Richard L. Brown, MD, MPH

Value-based reimbursement programs are creating strong incentives for primary care settings to deliver evidence-based behavioral screening and intervention (BSI). BSI for unhealthy drinking, depression, and smoking improves outcomes and generates substantial cost savings. Addressing other behavioral risks and disorders may also reduce costs. From experience helping dozens of primary care clinics implement BSI, the author makes several recommendations: expand the healthcare team, because BSI requires ample time; hire full-time bachelor’s-level health coaches selected for their personality attributes rather than prior training and experience; initially, deliver BSI for patients of one clinician, and optimize workflow, then serve patients of other clinicians; obtain buy-in from all clinicians and staff with strong support from healthcare organization and clinic leaders, data demonstrating patients’ need for BSI, initial success stories, and local data documenting effectiveness; train coaches rigorously in motivational interviewing, an empathic, respectful, patient-centered, evidence-based approach to promoting healthier behaviors; provide ongoing support for coaches after 2 to 3 weeks of initial training; have receptionists ask patients to complete screening forms in waiting areas, then have medical assistants review completed screens and summon coaches to see patients with positive screens; use a data-driven quality improvement program to refine the workflow, maximizing the proportions of patients screened and of patients with positive screens who see health coaches; and adapt electronic health records to support BSI. Administering BSI in this manner can help primary care clinics make substantial progress toward achieving the quadruple aim and thriving under value-based reimbursement.

The American Journal of Accountable Care. 2018;6(2):e9-e14

The Changing Playing Field of Value-Based Reimbursement

Under fee-for-service reimbursement, primary care clinicians in the United States have had little financial incentive to help reduce healthcare costs. Primary care practice leaders have understandably viewed the expenses of many recommended cost-saving measures as indefensible investments, as savings would accrue to payers or purchasers, not to the primary care practices.

Value-based reimbursement is creating a new playing field. Payers are increasing expectations and incentives for primary care clinicians to deliver cost-effective care. Under the Medicare Access and CHIP Reauthorization Act, Medicare will modify each practice’s reimbursement rates, starting in 2019, based on claims submitted throughout 2017. Whether primary care practices choose to participate in alternative payment models (APMs) or the Merit-based Incentive Payment System (MIPS), an increasing portion of their earnings will depend on quality metric performance and cost control.

Especially after commercial payers start to emulate Medicare and most payment is based on value, a Darwinian vicious-circle scenario will likely ensue. Better-performing and more highly compensated practices will invest more in innovation, improve their performance further, generate more profit, and entice clinicians and staff away from other practices with better working conditions and higher compensation while other practices will wither. Now is the time for primary care practices to start making bold changes if they aim to thrive under value-based reimbursement.

One recommended cost-saving practice not widely implemented by primary care practices in a robust, evidence-based manner is behavioral screening and intervention (BSI).

 The core components of BSI are smoking screening and cessation; alcohol screening, brief intervention, and referral to treatment; and depression screening and collaborative care. These services carry Grade A or B recommendations from the United States Preventive Services Task Force, are recommended by many other authorities, and generate ample net cost savings, as shown in 

 BSI can also include screening, intervention, and referral for drug use, intimate partner violence, diet, physical activity, and obesity—services that are recommended by many authorities but lack documentation of cost savings. BSI can be extended to address additional behaviors relevant to various chronic diseases, such as adherence to medication, dietary, and physical activity regimens.

The author and colleagues have helped dozens of primary care practices implement BSI. From 2006 to 2011, with funding from the Substance Abuse and Mental Health Services Administration, they helped 33 clinical settings to conduct alcohol and drug screening for more than 100,000 patients; deliver more than 20,000 interventions; and garner high patient satisfaction. Six-month follow-up phone calls to hundreds of patients found average reductions of 20% in binge drinking and 15% in marijuana use.

 An analysis of Medicaid claims data also found average net 2-year cost savings of $782 per patient screened.

 In a small pilot project in which 3 of the 33 settings also implemented depression screening and collaborative care, depression symptom scores declined by 55% over 8 to 12 weeks.

