Models of Integrated Care for Adolescent Alcohol and Drug Use in Pediatrics Predictors and Implications for Practice and Policy STERLING
Models of Integrated Care for Adolescent Alcohol and Drug Use in Pediatrics: Predictors and Implications for Practice and PolicyStacy Sterling, MSW, MPH,1 Andrea H. Kline Simon, MS,1 Constance Weisner, DrPH, 1,21 , KPNC Division of Research, Drug and Alcohol Research Team2 , UCSF, Department of Psychiatry HMO Research Network Conference May 1, 2012 Studies funded by NIAAA, NIDA, and Community Benefits, Kaiser Permanente Northern California
SBIRTDART studiesNIAAA Study of Pediatricians’ Alcohol and DrugScreening PracticesFeasibility Pilot of Teen SBIRTNIAAA effectiveness and cost-effectiveness trial ofAdolescent SBIRTPrimary Care as Medical Home
Adolescent Alcohol and other Drug (AOD)Problems AOD problems are major causes of mortality and morbidity for adolescents. Three-quarters of adolescent deaths are behavior-related: accidents (51.7%); assaults (13.7%) and suicide (11.0%), and a high percentage of these involve alcohol and/or drug use (Chaisson, 2005). In many cases, addiction is a pediatric-onset disease.(Riggs, 2012) The earlier a person initiates substance use, the more likely they are to develop problems.(Grant, 1997 & 1998; Placzek, 2009) AOD problems which persist into adulthood often begin in adolescence (NIAAA, 2006; Brindis, 2002; Blum, 1987; Shrier, 2003). Integrated health care (Primary Care and specialty AOD and Mental Health treatment) can improve access and outcomes (Sterling, 2005; Sterling, 2009; Bickman, 1996; Kaplan, 1998; Summerfelt, 1996).
Screening, Brief Intervention and Referral to Treatment (SBIRT)
Where the field is with SBIRT Despite recommendations, most studies show that relatively few PCPs screen, and even fewer screen according to guidelines or use standardized instruments (Denny, 2003, Center for Substance Abuse Treatment, 2000). It is difficult to get adolescents to specialty treatment when needed (Institute of Medicine, 2006) Special concern to clinicians There is little coordination with Substance Use treatment by PC after treatment, (Institute of Medicine, “Improving the Quality of Health Care for Mental and Substance-Use Conditions,” 2006)
Brief Intervention and Adolescents BIs with adolescents - efficacious and/or effective on a range of outcomes including: reducing AOD use, binge drinking, drinking and driving, smoking, AOD- related consequences, marijuana use, and ER utilization (Bernstein, 2009; Knight, 2005; Marlatt, 1998; Martin, 2005; Lawendowski, 1998, De Micheli, 2004; DAmico, 2008; Ozer, 2003). BIs have been shown to be as effective as more traditional therapies for AOD use for those with lower-severity problems (Tevyaw, 2004). Pediatric Primary Care Providers may be especially effective agents of SBI (Levy, 2002) – teachable moments, health risk context rather than moral or legal issue. High receptivity to screening and intervention by PCPs (Yoast, 2007; Steiner, 1996; Stern, 2007; Stern, 2006). Teens have more positive perceptions of care when their PCP discussed AOD use (Brown, 2009). Developmentally, young people are likely to be receptive to self-guided behavior change strategies, a cornerstone of brief interventions (Miller & Sanchez, 1993).
Brief Intervention by Non-Physicians BIs with adolescents have often been conducted in non-medical settings, and by non-MDs, and have shown promising results on a range of outcomes: AOD use, binge drinking, tobacco use, AOD- related consequences (Burke, 2005; Gil, 2004; Grenard, 2007; Martin, 2005; Winters, 2007; McCambridge, 2004). Non-Physician Interventions in adult PC had better adoption than (and similar effectiveness to) physician-delivered SBI. (Babor, 2005), but such comparisons have not yet been studied in pediatrics.
