Evidence Based Policy Making:

Oregon Health Authority, the Medicaid 1115 Waiver, and Opioids

Nearly three-quarters of states (37 as of November 2016) have CMS approved Medicaid section 1115 demonstrations[p1] , that allow states to test new approaches to coverage and to improve quality and access or generate savings or efficiencies.[1] CMS has approved demonstrations for a wide variety of purposes, allowing extended coverage to populations or for services not otherwise eligible for Medicaid, made payments to providers to incentivize delivery system improvements, and, more recently, expanded Medicaid to certain low-income adults by using Medicaid funds to purchase private health insurance coverage. While state demonstrations vary in size and scope, many are comprehensive in nature, affecting multiple aspects of states’ Medicaid programs simultaneously. In applying for this waiver, CMS criteria for award specifically requires (1) Care for conditions of causal inference[p2],[2] (2) use of ‘best available evidence based on sound research,’[3][p3] and (3) stakeholder input and participation into the construction of the waiver.[4][p4]

Drafters of the Oregon Medicaid 1115 (2012–2017) demonstration waiver adopted explicit language to indicate that one of the state’s goals is to ‘improve the health of the chronic care population and prevent premature mortality.’ The utilization of a network of Coordinated Care Organizations (CCOs)[5][p5] was contracted to deliver care to persons with complex care needs who are eligible for both Medicare and Medicaid (referred to as dual-eligibles). Evaluation of the five-year demonstration by OHSU Center for Health Systems Effectiveness (2017) provides evidence that selected (CCOs)[6][p6] failed to meet their expectations for any stated objective related to persons with chronic care or persons who are dual-eligibles. These failures occurred in the areas of data collection, management and outcome tracking; care coordination for persons with complex illnesses; confusion about coordinating flexible services with diagnosis and billing codes; management of flexible service contractors; standing up necessary infrastructure, and making complex care a priority.

Pain is the problem that brings the greatest number of individuals into contact with the healthcare system. Back pain is among the most prevalent of complaints. Opioids are frequently the drug of choice applied to treatment of pain of all types after many other methods have failed. Many of these individuals appear in the health system as dually eligible individuals who are enrolled first in Medicare by virtue of age or disability and then qualify for Medicaid on the basis of income, or vice versa. Dually eligible individuals experience high rates of chronic comorbid illnesses, with many experiencing multiple medical and social risk factors. Forty-one percent of dually eligible beneficiaries have at least one mental health diagnosis, and about half use long term services and supports (LTSS) (CMS correspondence to State Medicaid Directors, December 19, 2018).[7][p7]

Despite documented evidence of insufficient implementation of services directed into chronic care that includes pain management supports, Oregon Health Authority (OHA) proposes to widen and promote ‘flexible services (adding acupuncture, cognitive behavioral therapy, chiropractic, massage)’ while removing complex care patients with centralized (back and fibromyalgia) pain from medication assisted supports that employ a whole class of medications (opioids). To meet the stated objective of improving the health of the chronic care population while preventing premature mortality, plan administrators appear to be adopting revisions that will sever persons with progressive, complex, comorbid illnesses who demonstrate centralized pain, from opioid-assisted medication supports while widening availability to services without proven benefit like acupuncture and yoga. In this revised plan, there is no handoff of persons with chronic illness to integrated, coordinated chronic care as would be found in accordance with integrated palliative care unless one has a diagnosis of cancer. A second set of regulations promulgated as Oregon SB608 specifies coordinated care (including pain care) for serious illness which is life-limiting in its impact. Yet in implementing this legislation, the OHA delimited implementation to cancer care only in any setting, disregarding legislated intent to serve persons with ‘serious illness’ in any community environment. The difference between acute care, care for illness that impose lifelong limitations, and illnesses that end one’s life is a matter of care delivery models and should be addressed within a planned continuum of public health resources.

As a group, the Medicaid 1115 Waivers employed across the states have documented weaknesses among their various implementations. The General Accounting Office (GAO18–220, 2018) noted that a weakness of the 1115 waiver program is that the states are remiss for relying on Medicare Managed Care Organizations (MCOs) or Coordinated Care Organizations (CCOs) for data collection without verifying either data adequacy or relationship to outcomes. Reliance on third party data collection incurs a number of limitations evident in the Oregon data -

· Waivers that rely on CCOs operate within a system of narrow clinical networks as a matter of cost containment. Narrow networks are associated with fewer contracted physicians and longer eligibility determination and patient wait times. The states have relied on actuarial data methods to plan for implementation due primarily to their reliance on MCOs and ACOs. Actuarial data methodology under-represents the chronic care population and their needs because it reflects only utilization and rationing of care. The state of Oregon is experiencing significant cost over-runs associated with missed Medicaid cost and revenue projections that relied on this methodology.

