1. family and parenting
2. children

JOURNAl. DP CrllW AND ADOl.ESCENT I'SYCHOPHARMACOLOGY
VO!UH1C 15, Number 1. 2[105
<0 MLtfY Ann Li.eb!30;t.. lnc.
Pp,97-106
Psychotropic Medication Use in a National
Probability Sample of Children in the
Child Welfare System
F.~I)lesh R~ghavan, M.D., Ph.D.,! Bonnie T. Zima, M.D., M.P.H.,'
Ronald M. Andersen, Ph.P.,~ Arleen A. Leibowitz, Ph.D.,'.5
Mark A. Schuster, M.D., Ph.D.,'·6.7 John Landsverk, Ph.D.,~.'
ABSTRACT
Objectives: The aim of this study was to estimate the point prevalence of psychotropic med.
icaHon lise, and to descrIbe relaHonships between child-level characteristics, provider type,
and medication use among children in the child welfare system.
Methoels: The National Survey of Child and Adolescent Well-Being is tlte first nationally
representative study of dlildren coming into contact with the child welfare system. We used
data from its baseline and lZ-month follow-up waves, and conducled weighted bivariate
analyses on a S0111ple of 3114 children ond adolescents, 87% of whom were residing in-home.
Results: Overall, 13.5% of children in child welfare were taldng psychotropic medications
in :2.001-2002. Older age, mole gender, Caucasian race/ethnicity, history of physical abuse,
publie insurance, and borderline score6 on the internalizing and externalizii'g subBe,lcs of
the Child Behnvior Checklist were aS60ciated with higher propol'Uons of medication use.
Africnn.Americon and Latino ethnicities, and a history of neglect, were associaled with lower
proportions of mQdieatirlll u~e.
Concllls;o"s: TI,ese notional estimntes suggest that children in child welfare settings are receivIng psychotropic medications at a rate between 2 and 3 times that of children trented in
the community. This Bllggests a need 10 fllrtlter llnderstand the prescribing of p6ycllotropic
medications for child welfal'e children.
J'fhe Nali(Jnnl Center for ChUd Trnumntic Stress, University of California, Los Angeles. CalifomiEl.
20 epnrtment of psychiatry and Diobchaviorill Sdencj!s, University of Cnlifpmin, Los Angeles, CD1Uornia.
lDepnrtmenl'i of Health Services and Sociology. University of Ci\lifomin, Los Angeles, C"lifornia.
4Dl!plll·tlll!!nt nf Policy Shirlie!;, University ofC;\lHornin, Los Angeles, CillifomiLl.
5RAND. Simt<l Monico., California.
('[)cporlmUllls of Pedintrics and Health Service-lI, University of Califomla, Lo~Ang(!lc51 Cnlifomin.
7UCLA/RAND C~t\lcr rOt Adolescent HCilith Promotion, Univ~rHity uf California, Ll19 Angeles, Cnli((lrniil.
RChild ilnl:! Adl'l)clit!cnt Servi(;tl~ Rosearcl\ C'~nter, San Diego, Cgllfornin.
',ISnll Diego State UniverDity, SilO Diego, CnHfumiil.
Th!!; study WilS supported by the AgQllcy for H~ahhcare ReS(!ilfCh ilnd Quality (1 R03 HS01361l-Ql) (lnd the Nutioni'll Inslihlte L1f Mcntnj l-lt!nlth (P30 MHO 68(39)/ nmi the Centers for Di$eilse Contl'ol ill1d Prevention
(U40/CC1,JCj577~ and UoUI/DPOP0056).
97
ua
INTr,OPUCTION
.c. . .
,ULORE1'J AND Y0lJl11
in the child welfare sys-
tEm have disproportionately high rates of
emolional and behavioral problems (Clausen
et ,,1. 1998; PUowsky 1995; Halfon et al. 1995;
Landsverk et aI. 2002; Glisson 1996; Zima at al.
2000a) and mental haalth servke use (Halfon
et a1.1992~; Halfon 01 aI. 19nb; Takayalua et aI.
1994; Hannan et al. 2000; Gadand et a1. 1996).
Consequently, they may be more like!)' to be
1i'repcribed poychol:ropic medicalions. Estimates
of children in child welfare settings wHh behavioral, emotional, or developmental symptoms warranting services range widely between
27% and 80% (BaHon et al. 1995; Cl.usell el
al. 1998; Zimil et aI. :;WOOa), and 40%-60% of
childJen in foster care meet erHel'ia for at least
one Dingllo,lic and Stntistical Manual of MellIn/ Disorders, ,!th editiOll (DSM-IV) diagnoeis
(Landsverk et al. 2002). In a Los Angeles county
study of school-aged children in foster care, almost 20% had a clinical presel\tation for which
medication h""tmel\t is either of demonstrated
efficney or reeolmnended by dillica] consensus
and, of these children, slightly less tban one
half had ever received medication treatment
(Zi1l1a Gt al. 1999a).
IJU)'ing the past decade, psychotropic medica tion prescriptions nmong children in gen,
eral have increased Iwo- to three-fold (Rtlshton
aL,d Whitmire 2001; Olfson cl 01. 2002; Zito et
al. 2002; Zito et al. 2003), witl, a particularly
rapid ri.e among youth enrolled in Medicaid
(Goodwin et 01. 2001; Martin et al. 2002; Martin
et al. 2003; Olfson et al. 2003). Prescription
of combinations of psychotropic medications
has olso appreciably increased during Ibis
lime, outpacing the scientific evidence to support tbeir effectiveness and safety Gensen et 01.
1999; Olfson et a1. 2002; Zito et al. 2003). In adclilion, findings from earlier studies of prescription medication trends suggeet that girls
ancr children from Africall<American and
Latino ethniciti@6 are less likely to receive psychotropic ll,ecllcation treatment Ucnsen et aJ.