Guidelines Comparisons Among Recommendations for Adolescent Preventive Services Developed by National Organizations * AAFP AAP AMA BF USPSTF Screening/counseling Obesity Yes Yes Yes Yes Yes Contraception Yes Yes Yes Yes Yes Substance use Yes Yes Yes Yes Yes Alcohol use Yes Yes Yes Yes Yes Tobacco use Yes Yes Yes Yes Yes Hypertension Yes Yes Yes Yes Yes Depression/suicide No Yes Yes Yes No Eating disorders No Yes Yes Yes No School problems No Yes Yes Yes No Abuse No Yes Yes Yes No Hearing Yes Yes No Yes No Vision No Yes No Yes No Periodicity of visits Tailored Annual Annual Annual Tailored Target age, range, y** 13-18 11-21 11-21 11-21 11-24American Academy of Family Physicians (AAFP), American Academy of Pediatrics (AAP), American Medical Association (AMA), Bright Futures (BF), US Preventive Services Task Force (USPSTF)
Provider and Organizational FactorsProviders do not screen according to recommendations (Denny, 2003; Bethell, 2001)AAP survey (AAP, 2002) found that: < 50% of fellows routinely screen all adolescents for substance use; only 16% reported using standardized screening instrumentsProvider factors which may affect screening practices: Inadequate training and education (Emans, 1998; Gassman, 2003), Knowledge and competence in AOD area (Friedmann, 2000) Self-efficacy about sensitive health issues (Ozer, 2004; Gottlieb, 1987) Attitudes toward patients with AOD problems (Ogborne, 1986; Westermeyer, 1978; Roche, 1991; Miller, 2001) Specialization in adolescent medicine (Ellen, 1998) ConfidentialityManaged care organizations may have higher AOD screening rates than other health systems (Halpern-Felsher, 2000), but few health plans require AOD screening (Garnick, 2002).
Setting: Northern California Kaiser Permanente Staff-model integrated health care delivery system Serves 4 million members (about 40% of insured population in the region) ~ 400,000 members 12-18 18 hospitals, 27 outpatient clinics Integrated health care system (medical, psychiatry, AOD services)
DART Adolescent Studies (NIDA, NIAAA, CSAT & RWJF): KPNC Teens entering AOD Treatment and 3, 5, 7 & 9-year follow-ups KPNC Teens with Depression and Suicide Behavior“Dually-Diagnosed” KPNC Teens Teens typically arrive at AOD treatment with a clustering of problems: alcohol and drug use, psychiatric and medical problems, risky sexual behavior, school and legal problems, childhood trauma (Mertens, 2007; Sterling, 2004; Sterling 2005; Ammon, 2005). Integrated treatment (Co-location, concomitant AOD and Mental Health Tx, etc.) is beneficial (Sterling 2005; Chi, 2006). They have had frequent contacts with the health plan, often in Primary Care, yet few were identified as having problems and referred to specialty treatment before their problems became severe.
Utilization Patterns during 24 Months prior to Treatment Intake 24 months prior to 12 months prior 3 months prior intake (%) to intake (%) to intake (%) Primary Care* 89.5 79.2 48.7 Psychiatry 50.1 42.0 30.8 ER 26.3 17.9 11.7 *Includes visits to the following departments: Family Practice, General Medicine, GYN, Medicine, Pediatrics, Physical Medicine and Urgent Care.•Only 49% of those who had a visit in Psychiatry in the year prior to theirAOD Tx. intake had received a diagnosis in Psychiatry related to alcohol ordrug problems.
Referral Sources Parents - 83% Health care provider - 18% Legal system - 33% (20% Court Mandated) Friends - 19% Mental health providers - 35% Schools - 13%
Results: Study Participants versus Controls Study participants had higher ER utilization and cost in the one year pre- and one year post-intake (p < .01). There was no decrease in ER utilization or cost over time.
Study of Pediatric Primary Care Providers and Behavioral Health Screening Practices and Beliefs (NIAAA)
Survey of Pediatricians Conceptual Model(Data from Administrators, Providers, Health Plan Databases)
NIAAA Study of KPNC Pediatric Primary Care Providers andAlcohol and Drug Screening Practices Web-based survey, sent to all KPNC pediatric primary care providers with >50 adolescent patients in panel (80% response rate, N=437). 14% reported any recent (≤ 5 years) continuing education about AOD screening, assessment, treatment or referral. 11% reported receiving any AOD training in medical school. 48% were satisfied they were staying current on AOD problems and treatment. Fewer than 10% reported using a standardized or evidence-based screening instrument. Sterling S, Kline-Simon AH, Wibbelsman C. (2011). SBIRT for youth alcohol and drug use in primary care: Predictors and implications for practice and policy. INEBRIA, Boston, MA, Sep 21-23.
NIAAA Study of KPNC Pediatric Primary Care Providers andAlcohol and Drug Screening Practices Time/resources Clustering of problems Linkages with specialty care Discomfort with addressing alcohol and drug problems, particularly compared to other behavioral problems (e.g., risky sexual behavior, depression). They rated alcohol use as more difficult to discuss, or diagnose, than depression (19% v s 15% and 70% vs. 56%, respectively), and were more comfortable talking about risky sex than alcohol (32% vs. 22%). Sterling S, Kline-Simon AH, Wibbelsman C. (2011). SBIRT for youth alcohol and drug use in primary care: Predictors and implications for practice and policy. INEBRIA, Boston, MA, Sep 21-23.