 In other projects since then, the author and colleagues have helped other practices implement BSI with various combinations of behavioral focuses, including cardiovascular risk reduction for patients with hypertension, lipid disorders, and diabetes, as recommended by HHS’ Million Hearts Initiative.

BSI would help primary care practices that join APMs excel on 15 Medicare Shared Savings Program metrics, as shown in 

, and practices that choose MIPS increase scores in 3 of 4 areas: quality metrics, cost control, and practice improvement.

For practices that wish to implement BSI, this paper will report on lessons learned from helping dozens of primary care settings implement BSI since 2006.

Lesson 1. Plan for Robust Service Delivery and Expand the Healthcare Team

The robust services required for optimal behavioral outcomes and cost savings require ample patient contact time. Optimal smoking quit rates are obtained with motivational interviewing and at least 9 one-on-one support sessions.

 Collaborative care for depression involves contact with patients over several months.

 Although some patients with unhealthy drinking patterns respond to a single-&shy;session brief intervention, follow-up contacts elicit greater and more sustained change.

In planning for BSI estimate patient contact time requirements. 

shows such calculations based on the prevalence of smoking, unhealthy drinking, and depression in the US adult population. Practice administrators can easily obtain population prevalence figures for their states.

 These figures should be regarded as conservative because behavioral risks and disorders are usually more prevalent in clinical populations than in general populations. As shown in Table 3, robust BSI for smoking, unhealthy drinking, and depression for an average panel of 2300 patients

 would require more than 900 total hours of patient contact time per year. This calculation makes clear that busy clinicians or other staff cannot deliver robust BSI. The healthcare team must be expanded to enable delivery of the high-quality services necessary to improve outcomes and reduce costs. The calculation suggests that practices must plan to add 1 health coach for every 2 full-time primary care clinicians. A higher health coach-to-clinician ratio would be needed if coaches address additional behavioral risks and disorders and chronic diseases.

Lesson 2. Configure Teams of People Who Practice at the Top of Their Capability

Some healthcare planners assume that several experts are needed for BSI: a smoking cessation specialist, an alcohol and drug counselor, a mental health professional, a dietician, and others. Such staffing is neither feasible nor advisable because it would badly fragment care. Patients need 1 trusted individual to serve as their health coach. Interestingly, there is no established healthcare profession that trains its members to provide coaching on all behavioral issues.

We have observed that individuals with a bachelor’s degree—with appropriate training, support, and guidance&mdash;are the best fit for BSI coaching positions. In our initial 5-year project, bachelor’s-level coaches elicited greater reductions in binge drinking than their counterparts with master’s degrees in counseling or social work.

 Some master’s-level coaches can be very effective; however, some are more difficult to train in motivational interviewing (discussed later) because of ingrained communication patterns, are uncomfortable addressing behavioral issues outside of the realm of mental health, prefer to ply their more advanced counseling skills, or expect higher compensation than bachelor’s-level coaching counterparts. Consistent with the concept of having staff serve at the top of their capabilities and licenses, counselors and social workers with a master’s degree are best utilized as on-site referral resources for counseling. If coaches can make personal face-to-face introductions to on-site counselors rather than referrals to remote agencies, referrals for counseling tend to be more successful.

In some situations, there may be advantages to hiring licensed professionals to administer BSI. In small practices, licensed professionals may be able to serve as part-time coaches and counselors. Medicare, some Medicaid programs, and some commercial payers reimburse for some components of BSI only when delivered by licensed individuals. Practices with certain payer mixes may benefit from conducting financial simulations to help guide decisions on coach hiring requirements. Such simulations should account for coming declines in fee-for-&#8203;service reimbursement and increases in value-based reimbursement.

Lesson 3. Emphasize Efficient and Sustainable Workflow

BSI is not reserved for scheduled preventive visits, because many patients do not schedule such visits. In the outpatient settings where BSI works best, receptionists are prompted by electronic health records (EHRs) to ask patients to complete written screening forms each year, perhaps along with other annual forms. Medical assistants (MAs) review the completed screening forms and summon coaches to see patients with positive screens in exam rooms.

