Chronic disease management solutions have been difficult to implement at scale and in real-world clinical settings. We report on the preliminary results of an intervention designed to address these issues.

 We performed a prospective analysis on data collected during the early trials of our intervention to assess its effectiveness and feasibility.

 We partnered with 2 physicians to recruit adults with diabetes or prediabetes.  A health advisor from our team engaged each participant for 12 weeks with print information and individually tailored telephonic and text message—based counseling according to the participant’s preference(s). The primary measure recorded was glycated hemoglobin (A1C), in addition to several secondary clinical and behavioral measures and participant satisfaction. All measures, except participant satisfaction, were recorded at baseline and 12 weeks, and changes were measured with non–variance-assuming 

 We recruited 167 participants and received participant-reported data from 122 (73%). For A1C, we recorded data for 56 participants (34%); of these, we saw a mean decrease in A1C of 0.91% (95% confidence interval [CI], 0.44%-1.38%; 

 = .0053). For participants with diabetes, the decrease in A1C was 1.53% (95% CI, 0.77%-2.28%; 

 = .0009), and for participants with prediabetes, we saw no progression toward diabetes (

 = .2282). We saw significant improvements in most secondary measures (for most, 

 <.0001) and favorable participant feedback (mean satisfaction rating = 2.97 of 3 [95% CI, 2.91-3.00]).

This management model shows some promise in improving disease management in a scalable manner.

The American Journal of Accountable Care. 2016;4(4):16-20

Chronic disease in the 21st century United States has increased in prevalence for the past 15 years, and its management and treatment now consume the majority of healthcare spending.

 Intensive behavioral counseling is one strategy that has been deemed both necessary and cost-effective in the effort to combat this trend. Per the guidelines of the US Preventive Services Task Force and the Affordable Care Act, intensive behavioral counseling can be covered as a preventive measure against cardiovascular disease for obese or otherwise at-risk individuals.

 Although cost-effectiveness estimates vary widely, they are all well within the commonly accepted intervention threshold of $50,000 per quality-adjusted life-year added—below which, an intervention is considered to be worth the cost.

However, several issues threaten its scalability and efficacy. One, is availability of time on the part of physicians and patients. Team-based coordination has been suggested as a solution to increase the time available to physicians, but only the most qualified clinicians (ie medical doctors, physician assistants, certified registered nurse practitioners, certified clinical nurse specialists) can be reimbursed for counseling; furthermore, patient time constraints remain a barrier even when clinician time constraints are accounted for.

 Telephonic counseling has also been suggested as an option, but recent results have not shown promise.

To address this issue, we set out to develop, implement, and test an innovative remote intensive counseling intervention for type 2 diabetes (T2D) as a model for other chronic disease management interventions. Utilizing “processes of change” from the Transtheoretical model and motivational interviewing techniques, we delivered a multi-channel (mail, phone, text, and/or e-mail), 12-week intensive behavioral counseling intervention to individuals with diagnoses of diabetes or prediabetes. The aim was 4-fold: 1) deliver the intervention to the participant in a comfortable environment and timeframe to improve participant engagement, 2) act as an extension of the patient’s primary care provider to provide a coordinated care intervention (not an intervention delivered in parallel), 3) deliver an intervention that produces lasting results in value-based outcomes, and 4) deliver an intervention that is scalable to a large patient-per-advisor ratio.

Here, we present our preliminary results and implications to those attempting similar interventions or to those searching for novel management interventions.

Our design was prospective and followed a cohort receiving the intervention over 12 weeks. We had 3 main hypotheses: 1) participants would engage and find the experience favorable due to their ability to select the time and method of engagement; 2) clinicians would find the experience favorable due to our ease of enrollment, maintenance of contact with them, and our contributions to their patients’ value-based outcomes; and 3) it would produce significant positive results in clinical and behavioral measures of diabetes management.

Participants receiving the intervention were being served by 2 Alabama primary care physicians—one in an urban center and the other in a rural area&mdash;so as to test the intervention independent of geography. The physicians served as the enrollers and recruiters, and eligibility criteria for participants were as follows: 1) the clinician felt that the participant would benefit from intensive behavioral counseling for diabetes management or prevention; 2) the participant felt that he or she would benefit from intensive behavioral counseling for diabetes management or prevention; 3) the participant’s last glycated hemoglobin (A1C) measurement was at, or above, 5.7; and 4) the participant was aged at least 18 years. The principle of participant opt-in was paramount, as we hypothesized that individuals who wanted intensive behavioral counseling would benefit more than those who simply followed the advice of their clinicians. All identifying information was stored in a HIPAA-secured database. Enrollment for this sample began on June 16, 2014, and ended June 17, 2015, with the last postintervention assessment occurring on September 29, 2015. In all, 167 individuals were identified and consented to participate in the intervention.

The duration of the intervention was 12 weeks. Every participant was given standard print information and management-assisting tools in addition to being assigned to a personal health advisor, who was in weekly communication with the participant to provide support, motivation, additional education, and accountability.

The provided print material and tools were as follows: 1) 2 informational booklets—1 on diabetes self-management and 1 on healthy eating principles; 2) 1 interactive booklet encouraging participants to contemplate their goals, external facilitators, and external barriers; 3) booklets in which individuals were asked to track medication adherence, diet, and physical activity for personal and clinician viewing; 4) a blood glucose logbook; 5) a pedometer; 6) an informational card (to be carried in a purse or wallet) on suggested procedures for others to follow in the event of hypoglycemia; 7) a refrigerator magnet reinforcing healthy eating principles discussed throughout the program; and 8) 1-page informational cards, termed “mailers,” focusing on individual topics of diabetes management, sent once every 1 to 2 weeks depending on the depth of the topic.

The health advisor engagement procedure was as follows. First, the participant would be engaged through mail and phone, and could choose text message and/or e-mail based on availability and preference. The participant would receive telephone sessions once per week, at a time chosen by the participant, to discuss the topic of the week, in addition to the information they had already received. The health advisor would use motivational interviewing techniques to discover participant barriers and foster the participant’s creation of solutions that worked for them.

 Various processes of change from the Transtheoretical model would be taught and/or suggested as strategies to overcome barriers and achieve health behaviors, as needed.

