Welcome to Population Health unit seven, Population Healthy Management Interventions. This is lecture C. The learning objectives for lecture C are to, one, articulate successful strategies for human resource recruitment, retention, and training for population health staff. Two, describe the necessary health information technology for documentation of population health interventions. Behavioral health support has become more prevalent in primary care practices throughout the US. And sometimes this is embedded in primary care clinics. Sometimes it's embedded in community based organization. But behavioral health support would include licensed clinical social workers, psychologists, and even psychiatrists who are embedded in primary care offices. The clinic based services often include training for patients, assessments, embedded behavioral health specialists who provide counseling, and then facilitated referral. So if a patient needs a specific, for example substance use treatment, the behavioral health specialist in the clinic will often get a referral from the primary care provider. And then find a placement for the patient to get ongoing treatment from a substance use treatment or mental healthcare provider. Often in the community, we provide additional support through community health workers, community support specialists. Often who are volunteer, neighborhood navigators, disease specific peers and buddies, and technology enhanced support. Many patients have been given technology that allows them to use their cell phones to call into sessions with their care manager, or to report some of their physiologic parameters. Here we have the ABC approach. The first step, giving us the A in the ABC approach, is accelerated referral, or advanced specialty condition treatment. At J-Chip, the behavioral team which in embedded in the community and in the clinics, takes care of that referral when there is an intensive need for subspecialty psychiatric services. The B in the ABC approach refers to embedded services. We've talked about other embedded services on previous slides. On-site psychiatric treatment is another one. So primary care positions will often consult with psychiatrist about appropriate medication choices for depression or anxiety. Whole culture, the C in this ABC model, refers to caring and creating a culture of caring within the team caring for these complex patients. The team needs to be skilled in behavior change management. They need to be able to recognize psychiatric illnesses and basic substance use disorders. And they need to have a team attitude. They need to be willing to do whatever it takes to help patients with their behavioral, substance, and mental health issues. And the team needs to communicate that attitude to one another, as well as to the patients. We find that our patients often have a continuum of behavioral needs. And patients with complex medical needs, like diabetes may also have complex behavioral needs. They might for example meet diagnostic criteria for behavioral disorders. They may have an eating disorder, depression, substance use. Medically complex cases require tailored interventions and monitoring, as well as focusing on their behaviors. Patients often exhibit high risk behaviors and self harm. These patients must be given special attention in referral. Patients who are unable to change their behavior without more specific therapeutic intervention, have high assessed behavioral needs. However, this does not mean all patients with complex medical needs have high assessed behavioral needs. Some patients with complex medical needs actually have very low assessed behavioral needs. Some patients with low assessed behavioral needs maybe relatively healthy. Or maybe they're able to self initiate, self identify for services, or are able to learn and change their behavior, applying general feedback to their lives in a positive way. They benefit from support from coaching, from motivational interventions. J-Chip completes behavioral assessments in person and they compile those assessments in an electronic database. The following are some of the data points that they collect and have available for designing their interventions. General health status, medication adherence, cultural and linguistic needs, activities of daily living, nutrition, pain, stress, and sleep. All of these different factors contribute to health outcomes, sometimes in a very significant way. J-Chip also asks patients about their emotional status and depression, domestic violence and neglect. Because those things often need to be a part of an overall population health and care management intervention. The database produces a summary of the assessment. So when the case manager or health behavior specialist has completed and entered the assessment, the software will produce some assessment domains and possible interventions for each of those domains. So for example, in weight management we see that a patient has reported that their BMI is 32.6. Some possible interventions are that the case manager may be looking to the team for recommendations. The behavioral health specialist might present the patient with weight management interventions, the specialty clinic for a dietitian, medical weight loss, or bariatric surgery should that be required. So, there is a list of potential interventions related to that assessment of a high BMI. In the next example, if the patient is a current daily tobacco user and they're willing to quit, then the case manager may interact with the patient using motivational interviewing techniques. Or the help behavior specialist may be brought in for smoking cessation classes. So you can see that depending on the assessment results, there are a number of