Clinicians always have priority with patients in the exam rooms. If clinicians are running behind, coaches can see patients first and seek the patients’ permission to remain if clinicians interrupt. If clinicians are on time, MAs typically ask patients to remain and meet with coaches afterward. If patients cannot stay, or if coaches are busy with another patient, the EHR is flagged to remind the MAs that patients need to see a coach at their next visit.

How receptionists and MAs ask patients to complete screening forms and see coaches is critical. All should be instructed to use scripts framing the process as an important routine preventive service that clinicians want and expect their patients to receive.

Clinicians should not serve a critical role in the workflow, such as referring patients to coaches. Many clinicians do poorly with routine, in part because they focus on addressing patients’ other needs, generating documentation, and completing other administrative tasks in short time periods. Consistent with configuring teams to work at the top of their capabilities, clinicians are best reserved for delivering pharmacotherapy for depression and nicotine, alcohol, or opioid dependence; for reinforcing patients’ work with coaches; and for stepping in when patients’ needs surpass coaches’ training, such as further assessing patients who indicate suicidal ideation on depression assessment questionnaires.

The best coaches have high emotional intelligence, an engaging manner, therapeutic attitudes, enthusiasm for their work, strong abilities to work in teams, and bachelor’s degrees. Prior healthcare training and experience are less important. Ideally, coaches represent the same racial and ethnic backgrounds as their patients and, in some practices, speak multiple languages. Unfortunately, in many communities, it is difficult to find such coach candidates with bachelor’s degrees. Fortunately, BSI can be also delivered well by high school graduates who were qualified to attend college but did not have the opportunity. BSI programs often attain suboptimal results when coaches are selected from current staff rather than wider applicant pools.

For BSI to optimize outcomes and cost savings across populations of patients, coaches must be present at all times during workdays when patients are completing screens, because attempts to telephone patients with positive screens who were not seen face-to-face in clinical settings usually fail. When coaches serve part time in other roles, those roles tend to dominate their time, because BSI is rarely urgent. In addition, having full-time coaches enables more consistent routines and better performance by receptionists and MAs at distributing screens and summoning coaches.

The wisdom behind starting small, working out kinks, and then expanding is incontrovertible, but understanding how to start small is very important. Some healthcare organizations start implementing BSI one behavioral issue at a time, perhaps avoiding new hires until there is certainty that BSI can work. Even if the coach is a new hire, this approach is usually not optimal. One problem is that some clinicians and staff perceive only a tentative commitment to BSI by the leadership. Another problem is that the decisions made to optimize the effectiveness of a BSI program with limited scope make it more difficult to expand the scope later. For example, a program that starts addressing only depression may hire a coach with a special interest in mental health, but that coach may be uncomfortable expanding into smoking, alcohol, and other behavioral issues despite additional training.

There are other reasons to address multiple behavioral issues at the inception of a BSI program. Many patients have interrelated behavioral issues. For example, a patient who quits smoking is more likely to relapse if depression and frequent alcohol intoxication are not addressed. Another reason is that clinician and staff support for BSI tends be stronger when it addresses multiple issues. Some clinicians may not be supportive of a program that addresses only alcohol and drugs but may be glad to support a program that also addresses smoking and depression. A third reason is that patients respond more favorably to programs that address multiple issues. Although some may feel singled out to be screened for 1 or 2 sensitive issues, screening forms and coaches that mention sensitive issues (eg, substance use and depression) at the same time as other general health issues (eg, fruit and vegetable consumption and smoking) engender more accurate reporting and more cooperation. In addition, for patients with multiple behavioral risks or disorders, success in addressing one issue can often engender successes with other issues.

We recommend starting with a robust BSI program that addresses at least smoking, alcohol, and depression, the 3 issues for which return on investment is well documented. Brief feedback and referrals can be suggested for patients with unhealthy diets, physical inactivity, and obesity.

The recommended way to start small in a primary care setting is to begin with 1 coach who addresses several behavioral issues for the patients of 1 supportive clinician. That clinician ideally will have ideas and patience for working out any glitches, and then patients of other clinicians can be served.