One small goal for the upcoming week would be decided by the participant at the end of each session. In order to assist participants in staying on track, 3 text messages would be sent to the participant per week; all of which would be based on standard aspects of the week’s topic or on personal matters discussed with the health advisor in the previous counseling session. The aim was usually to encourage, remind, or hold the participant accountable to their goal for the week and in the long term. Alternatively, if the participant chose e-mail but not text message as their preferred method of communication, 2 to 3 e-mails would be sent each week to accomplish the same purpose as the text messages. Any inbound communication (eg, an additional question outside of the scheduled interactions) from the participant to their health advisor was received and responded to. For matters potentially of clinical concern, the participant was encouraged to contact their enrolling clinician, and the health advisor followed up on the matter the following week.

All primary and secondary measures were recorded via participant report at baseline and again at the 12-week program completion date. The primary measure used to assess the efficacy of the intervention was A1C, measured in percent glycation.

Secondary measures were based on individual aspects of diabetes management, which were covered throughout the intervention. We were interested in testing whether comprehensive education and counseling on diabetes management made a clinically relevant difference, as well as whether counseling on individual management behaviors changed those behaviors. There were 12 secondary measures, with the first 2 being: 1) body mass index (BMI) and 2) knowledge regarding diabetes management. Knowledge was identified by our team in 4 key components of diabetes understanding: that diabetes involves higher than normal blood sugar, that diabetes is a chronic condition, the causes of elevated blood sugar, and the potential complications if left unmanaged. Participants were asked to describe their understanding of diabetes at baseline and 12 weeks, and whether a participant accurately understood all 4 concepts was measured binarily.

Other secondary measures included: 3) health self-efficacy, which was measured—for participant convenience&mdash;by 1 item, which read thus: “On a scale of 1-5, with 5 being the highest, how confident are you in your ability to improve your health?” We measured binarily whether a participant rated their confidence as 5; 4) skipped doses of medication per week (number); 5) if a participant skipped any number of doses of medication per week (binary: 0 times per week or ≥1 time per week); 6) if a participant was engaging in moderate to vigorous physical activity 2 or more times per week; 7) if a participant was eating at least 7 healthy meals per week, defined as half nonstarchy vegetables, a quarter starchy vegetables and/or grains, a quarter protein, an optional small portion of fruit, and an unsweetened beverage; 8) whether a participant had received at least 1 formal eye exam in the past 12 months; 9) whether a participant had received at least 1 formal foot exam in the past 12 months; and 10) the number of emergency department (ED) visits a participant had in the past 3 months.

The final 2 secondary measures were: 11) the participant’s understanding of healthcare cost management, identified with 4 key strategies for maintaining low healthcare costs: managing diabetes (lowering A1C), utilizing generic medications whenever possible, getting appropriate insurance, and appropriate use of the clinician’s office and ED—with “appropriate” defined as understanding what was a wise use of each and what was an unnecessary use of each. We measured binarily whether a participant understood all 4 concepts by asking them to explain what they understood about these strategies. Finally: 12) satisfaction&mdash;each participant was asked at the end of the intervention to rate their satisfaction with the program by circling a smiling face, a neutral face, or a frowning face, then providing personal comments.

For all numerical measures, the mean and 95% confidence intervals (CIs) were calculated at baseline and at 12 weeks. For all measures described above as binary, a score of 1 or 0 was given to the participant, and the percentage scoring a 1 was presented. Then, all 12-week means were compared with baseline means using 1-sided 

 tests, not assuming equal variance, to discover or disprove statistically significant differences. One-sided 

 tests were used in place of 2-sided because all of our outcomes had an ideal direction of improvement (either up or down), except for prediabetes A1C, in which we wanted to either affect a decrease or affect the prevention of an increase. Weight change, although not necessarily directionally tied to diabetes management, was also measured directionally for a decrease because the average BMI of our participants was above 30. In the future, we would like to perform and report analyses of variance and covariance between our behavioral and clinical measures; however, for this preliminary report, we only wanted to examine the efficacy of the intervention on each measure individually. All analyses were conducted using Microsoft Excel and Google Spreadsheets.

Of 167 individuals consenting to participate in the intervention, 122 were retained and provided data on 1 or more secondary measures throughout the program (73.05% participant retention), with 101 providing baseline data on the primary measure (A1C). Of those 101, 56 self-reported the 12-weeks’ primary assessment (55.45% primary reporting retention). Participants were inconsistent in their responsiveness to secondary measures. Numerical measures are presented in 

 and management threshold measures are presented in 

. Where CIs passed the minimum or maximum value possible, they were presented as the minimum or maximum.

Intervention participants saw significantly improved behavioral and clinical measures. In the context of our model condition, A1C and diabetes management behaviors significantly improved, signifying better glycemic control. Although this decrease was seen mostly in participants with diabetes, we did not detect a significant change (2-sided 

 = .2282) in A1C for participants with prediabetes, signifying that the progression to diabetes may have been delayed or prevented. Also of note is that intervention participants saw a meaningful improvement in A1C, but not in weight. We believe this to be due to the fact that we did not focus on rapid weight loss, but rather on small and sustainable goals that would increase self-efficacy and reduce blood sugar, such as appropriate medication adherence and moderate physical activity. Longer-term iterations of this intervention may indeed yield weight loss, but that remains to be seen and was not a focus of our intervention.

For our secondary measures on individual aspects of diabetes management, results were quite favorable. We saw significant differences post intervention in the percent of participants meeting management thresholds for all related aspects except cost management understanding and ED use reduction. These results seem promising to our hypothesis that our intensive counseling intervention would produce positive behavioral and clinical results. Although we hypothesized that self-efficacy, medication adherence, diet, and physical activity were the arms of our intervention most correlated with clinical improvement, in a future formal study, we hope to test the covariance between all secondary values and A1C.

Our hope was that, in providing the participant with the authority to choose his or her time of contact and making the intervention 12 weeks (instead of the common 24-week intensive behavioral counseling intervention), we would see high engagement and retention throughout the intervention. However, health advisors were unable to systematically connect with participants at previously agreed-upon times and by previously agreed-upon means; thus, some participants consented with the physician to enroll, and then were unreachable by health advisors. In these cases, information was delivered and participant-reported measures were requested by mail, but response rates were low. Qualitatively, health advisors determined that common reasons for this were the lack of a static schedule and that baseline measures were collected during the same call in which the program was explained, resulting in sometimes long initial calls. Although we thought this would make the data-collecting process more personal, and we explained that ensuing calls would be much shorter, this could have been a factor in the loss to follow-up.