1999; Rushton 'md Whitmire 2001; Martin et
<11. ;>'002; Ollson et al. 2002; Safer et al. 20Q3;
:lito et al. 2003), but national dAta suggest such
disparities in medication use may be nUI'rOwillg (01£son et a!. 2002).
!'Qrthermore, psychotropic H1edicMion Pf~­
scribing milY vary by pFpvider lype. 01 aU
physici"" office visits that included il psyglOtropic medication prescription n~tio(laUy, 61%
were to a priJnary~care pl'ovicter, 25% were to a
psychiatrist, and the rest were to other physicians (Zito et al. 1999). This higher tQtio of pIimary,care physician visits is also seen in a<lult
samples of inctivicluals with a p.ychotFQpic
medication prescriptiOll (pincu. at al. 1998). In
a nationalmall"ged,care organization sample,
pediatricj~ns were more Ii~ely to be Ihe fh'st
prescribers of stimulant medication an<l psychiatrists were more likely to presc.ibe :;;SFJ
medication (Shatin and Drinkard 2002).
Few sh.\dies, however, have exanlined psychotropic medicHlion use among chi.ldren in
the child welfHre system, or explOl'ed how
such use vades by child.level factors or provider type. A study of school-aged chilclren in
out,of-home placement fOltnd that 16% hQc;l
ever taken psychotropic medication, and Caucasian and biracial children had very high
odds of receiving such treatment, compared 10
their Latino and Alrican-Al'ncl'ican COUl1!erparis aller adjusting for clinical neod (Zim~ et
al. 1999b). Another study of Medicaid,emolJed
foster-care youths found thaI' aO% had re,
ceived psychotropic medicatioils (dosReis et
al. 2001). Nevertheless, few definitive c"nelusiQl1s caft be drawn, as these studies were limitecl to single counties. Also, Ih~se studies
focllsed on children in foster eare. The majority of children who enter the child welfare system and receive health and ci,Ud welfare
services are, however, maintained within theit'
own homes (are pltlced "in-home"). Less than
one fifth of "Ii children hwestigilted for child
maltreatment are placed lI out-of-home"-"
placcment category th»t includes children
placed in foster csre (U.S. Department of Health
and Human Servkes 2004). rlance, while the,e
studies inCl'ease our under.tanding 01 children
in foster care, they are not rep~eseJ't"tiveof tho
overall population of maltreated children in
the child welfare system,
To begin to address this gilp, we eGtilf\~te the
prevalence rate of psy,llQtropic medication
tlee ~mong the first nationally representative
sample of children iJ' the child welfa,e sy~tem.
We do,scribe how medi~atiL1n use val'ies by pre-
l'&YCHOTROT'!C MEPICATION USElCHlLDWHFAllE
90
disposing (Le., sododemogrnphic characteris- study. These and Qthel' d,s tails reg~rl!ing the
tic~, m"lthlatment history), enabling (Le., in- design and information contained within ~h~
5uran5~, plaGoment), an<;l need (behavior NSCAW data are available elsewhere (Dowd
problems) vari'lbles, using a well-established Et a1. 2002; NSCAW ResEareh Group 2002).
help-seeking model (Andqrllen end Davidson
We used data from the baseline wove to gelt2001). We then ~~plore differel1Ces in medica- erate our independePt variaples, 'lnd clata IrQn'
tion use between children with a past-year the 12,month follow-up wave (fialded bel\vea/l
hi$tory of tltree types of provider visits-spe- OctobEr 2000 and MaJl:h 2(02) to ob!;Iin informadaHy mental health provider only, prim.ry- tion about our dependent v~riabl. of pSyChotl'Ocare pwvidel' only, or both types of providers. pic medkation use. We used data on aU children
for whon' an investigation had boen. Qjlanlid by
child welf'I'e agencies within NSCAW's 15.
month sampling tin1e frame, NSCt\.W also CORMJ'THODS
teins <111 additional sample of children in fosler
cara for approximately 1 yeer at the tim? of s.m.
Dafrl som'ces
pling. We excluded these chjl~el1 because their
The National Survey of Chil{l and Adoles- addition wO\I.ld have skewed the prev~lence of
cent WelH3eing (NSCAW) is the first notional psychoh'opics upward, and would hove limitEd
probability stuely of childrefl inveetigated for th~ generalizabillty of our findings. We al~o exchild abuse .nd neglect. NSCAW is a longitu- clvded d1ildren below 2 yeers of aBe beC"UijE
dinal panel study whose baseline "'lave sam- our indicator of dinical need is not normed for
pled 92 primary sampling unils (PSlls) spread this agE range. Our final sample consisted af
acroog 97 cUl1f1lies nationwide. NSCAW de- 3114 children,
fil'eel PSU. 'IS (laogrephic areas served by a
single child welfare agenGY. ThoGe PSUs were
genefully <ontiguo\IS with counties (although Stlldy vn/'inblss
~ome small cOlJnties were aggregated il1[o a
single PIOU), 'lnd were s'lmplcd proporlinnete DepewtCllt variable. From NSCAW's l~.month
to the size 01 the child welfere pflplJlatiQn resi- follow-up data, we extrected the v~riable of eUl'dent within tll011\. The origin~l design Qalled rent llS~ of psychotropic medkiltion. N6CAW
i"r s~ll1pling 100 PElUe, but child welf,.re regu- defines this varialJle flS " a p;rescripHon or rrt~ci.