In healthcare systems with multiple primary care sites, it is best to start with at least 4 coaches at 4 sites. Advantages include support and sharing of best practices among sites and coaches, economy of scale in training and start-up, and higher odds of success for at least some of the sites.

Lesson 7. Obtain Buy-In From All Clinicians and Staff

For BSI to succeed, all clinicians and staff must effectively serve the roles previously described (Lesson 3). To maximize their cooperation, they should understand the purpose of the program; the rationale for anticipating patient satisfaction, improved outcomes, and lower costs; and the importance of the program to the organization’s overall clinical and financial goals. Leaders of the organization and the setting must make clear to clinicians and staff their commitment to succeed and their expectation that all will do their part.

In the first several months of the program, internal marketing at each setting can help cement support. Coaches should share with clinicians and staff success stories in ways that respect patient confidentiality. Within the first month, leaders should publicize aggregate screening and assessment results, which will show patients’ need for BSI. And within several months, leaders can share data on the effectiveness of the program, as described later (Lesson 10).

Lesson 8. Motivational Interviewing Should Be the Lynchpin of All Coaching

Motivational interviewing is an empathic, respectful, collaborative approach to promoting healthy behaviors. The interviewers share information and advice sparingly and only when patients will be receptive. They elicit less defensiveness and more change by emphasizing patient autonomy rather than coercing and arguing. They help patients consider the advantages and disadvantages of change in light of their goals and values, and they create opportunities for patients to make and strengthen their own arguments for change. They help patients who are committed to change design and refine behavior change plans that help them meet their own goals.

A growing body of research suggests that motivational interviewing is more effective than other approaches to eliciting healthier behaviors.

 An advantage of motivational interviewing over other approaches to promoting change is its inherent patient-centeredness and cultural sensitivity. Some healthcare leaders and clinicians hesitate to implement BSI for sensitive topics because they fear that patients will be alienated. With motivational interviewing, even patients who decided against change at their initial visit expressed satisfaction with BSI.

Lesson 9. Coaches Must Be Well Trained and Well Supported

Motivational interviewing and other aspects of BSI require intensive training plus ongoing practice and feedback. We have achieved the best results with an initial 2 to 3 weeks of full-time training followed by weekly case conferences and regular feedback to coaches via audiotaped sessions with patients, with their consent. Some healthcare leaders recoil at such training requirements. However, undertrained coaches have lower confidence, see fewer patients, more frequently annoy or anger patients, and generate disillusionment by clinicians and staff. Relative to the duration of other clinicians’ and staff members’ training and the value coaches will generate, an initial training of 2 to 3 weeks and some additional development activities over several months are worth the investment.

Lesson 10. Faithful Continuous Quality Improvement Ensures Success

As for most new healthcare programs, a rapid-cycle quality-improvement approach bolsters and accelerates success. Key process metrics are (1) the proportion of eligible patients who complete screening questionnaires and (2) the proportion of patients with positive screens who see the coach at that visit. Most settings can achieve 90% or better on both metrics with 5 to 10 Plan-Do-Study-Act (PDSA) cycles over as many days.

 Slower PDSA cycles prolong dysfunction and generate disillusionment. Behavioral outcome metrics can be tracked so that coaches’ effectiveness in improving behavioral outcomes can be compared with that of other coaches and with outcomes obtained in prior studies. Multiplying these process and outcome metrics together can indicate the extent to which settings are attaining maximal possible improvement in population health.

Lesson 11. Health Information Technology Should Facilitate BSI

At a minimum, EHRs should score and store completed screens and assessments, offer coaches fill-in-the-blank templates to facilitate documentation, track key process and outcome variables, and generate real-time reports for internal and external quality reporting and population health management. Desirable add-ons include internet-based screening and assessment surveys that patients complete at home before appointments, kiosks or mobile devices for administering screens and assessment in waiting areas, real-time guidance for coaches, automatically printed session summaries for patients, and systems that proactively provide patients with reminders, check-in opportunities, and encouragement by email or text message. Until such technology is available, written materials can help guide coaching and maximize fidelity.