Another method to be improved on was that we did not ask participants to ensure they had up-to-date A1C scores preceding baseline and shortly following 12 weeks. This may have increased our sample size for the primary measure, and is now standard protocol for all interventions following this analysis. A final limitation was that we did not conduct follow-up assessments at 24 and 48 weeks (approximately 6 months and 12 months) to assess the sustainability hypothesis. Since this iteration of the intervention was completed, all of these changes have been implemented with the intent to strengthen the efficacy of the intervention and our power to measure that efficacy. 

Despite the aforementioned limitations, our feasibility study had several promising strengths. One is that, for the participants that did engage, we were able to provide extremely personalized counseling with multiple contact points and a conversational environment.  Participants routinely commented in an unsolicited fashion that they appreciated the accountability, the nonjudgmental setting in which to voice problems and brainstorm solutions, and the empathy and compassion of the health advisor, and they rated satisfaction with the program an average at 2.97 of 3 stars. Another advantage of the intervention was that it demonstrated successful outcomes for a meaningful percentage of participants. This is meaningfully different than fee-for-service reimbursements, such as the Chronic Care Management codes that focus on activity and process rather than effectively driving for better outcomes. Our cost per intervention per participant was approximately $300, which is 37.5% less than the annual cost of reimbursement for face-to-face chronic condition management programs sponsored by CMS.

 Furthermore, the clinicians received regular reports on participant outcomes, but put forth no extra time outside the participant’s normal visits. If we can replicate these results from remote counseling in a larger study group, we believe we can prove the scalability of remote counseling over face-to-face counseling in terms of clinician time spent and cost-effectiveness.

Our preliminary results from a remote intensive behavioral counseling intervention to improve diabetes management showed significant clinical and behavioral results at lower than usual financial cost while requiring less time on the part of the clinician. Although loss to follow-up and sustainability are being addressed in current iterations of the intervention, the preliminary results, in terms of clinical differences and scalability, are promising. A full-scale study comparing the intervention with usual care is underway.

The authors acknowledge with the utmost gratitude the following individuals/groups: esteemed colleague Hala Fawal, MBA, MPH, whose scrupulous work and keen perspectives were paramount to the development of this project; the health advisor team, including Barbara Schuler, MPH; M’Kayl Lewis, BS; Camilla Green, BS; Michael McMorris, MSW; Tamara Wilson, MS; Beth Vaughan, MPH; Tina Lu, MPH; Brass Bralley, MA; and Hazeza Kochi, MPH; the operations and development team, including Robert Ginter, MBA; Leigh Anne Gilbert, BA; Uma Srivastava, MS; and Maya Madden; all physician partners; and all the program participants who were partners with us in the pursuit of better health.

 Pack Health, LLC (MB, SB, DP, MR, DM, WW), Birmingham, AL; MCPHS University, School of Pharmacy Practice (DP), Boston, MA.

The implementation was supported in part by an unrestricted educational grant from Eli Lilly and Med-IQ.

 Drs Patel and Rasulnia, Ms Belshe, Mr Bowman, Mr Masom, and Mr Wright are employees of Pack Health, LLC, which uses the evaluated model in their practice. Dr Patel is on the advisory board for Novo Nordisk and Sanofi, and has received lecture fees for speaking at the invitation of commercial sponsors (Novo Nordisk, Sanofi, and Merck).

 Concept and design (DP, MR, WW); acquisition of data (SB, DM, WW); analysis and interpretation of data (SB, DM, DP, WW); drafting of the manuscript (SB, MR, WW); critical revision of the manuscript for important intellectual content (SB, MB, DP, MR, WW); statistical analysis (SB); provision of study materials or patients (MR, WW); obtaining funding (MR, WW); administrative, technical, or logistic support (MB, DM, DP, WW); and supervision (DP, MR, WW). 

 Sean Bowman MPH, Pack Health, LLC, 3613 6th Ave S, Birmingham, AL, 35222. E-mail: sean@packhealth.com.

1. Chronic care chartbook. National Association of States United for Aging and Disabilities website.  http://www.nasuad.org/sites/nasuad/files/hcbs/files/191/9519/ChronicCareChartbook.pdf. Published 2010. Accessed November 8, 2016.

2. Ward BW, Schiller JS, Goodman RA. Multiple chronic conditions among US adults: a 2012 update. 

3. Gerteis J, Izrael D, Deitz D, et al. Multiple chronic conditions chartbook: 2010 Medical Expenditure Panel Survey data [AHRQ pub No 14-0038]. Agency for Healthcare Research and Quality website. http://www.ahrq.gov/sites/default/files/wysiwyg/professionals/prevention-chronic-care/decision/mcc/mccchartbook.pdf. Published April 2014. Accessed November 27, 2015.

4. The Patient Protection and Affordable Care Act [HR 3590]. Government Printing Office website. http://www.gpo.gov/fdsys/pkg/BILLS-111hr3590enr/pdf/BILLS-111hr3590enr.pdf. Published 2010. Accessed November 27, 2015.

5. Final recommendation statement: healthful diet and physical activity for cardiovascular disease prevention in adults with cardiovascular risk factors: behavioral counseling. US Preventive Service Task Force website. http://www.uspreventiveservicestaskforce.org/Page/Document/RecommendationStatementFinal/healthy-diet-and-physical-activity-counseling-adults-with-high-risk-of-cvd. Published August 2014. Accessed November 12, 2015.

6. Lin J, Zhang P, Bardenheier B, et al. Cost-effectiveness of the new U.S. Preventive Services Task Force (USPSTF) recommendations for behavioral counseling interventions for adults with cardiovascular disease risk factors. Presented at: American Diabetes Association, 75th Scientific Sessions; June 5-9, 2015; Atlanta, GA.

7. Hoerger TJ, Crouse WL, Zhuo X, Gregg EW, Albright AL, Zhang P. Medicare’s intensive behavioral therapy for obesity: an exploratory cost-effectiveness analysis. 