· Oregon has significant personnel shortages throughout their health system which further exacerbate problems of care delivery and confound data collection. Many counties fall within labor shortage areas documented by OHA’s workforce planning division (see Figures 2 and 3).

· Oregon tracks required NQF indicators but does not track patient data (results, outcomes, or impact) related to specific chronic care codes sufficient to inform decisions about managing multiple chronic comorbid conditions and their associated outcomes as a feature of reducing prescriptions for opioids. And, because it relies on third parties for data collection (MCOs and CCOs) with a documented history of gathering insufficient data due to underservicing of the chronic care population, there is no baseline from which to ascertain that there is even a relationship of centralized back pain to social problems of prescription abuse, addiction, overdoses, or suicides. The evidence of a problem in this case is absent.

Evidence based practices (EBP), cited frequently by members of various appointees to OHA advisory boards as the basis for their pending recommendations, require the address of specific features: 1) A clear statement of what EBP means; 2) A description of the minimum skill set required to practice in an evidence-based way; and 3) A curriculum that outlines the minimum standard educational requirements for training health professionals in EBP. To directly quote the framers of the Sicily Statement on Evidence Based Practice in Health care (2005), “Health care delivered in ignorance of available research evidence, misses important opportunities to benefit patients and may cause significant harm [2–4]. Providing evidence-based care is recognized as a key skill for health care workers from diverse professions and cultures [5–10]. The ability to deliver evidence-based practice promotes individualization of care and assures the quality of health care for patients today as well as those of tomorrow [11].” Good practice including effective clinical decision making, requires explicit research evidence and non-research knowledge gained from accumulated wisdom and experience. However, in an environment where publications serve as markers of life-long learning by clinicians, much of this literature falls short on accounting for patient centered input and values, is bereft of information regarding long term outcomes, and lacks reproducibility outside the limitations of the case report or study setting. Generalizability is a serious limitation when relying on current published resources. EBP requires[8][p8] -

1. Translation of uncertainty to an answerable question

2. Systematic retrieval of best evidence available

3. Critical appraisal of evidence for validity, clinical relevance, and applicability

4. Application of results in practice

5. Evaluation of performance

Practitioners and policy makers should be able to distinguish evidence from propaganda (advertisement), probability from certainty, data from assertions, rational belief from superstitions and science from folklore.[9][p9] Oregon Health Authority (OHA) has provided insufficient evidence that their plan to widen (on paper) access to services associated with both personnel shortages, and capitated CCO contracts amounts to little more than an attempt to deal with uncertainty derived from untested assumptions, that it will actually result in improved access to care based on what is known, or that this plan is likely to reduce mortality and increase the health of the chronic care population based on best available reproducible evidence. Stakeholders, whose values and voice have not been incorporated into policy making decisions, are rightfully concerned that based on the paucity of evidence, plans to undertake revisions to services are being developed that fail to address and correcting the deficiencies noted in the waiver evaluation conducted in 2017, and that these revisions may result in less adequate care access and coordination than that which is available currently.

The Medicaid 1115 Waiver Context

Approximately 15% percent of Oregon Health Plan members (N = 108,500) were also eligible for Medicare coverage in 2011 when the current waiver was initiated.[10][p10] In 2016, the state estimated that 20.38% of the population in 36 counties was Medicaid eligible. Of these ‘dual-eligible’ members represent a unique segment of the Medicaid population because they have a high prevalence of complex chronic conditions which account for a large proportion of total spending on Medicaid members or about $16,500 per person annually depending on their combination of identified conditions.[11][p11] Vulnerable dual-eligibles are among the most economically disadvantaged Medicaid beneficiaries, with more than half making an annual income of less than $10,000 in 2018.

Compared to other Medicaid beneficiaries, they are substantially more likely to have multiple (two or more) serious, chronic physical conditions and/or co-occurring behavioral conditions which are life-limiting.[12][p12] This population is most likely to be associated with the long-term use of controlled pharmaceuticals that fall into the class of medications that includes opioids. Table 1 reflects the top 19 billed chronic conditions represented in Medicare claims data through 2015 and reported in 2016.[13][p13] Although they represent only less than 20 percent of the Medicaid and Medicare population, they account for 39 and 31 percent of total Medicaid and Medicare expenditures overall. Prior to Part D, dual-eligible beneficiaries had drug coverage through the Medicaid program. They are now enrolled in Part D.