By improving performance on various quality metrics and reducing costs, robust BSI could help healthcare settings thrive under value-&shy;based reimbursement. The previous success of BSI in a variety of healthcare settings suggests that organizations not undertake pilot projects with the intent of determining whether BSI can be effective in those settings. Healthcare organizations that aim to thrive under value-based reimbursement should make and enunciate an unwavering commitment to effective BSI, expand their healthcare teams with well-trained and well-supported coaches, consider lessons learned from prior implementations in planning an initial roll-out, optimize processes during initial implementation at several initial settings, and rapidly expand to other settings.

When this paper was written, Dr Brown was a professor of family medicine at the University of Wisconsin School of Medicine and Public Health (Madison, WI). He is currently Southwest Michigan Medical Director at Concerto Health (Kalamazoo, MI).

US Substance Abuse and Mental Health Services Administration.

 Concept and design; acquisition of data; analysis and interpretation of data; drafting of the manuscript.

Richard L. Brown, MD, MPH, Concerto Health, 6563 W Main St, Kalamazoo, MI 49009. Email: drrichbrown@gmail.com.

1. Brown RL, Smith MA. Population-level quality measures for behavioral screening and intervention. 

. 2016;31(4):323-330. doi: 10.1177/1062860615577131.

Articles

This article describes lessons learned over the past 10 years while helping several dozen primary care settings implement evidence-based, cost-saving behavioral screening and intervention.

Lessons Learned in Implementing Behavioral Screening and Intervention

 Unhealthy behaviors are responsible for most chronic disease, ample healthcare costs, and around 40% of deaths. This study assessed patient satisfaction and behavioral outcomes for a program that enables healthcare settings to deliver evidence-based, cost-saving behavioral screening and intervention (BSI) services, as recommended by the US Preventive Services Task Force and others.

Thirty-one diverse primary care clinics and 2 other sites participated. Over 110,000 patients completed multibehavioral screening questionnaires. Those with positive alcohol or drug screens met with dedicated, on-site health educators for further assessment, and then received either reinforcement, brief intervention, or referral. Over 300 patients completed satisfaction questionnaires during year 3 of the program. A pseudo-randomly selected group of 675 patients participated in a 6-month follow-up telephone interview. In addition, for a short time in a pilot project at 3 clinics, 29 patients received depression screening, collaborative care, and behavioral activation, and 22 completed a 3-month follow-up telephone interview.

Mean patient satisfaction scores for all services exceeded 4.2 on a 5-point scale. Over 6 months, binge drinking episodes declined by over 20% for most subgroups. Recent marijuana use decreased by 15%. Depression symptom scores decreased by 55%.

With intensive training and ongoing support, cost-efficient paraprofessionals can deliver effective alcohol, drug, and depression screening and intervention services in busy healthcare settings. The approach holds promise for systematically addressing on a population-wide basis a variety of important behavioral health determinants and reducing related healthcare costs.

This article describes a program that successfully delivers evidence-based, costsaving alcohol, drug, and depression screening and intervention services in primary care settings.

Ample research has documented the effectiveness of these services and their return on investment, yet they are seldom delivered, largely because of provider and staff time pressures.

Trained and supported, cost-efficient paraprofessionals can deliver these services, elicit high patient satisfaction, and attain substantial declines in risky drinking, illicit drug use, and depressive symptoms.

Over time, clinical settings were able to modify work flow so that most patients received recommended services.

Several federal agencies and other authorities recommend that primary care providers conduct systematic tobacco, alcohol, depression, and obesity screening and intervention

: henceforth, behavioral screening and intervention (BSI). Dozens of randomized controlled trials and meta-analyses have demonstrated the efficacy of these services. Most such studies addressed 1 behavioral issue, involved selected consenting patients, and delivered services with more resources and tighter quality controls than most clinical settings could typically offer. Because of such studies, these services carry Grade A or B ratings from the US Preventive Services Task Force (USPSTF) and are consequently being advanced by the Affordable Care Act (ACA). Indeed, BSI could help reduce the $500 billion annual economic loss from tobacco use, alcohol use, and depression,

 as for every dollar spent, $4 is saved in 1 year through BSI for alcohol,

BSI starts with a few screening questions to identify patients who likely have behavioral risks or disorders.