. 2015;48(4):419-425. doi: 10.1016/j.amepre.2014.11.008

8. Østbye T, Yarnall KS, Krause KM, Pollak KI, Gradison M, Michener JL. Is there time for management of patients with chronic diseases in primary care? 

9. 29 CFR 2590.715-2713 - Coverage of preventive health services. Legal Information Institute website. https://www.law.cornell.edu/cfr/text/29/2590.715-2713. Accessed November 17, 2015.

10. 2016 obesity counseling reimbursement fact sheet. Ethicon website. http://www.ethicon.com/sites/default/files/managed-documents/045443-160323_obesity_counsel_reimbursement.pdf. Published 2016. Accessed November 8, 2016.

Articles

The American Journal of Accountable Care

A chronic disease management company presents promising preliminary results from their remote intensive behavioral counseling intervention, aimed at addressing type 2 diabetes.

Evaluating the Efficacy of a Bundled Chronic Condition Management Program

Mental health conditions and substance use disorders (SUDs), referred to as behavioral health conditions, are a leading cause of global disability.

 In the United States, an estimated 1 in 3 adults suffers from one or both of these disorders.

 These individuals die, on average, 25 years earlier than the general population as a result of suicide or comorbid physical conditions, such as cardiovascular disease, diabetes, respiratory distress, or infectious disease (HIV/AIDS). They also incur significantly higher medical and societal costs.

Estimated spending on behavioral health conditions varies depending on the study. One analysis based on 41 million individuals covered under Medicare, Medicaid, or commercial health plans who were treated for a behavioral health condition in 2012 estimated a cost of $525 billion, nearly half of the $1.7 trillion spent that year on all health-related expenditures.

 Another analysis estimated a lower cost: $201 billion in 2013.

 Both analyses, however, noted that spending on behavioral health conditions was the highest category of any other medical condition in the United States, topping cardiovascular disease and trauma (

The number of people seeking services for behavioral health conditions is expected to increase over the next decade due to the Affordable Care Act, which mandates that insurance companies cover screening and other services for mental health and substance abuse conditions, and the Mental Health Parity and Addiction Equity Act, which requires that insurers provide equal coverage for behavioral and physical health conditions.

Despite the physical, economic, and societal consequences of behavioral health conditions, about one-third of individuals with these disorders receive no treatment, and the vast majority of the rest receive substandard treatment. This gap between needed care and received care increased by about two-thirds between 1997 and 2010.

 Indeed, 70% of Americans in a recent poll from the Kennedy Forum felt that the country needed significant changes in the way it manages behavioral health conditions.

Today, many individuals with behavioral health disorders receive care in the primary care or medical specialty, not behavioral health, setting.

 Of those, up to 80% receive no treatment or substandard treatment for their behavioral health disorder.

 This includes prescribing antidepressants for mild depressive symptoms, which are relatively ineffective, and the use of psychotropic medications with no documented behavioral health diagnosis.

 Psychosocial approaches, which studies find can be just as effective as medication, are also underutilized.

Prevalence of Comorbid Physical Conditions

Approximately 68% of those with behavioral health issues have comorbid physical conditions, typically chronic conditions such as asthma, low back pain, and diabetes.

 These individuals have higher morbidity and mortality rates and are more likely to be nonadherent with medication than those with only a behavioral or physical health condition.

 For instance, 21% of patients with chronic kidney disease (CKD) have comorbid depression regardless of their disease stage.

 These patients are twice as likely to be hospitalized and have a 41% increased risk of all-cause mortality compared with patients with CKD who do not have depression.

 Patients on hemodialysis have a 2-fold increased risk of death.

Patients seen in the behavioral health setting who also require services for comorbid physical conditions report difficulties accessing medical care.

 In one survey of 1670 adults with mental illness, one-third had difficulties accessing primary care, with 13% attributing this to stigma around their behavioral health condition.

 Further, the care received is less likely to include prevention and screening, and this limited clinician time may lead to less time spent on psychosocial issues.

 A review of the Veterans Affairs National Psychosis Registry showed poor adherence to medications for both psychiatric and medical conditions in patients with serious mental illness.

 Not surprisingly, individuals with comorbid mental and physical health conditions use more health-related services than those without, even when controlling for the higher prevalence of physical health conditions among those with behavioral disorders.

Behavioral health disorders require long-term, chronic-care management similar to that needed for chronic physical conditions such as hypertension, diabetes, and asthma.

 Such conditions respond best to management under a chronic-care model, which promotes enhanced access and care continuity, uses clinical information systems and decision support tools to identify and manage patient populations, provides self-management support to patients, and links patients to community resources. Under this model, providers also track and coordinate care and measure performance changes over time. Study results suggest this model can be implemented cost-effectively and even demonstrate some savings.

In the behavioral health setting, a chronic care model requires integrating mental and substance abuse treatment with physical health management. This “integrated” model is described as “the care that results from a practice team of primary care and behavioral health clinicians, working together with patients and families, using a systematic and cost-effective approach to provide patient-centered care for a defined population.”

Settings may involve embedding behavioral health professions in a primary care setting or primary care providers in a behavioral setting, or developing a close relationship between behavioral and primary care practices despite different physical locations, even using telemedicine (

Despite the robust literature demonstrating the benefits of integrated care models,

 behavioral and physical healthcare delivery have traditionally operated in separate spheres.

 Bringing the two together could not only improve outcomes, but also reduce costs.

 Indeed, analyses suggest that integrating medical and behavioral services in Medicaid populations could save states between $3 and $9 billion, while integration could save all payers (including commercial) between $26.3 and $48.3 billion (2012 dollars)

 Updating benefit designs to reflect these improvements needs to be part of the move to accountable care organization (ACO) behavioral health integration.

These savings are already occurring at the state level. For instance, Missouri’s Chronic Care Improvement Program, an integrative model designed for individuals with severe mental illness, such as schizophrenia, saved $8.3 million in its first year managing 6757 members, even with a $775,000 increase in outpatient costs.

 In addition, the state’s Community Mental Health Center healthcare homes (similar to patient-centered medical homes) for Medicaid-eligible individuals with severe and persistent mental illness, comorbid SUDs, and certain chronic health conditions reduced overall healthcare costs by 8.1% while significantly improving individual and societal outcomes.