different interventions that can be employed. Now the interdisciplinary team process is important. Because when there's an assessment summary and a care plan, it's important for the team to do a process called rounding. So using these documents, the medical record, the assessment summary, the care plan, the team will sit down once a week, with or without the patient, it's the patient's preference. And they'll talk about the care plan, what interventions should be included and what goals are appropriate. The case manager will update the care plan on a laptop. So there's one care plan, although there are multiple people involved in the care of the patient. That care plan is generally kept up to date by the case manager. And that care plan outlines how the physician, how the community health worker, how the case manager might be interacting with the patient. Here's a look at how a care plan may look. This example is from Johns Hopkins Medicine. For this patient, detailed information about the patient and his demographics is included in previous pages. This page shows the details of the patient's care plan, such as the names of the case manager, The community health worker, and the primary care physician, details typically identified as part of an overall care plan. Also, here is what we described on the previous slide, the most recent rounding summary. And this rounding summary includes the identification of some of the global barriers that need to be addressed as well as some of the problems to be addressed. Such as weight management, the patient's BMI is over 35, and smoking cessation. The care plan also includes the specific intervention goals. This is a very quick summary, but at least it gives you a sense of everyone on the team looking at the same care plan and planning interventions with the patient accordingly. Developing goals, self-management goals, and then tracking progress toward those goals. Here's a short summary of personnel who play a role in population health interventions. Health coaches are often Masters-prepared coaches. There is no licensure for health coaches in the US, although there are several certification programs. Services provided by health coaches are often fairly structured. These services are generally one on one and focus on specific goals for specific patients. These goals are usually related to issues such as weight, nutrition, fitness, smoking cessation, etc. Health educators tend to provide educational classes or group classes within clinic sites. They can perform individual sessions, but most of that is done within a coaching framework. And these are essentially public health messaging and general health information messages that are sent via health educators within community clinics. Community health workers we described in previous slides. They arrange transportation, enroll members in a variety of community programs. Refer members to health educators or health coaches and provide telephone and community support. Case managers, whom we've described in some detail, conduct assessments, immunize patients, assist in the securing of housing and supplies, monitor adherence to medication and provide education on diet and nutrition. Case managers are often certified diabetic educators and focus specifically on diabetes services. In the Johns Hopkins Community Health Program, 50% of patients were diabetic and needed extensive diabetic education and training. J-CHiP provided that service through certified diabetic educators who are also nurses. And then health behavioral specialists are, as we described in previous slides, those who are able to identify available treatment options and programs and facilitate referrals, and provide some on-site counseling at clinics. Staff who perform in population health interventions need to be seasoned professionals. So at Johns Hopkins, J-CHiP doesn't hire nurses or licensed clinical social workers directly from college or from a professional program. They generally look for people in the clinical area who have had at least five years experience. And they prefer that some of that experience be outside of inpatient acute facilities and in the community or in the ambulatory setting. For health coaches, they often recruit from local colleges and university staff who are coming out with a Masters degree. That Masters preparation is sometimes in psychology, sometimes in health behavior programs. One strategy, in terms of retention, is to offer payment for certification programs. So while there is no required licensure for a health coach, there's several certification programs and some that are stronger than others. When J-CHiP hires a health coach, they often offer to help them prepare for the exam, and then pay for the exam. The same is true for health educators. Health educators are often coming from colleges and universities locally with a Masters degree in health education. J-CHiP offers ongoing education opportunities for them to obtain some continuing education credits and to obtain some advanced certification. J-CHiP likes to recruit community health workers from the neighborhoods where the patients live. So they require community health workers to live in one of the zip codes of the patients that they serve. These locally recruited community health workers have a unique ability to find patients, and they have established relationships in the community. And because they come from the same demographic, they can more easily develop a relationship of trust with community members. In the state of Maryland, there is not a certification requirement for community health workers. Some states, however, do require community health workers to have a certification, but not a licensure. And so it's important before starting