DISEASE MANAGEMENT: A SPECIAL REPORT
Steps used in creating a system that markedly improved asthma outcomes and cut costs at Hartford Hospital are typical of what's involved.
Some argue that the extreme health problems in urban areas result from inability to deliver care. Others feel this inequality is inherent in the population's attitude toward seeking appropriate services. The so-called urban health penalty, the confluence of such circumstances as poor nutrition, poverty, and unemployment — with deteriorating housing and violence — has created a deepening health crisis.
Jessie Wing, M.D., from the Centers for Disease Control, wrote: "To begin addressing some of the issues that affect the health of our inner city communities, we must be willing to develop solutions that are timely, culturally sensitive, and acceptable to these communities. These complex health problems will require thoughtful, multidisciplinary approaches and interventions at multiple levels. Partnerships between inner city community leaders, health officials, and policy makers need to be fostered to develop successful interventions."
Potential solutions must account for reality. To operate effectively in the emerging "managed lives" environment, hospitals must redesign themselves as providers of effective, efficient care with predictable and beneficial outcomes — not only in the hospital setting, but wherever care is given — and emphasize prevention and health maintenance rather than treatment of disease or disability.
As the complexity of health problems increases, it will mandate a thoughtful, multidisciplinary, team approach with access to interventions at various levels of intensity. If costs are to be controlled and better-quality care is to be provided, there must be a focus on market-based-system management for a particular disease. Disease management is a comprehensive, multidisciplinary, collaborative, approach to health care delivery and prevention.
There are three main components of DM:
- A knowledge base that defines the natural history and economic structure of a disease for each particular juncture in the disease process and includes guidelines regarding the care to be provided, by whom, and in what setting;
- A health care delivery system of partnerships between primary care providers, subspecialists, social organizations, and others that provide coordinated care throughout the disease process, breaking down the traditional boundaries that fragment our system; and
- A continuous-improvement process that measures and evaluates clinical, financial, satisfaction, and health-status outcomes; refines treatment standards; and continuously ensures the highest quality of care.
DM and hospitals
Hospitals are the single largest provider of health services in their communities. In 1990, 38 percent of all health care expenditures — $269 billion — was related to hospital care. Not-for-profit hospitals often provide a disproportionate amount of critical care, such as intensive care, to lower-income citizens. Often, these services are not reimbursed.
To thrive in this climate requires identifying patients who cannot manage their needs independently outside the hospital, and helping them obtain the most appropriate level of care. The increasing complexity of health problems requires a thoughtful, multidisciplinary team approach that allows for interventions at various levels of care.
That is what we helped to implement — we think successfully — at Hartford Hospital, an 800-bed tertiary medical center in Hartford, Conn., which provides health care to a culturally and economically diverse population. The Adult Primary Care Practice serves a predominantly inner-city Hispanic population of approximately 15,000 patients. The payer mix is quite diverse. Approximately 60 percent are Medicaid (25 percent of these are managed Medicaid for whom we assume 100 percent risk). The remaining 40 percent are fee-for-service patients. Half of those are Medicare; the other half are "self pay," which is essentially lost revenue.
The initial step is choosing a disease to manage. DM is most applicable to diseases for which much information exists. Evidence-based protocols are easily developed, and outcomes that are sensitive to change can be measured. The diseases should be high volume, high cost, high morbidity and mortality, and a financial risk to the organization.
With these criteria in mind, we chose to target our asthmatic population, thus the name of the initiative, Asthma Control & Education, or ACE.
Once the disease population is identified, careful evaluation of the organization's administrative database on the specific disease is essential. Costs, both direct and indirect, regarding emergency department visits, hospitalizations, physician office visits, and disease-specific charges are a few of the elements one needs to examine. The data should be collected and tracked consistently, according to such modifiers as ICD-9 codes, discharge statistics, or claims data. Whichever you choose, ultimately based on the ease of acquiring data, continue to track the data consistently, preferably monthly or quarterly.
Next, we conducted a survey of benchmark programs and performed an extensive literature search for best practices, accepted guidelines, and thorough disease overview. Numerous resources (Medline and other Internet sites and discussion forums) can provide valuable information on existing programs and clinical-practice models. This information will serve as a guide when designing educational materials and content of specific programs.
We visited regional and national health care centers that had demonstrated success in initiatives with similar populations. We also found national conferences on disease management to be invaluable for obtaining information about programs and allowing us to network and build collegial relationships with others in the field. The National Managed Health Care Congress, the Zitter Group, and the Disease Management Association of America are just a few of the organizations that hold such meetings throughout the year.