 When Kaiser Sacramento integrated medical and substance use treatments in primary care clinics for individuals in an outpatient chemical dependency recovery program, per-member-per-month (PMPM) costs dropped more than 50%, with significant declines in hospitalization rates, inpatient days, and emergency department use.

The Affordable Care Act created a pathway for greater integration of physical and behavioral health services when it expanded the development and use of ACOs. These integrated models of care are typically built around patient-centered medical homes. Payment is typically linked to the quality and cost of care, with value-based, rather than fee-for-service—based, reimbursement. A common reimbursement model is shared savings, in which the ACO shares in any savings with the payer over a defined timeline. In some instances, ACOs assume the risk for spending more than the financial target. Other ACOs take on even greater risk under capitation: such reimbursement models provide a financial incentive to hire case managers, social workers, pharmacists, and other allied health professionals to work with patients with comorbid behavioral health issues.

ACOs are charged with managing the health of a patient population, which requires robust data systems, predictive analytics, and coordinated care. The goal is to achieve the Triple Aim of healthcare today: improved outcomes, improved patient experience, and reduced cost.

 In 2015, 70% of Americans had access to an ACO, 44% to 2 or more, and between 15% and 17% (49 to 59 million) received care from an ACO.

Integrating behavioral health management with physical health management in an ACO model could significantly improve population health management and outcomes, contributing to an ACO’s ability to survive and thrive under risk-based reimbursement models.

 This approach also fits with the ACO’s team-based, coordinated care approach.

Some ACOs are improving their delivery of behavioral health care. For example, Crystal Run Healthcare ACO in New York, which participates in the Medicare Shared Savings Program (MSSP), has 3 psychiatrists in its medical building. These psychiatrists share a waiting room with their medical colleagues and use a connected electronic health record (EHR) system. They also formed a mental health assessment team comprising of primary care and specialty physicians who meet with mental health specialists to discuss cases requiring comanagement.

The norm, however, is a continuation of siloed care. A survey queried 257 nationally representative Medicare, Medicaid, and commercial ACOs between 2012 and 2014, and was augmented with qualitative data from structured interviews with clinical leaders at 16 ACOs. It found that just 14% had fully integrated behavioral health and primary care teams and just 42% included behavioral health specialists among their providers.

 Another survey found that more than one-third of ACOs had no formal relationship with behavioral health providers despite the fact that the majority of contracts included behavioral health metrics. Although 84% of ACOs had at least 1 contract with a payer that included responsibility for behavioral health services, just 66% of the largest commercial contracts included behavioral health services in the total cost of care.

In another study, researchers analyzed Medicare claims for 20% of traditional beneficiaries with a mental health condition who received care in a Pioneer or MSSP ACO between 2008 and 2013 (they could not analyze SUD claims given federal confidentiality laws). They found cost savings in 2012 for Pioneer ACOs, primarily from a reduction in hospitalization, but the savings did not continue in 2013, nor were any savings observed in the MSSP. Neither ACO program showed significant differences in outcomes.

 However, the authors also found little evidence of integration between behavioral and medical management. Instead, the majority of ACOs contracted out behavioral health services. Improving outcomes for individuals with mental health issues, they concluded, required that ACOs adopt evidence-based approaches associated with greater effectiveness in this population, such as integrated services.

Even if ACOs are providing mental health services, far fewer are providing SUD services.

 A survey of 635 substance abuse treatment centers found just 15% had a signed agreement to be included in an ACO, while just 6% and 4%, respectively, planned to be connected with an ACO or were currently negotiating to be connected.

Publicly owned and private nonprofit SUD treatment facilities, as well as those in more competitive markets and those accredited by the Joint Commission, were most likely to have such contracts.

 However, these contracts don’t necessarily result in fully integrated services; they may only cover referral to behavioral health specialists. In addition, treatment centers in the Northeast were more likely than those in the Southeast and Midwest to sign contracts with an ACO, with those most likely located in states with 50 or more ACOs.

 The authors concluded that the results of the survey “suggest that ACOs are not effectively integrating treatment and services for individuals with SUDs into medical settings.” This, in turn, continues the fragmented, suboptimal, high-cost care received by this high-risk population, most of whom suffer from multiple chronic conditions.

Yet, most ACOs understand the interrelationship between behavioral and physical health on overall outcomes and costs. An analysis of data from 90 Medicare ACOs between December 2012 and June 2015, including 72 site visits, found that nearly all of the ACO staff interviewed understood that behavioral health disorders contribute to overall health outcomes and spending and that most were working to better coordinate behavioral and physical health services.

 They were working to integrate behavioral health and primary care, increase access to social workers, and enhance referral networks. Some embedded primary care providers in behavioral health facilities and included pharmacists and community resource specialists on treatment teams. One ACO even developed a mental health “center of excellence” for primary care referrals of complex patients who required significant behavioral and physical health services.

 However, the study also found significant barriers to greater integration of behavioral health in an ACO, including a lack of behavioral health care providers, access to data, and sustainable financing models.

Successfully Integrating Behavioral and Physical Health Services in an ACO

Successful integration of behavioral and physical health services in an ACO should focus on 5 areas: financial incentives, data sharing, legislative changes, quality measures, and alignment with existing initiatives.

Financial incentives in any healthcare delivery system must be aligned with expected outcomes. Thus, the value of integrating behavioral and physical health services is low under a fee-for-service system, which pays for the episode of care provided regardless of outcomes and provides no reimbursement for the additional time and effort required to coordinate care.

 The value is much higher, however, under a capitated system in which providers are essentially paid for keeping their population as healthy as possible.

Thus, value-based reimbursement models have entered the behavioral health sphere in the hope of spurring greater integration and improved outcomes. For instance, Minnesota’s Medicaid program capitates Hennepin Health ACO for behavioral and medical services, which encourages greater coordination between providers.

Capitation is but one reimbursement model. States may also require that ACOs share savings with behavioral health providers or leave it up to the ACO itself to compensate behavioral health providers. For example, Maine includes behavioral health services within the total cost of care (TCOC) calculations for its ACOs, leaving it up to the organizations to determine how to financially compensate the behavioral health provider. Given that the TCOC impacts the organizations’ receipt of shared savings, this is designed to promote greater accountability across settings.