1~lie!1s within 6 of tho.e PBUs proeluded Sl!f- kation for emotional, behavior"l, Jenrnii1g, iltvoy edministmtion, resulting in the lin"l sall'ple tanlional, or sub&tance..abuse probllJm~.u We
enosc mrrent use in order to be (""I'taifl that the
0192 PSUs.
Within each VSU, c:hildren who were sub- provider visit precede,! io time ti,e consu.mplion
jects of child al;llIse and neglect investigations of the psycl,otropicmedicalion,
by the child welfare system within a 15·month
time perio(i, beginning in October 1999, were II/dependent varialJles. Prc'liSl'o~illg cI"JI'ncleris/ics
sample~L These children were aged from birth iflcluded child age, gender, 'lnd race 01' ethniqto 14 years (nt the time flf sampling). The ma- ity (extracted from NSCAW's ("r"giver interjflri.ty of theae children (87%) resided within views), and maltreatment history (extracted
their OWl' homes, and Ihe rest were placed hom NSCAW's child welfare worker intel'out-flfch~l'"e (a celegory thet inclllded ehildren views). Categories of phy.ical ab\l~a, sexl,lal
in family fo.ter m'~). Dftta within NSCAW is abtlse, neglect, al1d abandonment were diobt~ined from interviews of c;hildren, tl,eil' chotomized stich th~t a chilel cpuld have more
cftreqiVel'6, their ~hild wellare workers, and Ihan one type of abuse coded.
teechers. Unweighted response rates were 84%
Enabling "esOill'ces Induded child plneemel't
for child weHQre workers, 'lnd 72% for pri· stalvs (In-home Ver5\1S out-of-home), as idenmary caralliv\ll's. NSCAW ll.e~ w~ighting pro. tified by the child's welfare worker, and the
cedure~ to enBure th~t reopens" biases are inaurance categol'ies of PLlblic inSllrnnce (Mea·
n>iniln~l ~nd do not affeel findings fro]" the ic"iel), pl'ivate insurance, federal (CI-lAMPUSl,
ll!u
[tHd uJ,1if.l6ured obtained {l"Om caregiver interviews. Beca1)Se only 39 children had a lederal
health insurance policy, and because fed.el1l1 in~l1rMCe programs resamble private insurapc:e
Hlpre than lv!erlicaid, we grouped t.'lem along
Wi1h privately insyred chjlqren into a single
Q'ltflgm;y of "1'rivate+.J;iedera)."
()~~ 11~et/ vnrialJlc w~s t]W idenl:ificotion of a
pl'qqilble beha'liior~l disorder if th~ ohild had
SCQlq~ in Ih~ ~QFdeElllW clinical ran!;e for an llltcrnij!jzing or [email protected] J:;eha.vior ]1lUplem,
as p~fined as a st~ndilr<ji2ed SCD"e a!Jove 60, on
Uw respective subscale of the Child Dehavior
Chec\$!ist (CDCL) (Achenbach (\<,d Edalbl'Ock
19B5). The CllC!, is a well-established sq:eening
measure of childhood behavior problema that
h~s 1)e\!II succe~fully used in fosler-care child
"aJ,l1ples (Zima at a!. 2000a; Garland et al. 1996),
and has modefiltaly high reliability and validity
(Adlef1bach ~nd Edelbrock 1993). The CECL
is norm-l'l!ferencod fQr large pQpl,llaHons, .nd,
All analysEs were performed in versio[l 8 of
Stata (StataCorp 2003). This secongary dola
analysis was conducted under a eertifi,ate of
axemptiSln hOUl review obtaine(i from OU.f institutional Review6oarct.
RESULTS
Sample alwac/erie/ics
Children in the sarllpJ~ rom.! " me.'l ~ge of
8 years (95% cr, 7.7, 6.2), 48% (II '" 14?9) W6r~
boys, 41% were Caucasiall, 26% Wel'6 Afrienn.
AnWl'iCiln, ar,d 18% were Latino/Latina (all
percentages are w~ghted). Physic~1 ~b\lse W<l,
the most frequent type 01 abuse experienced
(33%), followed by neglect (17%) an<;l sexual
abuse (12%). The mo.t hequent inourance covel'age was Medicaid (62%), and IDt'st chHdren
were liVing at hon1e (87%). On Ihe CnCL, 39%
therefore, socioeconomic Gt~tus and race have of children had a score above 60 on the exter.
litHe effect On the measured scoring indices nalizing scale, and 28% of children had a scare
above 60 on the internalizing SCale. Fourtaen
(Achenbach <lnd Edelbrock 1983).
p~reent of
Pl'uvidel' vieil vnl'illble. From NSCAW's Caregiver interviews, we obt"ined data on whether
the ehild h~d "aell a specially I1lenlal health
pXQviq.er (payehintrist, psychOlogist, il psychiatfic !'IUTS~, or a psychiatric ~ocl.1 \VQrk~r), a
nQJ\sl"eci.lly provider ("family doctor or any
other m~dica! doctor"), Df both types of provi~~rs (or an "emotiorl'!l, bahayipral, looming,
attention.l, or subst'!nce abllse p~oblelll" In
the 12·month period following the baseline interview. Childrell who had been to a " ... mell.
lal health or CQmllllUlity health center for any
of lhose problems" wem also cl~ssified as having sOl.lght speei.!Iy care,
A.nalysQs
Data were weighted to "ccount for the sampling design !hat lllVolved stratification and
~empling wHl1in PSUs (stratum and PSU
weights), as well '16 probability of selection.
Hivari.te analyses were conducled using
ehi.square test. for cfitegorJcal variables. All p
value~ were calculated en design-based F tests
tfl<lt eccounted for the w~ightil1S'
the childl'en saw a ~pecially lnfJl1t,,1
health provider only, 3% saw a nonsp~el'1Hy
n<ent'll health provider only, and 5% Saw bOI.h
e specialty and nunspecllllty proviaer for a
mental health problcm.
ClJo1ractdristics of children Inkirlg
llsyal!Otrpl'it; medir:aliol!s
A total of 13.5% of childrel1 (II '" 473) were
taking psychotropic medication ijt NBCAW's
12-month follow-up (Table 1). 111~ mea" age
of children Qn medication was 10 l'earS (95%
Cl, 9.5, 10.5) while that 01 children who were
not on medication was 8 years (95% cr, 7.3,
7.9; I' < 0.00'1). The proportion of medication
use alUong children aged betwaen 12 om! 11\
years was seVell Hmes gl'eater than that
among children aged between 2 to 5 yem's.