 Positive screens require further assessment, which guides further service delivery, as per prior research. Low-risk drinkers or tobacco and alcohol abstainers receive reinforcement. Tobacco users and high-risk drinkers receive on-site interventions and ongoing support. Nicotine-dependent patients are offered pharmacotherapy. Patients with possible alcohol dependence, and patients with the desire but not the ability to modify tobacco or alcohol use, are referred for specialty care or other resources. In addition to usual pharmacotherapy and/or counseling, depressed patients receive collaborative care, in which typically a nurse or mental health professional coordinates management, promotes patient engagement in treatment during and in between visits, tracks depression symptom scores, and alerts treatment professionals to reconsider treatment plans if scores are not improving as expected.

 Collaborative care can also include behavioral activation, in which patients are engaged in behaviors that ameliorate depression symptoms.

Two major sets of barriers impede BSI implementation. Financial barriers are waning, as fee-for-service reimbursement for BSI is expanding under the ACA, and shared savings models are taking hold. Staff time is a more stubborn barrier. A primary care clinician would need 7.4 hours every work day to deliver all recommended preventive services, including BSI, leaving little time to address acute concerns and chronic illness.15 A potential solution is team-based care. Indeed, if the nation’s undersized primary care workforce will be able to serve the growing numbers of insured, elderly, and chronically ill patients, then primary care clinicians may need to delegate preventive services to other team members.

The US Substance Abuse and Mental Health Services Administration (SAMHSA) administers grants for demonstration projects to advance delivery of its alcohol and drug Screening, Brief Intervention, and Referral to Treatment (SBIRT) services. Successful projects develop a feasible service delivery model, serve a large number of patients, demonstrate patient satisfaction, and document reductions in substance use. Between 2006 and 2011, with SAMHSA grant funding, the Wisconsin Initiative to Promote Healthy Lifestyles (WIPHL) helped 33 clinical sites deliver SBIRT by expanding their healthcare teams with trained and supported “health educators” of varied backgrounds. WIPHL obtained separate funding to support delivery of depression screening and collaborative care for a short time at 3 sites. This paper reports on service delivery, patient satisfaction, and pre-/postintervention changes in behavioral outcome measures at the participating sites. The intent of this paper is to inform the design and implementation of programs that are intended to systematically identify and address multiple behavioral issues among large numbers of unselected patients, respect healthcare provider time constraints, help clinical settings meet several USPSTF recommendations, and serve the triple aim of improved health, patient experience, and cost control.

The Wisconsin Department of Health Services administered the project, contracted with the University of Wisconsin (UW) School of Medicine and Public Health and the Wisconsin Medical Society to administer a coordinating center, and contracted with the UW Population Health Institute to conduct program evaluation. SAMHSA provided funding and direction.

The Coordinating Center recruited sites through networking and selected them for their diversity in geographic location, population density (urban, suburban, and rural), payer mix, and their readiness to take advantage of the program. Sites were launched in waves to fill 18 slots. Additional sites were recruited as vacancies occurred. Initially, only primary care clinics were recruited. In later years, an emergency department (ED) and a hospital were added.

Interested sites completed a checklist to assess readiness and guide preparations. Sites instituted annual, universal screening for their entire patient population or for a subpopulation defined by provider or visit purpose. At most sites, receptionists distributed brief screening questionnaires. Sites were encouraged to include questions on multiple behavioral risks, in part to enhance acceptance of alcohol and drug questions.

 Staff members who checked vital signs reviewed questionnaire responses and referred patients with positive alcohol or drug screens to health educators. Consistent with SAMHSA requirements, health educators primarily conducted alcohol and drug assessment, then education, intervention, or referral as appropriate; they provided brief feedback and referral for other behavioral risks. They reported back to providers, who were encouraged to provide additional support and pharmacotherapy.

Sites were encouraged initially to identify champions, establish a quality improvement team with representatives from various staff segments, and design an initial work flow. The coordinating center provided consultation on best practices. Sites were encouraged to engage in monthly plan-do-study-act cycles to optimize 2 metrics: (1) the proportion of eligible patients who completed screens and (2) the proportion of patients with positive alcohol or drug screens who completed health educator— administered brief assessments.