Massachusetts embeds behavioral health services within its 3-tier payment system: comprehensive PMPM payments for an optional set of behavioral health services; quality incentive payments based on 23 quality measures, including 4 related to behavioral health; and shared savings payments based on cost savings on non—primary care services, including behavioral health. The higher the level, the higher the potential compensation.

Successful ACOs use robust information technology systems to collect and analyze data on their patients. These systems are typically tied into scheduling and revenue cycle systems to provide a holistic view of the state of the practice and patient population at any given time.

 However, medical providers have been upgrading their information technology, particularly their EHR systems, for years thanks to the Health Information Technology for Economic and Clinical Health (HITECH) Act, which offered incentives for the development and meaningful use of such systems. The act, however, excluded mental and behavioral health providers and treatment facilities from this incentive program.

 Thus, behavioral health providers are far behind their physical health counterparts in the collection and use of data.

This could change if the Behavioral Health Information Technology Act and other legislations pass that are currently pending in Congress. The Behavioral Health Information Technology Act would extend incentives for meaningful use of EHRs to psychologists and mental health professionals who provide clinical care at psychiatric hospitals, mental health treatment facilities, and substance abuse treatment facilities.

 In addition, the Office of the National Coordinator of Health Information Technology has released grants through the State Innovation Models Initiative to enhance IT integration into behavioral health.

 ACOs could also require that their behavioral health providers participate in a joint EHR system and even offset some of the costs.

The lexicon for behavioral health diagnoses that is typically used for structured, coded information in health IT is also lacking. These structured data are required for the type of data analysis and clinical decision support necessary for successful population health management.

 Just as challenging is the lack of interoperability among existing health information systems and the lack of behavioral health data fields in medical EHRs or physical health fields within behavioral health EHRs. Some states are beginning to provide support for more integrated systems, however, while larger ACOs may have the resources to modify existing systems to facilitate greater coordination.

Legislative changes in the federal regulation that prohibits sharing patient information related to alcohol and drug treatment without additional patient consent (beyond the standard HIPAA form) are needed. Without these data, ACOs are unable to provide the level of analytics required to manage the health of a population and identify patients for targeted outreach.

Billing issues also require changes. Just 28 state Medicaid systems allow providers to bill for primary care and behavioral health services on the same day even though there is no federal restriction. This creates a significant barrier to integrated and coordinated care.

 Some legislative actions may seem minimal, but they can send a powerful message. For instance, in Arizona, it took legislative action to strike down a law that required separate waiting rooms for patients receiving mental health services and those receiving medical care.

States should also encourage the training of additional behavioral health care specialists. A survey of 90 Medicare ACOs found that a scarcity of mental health professionals posted a significant barrier to the greater integration of behavioral and physical health. The ACOs cited poor Medicare reimbursement as one reason for the low number of providers willing to see Medicare patients.

 Another survey of 2900 primary care providers found that 67% reported difficulties connecting their patients with behavioral health specialists because of a shortage of providers, as well as insurance barriers.

 The survey was conducted in 2009, before the full impact of the Affordable Care Act and expanded access to insurance occurred. We do not yet know if this expansion changed access to providers or if benefits featuring narrow networks, behavioral health carve-outs, and high patient cost share blunted the benefit of this access.

Quality measures play an integral part in the effort to improve delivery of behavioral and physical healthcare services, ensure appropriate access, and align incentives under value-based reimbursement.

 Although large national databases show 510 quality measures that address behavioral

health, just 5% to 10% are included in major quality reporting programs, such as the Inpatient Psychiatric Facility Quality Reporting Program, the Physician Quality Reporting System, the National Quality Forum, and quality measures for Medicare and Medicaid programs maintained by the Centers for Medicare & Medicaid Services.

Indeed, until recently, just 1 of 33 quality measures required for Medicare ACOs, depression screening, was directly related to behavioral health.

 Yet, the depression screening is much more of a comprehensive primary care measure than measuring behavioral health performance. New measures for 2017 add rates of depression remission and response to treatment at 12 months. Measures involving screening for and treatment of SUDs are under consideration. In addition, several measures, such as shared decision making, medication reconciliation, and patient ratings of physicians, also apply.

Adherence to behavioral health-related quality measures is poor, however, with reports demonstrating that patients receive recommended care based on quality initiatives about half the time. Conversely, recommended care is provided about two-thirds of the time for other chronic conditions, including diabetes and cardiovascular disease (

In a recent article in Health Affairs, Pincus et al highlighted 5 key areas for improvement in the realm of behavioral health quality measures

• Expanding outcomes measurements built around the concept of “recovery” and including patients and families in their development.

• Developing structural approaches to enhance the capacity of organizations and providers to provide effective, quality care. This could include accreditation, certification, recognition, and payment programs, such as Medicaid Health Homes and the National Committee for Quality Assurance’s patient-centered medical home and specialty recognition programs, which incentivize outcome-focused care.

• Encouraging the integration of behavioral and physical health services, as described earlier in this article.

• Expanding quality measures related to the use of effective patient-centered psychosocial interventions, such as talk therapy.

• Adding quality measures around the management of SUDs.

 highlights state efforts in these areas.

Integrating behavioral and physical health services should build on existing programs within ACOs. For instance, some ACOs already employ social workers and other behavioral health specialists for short-term support. New York state is working with existing “health home” programs, which provide care for complex patients, to grow them into ACOs and is providing grants to behavioral health providers to encourage them to better collaborate with health homes. Those health homes are already required to support care management across physical and behavioral health services and create links to community support and housing.

ACOs should also investigate the numerous state and federal grants available for behavioral health integration. For instance, the SAMHSA-HRSA Center for Integrated Health Solutions has awarded more than $26.2 million in grants to 100 community-based behavioral health organizations to support integration of primary care services into these settings.

A 2010 report from the Agency for Healthcare Research and Quality highlights several essential measures needed to facilitate an integrated mental health model in the primary care setting

• Normalize mental health into mainstream medical practice. This requires cultural shifts to move away from the stigma of behavioral health problems and recognize these conditions as chronic health conditions no different from diabetes and asthma. It also requires redesigning workflows and providing physicians with the technical and leadership skills they require for full integration.

