregnancy, and even fewer deliver while detained. The finding that hospitalizations for physical trauma are more common among detained male youth is consistent with the existing literature and is consistent with the observation that poor mental health may influence risk behaviors associated with trauma [
      • Forrest C.B.
      • Tambor E.
      • Riley A.W.
      • et al.
      The health profile of incarcerated male youths.
      ,
      • Conseur A.
      • Rivara F.
      • Emanuel I.
      Juvenile delinquency and adolescent trauma: How strong is the connection?.
      ]. Although emergency room and hospital-based interventions have shown promise for connecting high-risk youth to mental health services [
      • Cunningham R.M.
      • Vaidya R.S.
      • Walton M.
      • Maio R.F.
      Training emergency medicine nurses and physicians in youth violence prevention.
      ], such efforts may be too late. Community-based efforts to address the interaction of mental illness, violence, and trauma among youth must play a larger role in enhancing access to mental health services before entering the juvenile justice system.

      Hospitalization patterns for detained youth

      The transfer patterns from hospitals to facilities specializing in treating mental illness and substance abuse highlight the role of gateway into the mental health care system that the juvenile justice system can sometimes play. Recognition of this role, and the responsibility it implies, strongly suggests the need for enhanced screening protocols and coordination with outpatient, social, and educational services for detained youth.

      Resource utilization

      Despite recent decreases in violent and property crimes among adolescents [

      Office of Juvenile Justice and Delinquency Prevention Census of Juveniles in Residential Placement for 2011 year. Available at: http://ojjdp.gov/ojstatbb/ezacjrp/. Accessed January 21, 2015.

      ], it is possible that societal norms and political priorities during the last several decades have favored a shift of resources from community mental health systems toward the justice system—an inverse relationship that Penrose first described in the 1930s [
      • Penrose L.
      Disease and crime: Outline of a comparative study of European statistics.
      ]. Given that most detained youth are publicly insured [
      • Gupta R.A.
      • Kelleher K.J.
      • Pajer K.
      • et al.
      Delinquent youth in corrections: Medicaid and reentry into the community.
      ] and that a majority of hospitalized detained youth in this study had a public payer at time of hospitalization, the longer inpatient stays imply increased public expenditures. More broadly, policies that enhance the equitable provision of effective mental health and substance abuse services for high-risk youth in the community and in juvenile detention facilities could potentially reduce both health disparities and expenditures related to youth crime, juvenile detention, and hospitalizations.

      Limitations

      This study has several limitations. First, the reliance on a large administrative database such as the OSHPD discharge dataset does not provide detailed clinical information that would allow greater definition of the psychosocial, behavioral, and management attributes associated with hospitalizations among detained youth. However, this dataset has been used successfully in analyzing hospitalization patterns for a variety of population groups [
      • Chamberlain L.J.
      • Chan J.
      • Mahlow P.
      • et al.
      Variation in specialty care hospitalization for children with chronic conditions in California.
      ,
      • Pineda N.
      • Chamberlain L.J.
      • Chan J.
      • et al.
      Access to pediatric subspecialty care: A population study of pediatric rheumatology inpatients in California.
      ,
      • Huffman L.C.
      • Wang N.E.
      • Saynina O.
      • et al.
      Predictors of hospitalization after emergency department visit for California youths with psychiatric disorders.
      ,
      • Chamberlain L.J.
      • Pineda N.
      • Winestone L.
      • et al.
      Increased utilization of pediatric specialty care: A population study of pediatric oncology inpatients in California.
      ] and provides important insights into the special characteristics of hospitalizations among a large population of detained youth. Second, although many of our findings were associated with very large effects sizes, some were relatively small and may have reached statistical significance because of our large sample size. However, with LOS in particular, a seemingly small 1 day difference between Mental Health hospitalizations in detained versus nondetained youth may actually represent significant health system-wide utilization and expenditures, particularly among the publicly insured. Finally, as in prior studies using this dataset, the present study only examined hospitalizations, not individual patients. As a result, we can neither account for multiple admissions by individual patients nor calculate detained youths' hospitalization rates for comparison to the general adolescent population. Instead, this study documents overall utilization patterns and the service demands this special group of California youth generates on both the juvenile justice and hospital systems in the state.
      This total population study of hospitalized adolescents in California documents the serious health and mental health needs of youth involved in the juvenile justice system. These findings underscore the complex interaction between the health care and juvenile justice systems and outline potential opportunities to more effectively and compassionately address the profound needs of this special group of high-risk youth.