Table 1. Top 19 diseases associated with dual eligible status[14]

___________________________________________________________________

Alzheimer’s Disease and Related Dementia

Heart Failure

Arthritis (Osteoarthritis and Rheumatoid)

Hepatitis (Chronic Viral B & C)

Asthma

HIV/AIDS

Atrial Fibrillation

Hyperlipidemia (High cholesterol)

Autism Spectrum Disorders

Hypertension (High blood pressure)

Cancer (Breast, Colorectal, Lung, and Prostate)

Ischemic Heart Disease

Chronic Kidney Disease

Osteoporosis

Chronic Obstructive Pulmonary Disease

Schizophrenia and Other Psychotic Disorders

Depression

Stroke

Diabetes

___________________________________________________________________

In a study conducted by Bynum, Austin, Carmichael and Meara (2017),[15][p14] understanding how different types of complex patients use Medicare and Medicaid services over time can inform the policies that support this population. Two distinct groups of high-cost dual eligibles are found in the data: older beneficiaries who are nearing the end of life, and younger beneficiaries with sustained long term needs for functional supports. Both groups have high hospitalization costs. For those high-cost dual-eligibles living in the community, those who are older spend less on home and community-based services than those who are younger. Greater use of such home-based services might provide stable support in the last year or two of life, when illness and functional decline accelerate. Tailoring approaches to each population’s distinct needs could increase the value of care provided to patients and their families, with the potential to lower costs if patients’ needs can be met with fewer stays in short-term inpatient facilities. Given their high needs and high costs, care for dual-eligibles is a priority for policymakers.

Figure 1. 2018 Medicare Fact Sheet

Because improving care and reducing costs for all Medicaid users, including dual-eligibles, is especially important for Oregon’s Medicaid program, Oregon initiated an application to HHS-CMS for a Medicaid 1115 waiver to transform its Medicaid health care delivery system through the period 2012–2017 by contracting with coordinated care organizations (CCOs).[16] CCOs put a particular focus on coordinating care among different types of health care providers through implementation of patient-centered primary care homes, health information technology, integration of physical and behavioral health care, and other interventions. The care coordination provided by CCOs is thought to be particularly important for improving care and reducing spending for persons with complex needs who are dually eligible. However, an analysis by Oregon Health Services University (OHSU, 2017)[17] reflects that overall, the CCO approach improved some aspects of care quality but did not lead to any meaningful changes in health service utilization among dual-eligibles (p53). In fact, most measures of access to care decreased slightly. For an explanation of structural and personnel shortfall factors that affect this see discussion, see Oregon Health Workforce planning report to the legislature 2018.[18]

Improvement on quality measures of importance to the chronic care population was mixed. Quality measures generally improved in three domains: Prevention and wellness for Children and Adolescents, emergency department and hospital use, and avoiding low value care. But they noted that more work was needed to improve quality in four domains: Prevention and wellness for adults; care coordination; physical, behavioral, and oral health integration; and care for people with chronic conditions. For example, the percentage of members with diabetes who had an HbA1c test, a measure in the prevention and wellness for adults domain, decreased slightly among CCO members, while increasing slightly among Medicaid members in the comparison the state of Washington (pp53–54). As OHSU evaluation findings noted, CCOs have a limited ability to coordinate care, are poorly incentivized to coordinate care, time required for large changes to occur under the CCO model may be longer than the time period examined in this study, and pay for performance may appreciably impact utilization and quality (p124).

Oregon’s 1115 Waiver Outcomes

The Oregon health waiver adopted explicit language that indicates that one of the waiver goals is to improve the health of the chronic care population and prevent premature mortality (measure 3). The outcomes or endpoints adopted by OHA address frequency of visits to health providers and measures imposed by CMS based primarily on cardiac, respiratory, diabetes, and nutritional status, which are National Quality Foundation (NQF) indicators.[19] The data collected by the CCO reflect utilization for the following endpoints:

a) Any emergency department (ED) visit

b) Emergency department visits for any reason

c) Avoidable emergency department visits

d) Mental health-related emergency department visits

e) Any primary care visit

f) Any outpatient specialist visit (including cardiology, gastroenterology, nephrology, pulmonology, and urology)

g) Any physical therapy, occupational therapy, or speech therapy visit

h) Any inpatient hospitalization (excluding psychiatric hospital services)

i) Any inpatient hospitalization for mental health conditions

j) Any outpatient mental health visit

k) Any use of post-acute care services (including home health, skilled-nursing facility, inpatient rehabilitation facility, and long-term care hospitals)

l) Comprehensive diabetes care: HbA1c testing: Did a member with diabetes receive a recommended A1c blood sugar test? [higher is better] Modified NQF 0057 Yes

m) Comprehensive diabetes care: LDL-C screening: Did a member with diabetes receive a recommended cholesterol test? [higher is better] Modified NQF 0063 Yes