The coordinating center guided each clinical site in hiring a health educator—a term that carries no legal implialcations in Wisconsin. Of the 44 health educators hired, 9 were master’s-level counselors or social workers, 33 held bachelor’s degrees, and 2 were high school graduates with special language or cross-cultural skills. Seven health educators had a bachelor’s degree in health education, 4 were certified health education specialists, and 1 was a certified chemical dependency counselor.

One week of distance training and a subsequent week of face-to-face training emphasized alcohol and drug screening, SBIRT, motivational interviewing, and cultural competence. Learning activities included interactive lectures, discussions, demonstrations, and role-play exercises, sometimes with standardized patients. Health educators had to pass a written examination and skills/ performance assessment. At their clinical sites, they received ongoing support through weekly conference calls and quarterly retreats. Monthly, with patients’ written consent, health educators submitted audiotapes of sessions and received structured feedback from trainers, who completed a modified skills checklist.

 Custom software guided health educators in realtime in delivering high-fidelity, evidence-based services. As they saw patients, health educators entered data, which allowed tracking of service delivery.

, the alcohol and drug screen included 3 validated questions on substance use

 and a direct question about illicit drug use or nonmedical use of prescription drugs. Questionnaires were available in several languages. For patients with positive screens, health educators administered the Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST)

 plus 3 questions on quantity and frequency of alcohol use.

 Based on their responses, patients’ substance use was classified as low risk, risky, harmful, or likely dependent. Health educators asked additional questions as required by SAMHSA and the Governmental Performance Results Act.

For low-risk patients, health educators affirmed and reinforced low-risk behaviors. For patients with harmful or risky use, interventions ranged from brief feedback, recommendation, and negotiation to motivational interviewing. 

 Health educators varied their interventions in response to severity of risk or disorder, patient preference and response, and time constraints. For likely dependent patients, health educators recommended referral for specialized assessment or treatment. Most educators referred interested patients to a treatment liaison who was based at the coordinating center and gathered additional information by phone, provided encouragement and support, and negotiated referrals considering availability, financing, and patients’ needs and preferences. Grant funds were available to fund treatment for patients without other means.

Evaluation included follow-up telephone interviews 5 to 8 months after initial service delivery. Subjects included a randomly selected sample of patients who received interventions or referrals. The chief purpose of the interviews was to assess for changes in behavior. Questions on substance use were identical to those asked by health educators. Of 1099 patients eligible to participate, 874 (80%) provided consent, and 675 participated (77% of those consenting).

During the third project year, health educators were asked to distribute “patient satisfaction” surveys to all patients in 1 month. Patients anonymously submitted the surveys, which were based on the 4-item task subscale of the well-validated Likert-type Working Alliance Inventory, 

 which assesses the strength of psychotherapeutic relationships. The items, which originally focused on “problems” and “change,” were modified to focus on alcohol and drug use. The impact of these modifications on item validity is unknown.

Of the 33 participating clinics, 3 with bachelor’s-level health educators added the Patient Health Questionnaire-2 (PHQ-2) depression screen

 (Table 1) to their written questionnaires, and delivered BSI for depression in up to 4 sessions per patient. Their health educators attended 3 days of additional training involving lectures, role-play exercises, feedback, and coaching. For patients with positive screens, health educators administered the PHQ-9.

They referred patients with possible suicidality or major depressive episodes for further assessment and possible pharmacotherapy and/or psychotherapy. For patients with PHQ-9 scores of 10 or greater, they delivered collaborative care as guided by a workbook derived from protocols from several prior randomized controlled trials. Such care included education on etiology, treatment modalities, and treatment effectiveness, and motivational interventions to maximize adherence to pharmacotherapy and psychotherapy and to engage patients in behaviors that lift depressive symptoms, including exercise, sleep hygiene, socializing, enjoyable activities, healthy nutrition, and relaxation. At each subsequent visit, patients responded to the PHQ-9 and anonymously completed a brief satisfaction questionnaire. During 3-month follow-up telephone interviews, research assistants readministered the PHQ-9 and the satisfaction questionnaire.

