Primary Prevention Mental Health Programs for Children and Adolescents: A Meta-Analytic Review<br />Paola Arteaga<br />Iván Barragán<br />Helena Blanco<br />Beddy Elizondo<br />Norma Garza<br />Laura Gómez<br />
Primary Prevention Mental Health Programs for Children and Adolescents: A Meta-Analytic Review<br />Primary prevention is not a new idea.<br /> Although interest in primary prevention has been present in this country, translating this interest into effective action is more difficult.<br />It is not know exactly how or when currently healthy children eventually develop specific psychological problems, making it difficult to plan interventions to prevent future specific dysfunctions.<br />
Primary Prevention Mental Health Programs for Children and Adolescents: A Meta-Analytic Review<br />Many researchers have widened their goals beyond the prevention of specific disorders to include the general modification of emotional and behavioral problems.<br />It is important to document that the intervention has an immediate positive impact.<br />
Primary Prevention Mental Health Programs for Children and Adolescents: A Meta-Analytic Review<br />Preventive interventions may also seek to enhance positive behaviors or features of the environment that lessen the likelihood of negative outcomes or increase the possibility of positive outcomes.<br />An enhancement model assumes that as individual become more capable and competent, their psychological well-being improves and thus they are better able to withstand or deal with the factors or influences that lead the maladjustment.<br />
Primary Prevention Mental Health Programs for Children and Adolescents: A Meta-Analytic Review<br />Important issue is the relationship between health promotion and levels of maladjustment.<br />Primary prevention in mental health may be defined as intervention intentionally designed to reduce the future incidence of adjustment problems in currently normal population as well as efforts direct at the promotion of mental health functioning.<br />
Conceptualizations of Primary Prevention Programs<br />At a broad conceptual level, two major dimensions characterize primary prevention:<br /><ul><li> The level of the intervention.
And the way population are selected for intervention.</li></li></ul><li>Conceptualizations of Primary Prevention Programs<br />In terms of the level of intervention, programs can be categorized as either person or environment centered.<br />Person-centered programs work directly which children and often use or adapt change techniques drawn from the clinical and counseling literature, such as social learning and direct instructional approaches.<br />Most current environmental programs have sought to modify the social context of the child’s home or schools situation.<br />
Conceptualizations of Primary Prevention Programs<br />Ways population are selected:<br /><ul><li>First, in an universal strategy all members in an available population receive the intervention.
Second, groups considered at risk for eventual problems, but who are not yet dysfunctional.
Final, those groups about to experience potentially stressful life events or transitions.</li></li></ul><li>Method<br />StudiesReviewed. Studieseligibleforreviewconsisted of primarypreventionoutcomestudiestargeted at youth: ages 18 orunder and reportedbytheend of 1991.<br />Primarypreventionwasdefined as aninterventiondesignedspecificallyto reduce thefutureincidence of adjustmentproblems in currently normal populations, includingeffortsdirected at thepromotion of mental health. <br />
Method<br />Tobeincluded in therevieweachstudyhadtomeetthefollowingcriteria:<br />Adheretotheprevious concept of primaryprevention. <br />Involve a control condition of somesort.<br />Be reportedbytheend of 1991. <br />Be a programwith central mental healththrust, thatis, bedirected at childeren’s and adolescent’sbehaviorial and social functioning. <br />
Method<br />SearchProcedures. Threeprocedureswereusedtolocaterelevantstudies. <br />Literaturesearchprocedures. <br />Computer-generatedsearch of Pych Lit. <br />Manual study-by-studysearch of 15 journals.<br />A combinationcomputer and handsearch of DissernationAbstracts. <br />Thecurrentreviewisthusbasedonfindingsfrom177 interventions, 150 frompublishedreports and 27 fromunpublished doctoral dissertations.<br />
Method<br />Coding. Each study was coded on 51 variables falling into 7 major categories. <br />Basic identifying data<br />Methodological features<br />How effect sizes were caculated <br />Characteristics of the subjects<br />Change agents<br />Interventions<br />Outcomes measures<br />
Method<br />Reliability of CodingProcedures. <br />Threeresearchassistantsemplyedoverdifferent time periodscodedstudies. <br />Toassessinterjudgeagreement, 20 studiescodedbyeachraterwereselectedrandomly and comparedtothoseindependentlycodedbythefirstauthorwhohadtrainedtheassistant. Theassistantdidn’tknowwhichwereselectedforreliabilitychecks. <br />
Method<br />Calculation of Effectsizes.Effectsizes (ESs) werecomputedusingthepooledstandarddeviation of theintervention and control groups.<br />TheESswerecalculatedsuchthat: Positive scoresindicatethattreatedgroupwas superior tothecontrols. Negative scoresindicastedtheopposite. <br />Whenmeans and standarddeviationswerenotreported, procedurestoestimateESswereused.<br />Whenthe ES couldn’tbecalculatedbecausetheauthormerelyreportedresults as being “nonsignificant”, the ES wasconservativelyset as zero, whichocurred in 12 studies.<br />Since 88% of ESswere positive whenevertheycouldbecalculated, thepresence of zero ESS underestimatestheimpact of primarypreventions.<br />
<ul><li>The mean age of participantswas 9.3 years. 13% of thestudiesinvolvedadolescents.