Male gender, C"lcasian race/ethnicity (f "
19.7; P < 0.001), history of physical abuse, Plll>lie insurance status, and having a soore nl;1ove
60 on either subscale of the C8CL were significantly associated with medication use. Afr.icanAmerican (I' " 6.7; P " a.Ol) and Latino
race/~thnicity (F '" 9.0; fJ " 0.003), hi. tory of
I
FSYCHOTROPIC MEDICATION USEIq-IILD
W~LFARE
101
TAUL~ 1: CHt\II~CTEIUSTICG OF CHH,D WeLF.\R~ Ci·J1LDRGlJ UY MBmCATIt.'N STATUS
V,lri/lulc
"iilllC
'n,Ft;U8 p:oyr:lmlrvpic J1tfdiC!lliOlI
(1l11?1..'dg/ltl!J IIJ
WdSh1Cd ~i.
tnkiJJ8. illl!diCill!qiJ
oj'iii:i;,
p iJ"!"~
No
204.1
13.5
3.6
1M
21.p
~B,OOl
160
901
1153
571
308
165
1171
1470
19·.
7.7
<B.OIl!
115
273
49
35
7P6
9.2
ll~l
1~.2
')63
197
14.,
760
'145
411
11t
19·6
13.9
3,3
15.6
<0.001
0,9
O.O·l
0.9
385
73
15
1763
6;30
16.0
0.007
0.2
~3~
5,4
Q.Q~
c:llll'9rr hnJ11e
F8
(!'pR
'195
603
2036
la.l
In~Jll:m,e
011 tlw Qxl/.!vnali7.illR
350
8R3
:<?2
<O.pOl
1,ll;1l1~
>.60 Ofllhe illtofl1i1lizii1g
265
u60
30,8
<0.001
190
396
1).6
1.1-2
'13
62
'f).(
<0,001
<0.001
69.6
<O·POI
Yes
Toti11
'173
Pr~dla/ltl5j'lg. Cl/t11'llOlm'~tic;s
~g.
Gl)lId
<~~
¥u"rp
6-11 y~prs
12..,16 ¥~ilrs
·13
265
O.O~
fJ.Q02
Chihl B~n~ur
~qr
Girl
C;Nld J."qlle{ethniclly
~lt..friC;i'lJL-AmedCill\
Whit~
Li\tin9/~l!tinLl
Olhur m(!~
<0.001
7.•
E;1If1fliillg Resource;s
lyp9 of .hu~.
P.h}·~i~1tl
181
S~xu~l
84
Ne31Q~t
Il7
Atmndonment
~fi
lYp:,: nf intmHll-\C;:;t!
rl.lblic
P.r-ivi'lte + (i~eri\1
U~1hl!ll.lr~d
In
PI{lCel\l~nt
19,.a
IV",'
E;aC~"Q,"
~60
;cillc;
l?rijuj,/l?f T1!I!e
Fi:I!~~ Y~i\r visit to
!:jtJ~qiij.lty pWllldt:r
Ntm1i:E1~d;llt}r
prol,.'idcl'
O\lth typt.:ti of prllvlder!)
139
Obsef\'a.1inn.'i with incmnpl\!lu dittil wtlrf! not '!l\d~lded in bi\'afinte analysc~, and 50 N~ may vnry :ilightly ilcrC)~il'
tWP-WilY ,milly~cs. CfJeL, Child l]ohillJlor Checldi:;t.
lU~
nQglecl, and being uninsured were negatively ensure that the pyouidcr vjsit pr~<;ec\ed rnecl~55ac~4la~ with mQctic~tion use. Histories pf italian u&e , we focused 9n c.uf)!ent US~ of
llrovid"r vlsitG were equally significant in psychotropic medication ,.t the li!J'le of the
1)@ing a~SPGiati!d with psychotropic medica- follpw-up interview. Our eSlima!e is, tlwrefpre.
tion ].\se, olthough a l;Ieater propol'Hon of co- an estimate of p01nt prevalenee, It is likely that
nWlwged children (i.e" those who saw both many more children in the child welfare sysQped",liy "-tId nonspe,ially pmviders) were tern will consume psychotfOpic medieations if
measured over a period of time.
taldng p-ay,hotropic medications.
ijBciiUl:!~ our clat~
CIWl'nctarist!cs ~f vlii/(lrel! t<lkill!,
/J1vdicqtlQIIS by prD~id6r typc
Table 2 shaws the cherndetistics of children
t~ldng l'sYcf\l!ltl'()pic medieau'm by type of
pro"idor visit, A gre'lter proportion of sexually
.b\lSe,;! children taking psychotropic medi,alions had seen spcdally provic\ers or had
cCllnal,agpc;l visits (68 0/0) than nonspedalty providers (3.2%). Nearly twice the proportion of
<:hildren taking psychotropic medications who
!'"d " berderlille clinical SCOre for an extemali.ing disor~~r were comanaged (26%) when
compared to chUdren who had nomla) .,ores
on that subsca!.e of the CBCL (only 13,9% of
these ehUdren \Vore co,nimagecl; p =0.0~9) (not
shown in rable 2). Children with a bOl"dOl'line
c1inic~1 s€ore for an internalizing di.order who
W@l'e taking p.ychotropic medica !ions had a
gri'aler praportioll of special~J provider visits
(4!1.6%) compared to children who had normal
sCl21l'es on thilt 9\1bscaJe of the CBeL (Z9.8%; I' =
0.04) (llOt shown in Table 2).
P!§CUl.lSION
Thi!i ill the fjrst ,t\\dy to dascribe the prevalertce 01 psychotropic lna,ifoation use in a national sample of chilchan in the child welfare
system. ~tl.tdies that hove analy.ed statewide
M~dic~id sample. have generally r~ported
prevalence IiI!es of 5P/~6% of psychotropic
l11eclie.ntion !lse among children (Martin et al.