      Funding Sources

      Supported by Child Health Research Program Early Career Investigator Award, Stanford University.

      References

        • Joseph-DiCaprio J.
        • Farrow J.
        • Feinstein R.A.
        • et al.
        Health care for incarcerated youth. Position paper of the Society for Adolescent Medicine.
        J Adolesc Health. 2000; 27: 73-75
      1. American Academy of Pediatrics. Health care for children and adolescents in the juvenile correctional care system.
        Pediatrics. 2001; 107: 799-803
        • Feinstein R.A.
        • Lampkin A.
        • Lorish C.D.
        • et al.
        Medical status of adolescents at time of admission to a juvenile detention center.
        J Adolesc Health. 1998; 22: 190-196
        • Thornberry T.P.
        • Huizinga D.
        • Loeber R.
        The causes and correlates studies: Findings and policy implications.
        Juv Justice. 2004; IX: 3-19
        • Morris R.E.
        • Harrison E.A.
        • Knox G.W.
        • et al.
        Health risk behavioral survey from 39 juvenile correctional facilities in the United States.
        J Adolesc Health. 1995; 17: 334-344
        • Lederman C.S.
        • Dakof G.A.
        • Larrea M.A.
        • et al.
        Characteristics of adolescent females in juvenile detention.
        Int J Law Psychiatry. 2004; 27: 321-337
        • Forrest C.B.
        • Tambor E.
        • Riley A.W.
        • et al.
        The health profile of incarcerated male youths.
        Pediatrics. 2000; 105: 286-291
        • Teplin L.A.
        • Abram K.M.
        • McClelland G.M.
        • et al.
        Psychiatric disorders in youth in juvenile detention.
        Arch Gen Psychiatry. 2002; 59: 1133-1143
        • Domalanta D.D.
        • Risser W.L.
        • Roberts R.E.
        • et al.
        Prevalence of depression and other psychiatric disorders among incarcerated youths.
        J Am Acad Child Adolesc Psychiatry. 2003; 42: 477-484
        • Cauffman E.
        A statewide screening of mental health symptoms among juvenile offenders in detention.
        J Am Acad Child Adolesc Psychiatry. 2004; 43: 430-439
        • Balla D.
        • Lewis D.O.
        • Shanok S.
        • et al.
        Subsequent psychiatric treatment and hospitalization in a delinquent population.
        Arch Gen Psychiatry. 1974; 30: 243-245
        • Cropsey K.L.
        • Weaver M.F.
        • Dupre M.A.
        Predictors of involvement in the juvenile justice system among psychiatric hospitalized adolescents.
        Addict Behaviors. 2008; 33: 942-948
        • Chamberlain L.J.
        • Chan J.
        • Mahlow P.
        • et al.
        Variation in specialty care hospitalization for children with chronic conditions in California.
        Pediatrics. 2010; 125: 1190-1199
        • Pineda N.
        • Chamberlain L.J.
        • Chan J.
        • et al.
        Access to pediatric subspecialty care: A population study of pediatric rheumatology inpatients in California.
        Arthritis Care Res. 2011; 63: 998-1005
        • Wang N.E.
        • Chan J.
        • Mahlow P.
        • Wise P.H.
        Trauma center utilization for children in California 1998–2004: Trends and areas for further analysis.
        Acad Emerg Med. 2007; 14: 309-315
        • Huffman L.C.
        • Wang N.E.
        • Saynina O.
        • et al.
        Predictors of hospitalization after emergency department visit for California youths with psychiatric disorders.
        Psychiatr Serv. 2012; 63: 896-905
        • Hawton K.
        • van Heeringen K.
        Suicide.
        Lancet. 2009; 373: 1372-1381
        • Groholt B.
        • Ekeberg O.
        • Haldorsen T.
        Adolescent suicide attempters: What predicts future suicidal acts?.
        Suicide LifeThreat Behav. 2006; 36: 638-650
        • Marcin J.P.
        • Schembri M.S.
        • He J.
        • Romano P.S.
        A population-based analysis of socioeconomic status and insurance status and their relationship with pediatric trauma hospitalization and mortality rates.
        Am J Public Health. 2003; 93: 461-466
        • Shelton D.
        Failure of mental health policy: Incarcerated children and adolescents.
        Pediatr Nurs. 2002; 28: 278-281
        • Ståhlberg O.
        • Anckarsäter H.
        • Nilsson T.
        Mental health problems in youths committed to juvenile institutions: Prevalences and treatment needs.