Therewas a widevariability in samplesizes. 34% 50 orless. 29% 100 or more.
Figures 1ª, 1b, 1c and 1d illustratethestagesusedto divide studiestoachievehomogeneity. Unfortunately, thefeatures of currentprogramslimitedthetypes of analysesthatcouldbeconducted. <br />Forexample, Figure 1ª presentsthenumber of studiescontained in eachcell of the 3 x 2 conceptual model of primarypreventiondescribed in theIntroduction.<br />Itwasnotpossibleto test thissix-cellmodelcompletelybecause of theinsufficientnumbers of programsusing a high-riskselectionstrategyorusinganenvironmental-levelimterventionfortransitionprograms.<br />Thelargemajority of interventionswereperson- ratherthanenvironment-centered (ns = 150 and 27, respectively) and used a universal ortransitionapproach in selecting target groups (ns = 118 and 46, respectively). Amongthe 46 transitionprograms, allbut 2 wereperson-centered.<br />
In Stage 1 of theanalyses, depicted in Figure 1b, weretained as manyfeatures of the original conceptual model as possible. Thatis, wedroppedthedistinctionbetween universal and high-riskselectionstrategies, butretainedthetransitionapproachwhichnowbecameonecategoryconsisting of 46 programs. Wealsocreated single categories of person- and environment-centeredinterventionswithrespectivens of 106 and 25. None of thesebrodstudygroupingswashomogeneous, however, indicatingtheneedto divide studiesfurther.<br />
Figure 1c representsStage 2 in theanalysis. Transitionsprogramsweresubvidedaccordingtofourtypes of transitions, and environment-centeredprogramsweredividedaccordingtotheirprimarysetting: home orschool.<br />Person-centeredprogramsweredividedintothosefocusingon mental healthpromotion versus allotherprograms (benceforthcalled “Other”). Thesesubdivisionsweresuccesful in achievinghomogeneityforbuttheperson-centeredprogramswhichneededfurthersubdivisions.<br />
Figure 1d indicatesthe final grouping of person-centeredprograms.<br />Programswerefirstdividedintotwocategoriesor mental healthpromotion: interpersonal problem-solvingoraffectiveeducation. Homogeneitywasnotachieved, however, untilbothprogramcategoriesweresubdividedfurtheraccordingtothechild’sdevelopmentallevel. TheOtherperson-centeredprogramsweredifficulttocategorize. Althoughsomewerehealthpromotionprograms, thesestudiesfocusedondifferentskillsorcompetencies and couldnotbecombinedinto a single coherentgrouping.<br />Homogeneitywasachievedbydividingstuidesaccordingtothetype of changestrategyused: behavioralornonbehavioral. Weshouldpointoutthatdevelopmentallevelwasnotsuccesful in achievinghomogeneityfortheOtherprograms and changetechniquesweresimilarlynotsuccessful in reachinghomogeneityforproblemsolving and affectiveeducation.<br />
Environment-Centered Programs<br />Fifteenenvironmentaleffortstargetedchanges in schoolsettings and producedsignificant positive effects (mean ES = 0.35; seeTable III).<br />Theseprogramseithermodifiedexistingsettingsforschool-agedchildrenorcreated new settingsforyoungerchildren. Seeveral of thelatterprogramsalsoofferservicestoparents, butthemainfocuswastheeducationalsetting. Alltheinterventions in whichschoolsettings are changed are multidimensional, buttheprimaryfeatures of somerepresentativeprograms can besummarized.<br />Mostinvestigatorssoughttochangethepsychosocialaspects of thetypicalclassroomenvironmental. Forexample, Hawkins, Von Cleve, and Catalanio (1991) trainedfirst- and second-grade teachers in effectiveclassroommanagementprocedures, and introduced new interactiveteachingstrategiesthatencouraged more supportive and reinforcingcontactsbetweenteachers and students; teachersalsointroduced social skills training proceduresintotheclassroom. Thisinterventionwaseffective in reducingaggresivebehavior in boys and self-destructivebehavior in girls.<br />Oneintervention at thehighschoollevel (Weinstein et al, 1991) modifiedseveralclassroomfeaturessucha as curricula, studentabilitygruopings, evaluationprocedures, teacher-studentrelationships, and parentinvolvement in schoolactivities. Theseproceduresproducedsignificantbenefitsfor at-riskmuñtiethnichighschoolstudents in terms of grades, disciplinaryreferrals, absenteeism, and schooldropoutrates.<br />