a program of community health work that you check state regulations. There is no national certification or requirement. In Maryland, as I said, there is no certification requirement. So J-CHiP often hires community health workers from the community, people who may have been nursing assistants or worked in geriatric care. And then they offer them initial training in community health work, social determinants of health, and give them a certification. They do offer college tuition assistance as a recruitment strategy and retention strategy. So that the community health workers can advance their careers. And many of them are interested in doing so to become licensed practical nurses, registered nurses, or social workers. The case managers are either registered nurses or licensed clinical social workers. And the same is true of health behavior specialists. J-CHiP requires that the licensed clinical social workers are certified clinicians, and have had at least five years of experience. Some of which should be in an ambulatory setting. They offer payment for case management certification. There are at least two national case management certifications. And when that is obtained, case management certification requires ongoing, continuing education credits. And J-CHiP pays for nurses to continue their accreditation and certifications by getting those additional credits and going to both local and national conferences. For health behavior specialists, J-CHiP found that it's helpful to recruit Masters in social work students who need placement and supervision. And they offer them, again, case management certification when they're eligible for that, and ongoing, continuing education. Generally, with people who are embedding in primary care clinics, they have a collaborative relationship with the clinics, in that they allow the clinics to screen and help them hire candidates. It's very important that the interdisciplinary team have a bond, and have some degree of a collaborative working relationship. So we've spent a great deal of time in our health systems, community, social services, and public health interventions parts of our conceptual model. And now, we move to the very end of our conceptual model, to where we are engaging stakeholders, monitoring program implementation, and seeking to continuously improve programs to maximize health outcomes. And in this section, which is highlighted on the left, the stakeholders are our providers, our patients, and the community. And we try to engage with them, so that we can record and produce better health outcomes for the population, as well as for sub-populations. In the last part of our model, we are very interested in continuous quality improvement. So that if we do find it in our evaluation that there are some changes that we want to make to the model. We can take that data and feed it back to the intervention team so that they can use that data to develop better interventions with better outcomes. And finally, we do evaluate our programs for a number of different factors, such as, clinical improvements, patient self-management, and cost and utilization. Underlying our evaluation of the skills of our analytical team and IT infrastructure and support. We’ll spend a moment on the relationship of data and analytics to population health interventions. First of all, we do constant surveillance and assessment to determine population health needs and patterns. And then, over time, to track population level health changes or trends that result from the interventions that we've deployed. In terms of surveillance and assessment, we collect data and refresh data. Pharmacy data on the daily basis, EMR data on a monthly basis, and enter that data into our population health database and claims on a monthly basis. Predictive modeling scores are updated also on at least a monthly basis. We identify population subgroups in need of particular interventions using our data and analytics such as, risk stratification and risk scoring. And we monitor intervention processes, procedures, and implementation. So if you think back to the re-aim model, we talked about population health interventions. We said that we wanted to make sure that we were monitoring the implementation, and whether or not the population health interventionists were providing and following protocols, as had been described. And also, whether individuals were adhering to protocols. And whether there was persistence in keeping their self-management skills, and their behaviors in line with the protocols. And then finally, we want to evaluate intervention effects on designated clinical, behavioral, community, health system, and economic outcomes. So, we're very interested in evaluation that leads to population health intervention research. So what is population health intervention research? It's research that involves the use of scientific methods to produce knowledge about policy and program interventions, that operate within or outside the health sector. And have the potential to impact health at the population level. In our remaining time our objective is to look at the basic processes of population management and what IT is required to support population health management interventions. One of the things that population health much do is engage patients. And that includes identifying our populations, mapping and tracking populations, delivering care to populations and subgroups, and then coordinating care across the continuum of care. All of these different interventions and processes must be administered, monitored, and reported. And so we have processes in place, and IT is required to produce ongoing reports and ongoing evaluation of our program. In a framework for IT enabled population