n) All-cause 30-day readmission: Did a member who was discharged from a hospital stay have a readmission for any reason within 30 days? [lower is better] HEDIS Yes

o) Prevention quality overall composite: Did a member have an ambulatory care sensitive admission? [lower is better] PQI #90 No

p) Use of high-risk medications in the elderly (66+): Did a member receive at least one high-risk medication? [lower is better] HEDIS No

q) Annual monitoring for patients on persistent medications: Did a member taking selected drugs for more than 180 days have a therapeutic monitoring event for the medication? [higher is better] HEDIS No

r) Low-value head imaging for an uncomplicated headache: Did a member with a headache receive head imaging tests? [lower is better] CW No

s) Low-value head imaging for syncope: Did a member with syncope receive

head imaging tests? [lower is better] CW No

The Impact of Data Quality

Data considerations. The actuarial data quality standard[p15] clarifies that for an evaluation to be considered reliable, one must (a) identify who prepared the data — the actuary or a third party — utilized to conduct reviews (section 3.5, Review of Data); (2) the definition for each data element; and © a review of the obvious characteristics of the data.[20] For purposes of data quality, data are appropriate if they are suitable for the intended purpose of an analysis and relevant to the system or process being analyzed. For purposes of data quality, data obtained from inventory or sampling methods are comprehensive if they contain sufficient data elements or records needed for the analysis. An informal examination of the obvious characteristics of the selected data to determine if such data appear reasonable and consistent for purposes of the assignment. A review is not an audit of data. In most situations, the data are provided to the actuary by others. The accuracy and comprehensiveness of data supplied by others are the responsibility of those who supply the data. The actuary may rely on data supplied by others, subject to the guidance in section 3.5. Oregon Medicaid is operated by third party contractors (CCOs) which prepare all of the data supplied for evaluation by the OHA and GAO’s contracted external provider. CMS relies on the use of CMS-64, which serves as the basis for calculating the amount of federal matching funds for states, and the Medicaid Statistical Information System (MSIS) which is designed to report individual beneficiary claims data[p16] .[21]

Limitations of Medicaid Waiver 1115 Data. Oregon’s Medicaid Waiver Program evaluation reported challenges in extracting and analyzing information from electronic health records (EHRs) and using EHRs to exchange information with other clinics and hospitals (GAO 17–312, p122). Oregon Health Authority (OHA) relies on actuarial data processes. The General Accounting Office found that across the nation, Medicaid 1115 Waiver expenditure and utilization data do not provide sufficient information to consistently ensure that payments are proper or that beneficiaries have access to covered services. While both have the potential to offer an excellent overview of utilization, GAO[22][p17] found that the usefulness of CMS-64 and MSIS managed encounter data within the 1115 Waiver program is limited because of issues with completeness, accuracy and timeliness. This problem introduces elements of bias, and increases the likelihood of both under reporting and false negatives. Oregon’s audit division evaluation of Oregon data for flexible spending associated with chronic care identified specific deficiencies[p18] related to processing data associated with Care coordination and disease management:[23]

(1) CCOs often omitted programs to coordinate members’ care or help members manage diseases from flexible services reporting. Under the 2012–2017 waiver, CCOs were confused about the definition of flexible services and how flexible services fit into OHA’s process for setting capitation rates. This resulted in inconsistent reporting of flexible services spending among CCOs, and may have contributed to relatively low flexible services spending (p123). One interviewee said a CCO omitted its “overall” programs, including a community health worker (CHW) program, a peer support program, and a wellness center that provided education for pain management and other disease management. (p33). This raises the concern that integrated care that includes pain management supports may not be any better off after flexible services are widened.

(2) What gets measured matters. GAO found that within CCO reported data sets, services were not tied to medical diagnoses or services with billing codes. Some Oregon interviewees said CCOs stopped reporting health-related services that were not tied to a medical diagnosis or services with billing codes following the 2015 change in state administrative rules defining flexible services; however, the CCOs continued providing these services, in some cases from a different “pot of money”(p33). If the State of Oregon asks CCOs to demonstrate a link between health-related services and cost-effectiveness or quality improvement based on diagnostic codes, it should provide information on acceptable data and consistent methods for evaluating the link (p123).