NIAAA Study of KPNC Pediatric Primary Care Providers andAlcohol and Drug Screening Practices Differential screening? 13% reported being more likely to screen boys than girls; Male PCPs were even more likely to screen more boys than girls (23% vs. 6%; (p<.0001). 94% said that evidence of depression would trigger them to screen for AOD. What would it take to consistently screen every teen? More time – 77% Have MA screen and another clinician receive results for discussion with teen – 57%
Self-reported likelihood of referring patients to AOD treatmentLogistic Regression Models ≥ 10 years of experience being an adolescent medicine specialist having had recent AOD training (all p<.05).
Self-reported vs. Actual (From EMR) Screening Self-report Actual Actual 6-mo pre-survey 6-mo pre-survey 6-mo post-survey (%) (%) (%) Alcohol 92 65 66 Other drugs 88 65 66 Tobacco 92 66 64 Friends’ AOD use 76 66 64 AOD use while driving 47 66 64We examined by: experience, self-efficacy with AOD Dx, comfort level with AOD, attitudes about AODTx, confidentiality as a barrier, linkage with AOD program, awareness of AOD Svcs, training,specialization –> No differences
These findings informed an intervention:Purpose3)Identify adolescents before their problems become severe4)Provide brief interventions for those whose problems are notsevere5)Improve referral and engagement with specialty care if needed Better integrate care between Pediatrics, Substance Use treatment, and Child and Family Psychiatry for those needing referral
“Medical Home” Model for Adolescent Substance Use Problems Higher Lower severity severity problems problems, and acute careongoing disease management Specialty Care Pediatrics (CD and Psychiatry)
Basic principles in developing SBIRTintervention: Critical for adoption Evidence-based screener and intervention Resources (skilled helper/time) Fit into the existing workflow Using electronic medical record Time-efficient Generalizable to other health systems (FQHCs)
Pilot study of a Brief Intervention Model of Care forAdolescents in a General Pediatrics Clinic – InternalKaiser Funding Does an SBIRT model of care in Pediatrics increase identification of behavioral health problems compared to usual care? Is this model of care more effective than usual care at promoting behavioral health treatment utilization? What factors affect the implementation of an SBIRT model in Pediatrics?
Adolescent SBIRT Pilot Protocol Behavioral Clinician in General Pediatrics clinic, trained in Brief Intervention and Motivational Interviewing techniques. PCPs Screened teens and referred those identified with behavioral health problems to the Clinician. Clinician assessed patients further, and either: provided BI for lower-severity substance problems. facilitated Referral to specialty Treatment (CD or Psych) for higher- severity substance use or mental health problems. 77 teens referred (55 girls, 22 boys)
Adolescent SBIRT Pilot Findings Depression, Anxiety, school and family problems, and stress much more common than AOD as presenting problems. After further screening however, AOD use was frequently present and problematic. Very well-received by PCPs and patients and parents de-stigmatized: “Teen Healthy Lifestyle Check-Up”. Warm hand-off was very important. Providers reported that the model improved care.
Adolescent SBIRT Pilot Findings Behavioral Health Treatment Initiation PC Adolescent Well Visits 2008 – 2009 2009 – 2010 p-value (N=2,611) (N=2,708) Total Behavioral Health Visits (BI & Specialty AOD, MH) 228 / 8.7% 325 / 12.0% <0.0001 N/% Specialty AOD and MH only 228 / 8.7% 269 / 9.9% 0.0660 N/% Behavioral Health treatment utilization (overall, and specialty AOD and Mental Health treatment alone) increased during the pilot.Sterling S, Kline-Simon AH, Wibbelsman C, Wong A, Weisner C. (under review). Screening for adolescentalcohol and drug use in Pediatrics: Predictors and implications for practice and policy. Addiction Science &Clinical Practice.
NIAAA Teen SBIRT Comparative Effectiveness Trial Randomizes PCPs to different modalities of delivering SBIRT for adolescent behavioral health (PCP-delivered vs. BMS model of care). No studies have examined the effectiveness and cost-effectiveness of physician vs. non-physician delivery of SBIRT for adolescents. Little research on implementation of SBIRT in Pediatrics. Non-Physician Interventions in adult PC had better implementation rates than (and similar effectiveness to) physician-delivered SBI (Babor, 2005). Takes advantage of the health plan’s EMR: an important facilitator of screening and generalizable to the future of U.S. health care (Saitz, 2006; IOM, 2006; IOM, 2001).