health, the advisory board lays out the following processes for population management. One, population identification. Two, population tracking. Three, care delivery. Four, cross-continuum care management. Five, patient engagement. And six, administration, performance monitoring, and reporting. And I'd like for you to think of the model that we had displayed on the last page in terms of these processes. These overarching process of patient engagement and reporting. And then we'll look at each of these processes within population health, and what kind of IT is required to support those. We begin with population identification. And I've structured this so that we look at first, the requirements of population health management and then systems that could be used to support population identification. We have previously discussed the requirements for identifying a population. Data acquisition and aggregation. And we identified many data sources. The data needs to be normalized, brought together, and aggregated. So that we have what we call a 360 patient view or as much information about the patient on an individual level as we can. We need analytics-based population definition and contracting, predictive modeling, algorithmic population identification, patient and provider attribution. We need to know who the providers are who are interacting with the patient. And oftentimes, the patient can have many different providers. And we'd like methodology to identify who those providers are and who's providing the majority of the care. The kind of systems that we can use to support population identification, primarily, are data aggregation platforms. So is there an EMR clinical data repository in the organization? Clinical data warehouses, health information exchange data warehouses? In the state of Maryland, there is a very nice Chesapeake Regional Informational System for patients, which is an HIE. It's populated primarily with in-patient and acute data from hospitals, but will expand. And that data needs to be accessible by case managers to the extent possible, so that they can see the full scope of services that are being provided to patients. The claims data source is also important in identifying the population. Statistic modeling tools are necessary within that environment, and algorithms and analytics for inclusion and exclusion criteria by population. Those algorithms have to be written and coded so that they not only make clinical sense but that we are able to access the data sources that we do have. And then often we need processes and algorithms for attribution. For population tracking, there are several populations that we require to be tracked for the overall total population. For population tracking, there are several populations that we require to be tracked from our overall total population. We have subgroups such as high-risk, high utilization populations, who get very intensive services, from any number of high-risk interventionist. Such as case managers, health behavior specialists, or health coaches. For the chronic disease group, we need to identify which patients have particular chronic diseases. Such as asthma, diabetes, or heart failure. And for the total population, we want to track preventive care to see that health promotion and disease prevention care is being given as an opportunity for all members of a population. We map care against care guidelines. We map populations against care guidelines, and map against care over time. So how many times in the past five years has the patient, who's diabetic, received hemoglobin A1c? What is the trajectory of those scores? Are they increasing? Are they decreasing? The systems that are most used to track populations over time are disease and population registries. Many EMRs will provide that capability now, within the electronic medical record. And many population health software packages, also include disease and population registries. It's important to track care being provided by the population health and care management staff. We want to have a clinical data view. Because we are clinicians who are delivering service to patients. E prescribing. We love to see what prescriptions are prescribed, and what prescriptions are picked up by the patient from the pharmacy. Clinical documentation is important so that we can track the kinds of interventions that patients are getting in communication tools. We want to have a patient-centered view of interventions and also decision support for interventions. So if there are, for example, in our example for Johns Hopkins Community Health Partnership, if we have an indicator on our assessment of body mass index over a particular number, and the patient is obese or morbidly obese, we want to have embedded in that software the capability to support clinicians in terms of what interventions might be administered for that patient. Much of the clinical data is captured in electronic medical records or in population health systems and registries that have the patient-centered view. It is important for case managers, and care managers, and physicians to follow patients across the continuum of care. Meaning that while they're in primary care, we need to have documentation systems that document patients' progress and self-management, and self-care behaviors. When a patient has an exacerbation and is admitted to the hospital, it's important for the care managers to have access to patient data while the patient is in a more acute or intensive setting. We like to have the ability to have some real time video interaction at remote locations. So if one of our patients is traveling or one of our patients is far removed from Baltimore City, we like to have the ability to have either video or audio