(3) Oregon relies on CCO level data reporting for spending levels across members for each flexible services category. These datasets are not consistently reported across CCOs and the data is severed from individual users, making it impossible to determine how many persons are presented in the flexible spending datasets (p124). This raises the concern that outcome tracking and data tracking concerns will remain under addressed. The current endpoints may obscure the lack of services and poor service quality and evaluation of treatment effectiveness for persons with complex comorbidities. Without a process of linking medical data or checking with patients for outcomes, one is unlikely to find that there is either a data problem or a service delivery problem. Despite the sacrificing of patient privacy protections, conclusions about the role of individual prescriptions for opioids (scripts, dose, or utilization) with regard to downstream social problems cannot be verified. Assumptions that the PDMP, a prescription tracking tool, will serve this function may well prove to be erroneous.[24]

(4) CCOs were just beginning to provide flexible services as defined in state administrative rules…CCOs did not receive detailed guidance from OHA about the definition and use of flexible services until the fourth year of the waiver…when OHA came out with flexible services, there was absolutely no guidance provided from them as to what that meant, and the CCOs for the first two years defined what flexible services were and how they were utilized. In 2016, they actually pushed down some guidance and rules related to what flexible services could be and so that’s why we now have standard policies and procedures of what a flex service is and what those funds can be used for versus what a benefit is. (p34): Local plan resources may well influence how flexible services are interpreted and implemented. Where there are service provider shortages, or there are delayed access issues, there will be no data, no expenditure, and potentially, no care and no data. This may in turn be reflected as a satisfactory improvement in reduced cost of utilization when in fact it is a very unsatisfactory outcome. Reduction of utilization in this case indicates poor quality care.

(5) CCOs varied widely in their capacity to track and report on flexible services. Some CCOs used sophisticated systems to track flexible services spending for accounting and financial reporting purposes. CCOs lacked systems for tracking and reporting on members’ use of flexible services and outcomes associated with flexible services. While some CCOs used sophisticated systems for accounting and financial reporting, these systems were not used to track flexible services use or outcomes, such as members’ health status or satisfaction. Most interviewees said members’ use of flexible services and outcomes were recorded as part of the care management process and stored in CCOs’ case management systems or providers’ electronic health records, which are separate from systems for accounting and financial reporting. At the time of the interviews, most CCOs said data needed to evaluate the effect of flexible services on health care use and spending were unavailable. Several interviewees highlighted the challenge of demonstrating that flexible services cause decreases in spending or improvements in health outcomes …trying to do a pre and post-expense measurement is tough because it is difficulty to construct or find control groups. This problem will remain as the data system is currently designed.

(6) Another CCO described the need to train nurses and other health care workers in its community as illustrative of the need for infrastructure to support CCOs: “If we made one major mistake in standing up CCOs from day one, it is that we did not recognize the need in different communities for infrastructure (p40).” Widening the available flexible services without assuring that there are health care workers available to provide these services. will create a mismatch between expectancy and reality. For more about this as it reflects the state of Oregon, see OHPolicy’s Workforce assessment of needs produced for the legislature in 2018 which identifies workforce shortages through three lenses:[25]

A. industry/economic demand for health care workforce.

B. Measuring workforce needs by examining patients’ access to care and the array of health care services that should be available in communities of various types

C. Examining workforce capacity at the county level

Oregon is experiencing insufficient primary care capacity across the spectrum. Despite real gains made around recruitment and retention in many parts of the state, there remain geographic areas that lack primary care providers — including physicians, nurse practitioners and physician assistants — when viewed from any of the three lenses. Shortages fall within thirty specialties that are targeted for recruitment: Physical therapists, physicians and surgeons, family practitioners, internists, obstetricians and gynecologists, physician assistants, registered nurses, nurse practitioners, anesthesiologists and pediatricians; Mental health counselors, physical therapist assistants, mental health and substance abuse social workers, occupational therapists, dentists, medical and clinical lab technologists and pharmacists. This report recommends an approach to recruitment and retention where the goal is to increase capacity and the ability to meet geographic areas of greatest need. This applies to primary medical care, oral health and behavioral health (p23). Notably there is little attention paid in this plan to the need for recruiting the types of clinicians who specialize in chronic care — geriatricians, palliative care physicians, and pain management. In 2016/17, the OHA added licensing boards to widen access to additional kinds of care:

• Oregon Board of Chiropractic Examiners

• Oregon Board of Clinical Social Workers

• Oregon Board of Examiners for Speech-Language Pathology and Audiology

• Oregon Board of Licensed Professional Counselors and Therapists

• Oregon Board of Massage Therapists

• Oregon Board of Medical Imaging

• Oregon Board of Naturopathic Medicine

• Oregon Board of Optometry

• Oregon Board of Psychologist Examiners

• Respiratory Therapist and Polysomnographic Technologist Licensing Board

Figure 2. Primary Care Health Professional Shortage Areas in Oregon. https://www.oregon.gov/oha/HPA/HP-PCO/Pages/HPSA-Designation.aspx

Figure 3. Oregon Medically Underserved Areas and Populations. From https://www.oregon.gov/oha/HPA/HP-HCW/Documents/hcw-assessment-needs-oregon-communities-patients.pdf

What Should Happen Next?