Theapproachby Comer (1985) was more ambitious in terms of modifyingthestructure and functioning of anentireelementaryschool.<br />In this case, teachers, administrators, mental healthprofessionals and parentsworkedtogethertogoverntheschool, institute and monitor new policies and procedures, and assessstudentsprogress. Theoverallintentwastodevelop a positive schoolculturethataddressedboththeacademic and social needs of students. Begun in 1968, thisintervention has beensuccessful in changingtheacademicachievemientlevels of participatingschools, reducingseriousbehaviorproblems, and improvingstudentssense of personal competence.<br />Finally, the Houston Project (Johnson & Breckenridge, 1982) isanexample of howenvironmentalchange can occurthroughthecreation of a new setting: a childdevelopment center toserveentirefamilies. Early home-visitinguntilthechildwas 1-year-old preparedthefamilyforfullerengagement in childdevelopment center activities a yearlater. A the center, parentslearnchild-rearingtechniques and are offered general support. AdultEnglish-languageclasses are alsoavailablesincethe Project targets low-incomeMexican American families. The 2-year programinvolves up to 500 hours of familymembers time. <br />Processevaluationshaveindicatedthatparentschangetheirinteractionalstyles and child-rearingbehaviorswiththeirchildren. Children´scognitivedevelopmentwasenhanced, and theinterventionproducedsignificantlong-termefectsonbehavioraladjusment in earlyelementaryschool.<br />
In theothertype of environmentalprogram, parent training wasthe exclusive focus. These 10 programsprimarilysoughttomodifythechild’s home situationbyeducatingparentsaboutchilddevelopment and modifying parental attitudes and child-rearingtechniques as needed (e.g, Frazier & Matthes, 1975; Graybill & Gabel, 1981). Theseprogramswerehomogeneus in outcome, butnoteffective (mean ES = 0.16; seeTable III). The mean ES doesnotdiffersignificantlyfromzero.<br />
Transition Programs <br />Four kinds were identified:<br />Five programs targeted to first-time mothers beginning when infants were 2-4 months and until they were 3 ½ years. It included home visits and periodic group meetings reducing reports of child abuse and neglect during the first 2 yrs.<br />There were seven programs for children of divorced parents designed to ease the passage, understand and cope with the changes in their lives.<br />There were eight programs designed to help children during school transitions. <br />There were 26 programs for children experiencing potentially stressful dental and medical procedures. Modeling and desensitization conditioning helped reduce fears and anxieties. <br />
Person-Centered Strategies<br />Health Promotion Programs<br />Two main categories were:<br />Affective education- attempts to increase children’s awareness and expression of feelings and their ability to understand the possible causes of behavior. Combines stories, puppet play, music, etc.<br />Interpersonal problem-solving- teach children how to use cognitively based skills to identify interpersonal problems and develop effective means of resolving such difficulties.<br />Importance of Development Level<br />There was homogeneity until the participants were subdivided into according to development level. Children ages 2-7 yrs were considered as preoperational, 7-11 yrs as concrete operational, and 11 or older as formal operations.<br />Other Programs<br />Behavioral techniques were nearly as twice as effective than using non-behavioral techniques.<br />
Alternate Analysis <br />This alternate analysis was used to see if other models would produce an equal or better fit for the data. To do this, they combined the different kinds of variables in other orders and compared them. As a result, none of the alternative analytic strategies approached the initial model.<br />