2003; Zito at ill. 2003). The rates of medicatioll
use aJ;llOllg commerciaUy Insured children are
eVBn lewer=approxirnately 4% (Hong and
Sh~pherd 1996, Stein ~t .1 2001). Our l'Ate pf
1~,~%, thl1s, represents • two- to three.fold
grc.ter prevalence 01 psyeRl'trppic medication
\IBe among children being treated in the child
wellore ijysteln. Also, b"cau5~ of our desire to
sot hlelc:s a ;nWijSlu'a of l~ead
for psychotropic medication ljSe (such Q5 cl.l.qljnostie in/ormation), we are unable to cQ!mUeot
Oil the appropriatene:;s of such use. It is well
IInown that childl'cn in child welfare ore a
population with high needs for mental heAltir
care (Clausen at al. 1998; Halfon at al. 1995;
Landsverk et ill. 2002; ZirJla et a1. 2,OOOa), ancl
may have greater access to servkes thro\\gh
tire efforts of their assigned chile;! welfare
workers (caseworkers). Our findings, then.
may be a reflection of the ability of child weI.
fare systems to obtain mental heallh services
for children under their care.
On the other hand, there llWy still be high
un met needs occurring ameng this populatiOn; 28% pnd 39% of o\\r sample hod berderline dinkpl scores on the inteVMlil'lng and
extefrtfilizillg subs,ales of the CBeL, resp.clively. SOIDe of these children might h~ve disorders requiring medication. A11~ given thnt
2imp et al. (1999a) found tlla! move than half
of schooh.ged fostar care children in a cOllntywide sample who h~d a psychi~tric <;\i.oni"r
that merited a m~dieatiyn ev~luatiQn had not
r..caived medication, it '~~J1ls deal' thijt child
well~re children have Bignific"nt b.rri9l"S t~
medi.cation use. Further rcsa..cll thilt incQrpoPltes oiagnosi. and othol' clinical measures pi
need is required to better lInclerBt.]1<,j medk(l.
lion use among these ehildrell.
That a greater proportion of older ehUe\~en,
boys, and those of Cauc.sian Msc~/ethnicity
receive psychotropic 1l1ec1ieatiQrtS is consistent
with stUdies examining this ]ss\\e "mong le.ter
children (Zilna et aJ. 1999a), anll amOllg children receiving sel'vices throllgh pullJiE"sector
Agencies (Leslie at aJ. 2003). OUI' finding of
nleoic"tion use ilU10n!;i a greater propOrtioll of
physically abused children and a lower proportien Df neglected childr@n is ~1.Q eonsi.tellt
with studies that r~port greater seFvi<;c lise for
childrQIl who ilI'e "acCiveiy" l'l'Ia)tre~\~d (physi.
I'SYCUOTROjJIC MjIDICA1'lON USE/CHILD WELFAHE
Only Bl!cn"rrlt,V
H
(ljI,;igl'lerl %)
103
(lilly 11D/I~spq~inlly
II (weigh/cd %)
Drilll Jypi:f; !Vl'roai;Jltl's
II (Ndg1tIt.:~1 %)
'12 (10.6)
'POTAL'
[f,r.Ji,.1i~pp;;;iIJ~}
Cllnmr.tarfsticB
€hild 0BQ
12.,..16 yl3oP;
Ohil~1 gl.mdl:r
aD)'
c;irl
Child J'f.t,o/ cUmit;ity
Af¥lci\n"An-1eriCtltl
White
~a.HnQ/Li;lthln
Oth~l·l'ilCj.::
11 (a.,Il)
M ("3,9)
6 (13.9)
27 (13.3)
9 (M)
~9 ~O.Q)
~ (f6.1)
120 (42.71
70 ('10.6)
26 (9.5)
16 (1(3.2)
9p (~~.O)
46 (10,~)
46 (39.1)
lOA ('10.5)
22 (53.8)
17 (-)7,3)
10 (3.7)
28 (12.3)
2 (4.3)
2 (9.5)
32. (;11..8)
S3 (W.n
1~ p8.6)
10 (~a.o)
11 (30.7)
11l (42,3)
2...-5 Y;l1J:~
(j...ygiltB
';//(lII/ill8 Ri:5QmW5
Type of .,,~"~
80 (48.8)
35 (51.0)
18 (5p.7)
13 (37.0)
Phv,i"i
SC~1,Ii1l
Nesiecl
Ap.t1ndonnl tmt
10 (6.5)
8 (3.2)"
3 (6.3)
1 (<1.3)
54 (2M)
26 (17,0)
H OU)
8 (37.0)
Typo of h)s\lrnnq~
l~l,lb.lil:
Frivute -1'
F-GcI~n;1
U~liJ1~;urec!
157 ('11,5)
27 (,3,3)
6 (4d.B)
30 (11.6)
ID(M)
2. (<1,6)
IH (20.8)
., (2M)
3 (35:7)
71 ('lD.o)
12 (12,6)
30 (10,2)
6? (26.1)
'1-7 (~;!.~)
26 (10,2)
meW.7)'
20 (9.'1)
g, (42.3)
f1l<Hl~~n~t1t
e)~l.af~home
In:lwme
119 (4H)
Ne,d
CO<;J. "~'''~
~60 tlf\ th~ l!l{tf.!rn~lizi.ng
f/!:!illc
::..t';i) Qll tIlt.! inl§I'llilli7j!l1g
12(, (46,6)'
:ical~
'" <0,05,
~t<:a,Ol.
1"fhol!e pUl1ccntl\ges do nc}t ilcld \.lP tp 100 dlher b~CiltlSe !inme children werl! til king p~ycholropic m~di(mti()ns prlr>r
to t1l'\~fY inll;l NSCAW, Clr becnu!ic they reht,!iv£;d their lllcdic~liDn5 d\.ll'ing the course of n hospililli~Ziltlun fi'lthcf
fln (!Tnb\.llilt~wy visit.
