Biologics account for 70 percent of all U.S. Food and Drug Administration (FDA) medication approvals, with many more presently in clinical trials.4 The good news is that many of these new therapies, although extremely expensive, are extremely effective. The bad news is that for most applications, indications in treatment protocols remain ambiguous, or treatments are transacted ambiguously. While price per unit was a legitimate point of attack for our move to value, our primary goal here was to eliminate unnecessary or hurtful use of biologicals and to establish the most efficient treatment protocols and most accessible and optimal treatment venue when the indication for treatment was, in fact, unassailable.
Determining the right thing to do for patients in all ProvenCare Biologics value reengineering initiatives is primarily the job of disease-specific experts, enabled by pharmacist-led expert programming teams. Whether it’s something as straightforward as erythropoietin (EPO) for the treatment of anemia or something as complex as multiple sclerosis (MS) management algorithms and subalgorithms, design of the initial best practice pathway, as well as the more dynamic adherence monitoring and practices pathway updating, could not be accomplished without this team approach and sociology. At Geisinger, it doesn’t matter whether the insurance side or the clinical enterprise/provider side wins financially as long as patients benefit and the total cost of care is decreased. And more often than not, the two go together.
The clinical pathway development is as follows:
• Gather collective experts and support staff (physicians, nurses, pharmacists, allied health professionals) from across the system.
• Explain process and goals.
• Research the current literature.
• Obtain best practice clinical consensus for ensuring safety, reducing variation, and enhancing education and collaboration.
As with all of our ProvenCare processes, the team is responsible for developing the clinical pathways and maintaining and updating each of the reengineered best practice algorithms. While the composition of our reengineering teams varies depending on disease state and stage, they typically include physicians, specialists, and subspecialists who are subject matter experts; clinical pharmacists with particular expertise in the disease; care managers; nurse coordinators; physician assistants; primary care physician leads in community practice sites; health plan medical officers; data analytics and information technology staff from our Epic electronic health record (EHR) support group; and, in most cases, patients with the disease and/or their family members. They are joined by health economists from both the payer and provider sides of our organization. Team members understand that their accountabilities and involvement do not end when optimal care templates initially are put into place. Rather, the process is dynamic and continues because knowledge is changing constantly and rapidly, and we expect to learn and iterate as we go.
The primary component of the formalized reengineering process planning road map is defining which patients have the appropriate condition to begin treatment. For the most prevalent diseases, generally 20 to 30 percent of treatments are not indicated or are indicated only ambiguously. Unnecessary treatment has more than economic consequence; more important, it can be hurtful, causing complications and, in many cases with the use of biologicals, major quality-of-life compromises.
The second component of the planning road map is to determine the default optimal treatment algorithm. The third is to put the clinical pathway into action, and the fourth is to assess performance, both process adherence and patient outcome metrics. The final component is to determine how the template should be continuously improved based on new knowledge, treatment transaction issues, and outcomes.
ALTERNATIVES FOR TREATMENT FOR ANEMIA, MULTIPLE SCLEROSIS (MS), HEPATITIS C, AND PSORIASIS
At Geisinger, the most prevalently prescribed biologics are Harvoni for hepatitis C; Humira for rheumatoid arthritis, inflammatory bowel disease, or psoriasis; Enbrel for rheumatoid arthritis or psoriasis; Tecfidera for MS; and Revlimid for multiple myeloma. We have teams in place to optimize treatment templates and transaction for the following therapeutic areas: anemia, MS, psoriasis, hepatitis C, Crohn’s disease, ulcerative colitis, and oral chemotherapies for cancer.
Anemia
More than 20 percent of our patients with anemia associated with chronic renal failures were being treated inappropriately with EPO, which makes a huge difference both in terms of cost and the frequency of disease-related complications. For non-oncology anemia therapy (such as presurgical or chronic renal disease treatment), the minimum EPO dose costs $1,417. Oral iron, more appropriate treatment for up to 20 percent of anemia patients, costs about $3.80. If the better treatment were intravenous iron dextran, the cost would range from $325 to $514 per patient and would entail a single treatment. Iron sucrose is yet another EPO alternative indicated for some of the 20 percent of patients now being treated with a biological and would entail three to four patient encounters at a total cost between $272 and $347 for the entire course. More important than the financial considerations, though, is that EPO is known to exacerbate central and peripheral vascular disease symptoms, particularly in older patients. Even before establishing efficient treatment protocols, and prior to designating centralized sites for biological treatments (for example, no longer permitting EPO treatment in nephrologists’ offices), more than 20 percent of unnecessary costs and potentially hurtful outcomes can be avoided simply by applying stringent, standardized indications for treatment.
Under our reengineering of anemia treatment with EPO, there has been a 30 percent total cost savings while improving care. Some 20 percent of cost was eliminated immediately because it represented patients who didn’t need to be treated with EPO. An additional 10 percent cost efficiency occurred due to Geisinger’s reengineering to an improved treatment of the anemia using EPO, with no overshoot or undershoot of red blood cell optimization. Red blood cell optimization was extremely volatile when treatment was decentralized in nephrologists’ offices, and it improved dramatically with implementation of a centralized treatment location supervised by specialists, but transacted by pharma techs.