• Integrate reimbursement mechanisms. This includes eliminating separate coding and billing procedures.

• Create a roadmap for implementation. This includes research that identifies the most effective and cost-effective primary care models for this population and the development of decision support tools that identify patients who require integrated services.

Integrating behavioral and physical health services within an accountable care organization offers a significant opportunity to address both behavioral health conditions and substance use disorders, as well as to improve outcomes and reduce costs.

Treating Behavioral Health Disorders  in an Accountable Care Organization

: Virtual assistants formed through the union of artificial intelligence (AI) and natural language understanding (NLU) technologies (eg, Siri) are routine in our daily lives. The purpose of this study was to test patients’ acceptance of a virtual health coach (VHC) who emulates medical staff to activate the patient to engage in a conversation with their physician about healthy lifestyles.

: Eighty-nine primary care patients agreed to test “new technology” while waiting in the exam room for their physician. Patients used a handheld tablet computer to interact with an animated VHC who asked questions about common health concerns, including weight, smoking, drinking, stress, or medication adherence. During the brief intervention, the VHC advised the patient to have a conversation with the physician about the identified lifestyle concern.

: Patients were comfortable with the technology and said it positively impacted their feelings about the organization. They talked to the VHC, rated the technology as easy to use, and felt that they would use it again and that other patients would enjoy it. Many patients tried to talk to the VHC about the specific reason for their visit. 

: VHCs using AI and NLU technologies are acceptable and useable, and they positively impact patients during exam room “downtime.” The technology easily integrates into the clinic workflow and does not extend the length of the visit. Although patients were willing to talk with the VHC, in many cases, they confused the VHC’s line of questioning about general lifestyle concerns with the purpose of their visit.

The American Journal of Accountable Care. 2016;4(4):41-45

Patient “activation” is a recognized concept that describes a person’s knowledge, confidence, and skills to manage his or her own health and healthcare; thus, patients who are activated in their health and healthcare generally have better outcomes and lower costs.

 Yet, 75% of patients are not adequately informed or confident, nor do they have the necessary skills to manage their health. In particular, healthy lifestyle behaviors are problematic for most patients. Those with lower activation rarely participate in behaviors that require active self-management, such as exercise, eating a healthy diet, and managing stress.

Interventions to increase patients’ activation in their health have focused on the development of skills and advice tailored to the patient’s level of activation. For example, training patients to ask their doctors questions and giving them support to do so has been shown to increase both activation levels and participation in healthcare.

 Observations of patient—physician encounters suggest only 10% of the primary care visit is devoted to lifestyle topics and, further, that these 2 minutes mostly entail the physician bringing up a lifestyle concern. Such limited and ineffective interactions have prompted calls for additional communication training for primary care providers to help them engage patients in lifestyle issues in a time-efficient and effective manner.

A virtual health coach (VHC) offers a solution to the problem of patient activation and patient—physician discussions of lifestyle issues. Current technology enables the design of an animated VHC who can understand everyday language and respond to patients in a conversational manner. Patients are accepting of virtual agents and, interestingly, some disclose more to a computer interviewer than to its human counterpart.

 Standardized behavioral change and motivational interventions can be built into VHC interactions.

 The ability to standardize conversations allows challenging and time-consuming interactions to be brief, consistent in quality, and scalable. VHCs could ultimately cost less than real providers in specific contexts.

 Thus, they are an attractive alternative to training physicians to be proficient in effective lifestyle discussions by using motivational interviewing strategies, which has proven to be challenging at best.

The purpose of the current study was to test the feasibility of a technology innovation in the primary care setting, building on an earlier unpublished study of a prototype. We wanted to determine patients’ acceptance of a VHC—with artificial intelligence (AI), natural language understanding (NLU), and animated capabilities&mdash;by engaging in a verbal interaction with the coach while waiting in the exam room for the doctor. The VHC is intended to emotionally connect with, encourage, and support patients in preparation for a conversation with their doctor about a lifestyle issue that the patient identifies as concerning.

Physicians and their medical assistants (MAs) from women’s health and adult medicine clinics participated in the study. MAs gave patients a handheld tablet and provided the instruction, “We are testing new technology and need your help.” There was a prominently displayed “start” button. The Northern California Kaiser Permanente Institutional Review Board allowed a waiver to a signed consent document.

The VHC oriented the patient to the purpose of the tablet app and provided instructions for the microphone. As the VHC “spoke,” the tablet’s screen simultaneously displayed the VHC and the text of the VHC’s words on the screen to allow patients to read the conversation. The patient was asked to select a general health concern (eg, weight, smoking, drinking, stress, taking medications, or none of the above) and respond to a series of questions, interspersed with empathic and supportive statements, all with the purpose of encouraging the patient to bring up the concerning health topic as soon as the doctor entered the room. The VHC used motivational questions to prime the patient for the conversation.

 At the end, there was a “learn more” button for reading materials relevant to the lifestyle topic the patient selected. Upon entry, the physician asked, “What did you discuss with the health coach?”

Patients completed surveys disseminated by the authors and brief interviews at the end of their visit. Physicians were surveyed via e-mail.

The VHC (an AI-based cognitive software solution) was created by a contracted vendor, and it spoke English and Spanish. It was built on the Alme platform, which combines a healthcare domain model and NLU with an avatar to drive interactive conversations. Samsung tablets were connected to a secure, corporate wireless network.

Data were captured via the tablet app, including verbal interactions, the lifestyle concern selected, and the mode of communication (talking, typing, and tapping an icon). Surveys used a 1-to-7 Likert scale to measure aspects of the experience, and patients also provided verbal feedback. Means and standard deviations were calculated for survey responses, and the user experience data were also described.

Eighty-nine patient—VHC conversations were captured. One patient refused to participate, and 63 patients (12 male, 48 female) were available for surveys and interviews. Thirty-four patients were from women’s health clinics and 29 were from adult medicine clinics. The average age of males was 56 years (range = 33-82 years), and the average age of females was 41 years (range = 18-84 years). Eight physicians participated.