~"!ly
0, sexually abllsed) r~ther
~hlln
than "pas- fOUlld-like Ollson et ill. (2002)-that a smaller
sively" moltrealed (negie~ted or obandoned) proportion of uninsured chHdren (5.4%) we"e
(G~rlUIl<;l ot 01. 1996). While previous reseilr~h lilking psychotropic rnedic"tions Whell comh~s shown little or no itnpa~t of insur"n~e on p"red to those insured (14,4%; p = 0,02). Most
U68 of ambul~to<'y mentill health servi~es of the research cited above relies on d~tij cDI.
.moll!! children (Glied et ill. 1997; Giled et ill. leoted prior to 1995, and it is likely thilt our
1997; Burns ot.1. 1995; Burns el ~). 1998), we data reflect changes in the health.cilre nJ~J:ket-
place bog;,nning in the li)te 1990s, such as the
penetraliQJl of Medicaid managed care into
chil@,aerving systems.
!nterPl'ctation of these li"c"\ings should take
i'1to account several ]jmitation6, Partly owing
to small sample sizes. our shldy cloe.not per,
!"'1it
to identily mallY associatioFis between
type of proviger visi t and 1.\5e of psychotrop.ic
medication. Our prejinlin""y finding; that a
greqt~r jlFoportiQn of ohildren with a history
pf sexual ijb1.\se reeeive speoialty or ~omal1­
aged C;lre is conaistant wi,l:h lite,at\lre lilat
suggest. the exlstel1ce 01 defined referr-al pathways tOWilrd speciil1ty care for slloh qhi!<jren
(GtU"land 8t al. 1996). Similar pathways 1M}'
Op!rrate for children with higher behavi9,M
~ympto!11 soore•.
Als9. bec;lUGe our provider history is based
lIpon paront and caregiver report, we do not
have a WilY to determine with certainty whioh
eat~gory of providQr prescribed the psychotropic to Ire child. The lack of diagnostic information limits discussion on the appropriateness
of Cilre received by U'e children in Otll sample.
u.
Finally, our data set cloes not contain informi;'\tion on the type, d9sage, and duration of psychotropic medic~tions conGumed by th~ child,
£Wthe!' limWng conolusions regarding approp.iater;e6~,
CONeLUSlQN§
David Goldin, P~ltSY \-Vood, rJnd Iinj!n .?ll~I1g
for their assistance with data Illilnil~emen!.
The Caring for Children in Chikl Welfare
project (CCCW) is a collabol'aHve ,eff9rt between the Child and Adolescent £@r~i~os Researoh Group (CASRC) at Chilclr-el"s Ho,pital
San Diego, California, the Devwtment of Feycmalry at the Univer-sity of Pitta!JurgJ> (E'iHsbUl'gh, Penn,ylv~nia), ti,e C91\,lp.,IJu;; Children's
HQ~pital. Colmubus, Ohio, the r,;pi>!emi,,19gy
and Services R05ear;:h Group 'It D\.1ke Univer.
si~1 (P~ke), Dill'ham, Nmth ea,oli;10, aM the
Research Triangle lnstito te (RTl), D\lvham,
North Carolina. 'n,is study wae .[unde\!! by the
National Instihlte ofM~ntaJ Health (1vIH59672).
Aeomplete description of the sludy and a list (if
key personnel are available at WWW.casrc.org/
projects/CCCWlindex.htm.
It should be noted that this ,locum~nt ijlso
includes data from the National Survey of
Child and Adolescent Well.Being (NBCAW),
which was developed w1der Cont.,lct to RTI
from the AdminIstration on Chilc\.ren, Youth,
and Families, U.S. Department of Heallh ,mel
Human Services (ACYF/DHH9). The r:.CCW
also maintains ongoing collaboration with the
NSCAW Research GrOll p.
The informotion and opinions eKl're~sed
herein refleet solely the posiliol' of ihe author(s). Nothing herein should b~ ~t'J11sh'\1~~1
to indicat~ the sl.pport or ~ndOl's~m~nt of Hs
content by ACYF /D1?IHS.
In cQ/Wl\lsiol1, tIl!. stuc!y presenls the til'st
!(IlFERtlNCES
representative estir[l~tes for p~ychD­
tmpie l'f\edic~tiQn uae among child welfare
Achenbach TM, Ildelbrock C: Manual 101' the Child
pop'11aHotls. FI1rther rosearch examinin.g the
Behavior Checkli6t and Revi6cd Child Behavior
types of psychotropic m~dications pr~scribed
Pl'olile. Burlington (Verm9nt): University of Verto 'children, the diagno~lic indic~tors for sueh
mont Department of rsychiatty, 1983.
f)resel'iplion, and the characteristics of the pro- Andel'sen ltl",!, Davidson PL: 1111prQving aCC12~S 1c.
Ci!l"a in America: Inclividt.tal EPlt.1 eOl1textval j':l1di..
vider doing the pl'escribing is nee,ded to deental's. In: Chnngil'l!> the U.S. Health Care Syolelll,
velop p.'ogroms to improve the quality of cal'e
2J1a cd. Edited by Andersen RM, Rice 1'1-1,
for this highly vulnerable poplliation.
Kominski GF. San Francisco (Cali(ol'!1i~)' JpseeyBa», 2001, pp. 3-30.
Burns B), Costellp Ill, AngQld A, Tweed D, Stongl
0, Farmer EM, Erk.l~ni A: Child"en's IlleUIQI
,.,CKNQWLlJDOM'pt.JTS
health ,ewice u~e qcl'o.' service ~ecIQfS. Neillih
Aff 14:1d7-15~, 1995.
.