Current Study - NIAAA Teen SBIRT ComparativeEffectiveness Trial Addresses clinician (and research) concerns: Evidence-based Skilled help (BMS) Integrated into clinic flow Time efficient (smooth hand-off and easy script) Got in Electronic Medical Record!!!!! Based on pilot, advocacy by clinicians, business case of previous studies, and improvement in successful referrals Population-based Study design integrated with EMR, uses regular practice
PCPs are trained to BMSs are trained deliver SBIRT to conduct SBIRT Treatment as usual CMEs PCPs refer to BMSs
Research Questions Effectiveness: Provider outcomes: Which model of SBIRT produces the best screening, brief intervention and referral rates? Patient Outcomes: Which model of SBIRT produces better patient outcomes (AOD use and AOD-related-school, legal & family problems) at 1 year? Which model results in better specialty treatment (CD or Psychiatry) initiation and engagement rates?2. Cost Which model of care is most cost-effective? Process of/Barriers to Implementation
What we’ve learned and next steps: Adapt based on findings and other learnings from the study (quantitative and qualitative) Work with clinicians, health plan administrators, consult with FQHCs and other health systems and other new research Roll-out or larger implementation study as with adult SBIRT
DART Research GroupInvestigators/Staff Scientists Interview SupervisorConnie Weisner, DrPH, LCSW Gina Smith AndersonJennifer Mertens, PhD Research AssociatesCynthia Campbell, PhD Georgina Berrios Virginia BrowningHealth Economist Jessica Duhe Diane Lott-GarciaSujaya Parthasarathy, PhD Melanie Jackson Cynthia Perry-BakerAnalysts Barbara PichottoFelicia Chi, MPH Martha Preble Lynda TishAndrea Kline Simon, MS Sabrina WoodWendy Lu, MPHTom Ray, MBA Research CliniciansProject Coordinators Thekla Brumder, PsyDTina Valkanoff, MPH, MSW Sarah Ferraro, PsyD Amy Leibowitz, PsyDAgatha Hinman, BA Ashley Jones, PsyDAliza Silver, MA Clinical Partners Anna Wong, PhDAdjunct Investigator Charles Wibbelsman, MDDerek Satre, PhD KPNC Chemical Dependency Quality Improvement Committee KPNC Adolescent Medicine Specialists Committee KPNC Adolescent Chemical Dependency Coordinating Committee
In-person group trainingsWEBEXesE-mailed Power PointsBoostersQualitative conversationsFeedback reportsOne-on-one training
PCP-arm trainings – three, 1 hour trainingsDay 1Overview of problem & rationale for screening - Prevalence of problems,Recommendations (AAP, AMA, etc.)Study overview – research questions, randomization schemeHow-to: Screening and assessment tools in EMR Workflow Review of resources DocumentationDay 2 How to - Motivational Interviewing & Brief Intervention techniquesDay 3 Skills practice using cases from pilot How to make a referral to specialty (CD and Psychiatry) treatment
BMS-arm trainings – one, 1 hour trainingOverview of problem & rationale for screening Prevalence of problem Recommendations (AAP, AMA, etc.)Study overview – research questions, randomizationHow-to – Day 1 Screening tool and “trigger” questions in EMR Workflow
During the past year: 1. Did you drink any alcohol? 2. Did you smoke any marijuana or hashish? 3. Have you used any other drug to get high (such as prescription drugs, meth, ecstasy, glue of cocaine)?OR 1. During the past few weeks, have you OFTEN felt sad, down or hopeless? 2. Have you seriously thought about killing yourself, made a plan, or tried to kill yourself?OR In your clinical judgment, teen has risk for AOD or other behavioral health problems Ask CRAFFT Questions and enter into Health Connect Give praise and CRAFFT score? encouragement. Each yes=1 Are there no major problems & patient believes he/she will be successful in making a change? Counsel patient to stop using substances. Provide brief advice linking substance use to undesirable health, academic & social consequences. F/U at next visit. (1 - 2 min) Brief Intervention. Refer to CDRP for assessment 43 At follow up visit, confirm whether patient stopped using
Accessing the CRAFFT Questionnaire in Health Connect 44
Use this V Code (V65.42D) to document that a BriefIntervention for alcohol or drug problem was performed – this is how we will measure intervention rates 45
Patients’ progress over time can be viewed in this CRAFFT flowsheet 46