We also obtained a lot of valuable information from Lovelace Healthcare Systems, which offers in-depth seminars on disease management.
The resources available are considerable and those that have experienced early success in disease management have been extremely generous with their time and in sharing their experiences and lessons learned along the way.
Next, we identified the key players, assembling a team that represented all disciplines involved in the disease-fighting process. It is imperative that a well-thought-out vision be articulated, and be incorporated into the mission of the organization. Is the focus to be inpatient or outpatient? Will it be primary-care-based or subspecialty driven? Will it be based in community practices or at a large tertiary center? It is important to define the scope of the program to anticipate the resources that will be needed to execute and maintain it.
ACE is a primary care, outpatient-based educational program targeting high-risk, inner city, predominantly Hispanic patients with asthma. Although focused on the outpatient setting, this comprehensive program begins in the emergency department and continues throughout the hospital stay and into the home with an environmental survey and home assessment. The patients are seen initially for three educational sessions with follow-up at three, six, and twelve months.
A "physician champion" is necessary to lead the charge and provide the clinical expertise in regard to guidelines, best practices, and clinical outcomes. Using outcomes research, our own "champion" identified data that were clinically relevant, as well as feasible to measure. We found that it was important to complement clinical measures with economic and quality-of-life outcomes, while not limiting the patient and burdening the staff with laborious exercises in data acquisition.
Next, we recruited in-the-trenches players who were essential to the success of the program. The foundation of our ACE program is our asthma nurse coordinator. Since more than 90 percent of the patients in our program are Hispanic, we utilized bilingual and bicultural nurse educators.
Other team members necessary to ensure success include pharmacists, social workers, substance abuse and smoking cessation counselors, ER staff, respiratory therapists, inpatient nurses, and subspecialty consultants in pulmonary medicine, and allergy and immunology.
In addition, partnerships with the managed care organizations in which patients are enrolled are important. In our program, case workers representing the HMOs conduct home assessments and environmental surveys on all of our patients enrolled in their plans. They provide outreach services as well as enrollment data.
Local agencies can provide community based research, disseminate culturally sensitive educational materials, and help promote change in health and health care policy in the community. We worked closely with the Hispanic Health Council, which helped us identify barriers to care that are specific to our patients, and helped provide resources to overcome those barriers.
The council also provides a nonthreatening environment within which we can deliver effective education. We hold monthly group support sessions where members of the community can convene in a familiar setting and share experiences. The power of patient-to-patient education is tremendous and should be encouraged wherever possible.
Finally, local health departments can provide resources for new initiatives and help collect and coordinate data on the broader determinants of health that inform decision making and resource allocation throughout the community. Once we had assembled our team, we created a flow map of the care provided. We included all of the members of the team and actually walked through each individual's area of responsibility.
The most valuable participant in this process is a patient from the community. Who better to identify process issues than someone who has experience in each segment of the health care environment, from the emergency department to inpatient admission, to the follow-up appointment process in the primary care practice, to the subspecialty referral, and even to the visiting nurse at home? By mapping out these care practices, we were able to identify "critical junctures," or loopholes, in the system with potential for poor outcomes.
Outcome evaluation is an important component of any DM program. Ours came from traditional clinical endpoints — such as mortality and morbidity — as well as from patient-reported outcomes, such as functional-health status, quality of life, and patients' satisfaction with the care they received.
In addition, we incorporated process measures, such as clinical tests and service components, so we could link process to outcomes. The data from the evaluation identified opportunities for improvement and documented successes. We got answers to the questions: "Was this a worthwhile effort? What impact did it make? How can we improve the program?"
An examination of the literature told us that the advantage of employing widely used measures and tools was the ability to set targets and identify best practices with which to compare our progress. An example of functional health status measures that have been widely used in populations typical of DM programs are the SF-36 and SF-12.
For clinical outcomes, national associations have developed standards for measurements that provide a good starting point. Guidelines for the Diagnosis and Management of Asthma, from the National Institutes of Health Expert Panel Report 2, include a severity rating scale and outline treatment algorithms.
Patient satisfaction surveys are also commonly part of an evaluation of a DM program and again should utilize instruments that at least have been field-tested.
Because of the increased competition for dollars, we were called upon to address financial, as well as quality, issues. Quality improvements need to be converted to dollars saved and the cost of running a program has to be tracked so that an estimate of return on investment can be calculated.
We made projections before implementing the program, using best- and worst-case models. Risk vs. fee-for-service dollars need to be included in this "what-if" tracking.