According to World Health Organization (WHO),[26] a well-functioning health system responds in a balanced way to a population’s needs and expectations by:

· Improving the health status of individuals, families and communities

· Defending the population against what threatens its health

· Protecting people against the financial consequences of ill-health

· Providing equitable access to people-centered care

· Making it possible for people to participate in decisions affecting their health and health system.

To fully grasp the current public conversations surrounding the removal of opioid prescribing from the Oregon Medicaid Waiver program, one must examine the history and contextual factors in which the conversation is occurring. Oregon Health Authority adopted and implemented the Medicaid 1115 Waiver model without addressing critical infrastructure (personnel, clinic distribution, specialties), necessary to transform the system in a manner consistent with adequately service the more than 108,000 persons with multiple chronic comorbidities who are dually eligible and reliant on utilizing a CCO model. The OHSU Center for Health Systems Effectiveness Analysis (2017) suggests that from 2012–2017, the fourteen CCOs struggled to provide necessary care coordination to dual-eligibles and consumers with multiple chronic conditions who need care coordination for reasons closely aligned with a variety of GAO findings regarding Medicaid Waiver processes. Further, OHA failed to plan for designing in the protective, contractual adjusting factors that would assure revenue safeguards in case of changes in state economic status. Three years into the waiver, Medicaid reimbursement rates were adjusted downward by the federal government. This lack of protective planning imposed a significant revenue shortfall during the fourth and fifth years of operation which negatively affected other operational areas.

OHA workgroup members have asserted continuously that their processes rely on evidence based practices (EBM).[27] Evidence based practices rely on three areas of analytics working together — best evidence derived from research, clinical expertise, and patient values and preferences. When one considers the inadequate system data collection which underpins the evidence for state planning assumptions, and the limitations associated with actuarial methodologies, it raises questions of adequacy of state planning processes. The very fact that a shift in state median income resulted in a reduction of Medicaid revenues due to formulary changes, which in turn created a reduction in revenue necessary to operate the plan even as enrollment was increasing, could have been predicted and prevented through data verification and testing and planned adjusting safeguards available to the state in negotiating with the federal government.

Acute care, or services to ‘infrequent flyers’ has offered a rather healthy contribution to the program according to evaluation reports. Increasing enrollment of healthy, younger people was advantageous for plan operations during the first three years. However, the population most greatly affected by planning failures has been the chronic care population for whom insufficient data and inadequate assumptions led in turn to insufficient planning to reduce structural and access to care barriers. Moreover, the current OHA proposal to account for these deficits by reducing care levels for specific ICD-10 codes associated with centralized pain, while widening (on paper) access to services that are documented as unavailable due to personnel shortfalls, strikes this reviewer as decidedly disingenuous. If OHA were truly relying on evidence, there would be a planning process in place to close the gap for the chronic care population that is installed and communicated throughout the overall system — that of access to chronic care and palliative care supports which are well understood areas in which it is possible to achieve cost savings without denying critical care and while obtaining full functionality of the medical home model. SB608 as implemented, should be widened and structurally supported with the specialties necessary to serve the dual-eligible population with multiple chronic comorbidities in keeping with the National Consensus Standard on Palliative Care issued in (CAPC, 2018).[28]

As it stands now, there are legitimate concerns that OHA has conflated managing the costs of care for chronically ill with what look suspiciously like paternalistic, moral equivalencies about the use and effectiveness of opioids without accounting for characteristics of this population of users or a primary rule of evidence based practice. It should be noted that relying on NQF’s indicators may create an anchoring effect that misdirects attention because it generates no patient centered outcome or impact data. While the absence of evidence is not evidence of absence, the available data does not suggest that persons with chronic care needs are responsible for Oregon’s opioid crisis. On balance, the paucity of evidence to support the role of prescription opioids used by individual complex patients as a factor in Oregon’s community public health challenges is a function of ethical limitations of research, moratoriums, and FDA approval processes, not necessarily the ethical or personal behaviors of affected users or the performance of these medications. The influence of other community factors have not been ruled out.

Evidence based practice requires accounting for patient values and preferences. There is little evidence that OHA has systematically sought input from the chronic community and there is less evidence that they value the feedback they have received to date from the most affected persons and other professionals. As a result, the complex patient community, which has been most neglected by a plan ostensibly redesigned to improve their access to care, is rightfully alarmed by the current construction of alternatives that may well harm them through a variety of predictable and in some cases unintended consequences.