Tl,e alltl'OfS !oll'otofuliy acknowledge Roshan !lulnsll}, CosteUo E), Ilri<nnlj A, Tweed DL, F.rm~r
a~stiUli, Ph.D., for her helpful commentg durllM, Angola A: InsPI'.nce coverege a'ld mqntal
ing the cQnc~p\llaliziltion of this work, and
he~lth .ervice use by adolesrent, with seriOlls
!I~ti",nillly
P-SYCHOTROl'IC MEDICATI8N USI;/CBILD WELFARE
'IQn
amotional !Jistulbance. I Child Fam Studies Land.verl< I, Gllrland AF, LeBlie LK: Me"tol he.lih
6:89,-111,1997.
selvices for .hildren repoFted to Child P~ot~c'tive
Clallsen 1M, Landsverk I, Ganger W, Chadwick D,
Sarvjces. In: APSAC Handbook on Child MaLLilt'ownik A: Ment~1 health problems of children
lre.tm,,,!. Edited by Myer. IEB, Berliner'L,
in f~~lor CilIa. JChild Fam Studies 7:2S3~296, 1998.
Driere I, Henclrix CT, Jenny C, Reid TA. Thou.
do.Reis 5, lito nvt, SilfeI DI, Soeken KL: Mental
sand Oak. (Californio): Sage PllPlicutio,,~, 2QO~,
henlt;h s~vices fDr yQuths in foster c:are und dis·
1'1'.467-507.
nbl_d ycu~hs. A,11 I l'u"lie Health 91:1094-1099, Leslie LK, Weckerly I, L',nct,verk j, MOllg!> RL,
:l001.
liurlburt MS, wooa PA: Rnciili/etlmk differOllee
Oowd K, lSin,ey 5, Wheeless S, Thissen R, Richardin the usa of psychotropic jr,.diention in highs~n L l\.1if!r~wp. F. Riemer P: Na. tiQn~l Survey of
dsl< chijdren _nd odoleocenl', I Am ....e.a Child
Aclole,c Psychi.h')' 42,1433~144b :lD03. . .
el1ild and AdQles,~t Well-l?eing (JII"CAW), IntrOciliotiOl\ to the Wava 1 Ganer~l and Restrictad Martin A, Shcrwir. T, StllbbJ! Cl, V~nHoor 1~ S••hill
u,~ ~ale.6~s. IIhaoil (Now York): NatiOJln.l Dilla
L, Leslie D: Use of U\llltiple psy.hQlropie drug~
Archive on Child Abu.e and Neglect, 2002.
by Medic~id-insured al]d p,ivijtc!y illsured chi.!.
G~,land AF, LandsvGrl< I, Iio,-\gh RL, llllisdren. Psychiatr SelV 53:1508, ~O(l~.
ManLeod E' Type of n"lltrefitl1lanl 'lS _ predictor MartinA, Van HoofT, Stu1;lbe 0, Sher-win T, Scahill
of m~ntal hoallh service uoe lor children in fostor
L: Multiple psyellQfrOpic pharmacotherapy runoqg
Car•. Ch!ld Ab\,se Negl 20:675-688, 19V6.
child and adolescent cnrQl)e~;! in eUJll\C~liGl~l
Clied $, Glln.tt AI3, Hoven CW, Rubio-Stipec M,
Mecticil.id managed eare, P.~Yf;ttiptr Su~rv 5f1;7~-77,
2003.
R~gi",' CIA, Moore RE, GOodman S, Wu P, Bi,'d H:
Child oUlf"tient manto I henllh services tlse- NSCAW Re.earel\ Group: MelllodolQgic.l I."on,
Why <10.8n't i!15unmee m.tlev? I Ment H_filth
from the National Survey of Chilcl "nd Adol.,"
llohey Eco,' 1:'178~]B7, 1998.
eenl Well-6eillg: The fir.t Uwe yea,o of U\e UBA's
Colied 5, Hoven CW, Moore Rl), Garrett AB, Regier
fi!'st notional probnbility ol\tcty of ehildren and
DA: Chil<;iren's tlCCeS5 to mental hanHh Cflrefamilias for abllae and negleot. Children Youth
Doe. in,m'lI1c. !11ntter? HenHh Aff 16:167-171,
Serv Rev 2'1:513-5111, 2002,
1997.
Olfson M, Marcus se, Weissman MM, Jensen, PRo
Gli!1tlol'l c: Illdiciul flnd service dcdsiollS fC)1' chi!..
NotionElI trends in thB use of psychotropic l11od~
d1'ol1 entering stC1t~ custody: 111e lirnited role of
ieationa by child!'a". J Am Acod Child Adolesc
o\.ntq) health. SQc Sd Rev 257-281, 1996.
Psychiolry 41:51<1-5:/.1, 2002.
Goo(!will I{, Gould M6, 61'\I1co C, 0lf50n M: Pre- Olr,on 11.1, Comeroff MJ, Milr.us Be, Je"G~n 1'5: Nil.oriptinn of psychotropic me(!ications to yOtlth,
UO,nal trende in the treatment of attention delleit
in l1Wea-pa'ed practice. l'~ycbiatr Serv 52:1081hyperactivity disorlier. All\ I I'Byehi!11ry 160: 10711087,2001.
1077,2003.
H~lfon N, Berkowiz C, lOee L: Children in foster Pllow'ky D: I'sYc)lOpatho!~gy a,nong ehildran
care in ~ill1fot'nifl: An e:r.QfTlil1f\ti(Ul of 'Medicaid
pl<1C~ec;l in (ilmily fopter ~il(C. P6ychi(ltl' ~el'\t '16:
reiI11bul"tisd hEillN\ SQ1'VICCl'1 utiliziltion. Pe~jntricB
~06..;J'lO, 1995.
69!1:!SCH2S7, J992•.