Multiple Sclerosis (MS)
For MS, the most common disabling neurologic disease of young adults, biological disease-modifying treatments have proven beneficial only for the relapsing-remitting types of the disease. Geisinger cares for more than 2,600 patients classified as having MS. Documentation of which MS patients should receive a specified treatment is the first and probably most critical step in establishing best practice treatment protocols. We have determined how this complex information can be used to quickly identify appropriate patients, physician prescribing habits, and starting points for the various disease management algorithms. (See Figure 11.1.)
FIGURE 11.1 MS Protocol Example
Our reengineering teams not only developed the treatment protocols for MS at various stages in the patient’s symptomatological journey, but also created metrics to measure the reliability of treatment and reassigned various clinical tasks. Monitoring these redesigned roles and ensuring performance of the various protocols and subprotocols, as well as capturing the initial patient data, are done via the Epic EHR system and analyzed through Geisinger’s data warehousing and analytics. Geisinger payer data, as well as prospective management through the Geisinger provider data, are critical in both the redesign and the iterative process of evaluating adherence to the redesign and updating of the various protocols.
The overall intent of the MS ProvenCare Biologics program is that only the most appropriate candidates receive specific biologic therapy. This includes:
• Comprehensive drug therapy assessment and optimization
• Proper delivery of medications
• Monitoring of therapeutic response and toxicity
• Support of medication adherence
• Effective patient education
• Communication throughout the entire care team
Patient-reported data is captured
by iPads distributed through the neurology clinics or through touchscreen monitors at clinic physical locations. Identifying appropriate clinical content, building questionnaires into Geisinger’s survey software, and interfacing the survey software capture with the patients, their families, and the EHR are what enables initial identification of patients for specific biological therapies and also the effective monitoring of adherence to the various protocols.
A decision tree app was created to enhance the reliability of and adherence to specific MS treatment protocols. Features include:
• A program launched directly from within the EHR
• Evidence-based algorithms embedded in the program
• Electronic aids to help select medication, laboratory tests, imaging studies, and referral orders within the specific EHR-embedded protocol
• Ability to capture end-user interactions and incorporate adherence and possible toxicities into the EHR
• Interaction with existing data to provide program adherence metrics to the accountable provider, either the neurologist or the pharmacist
• Maintenance of all clinical content performed by non-discipline-based technical staff
For many patients, home-based monitoring of symptoms is sufficient for certain aspects of the algorithm-directed care. Structured patient-reported questionnaires, if not done at the time of the office visit, can be completed via the Epic-based patient portal, MyGeisinger. It has been an important interface in the communication between primary care physicians, specialists, subspecialists, and our patient population for the past 15 years. Most recently, physicians’ progress notes have been made accessible to patients and families, which has fundamentally changed the level of patient activation in relationship with providers. Responses to MyGeisinger questionnaires by patients diagnosed with MS are managed similarly to laboratory results, routed to appropriate EHR-based “in baskets” for clinical management by either the primary physician, the neurologist, or the appropriate disease-specific pharmacist expert.
The MS ProvenCare initiative is evaluated across a number of clinical and financial domains. Evaluation was designed in partnership with clinical leadership in neurology, pharmacology leadership with special expertise in neurologic disorders, and biostatisticians and health economists from our payer and provider sides. Clinical outcomes measured include:
• Radiologic remission focused on lesion burden and whether new MRI lesions are occurring
• Progression versus relapse of MS symptomatology
• Exacerbation of MS symptomatology
• Disability progression
• Adverse medication reactions
• Differential diagnosis of comorbid conditions (like infection) versus relapse of MS
• Progression of comorbid conditions
• Effect on other concomitant diseases and their management
The application of the intervention for MS is exemplified in Figure 11.2.5
FIGURE 11.2 Risks and Adverse Effects
The net financial benefit estimate for reengineering the MS biologics treatment process and the monitoring of the 2,600 patients in our MS population is considerable, not to mention the benefits in avoiding unnecessary complications, unknown increases in comorbid diseases including cancer, and unacceptable quality of life compromises. These additional benefits are difficult to quantify, but are without a doubt more valuable than the financial savings. (See Figure 11.3.)
FIGURE 11.3 Cost and Benefit Estimates for MS ProvenCare Biologics Reengineering
Psoriasis and Hepatitis C
As with our other ProvenCare reengineering efforts, we chose to start with cases that were high frequency, high cost, and high variability in both indications and in how and where the specialty drugs were administered. We focused first on what was most likely to succeed in improving patient outcomes. Not surprising, application of the ProvenCare approach to high-use, high-cost biologicals produced two outcomes: better treatment effects for patients and lower costs—both lower expense and lower toxicity. Two additional ProvenCare Biologics pathways are shown in Figures 11.4 and 11.5.