The state has appointed committees to meet the needs of persons served by the Medicaid 1115 Waiver, yet these committees are failing to proactively include representatives from this community in planning for the address of potential shortfalls and services. This raises concerns about bias and the role of personal member beliefs, agendas, and conflicts of interest being installed into this process as a substitute for the evidence that representatives claim to rely on and value. Variations in CCO characteristics, approaches to implementation, and the lack of administrative and quality guidance from OHA, supports the expressed concern of OHA’s most vulnerable plan stakeholders — the patients — about the replacement of current care services and supports with flexible services as proposed by the Oregon Health Authority’s designees. Governor Kate Brown would be wise to -

1. Direct her health planning agency to suspend their stated intention to remove opioid supports from patient care until there is evidence that an adequate safety net of practices are in place of sufficient evidence quality to inform program management decisions about the link between effective patient outcomes and pharmacy prescribing practices.

2. Direct CCOs to adopt uniform data fields and real time procedures within electronic health records systems to align with audit capability of the state oversight processes.

3. Adapt data links from individual billing records to incorporate ICD-10 codes and services, products provided, and billed.

4. Utilize the information gained to determine data end points that inform future decisions about the relationship of services provided and billed to effectiveness outcomes for persons served.

5. Develop a systematic continuous improvement approach to data quality and integrity assuring that not only are individuals are getting the care they require, but also that the link between care provided and outcomes is clearly discernable.

6. Ensure that stakeholder participation is consistently invited, incorporated, and valued throughout the entire process by interviewing program users, holding public meetings, setting up mechanisms for online feedback, and appointment of stakeholders to public committees so that disparate points of view are balanced and biases are reduced.

7. Offer training in evidence review practices to stakeholders so that they may confidently participate.

[1] GAO18–220 (2018, Jan). Medicaid Demonstrations: Evaluations Yielded Limited Results, Underscoring Need for Changes to Federal Policies and Procedures. General Accounting Office: Washington, DC

[2] Contrera, K., Bradley, K. & Chao, S. (2018, June). Best Practices in Causal Inference for Evaluations of Section 1115 Eligibility and Coverage Demonstrations. Mathematica Policy Research: Washington, DC

[3] Irvin, C., et al (2015, May 15). Medicaid 1115 Demonstration Evaluation Design Plan. Mathematica Policy Research: Washington, DC

[4] Colby, M., Bradley, K., Contrera, K. & Natzke, B. (2017, July). Premium Assistance, Monthly Payments, and Beneficiary Engagement: Design Supplement, Outcomes Evaluation. Mathematica Policy Research: Washington, DC

[5]State of Oregon Health Authority (2012, May). Proposal to the Centers for Medicare and Medicaid Services Medicare/Medicaid Alignment Demonstration to Integrate Care for Individuals who are Dually Eligible. Retrieved from https://www.cms.gov/Medicare-Medicaid-Coordination/Medicare-and-Medicaid-Coordination/Medicare-Medicaid-Coordination-Office/Downloads/ORProposal.pdf

[6] CCO network list

[7] Correspondence dated December 18, 2018, from Administrator, Seema Verma, Administrator, Department of Health and Human Services, Centers for Medicare & Medicaid Services 7500 Security Boulevard, Mail Stop S2–26–12 Baltimore, MD 21244–1850

[8]Dawes, et al (2005, January 5). Sicily Statement on evidence-based practice. Published: 05 January 2005

BMC Medical Education, 5:1 doi:10.1186/1472–6920–5–1 Retrieved from: http://www.biomedcentral.com/1472-6920/5/1

[9] Hurd, PDH: Scientific Literacy: New Minds for a Changing World. Science Education 1998, 82:407–416.

[10] Ballotpedia,(n.d.) Medicaid spending in Oregon https://ballotpedia.org/Medicaid_spending_in_Oregon

[11] CMS.gov Chronic Conditions. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[12] Center for Medicare and Medicaid Services. Chronic Conditions. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[13] Top 19 Chronic conditions in Medicare Claims data 2007–2015. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[14] Source: CMS.gov https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[15] Julie P. W. Bynum1Andrea Austin2Donald Carmichael3Ellen Meara4 (2017, June). High-Cost Dual Eligibles’ Service Use Demonstrates The Need For Supportive And Palliative Models Of Care HEALTH AFFAIRS, https://doi.org/10.1377/hlthaff.2017.0157

[16] Funding Opportunity Number CMS-1G1–12–001. US Department of Health and Human Services, Centers for Medicare & Medicaid Services

[17] Oregon Health Services University (2017). Evaluation of Oregon’s 2012–2017 Medicaid Waiver. OHSU Center for Health Systems Effectiveness: OR

[18] Oregon Health Authority (2018). Examining the Health Care Workforce Needs for Communities and Patients in Oregon, OHA Policy and Planning Division.