interactions with patients. And the kind of systems that support that of course, are care management systems that capture documentation. Mobile EMR and registry, we like to have access to the EMR while we're in the community. Home monitoring data capture and telemedicine capabilities which might allow us to have real time video interaction with patients in remote locations. In terms of patient engagement, we love to have documentation of patients' contact information and update that in real time and really have multiple communication modalities with the patient. Whether it's access to an electronic medical record, patient portal, a text- based communication system, a call center. For difficult-to-reach patients, or those who have not a steady or reliable home address, we often given patients cell phones with limited text. Or, with limited phone numbers so that we can get in touch with them, and they with us, if they need to. And then generally, for all of the interventions and all of the programs that have been mentioned, it's a requirement for us in terms of improving our programs, to have the ability to monitor our performance. And report back to our funders, or to those who are delivering the program. It's imperative to have the ability to produce performance data against care plans at the clinician, physician, and organization level. To be able to do E-measure calculations and reporting using systems such as dashboards, or contract management and financial systems, revenue cycle systems, and financial decisions supporting cost accounting systems. So what are the greatest challenges in having an IT-supported population health strategy? The first is data normalization. Claims data is very essential to population management. But claims data are often qualitatively poor and insufficient in accuracy and granularity for care delivery decisions. So a bill may come in, a claim may come into the organization that says a patient has diabetes, but on further review, the patient actually does not have diabetes. The patient has had other medical problems, and diabetes was coded incorrectly, for example. So while claims data is very helpful at the population level, before individual intervention is launched, it's important to do some real clinical screening with patients. EMR data, therefore, is essential. And to the extent that claims data and EMR data can live in a consistent database, a population data base, it's important to do so to help validate some of the claims data and diagnoses that are submitted. EMR data must be reconciled with the claims data around one single data model. And that actually maybe the single hardest task for population health management programs. Another challenge is risk stratification. It can be a very complex process. There are several factors that make it a complex process. Quantitative clinical and demographic data are sometimes easy to get from claims and EMRs. We talked about needing to have that in a single database but they are data that are available. Qualitative data is less available and issues such as social issues, housing, food availability, transportation, availability of community resources, are often things that you have to personally ask a patient about in a face-to-face assessment prior to starting an intervention. And so having that, is difficult data to have therefore, in the whole population. And so while social determinants are extremely important in terms of determining health outcomes, the data are hard to collect and often require a great deal of face-to-face interaction. A third challenge will be determining the most effective way to intervene. So the most effective mode may be, for lower risk patients, via telephone and text. Or maybe, for high-risk patients, in person. Delivering care in the most cost and clinically effective venue is a priority. And so we want to know that our interventions will be effective. And we have seen evidence in the literature for example, that telephone intervention with high-risk patients has not been shown to be clinically effective in reducing health outcomes and quality outcomes. IT tools can be used to track interventions and can be used in terms of documenting patient characteristics and outcomes. But again, we have to consider the cost of deploying that across a population versus deploying it within a subgroup of a population that may be more likely to use the tools that are available, and supported by IT. In summary, in lecture c of Population Health Management Interventions, we continue to articulate some successful strategies for human resource recruitment, retention, and training for population health staff. And finally, we describe the necessary health information technology for documentation of population health interventions. This concludes unit seven, Population Health Management Interventions. To briefly review, we described the population health data necessary for segmenting into risk-cohorts. Differentiated the key cohorts of a population by degree of risk. Analyzed the root causes of risk in a population by utilizing socioeconomic date, behavioral data, electronic medical record data, and other demographic data. We also explain the processes and key decision points by which interventions are prioritized for segments of the population. Delineated interventions and staff who were deployed for high-risk, rising-risk, at-risk, and low-risk populations. Described three types of deployment strategies, models, for population health management. Finally, we articulated successful strategies for human resource recruitment, retention, and training for population health staff, described the necessary health information technology for documentation of population health interventions. Thank you.