After identifying the measures to be incorporated into the evaluation, the methods and procedures for implementing the evaluation must be carefully explored. Grand plans that are not feasible only result in failure and frustration. This is particularly so in working with the managed Medicaid population, in which language barriers and complicated lifestyles do not lend themselves to lengthy self-completed surveys or additional attendance at sessions where clinical care is not offered or some other incentive is not provided.
We realized that resources for conducting the evaluation needed to be identified and, if incorporated into existing roles, had to be as simple as possible and follow the vein of the program closely. In designing the evaluation, the population needed to be clearly defined to include all points of entry into the system.
For example, if participants were scheduled to attend a site session quarterly, it seemed sensible to have participants complete surveys while on site.
If telemanagement is the cornerstone of a program, it may make sense to incorporate short data-collection interviews into the phone call. While mailed surveys are less labor intensive, one should not choose them without considering potential difficulties with response rates.
Despite technological advances, clinical data — such as laboratory tests — can be difficult to track unless part of a closed system. Identifying all steps and potential pitfalls up front can save a tremendous amount of time on the back end.
When we were ready to launch our DM program, we assembled an implementation team to help with systemwide communication. In-service workshops regarding the fundamental principles of DM, the care process that was to be implemented, the clinical guidelines that would be followed, and the value of this initiative needed to be delivered to all the clinical areas involved. (You may want to videotape this presentation to ensure that all staff members have a chance to experience it.)
We provided education on the criteria for enrollment to people at all points of contact, and created a customer-friendly process of access for both the patients and the clinical staff. New forms and data-collection tools were kept in visible areas with the necessary information highlighted.
We laminated the clinical guidelines and protocols and placed them in the areas where we knew they'd help improve compliance and consistency of care.
We knew that it was extremely important that all members of the organization were consistent in their interactions and interventions with patients to decrease practice variation and to minimize confusion.
Educating patients and their families was also necessary to ensure success. DM implies a proactive approach. We engaged patients and their families as active participants. We tried to make patients feel confident in managing their illnesses, so that they could recognize a change in status and respond appropriately.
Once all procedures were outlined and the staff trained, we began to collect data. Keeping in mind the culturally specific needs of participants, we employed bilingual interviewers.
Frequent meetings with personnel involved in data collection identified problems and solutions, and made us all more cognizant of the importance of complete data capture and accuracy.
Continuous feedback helped us identify issues early on and kept providers informed. Demographics of the population served, numbers enrolled, individual case studies, and aggregate analysis of key indicators — such as compliance with standards of care and best practices — were all useful.
Analysis of individual case studies helped to identify process problems that were subsequently avoided. Interpretation of the data turned out to be an iterative process that required several meetings between the outcomes research and clinical staffs.
Results were both descriptive (counts or percentages of tests completed) and inferential in nature. In the latter case, statistical significance was carefully interpreted.
How we fared
Our outcomes research initiative is three years old and going strong. To date, we have enrolled more than 500 patients. In this very high-risk population, we dramatically reduced ER utilization and hospitalization by 63 percent and 79 percent, respectively. We have demonstrated a statistically significant improvement in both physical and mental components of the SF-36, reflecting improvements in both functional status and quality of life. Most important, however, is the marked increase in patients' disease knowledge as demonstrated by responses to an open-ended questionnaire addressing key issues in asthma management.
Developing and implementing any DM initiative is a formidable challenge, and success depends on several factors. First, the focus must always be the needs of the patient. There must be organizational alignment with DM woven into its strategic plan. The concept must be endorsed and supported by senior management who are prepared to participate in nontraditional segments of health care.
Success also hinges on the availability of, and access to, sophisticated integrated information systems that must be user-friendly and provide access to clinical, financial, and health status data. Outcomes research staff should support the coordination and analysis of the data with continuous feedback to providers based on disease specific and population processes and outcomes.
Respecting the physician
Physician buy-in and support is of paramount importance to ensure the success of the DM movement. Gaining acceptance from physicians requires their participation in the process from development to implementation. Guidelines must be developed that will endorse, but not impede, medical judgment. Clinical models must be flexible enough to respect individual style but not encourage variation from best practice.
The landscape of the health care environment is ever changing. As DM evolves, a new prodigy is already on the horizon: management of population health. Emphasis on wellness and prevention will result in populations with much broader demands on the health care system. Attention to the psychosocial and environmental aspects of wellness and disease will mandate redesign of existing resources to deal with the challenges these dimensions will create.
Never before has our dependence on integrated networks of health care delivery and information been so strong. The foundation built by the pioneers of DM thus far has us well positioned to face these challenges and ensure the stability and success of our health care delivery system well into the 21st century.