[19] National Quality Forum (2012, May). Multiple Chronic Conditions Measurement Framework Retrieved from file:///C:/Users/pc/Downloads/MCCFinalReport.pdf

[20] General Committee of the Actuarial Standards Board (December 2004). Actuarial Standard of Practice №23., Data Quality Revised (Doc044). Actuarial Standards Board Updated for Deviation Language Effective May 1, 2011

[21] GAO 16–53 (2015, October) MEDICAID: Additional Efforts Needed to Ensure that State Spending is Appropriately Matched with Federal Funds. US Government Printing Office: Washington:D.C.

[22] GAO 17–312 (2017, April), Medicaid: Federal Action Needed to Improve Oversight of Spending US Government Printing Office: Washington:D,C,

[23] Correspondence dated December 1, 2017 from Patrick Allen, Director, to the Honorable Governor Kate Brown RE: Oregon Health Authority Bi-Weekly Update on Ongoing and Emerging Issues, Dual Eligibles.

[24] Robeznieks, A (2017, August 16). PDMP case pits patient privacy against law-enforcement intrusion. Retrieved from https://www.ama-assn.org/delivering-care/patient-support-advocacy/pdmp-case-pits-patient-privacy-against-law-enforcement

[25] Oregon Health Policy and Analytics Division (2018). Examining the Health Care Workforce Needs for Communities and Patients in Oregon. Oregon Health Authority. State of Oregon: Portland

[26] WHO (2010, May). Key components of a well-functioning health system. Retrieved from https://www.who.int/healthsystems/publications/hss_key/en/

[27] M Dawes, W Summerskill, P Glasziou, A Cartabellotta, J Martin, K Hopayian, F Porzsolt, A Burls, J Osborne Sicily statement on evidence-based practice. (2005, Jan). BMC Medical Education20055:1, https://doi.org/10.1186/1472-6920-5-1

[28] NCHPC (2018). Clinical Practice Guidelines for Quality Palliative Care, 4th Edition. Richmond VA: National Coalition for Hospice and Palliative Care, Retrieved from http://www.nationalcoalitionhpc.org/ncp[p1](https://www.gao.gov/products/GAO-18-220)

[p2]file:///C:/Users/pc/Downloads/medicaid1115demoevaldesignplan.pdf

[p3]https://www.mathematica-mpr.com/our-publications-and-findings/publications/medicaid-1115-demonstration-evaluation-design-plan

[p4]https://www.mathematica-mpr.com/our-publications-and-findings/publications/medicaid-1115-demonstration-evaluation-design-plan

[p5]https://www.medicaid.gov/medicaid-chip-program-information/by-topics/waivers/1115/downloads/or/or-health-plan2-ca.pdf

[p6]https://www.oregon.gov/oha/hsd/ohp/pages/coordinated-care-organizations.aspx

[p7]Correspondence dated December 18, 2018, from Administrator, Seema Verma, Administrator, Department of Health and Human Services, Centers for Medicare & Medicaid Services 7500 Security Boulevard, Mail Stop S2–26–12 Baltimore, MD 21244–1850

[p8]https://bmcmededuc.biomedcentral.com/articles/10.1186/1472-6920-5-1

[p9]https://www.researchgate.net/publication/253316043_Scientific_literacy_New_minds_for_a_changing_world

[p10]https://ballotpedia.org/Historical_Medicaid_spending_in_Oregon

[p11]https://www.cms.gov/research-statistics-data-and-systems/statistics-trends-and-reports/chronic-conditions/cc_main.html

[p12]Center for Medicare and Medicaid Services. Chronic Conditions. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[p13]Top 19 Chronic conditions in Medicare Claims data 2007–2015. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Chronic-Conditions/CC_Main.html

[p14]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5633373/

[p15]General Committee of the Actuarial Standards Board (December 2004). Actuarial Standard of Practice №23., Data Quality Revised (Doc044). Actuarial Standards Board Updated for Deviation Language Effective May 1, 2011

[p16]GAO 16–53 (2015, October) MEDICAID: Additional Efforts Needed to Ensure that State Spending is Appropriately Matched with Federal Funds. US Government Printing Office: Washington:D.C.

[p17]GAO 17–312 (2017, April), Medicaid: Federal Action Needed to Improve Oversight of Spending US Government Printing Office: Washington:D,C,

[p18]Correspondence dated December 1, 2017 from Patrick Allen, Director, to the Honorable Governor Kate Brown RE: Oregon Health Authority Bi-Weekly Update on Ongoing and Emerging Issues, Dual Eligibles.