Howparticipantschanged<br />To clarify how programs affected patients, we coded outcome measures in 2 different ways and summarize these data.<br />Problems and Competencies.<br />Domain of functioning.<br />Externalizing symptoms<br />Internalizing symptoms <br />Academic achievement<br />Sociometric status<br />Cognitive processes<br />Physiological assessments <br />
Practical Significance of Outcomes<br />The magnitude of an ES does not necessarily reflect its practical or social significance. <br />One way to assess the practical significance of an intervention is to indicate how outcomes for the experimental and control groups overlap.<br />
Practical Significance of Outcomes<br />Another method of depicting the practical significance of interventions on participants at a single time point is through the BESD (Binomial effect size display).<br />The BESD reflects the change in the success rate resulting from an intervention. <br />If a program has no effect, the success rates for intervention and control groups would both equal 50%. <br />
Future Research Directions<br />General Design Features<br />
There are both positive and negativefeatures in themethodologicalcharacteristics of currentprograms.<br />Onthepositivieside, a majority of studieswere true randomizedexperiments(61%), hadlittlesampleatrrition(10% orless in 80%of thestudies), and usedmultipleoutcomemeasures(90%)<br />Onthenegativeside, more studiesneedtocollectfollow-up informationoverlongerperiods. Only 25% of studiescollectedanyfollow-up data, and thefollow-up periodwasrarely 1 yearorlonger.<br />Overall, currentfollow-up data do notpermitanyfirmconclusionsaboutthelong-termimpact of mostinterventions.<br />
A fewinvestigators are beginingtoreplicatetheirearlierfindings, whichincreasesconfidence in theefficacy of certaininterventions.<br />It’simportantto note howcurrentstudycharacteristicslimitedthetypes of analysesthatcouldbe done.<br />As data onprimarypreventionaccumulate, itpossibletoassessvarious conceptual aproachestointervention in more depth and detailthancouldbeaccomplished in thecurrentreview.<br />Itwasalsonotpossibleto examine howprogramimplementationinfluencedoutcomessinceveryfewinvestigationsprovidedanyrelevant data.<br />
Sinceitisreasonabletobelievethesamefindingwouldoccurfortheprevention of behavioral and social problems, futureresearchersshoulddocumentthequality of implementation and itsrelationshiptooutcomes.<br />Currentresearch can bealsoimprovedbystablishingspecificprogramgoals, clearlyoperationalizinginterventionprocedures, and byusingtheoryto guide interventions.<br />
Futureinvestigatorsshouldoperationalizetheirinterventionscarefully, collectprocess data, and compare programscontainingdifferentcomponents in ordertogainanunderstanding of themechanisms of changeoperating in differentprograms.<br />Thevalue of theory-drivenresearchcannotbeoverstimated. Goodtheoriesprovide a coherentmodelforintervention and analysisbyspecifyingthe target population, suggestingthecomponentsthatshouldcomposetheintervention and predictingchange.<br />
Assessing Outcome<br />Twocharacteristics of outcomemeasureswereassociatedwithlowereffectsizes<br />A) use of multipleoutcomemeasures<br />B) use of at leastonenormedmeasure<br />Althoughlowereffectsmaybeobtained, investigatorsshouldcontinuetoassesshowpreventionmodifiesdifferentaspects of adjustmen, and use themostreliable and validmeasuresto do so.<br />Futurestudiesneedto determine howdifferentprogramparticipantsbenefitfromintervention, sinceitisunlikelythatallthose in the target populationdemonstratethesameamount of change.<br />
Miscellaneous Issues<br />CurrentfindingssuggestthattheInstitute of Medicine’s (1994) decisiontoexcludepromotion as a preventiveintervention’sispremature. Healthpromotionappearstohavevalue as a preventiveintervention, although more data are needed.<br />In summary, althouthseveralissueshavetobe resolved, outcome data indicatthatmostcategories of primarypreventionprogramsforchildren and adolescents produce significanteffects. Thesefindingsprovideempiricalsupportforfutherresearch and practice in primaryprevention.<br />