Patients talked out loud to the VHC in 82% of the interactions and typed responses 6.6% of the time. When an icon was available, patients “tapped” on it 74% of the time versus voicing their lifestyle concern audibly. Most patients rated the technology as easy to use, felt other patients would enjoy using it, and were willing to use it again, while the modal response was “neutral” about their own enjoyment. Many reported the positive impact of the technology on their feelings toward the healthcare organization. Patients’ survey data are presented in 

Patients said they felt encouraged by the VHC to talk with their doctor, and 51% of them reported they engaged in a conversation with their physician about the health topic they selected on the tablet. Twenty-six percent reported they had not talked with the physician about this topic previously. “None of the above” was the most frequently (66.3%) selected lifestyle concern. Of that 66.3%, 24 patients specifically stated, “I’m healthy,” 26 vocalized a specific reason for their visit, and 9 of the verbal responses of patients were unintelligible. Weight was the most frequent concern (18%), followed by stress (7%), medication adherence (6%), drinking (2%), and smoking (1%).

The verbal comments regarding feedback on the VHC included it was “fun,” “interesting,” “very cool,” “impressive,” as well as “It was great to have something to do while waiting” and “I love this about this organization.” A number of other comments were similar to “It would be great if it related to the reason I’m here.” Comments specific to the VHC included, “weird movements,” “robotic voice,” and 1 patient stated, “I didn’t like it.”

Sixty-seven percent of the patient—VHC conversations were completed, meaning patients reached the final screen. There were Internet stability issues in some cases, and technology malfunctions in others; for some, the physician entered the room before the patient–VHC exchange was completed.

There was a significant negative correlation between age and comfort with using “touch screens” (

 <.00003; R² = 0.26). No relationship was detected between the variables of age and comfort using voice recognition technology. Physicians reported minimal impact of the technology on the duration of the visit, and they were, on average, neutral about the usefulness of the VHC. 

The purpose of this feasibility study was 2-fold: first, to determine patients’ acceptance of a VHC while waiting for their physician, and second, to ascertain whether patients would engage in a conversation with their physician about a lifestyle concern identified through the virtual interaction. We believe this is the first report of using AI/NLU technologies with patients in the primary care setting and consider the findings worthy of further investigation.

Patient responses to the technology exceeded expectations. Regardless of age, patients were willing to use a VHC while in the exam room and to speak aloud to it. As in earlier reports, patients were positive about the technology.

 Patients also felt encouraged by the VHC to engage in a conversation with their physician. Interestingly, although patients think others will enjoy the VHC, their own enjoyment ratings were “neutral.”

Patient responses regarding the intent of the VHC were perplexing. A sizable number selected “none of the above” regarding lifestyle concerns, which contradicts current healthy lifestyle investigations.

 Many patients neglected the instructions about “overall health” and “not your specific reason” for today’s visit and told the VHC about a specific problem. Nevertheless, half of the patients reported they had a conversation with their physician about the topic they discussed with the VHC, and about half of this group said they had not discussed this topic with their physician previously.

Physicians were “neutral” about the VHC. However, although several expressed verbal concern about increasing the encounter time, this outcome was not reported. Physicians were more positive in the earlier prototype test. One said, “It makes me have a conversation relevant to the patient. I want to discuss smoking, but this lets me know the patient is worried about stress. It saves time.” In the current study, little time was spent orienting the physicians about the purpose of the VHC, and this may, in part, be responsible for their lower satisfaction ratings. In addition, the desired outcome of the VHC interaction was for the patient and physician to have a discussion about a lifestyle concern, but this did not happen consistently.

The technology can improve. The voice recognition technology did not always capture the verbalizations of the patient, and the secure wireless connection was unstable. Some patients described the voice of the VHC as “robotic” and the movements as “unnatural” and “jerky.”

Based on greater success of an earlier prototype, we speculate that the appearance, movement, and/or voice of the VHC could be distractive to patients. In our unpublished study of 11 patients that displayed a posturized image of a coach on the tablet, patients followed instructions, identified a lifestyle concern, and discussed it with the physician in every case. In the current study, the VHC was an animated, semi-realistic virtual agent (similar to those in a video game) with empathic gestures and expressions. Elements, or the combination of these features, potentially diverted patients’ attention from the instructions, and, although the multi-featured VHC might be acceptable in a video game, the healthcare setting is inherently different and patients have ingrained expectations

The application of behavioral and psychological strategies through technology (ie, behavioral intervention technologies) has tremendous promise for healthcare.

 We see the VHC as the inaugural attempt to integrate motivational and patient activation strategies with AI/NLU technologies. Future VHC iterations could include voice enhancement, modified movements, and altered appearance.

Strengthening the instructions about the purpose of the interaction will be important, as well, to ensure the desired conversation takes place. Additional advancements could leverage data from electronic health records to personalize the exchange, and the VHC’s questions could address specific visits, such as birth control considerations for women’s health and alcohol screening in adult medicine.

The introduction of a VHC in the healthcare setting could be considered “disruptive technology,” which is defined as “a technology that shakes up an industry, or a groundbreaking product that creates a completely new industry.”

 According to Clayton Christensen, it often lacks refinement, does not perform well because it is new, and appeals to a limited audience. We think this describes our VHC concept: it needs refinement, better performance, and broader appeal, and it has untapped potential to improve quality, efficiency, health outcomes, cost, and patient satisfaction.

The authors would like to thank their colleagues from the iFund, NextIT, and The Permanente Medical Group administration, as well as physicians and medical assistants who made this study possible.

Kaiser Permanente (SDP, DER), Roseville, CA.

This study demonstrates patient acceptance of a virtual health coach while waiting for the physician to enter the primary care exam room.

December 2016

Evaluating the Efficacy of a Bundled Chronic Condition Management Program

Treating Behavioral Health Disorders in an Accountable Care Organization

Can a Virtual Coach Activate Patients? A Proof of Concept Study

The Post-Election Future of ACOs

The Impact of Health Information Technologies on Patient Satisfaction

Beyond Regulatory Requirements: Designing ACO Websites to Enhance Stakeholder Engagement

Quality in Healthcare and National Quality Awards

High-Dose Flu Vaccine Prevents Symptomatic Influenza and Reduces Hospitalizations

Tobacco Control in Accountable Care: Working Toward Optimal Performance