The University of Wisconsin Human Subjects Committee exempted the main project from its purview, because it involved delivering evidence-based services and limited program evaluation to federal requirements. Nevertheless, health educators informed patients of the risks and benefits of participating, and patients gave written consent to participate in 6-month follow-up interviews. The Committee approved the research protocol for the depression pilot study.

Of 33 participating sites, 9 were located in or near Milwaukee, 10 in or near cities with populations of 30,000 to 250,000, and 14 in rural areas. The sites were 17 large commercial primary care or multispecialty groups, 6 federally qualified health centers or look-alikes, 5 independent commercial primary care sites, 2 tribal health centers, 1 health department clinic, 1 hospital-based ED, and 1 hospital trauma center, plus other inpatient units. Over the course of 51 months, average duration of participation was 25 months.

Between March 2007 and June 2011, 113,642 patients completed screens. Those screened were 63% female, 58% white and non-Hispanic, 15% Hispanic, 14% black or African American, 11% Native American, 1% Pacific Islander, and 2% multiple races. Of those, 15% were aged 18 to 24 years; 23% were aged 25 to 35 years; 50% were aged 36 to 64 years; and 12% were 65 years or older.

shows the flow of patients through the clinical process and the final categorization of patients’ substance use. Throughout the project, 68% of patients who were eligible completed a screen, and 63% of those with positive brief screens saw a health educator and completed a brief assessment. A plot of these metrics over time, as reported by health educators, is shown in

. A linear regression confirmed significant improvements over time for quarterly completion of screens (

 <.001). Metrics declined toward the end of the project, when new clinics replaced experienced clinics that had become self-sufficient and exited the grant-funded program.

 also shows the results of the brief assessments for patients who completed them. Assuming similarity between screened and unscreened patients, and between patients with positive screens who did and did not undergo brief assessment, the estimated population prevalence of abstinence or low-risk use was 73%; risky use, 22%; harmful use, 2%; and likely dependence, 3%.

Alcohol/Drug Intervention and Referral Services

All patients in the risky and harmful use categories received a brief intervention. All likely dependent patients were encouraged to accept a referral for specialized assessment or treatment. Of the 1855 likely dependent patients, 452 (24%) expressed initial interest in referral. Health educators’ and the treatment liaison’s records showed that 183 (10%) entered treatment, though data capture may have been incomplete.

We compared 675 patients’ initial, face-to-face selfreports to health educators with telephone self-reports provided 6 months later to evaluation staff. Results are shown in 

. Reductions in risky drinking, mean number of drinks per drinking day, and maximal number of drinks over the prior 3 months exceeded 20% and were consistent across most subgroups. Patients over age 65 years and patients with Medicare had lower base rates, and manifested small reductions. Bachelor’s-level health educators elicited greater reductions than those with master’s degrees (

Among those who screened positive for substance use, reports of past 3-month marijuana use declined from 28% at baseline to 23% at follow-up (

 <.001). Reports of past 30-day marijuana use declined from 23% to 19% (

 <.001). The low rates of other drug use precluded assessments of intervention effectiveness.

To assess for possible differences in the accuracy of self-reports of substance use to health educators at baseline and to evaluators at 6-month follow-up, we compared self-reports of lifetime substance use. Of the patients who provided complete information, 7 (1%) of 673 gave disparate reports of lifetime alcohol use, 4 reported use only to the evaluator, and 3 only to the health educator. Of the 128 with disparate reports of marijuana use, 85 (66%) reported use only to the evaluator, and 43 (34%) only to the health educator (

 <.0001). Of the 79 with disparate reports of cocaine use, 64 (68%) reported use only to the evaluator, and 25 (32%) only to the health educator (

Expanding primary care teams with trained and supported paraprofessionals enables systematic delivery of widely recommended, evidence-based, cost-saving alcohol, drug, and depression screening and intervention services.

Richard L. Brown, MD, MPH

Articles

Lessons Learned in Implementing Behavioral Screening and Intervention

A Team Approach to Systematic Behavioral Screening and Intervention

Correspondence regarding, "The Psychopharmacology of Depression in the Managed Care Setting"