        Eur Child Adolesc Psychiatry. 2010; : 1547-1557
        • Teplin L.A.
        • Welty L.J.
        • Abram K.M.
        • et al.
        Prevalence and persistence of psychiatric disorders in youth after detention: A prospective longitudinal study.
        Arch Gen Psychiatry. 2012; 69: 1031-1043
        • Schubert C.A.
        • Mulvey E.P.
        • Glasheen C.
        Influence of mental health and substance use problems and criminogenic risk on outcomes in serious juvenile offenders.
        J Am Acad Child Adolesc Psychiatry. 2011; 50: 925-937
        • Kjelsberg E.
        Adolescence-limited versus life-course-persistent criminal behaviour in adolescent psychiatric inpatients.
        Eur Child Adolesc Psychiatry. 1999; 8: 276-282
        • Bardach N.S.
        • Coker T.R.
        • Zima B.T.
        • et al.
        Common and costly hospitalizations for pediatric mental health disorders.
        Pediatrics. 2014; 133: 602-609
        • Odgers C.
        • Robins S.
        • Russell M.
        Morbidity and mortality risk among the “Forgotten Few”: Why are girls in the justice system in such poor health?.
        Law Hum Behav. 2009; : 1264-1279
        • Abram K.M.
        • Teplin L.A.
        • McClelland G.M.
        • Dulcan M.K.
        Comorbid psychiatric disorders in youth in juvenile detention.
        Arch Gen Psychiatry. 2003; 60: 1097-1108
        • Fazel S.
        • Doll H.
        • Långström N.
        Mental disorders among adolescents in juvenile detention and correctional facilities: A systematic review and metaregression analysis of 25 surveys.
        J Am Acad Child Adolesc Psychiatry. 2008; 47: 1010-1019
        • Zoccolillo M.
        • Rogers K.
        Characteristics and outcome of hospitalized adolescent girls with conduct disorder.
        J Am Acad Child Adolesc Psychiatry. 1991; 30: 973-981
        • Pajer K.A.
        What happens to “bad” girls? A review of the adult outcomes of antisocial adolescent girls.
        Am J Psychiatry. 1998; 155: 862-870
        • Williams R.A.
        • Hollis H.M.
        Health beliefs and reported symptoms among a sample of incarcerated adolescent females.
        J Adolesc Health. 1999; 24: 21-27
        • Breuner C.C.
        • Farrow J.A.
        Pregnant teens in prison. Prevalence, management, and consequences.
        West J Med. 1995; 162: 328-330
        • Mertz K.J.
        • Voight R.A.
        • Hutchins K.
        • et al.
        Findings from STD screening of adolescents and adults entering corrections facilities: Implications for STD control strategies.
        Sex Transm Dis. 2002; 29: 834-839
        • Conseur A.
        • Rivara F.
        • Emanuel I.
        Juvenile delinquency and adolescent trauma: How strong is the connection?.
        Pediatrics. 1997; 99: e5
        • Cunningham R.M.
        • Vaidya R.S.
        • Walton M.
        • Maio R.F.
        Training emergency medicine nurses and physicians in youth violence prevention.
        Am J Prev Med. 2005; 29: 220-225
      2. Office of Juvenile Justice and Delinquency Prevention Census of Juveniles in Residential Placement for 2011 year. Available at: http://ojjdp.gov/ojstatbb/ezacjrp/. Accessed January 21, 2015.

        • Penrose L.
        Disease and crime: Outline of a comparative study of European statistics.
        Br J Med Psychol. 1939; XVIII: 1-15
        • Gupta R.A.
        • Kelleher K.J.
        • Pajer K.
        • et al.
        Delinquent youth in corrections: Medicaid and reentry into the community.
        Pediatrics. 2005; 115: 1077-1083
        • Chamberlain L.J.
        • Pineda N.
        • Winestone L.
        • et al.
        Increased utilization of pediatric specialty care: A population study of pediatric oncology inpatients in California.
        J Pediatr Hematol Oncol. 2014; 36: 99-107

      Linked Article

      • The Cause and Consequence of Mental Health Problems Among At-Risk Youth
        Journal of Adolescent HealthVol. 57Issue 5
        • Preview
          Health care providers who care for minority teens and young adults witness firsthand the enormous impact that childhood poverty has on physical, mental, and behavioral health and well-being. The relationship between poverty, stress, and mental health is especially evident among the excessively high population of poor and minority youth in the juvenile justice system.
        • Full-Text
        • PDF