PiOcus FLA, Tanielian TL, Mare". Be, Olfson M,
Halfnn N, llefkowiz G, Klae L: Mentol he"lth serZarill DA, Tho!'11pSOn ], Zito 1M: 11relier-ibing
\'lc~
utiliziitlon by chi1dr~n in fost~r cQlle in Cali·
trends in psyehotropic illeclicnticms: Pr.nlQ.IlY Ci\!'O,
fQfIlia. Peoiatrlcs 89:J238-12d', 199~b.
psyehiillry, iInd oJher medieal speaialltie~. I Am
Nellon N, Mendonca A, Berkowitz G: HeaLth staniS
Med Assn 279:52~531, 1998.
of children i!1 foster cafe: The experience of the Rushton JL, Whitmire JT: Pedifltric 6limulillit nnd
Center for the Volner,ple Child. Arch Ped Ado$alective serotonin rf;!llplil)<e inhibitOl' pre.!icrip..
IGSC Mad 149:386-392, 1995.
titl11 ~!'ends. Arch lIedint, Adolesc Med 1~§:560­
§6~, 2001.
HOfl11n" Is, Childs Oll, Kelleher Kl: Ment"1 he81th
1!iJ,l'e utiIi~i!tiQn und C:l;.<pcmditures by children in Safer PI, Zito 1M, doaR.i. S: Concaplitont psyeho,
foeler core. Arch Ped Adole~c Med 151:1114-1117,
tropic medication for youti". Am I I'sychiat,y
100:438-449,2003.
2000.
I-rong SH, Shepherd MT:l: Psycho.ociel nnd demo- Shntin Dr Drinkard CR,: Ambulatory u~e of psy·
gNphie predletme of pedirotric psychotropic
chotropics by employer-Iv.urod children ~nd
medication use. Am I Heellh Sys I'h.,'m 33:1934adolescents in n. flilUonill managed c:~re erganiZil·
1939,1996.
lion. Ambulatory Ped 2'W-119, 2002.
jensen PS, Dholera VS, Viliello 13, Ho"gwood K, Feil Slein 130, Sturm R, K"ptlr K, Ringel j: Poychotropic
M, l3Ul'ke LI3: Psycho"ctive nlocliention prescrib.
mediciltion cost::; ilmong youth with private Inlog pr"clkes for U,S. children: Gop. betwaen ,eSUrance in 1998. lIsychiatr Sa..v §2l15~, 2001,
search nlld e1inieal pmetice. J Am Acad Child StatilCor!'. Stala Stoli'ticoL SoFtware: Relea.e B.O.
Adoles, l'syehiotry 3B:557-555, 1999.
College Station (T"-~ij.), St~t. Corp, 2003.
IQti
Tn};:ilym::n¢; JJ,
B~rgman AB, Connel FA: Childnm in
(Oliter Cijre in tne state of Waffiington: Health cnl'(~
~Uliza~ion and ei<penditures. j Am Med Assn
RACHAVAN 1)"): AL.
Zi!o 1M, Saf,," Dj, dosReis 5, M.gd~r LS, e"·.;4wr
IF, Zarin D,A: l"ychetller;'pe\llic medk~tion pal-
terns for youth with attention deficit hypcractiv~
271l165Q.,1855, 199~.
ity disorder. Arch Pedialr Adolesc Mec,l15~:1257­
1263,1999,
U.s. Department of Health and I-Iuman Services,
Administration for Children nnd Fmnilies, Chilo Zito IM, S.fer OJ, doaReis 5, G.,dner )F, Magder L,
Iiren'slJurenu: Child Maltre.lnllmt 2002: Reports
Soeken K, Boles M, Lynch P, Riddle MA: l's.ytho.
tropic practice patterns [or youth: A la-year perfrom tho Stntes to the N.lionnl Child Abus. "nd
Nagl.ct D.ta System•. W"shu,gton (DC), 2004.
<ipeetive. Arch Pedi"tr Adolesc Med 157:17-:U;,
2003.
Avail.hle al http://www.•d.dhhs.gov!program./
cb/pllbli,ation.!crr.OUCm02.pcll. Accessed on ZitQ IM, Siller Dj, dosRois S, G"rd.llor JP, Soa1.;en Kj1Jly 8, 2004
Eoles lvl, l.,y.llcl1 F: RisiI\g preva],,,"" a[ ""tide'
21ma BT, Blts,ing R, Cr.O\!lius GM, Kauftn.n A,
pressM!, amOnt> U.S. YQuth~. l'edi"triQs 10~:7~1­
727,2002.
Bali.n TIl: I'Syq\otropic medk.tion UDe alaOng
cl1il~ren in foster Cn(.: Rel.lion,hip tv SeVij1e
paydljn~ic dtsOPd~r•. Am j Public l{e~1U1 89:
l73"-1¥~E, 1999q.
.Ad(lress l~print rf:qtlest~ t(J:
2111.'0 ll'l~ BIlSeb'!l R, Cr~c~il"s GM, K.ulm,,,, A, Delin
Rmlles"
Rqg/iavOIJ, M.O., PIl.l).
1~: PsyehQh.'opic medhmtion lrellBnant patterns
Tile
Nationnl
ConteI'
for
Cllild 'Ii'al/malic SIress
.l'nQJ1g .<booh.ged children in [osler care. j Child
UCL4
Ncuropsycl/jQlric
Institute mid Hqlip/lnl
Adolesc Plo/<ihQpha..maeoI9:1~1il7, 199%.
Zil'nn Wf, Bussing R, Freeman S, Yang Xiaowei,
11150 WOlympic Boltl.val'd
13e.1ln TH, Forn~8P Sr{: Behavior problems, a,a~
Sltit~ 650
clami" ,kill ~.lny. and school f"ilur. among
~DS A1l8~les, CIl ~0064
aph~"l·"g~d ,hildr~n in [o,tar care: Their rei~·
tlollShip 10 pl.~ern~nt churact8!,iatics. j Child I'~Jll
S111Qlo. 9:B7-103, 20000.