FIGURE 11.4 Dermatology: Psoriasis Treatment Care Path
FIGURE 11.5 Hepatitis C Treatment Care Path
Compared to MS, the best practice algorithms are straightforward for both the psoriasis and hepatitis C pathways. Implementation of the care paths includes development of tools embedded into the transactional EHR; organization of suitable patient and family support; and appropriate patient flow into centrally located treatment sites. Most sites are managed by pharmacists who utilize supervised pharmacology technicians to transact the best practice algorithms. The technicians’ adherence to these algorithms is the basis for the care redesign benefit to our patients.
Applicable ongoing support includes defining the roles for clinical oversight, which is shared among pharmacists, care managers, and nurse coordinators, all focused on the patient postdiagnosis after physician referral. Prescriptions are directed to the Geisinger specialty pharmacy for medications dispensing and monitoring, including education, patient adherence and medication tolerance, compliance issues, and any required lab monitoring. Scheduling appointments and reviewing lab test results, answering basic patient and/or family questions, and providing overall support including transportation all are part of this wraparound effort, with follow-up scheduled monthly. “Missed to follow-up” monitors are generated by Epic, and patients on this list are reviewed and scheduled manually as needed. This “lost to follow-up” process consists of the following:
• Structured, uniform diagnosis-naming conventions
• Population identification processes
• Monthly monitoring of missed to follow-up reports
• Patients contacted and scheduled as needed
Additional implementation targets include achieving the maximum leverage possible to drive down unit price by coordination of contracting and purchasing through the specialty pharmacy supply chain on both our payer and provider sides. Finally, key input by patients and their families is transacted primarily through the web-based EHR portal. We queried convenience, general access, and patient satisfaction, and designed the ProvenCare Biologics insurance product to be appropriately incentivized and financially advantageous for patients as well as the purchasers.
Early evidence for patients with psoriasis is promising. Some 15 percent of psoriasis patients in Geisinger Health Plan could be treated just as well with daily exposure to ultraviolet light, with a start-up cost of $3,000, versus approximately $80,000 to $100,000 a year for a continuing course of biologics. (See Figure 11.6.) The difference in cost is projected over six years, the average life span of a UV light source. Our year one actual savings for treating psoriasis patients was $2,139,867.
FIGURE 11.6 Dermatology: Psoriasis
A significant issue encountered in planning for a more appropriate psoriasis treatment was a necessary review and modification of the current insurance benefits. Originally the UV light box was an out-of-pocket expense for the patient, while insurance covered the full cost of the more expensive biological treatment for the entire patient population, whether it was appropriate and indicated or not. We mitigated the out-of-pocket cost so it would not be a barrier to patient acceptance of the much more reasonable therapeutic alternative.
Even more important than the financial consequence, though, is the issue of long-term comorbidity in treatment with biologicals, particularly if the treatment is unnecessary. For the approximately 7.5 million Americans with psoriasis, those treated with long-term biologicals not only have an unjustified huge economic consequence, but also an increased cancer risk. During the period September 2014 to September 2015, among nearly 39,000 patients with all types of psoriasis taking biologicals and reporting adverse events 1,315 cases (3.4 percent) were identified as having cancer diagnoses. This contrasts sharply with no increased cancer risk in psoriasis patients using UV light boxes compared to the population at large with or without psoria
sis and having normal exposure to sunlight.6
BIOLOGICALS AND CANCER TREATMENT
The high-frequency use of biologicals for treatment of a variety of cancers represents about 70 percent of the biologicals pipeline now at the FDA. Spending on oncology drugs is expected to increase 20 percent annually for the next several years, reaching almost $173 billion by 2020.7 Use for cancer treatment is expected to quadruple over the next eight years, and development of algorithms for reengineering many cancer care treatments is highly complex. Much of this use not only is unjustified, it’s potentially harmful. But cancer patients often want whatever treatments are available to them, providers get paid more when they prescribe more, and pharmaceutical companies get paid regardless of whether patients benefit from the prescribed treatments.
In 2014, the following number of medicines were in development for cancer:8
• 168 for breast cancer
• 96 for colorectal cancer
• 176 for lung cancer
• 155 for lymphoma
• 120 for prostate cancer
• 121 for skin cancer
Attempting to standardize this array of biological treatments is fraught with peril regarding patient preference and the sociology of the oncology specialty providers, especially with inadequate evidence and significant specialty oncology advocacy for widespread application. Throughout the United States, there is direct professional incentive to use many of these high-priced and perhaps effective treatments. The current pharma business model enables the treatment of significant numbers of patients with a given drug, whether or not it’s biological, even though good results occur only in a minority of patients treated. All patients, however, are charged, and payment is from some combination of insurance or out-of-pocket expensing. Future business models for professional providers and pharma must change, as undoubtedly evidence will increase regarding which biologicals are effective for specific cancers and cancer stages. But at the present time, all remains very much in flux. Hence Geisinger’s decision to move into other disease-use cases to target more standardized and high-value approaches, leaving the compelling area of oncology treatments with biologics perhaps to a future ProvenCare program.
ProvenCare Page 16