III. Quality Assurance

A. Overview

Quality assurance is an essential component of the triangular relationship in health care involving cost and access. [FN]216 As states undertake efforts to control cost and increase access simultaneously, it is important that they do so without reducing quality. [FN]217 Although much discussion has focused on cost containment, the vulnerability of the poor and minorities make it absolutely essential that the system provide “reasonable quality of care” at an affordable price for all Americans, rather than offer low-quality care at cheaper prices to the poor. [FN]218

The most significant issue is determining the quality of health care. This determination cannot be measured merely on the quantity of care given since more is not necessarily better. [FN]219 Unfortunately, there is no single method of defining quality care. [FN]220 In fact, quality health care may be defined in a variety of ways. Furthermore, in defining quality, it is important to remember that quality health care involves both interpersonal skills and technical care. [FN]221

Quality care, as defined by Brook and Kosscoff, is the “ ‘performance of specific activities in a manner that either increases or *1111 at least prevents the deterioration of health status that would have occurred as a function of a disease or condition.”’ [FN]222 According to the Council on Medical Service for the AMA (the AMA Council), quality health care “ ‘consistently contributes to improvement or maintenance of the quality and/or duration of life.”’ [FN]223 Yet another definition of quality health care is “the ‘component of the difference between efficacy and effectiveness that can be attributed to care providers, taking into account the environment in which they work.”’ [FN]224 Thus, quality of care involves not only the selection of the right activity, task, or combination of activities, but also the actual performance of those activities in a manner that produce the best outcome. [FN]225

An approach to defining quality care would be the “unexamined practice” rather than the mistake. In the unexamined practice, the focus is on “direct monitoring beyond the detection of fraud and gross incompetence and toward redirecting ‘the practice of honest practitioners into more strictly appropriate channels.”’ [FN]226 The AMA Council recognizes several factors in the determination of quality: (1) the production of optimum improvement in the patient's physical condition and comfort; (2) the promotion of prevention and early detection of disease; (3) the cooperation and participation of the patient; (4) skilled use of necessary professional and technological resources; (5) concern for the patients' welfare; (6) efficient use of resources; and (7) sufficient documentation of medical records to ensure continued care for evaluation of the care by peer review. [FN]227

What these definitions ultimately have in common is the need to assess health care in terms of structure, process, and outcome. [FN]228 Assessment of structure involves the evaluation of the setting in which health care is provided. It involves, for instance, an assessment of “the qualifications of medical personnel, the adequacy of equipment and facilities, the administrative structures, and the operation *1112 of programs providing care.” [FN]229

The assessment of process involves “the manner in which patients are treated and evaluates the performance of [health care] personnel.” [FN]230 As Furrow notes, however, a process approach has “disadvantages when community practice--standard of diagnosis or treatment against which physician deviance is measured--is itself unexamined.” [FN]231

Outcome measures involve examining whether health care improves the health of the patient and are the most significant way to assure quality. On the other hand, it presents several problems, including the need to isolate factors extraneous to treatment that may affect the patient's condition, [FN]232 the difficulty in measuring some outcomes, [FN]233 the variation that can occur in outcomes based solely on the timing of the assessment, [FN]234 the expense involved in aggregating data, [FN]235 and the difficulty in determining outcomes that will “reveal deficiencies in treatment to which a harmful outcome might be traced.” [FN]236

In many ways, the issue of quality in the provision of health care services is no different than discussing quality in other consumer areas. Purchasers of health care services expect to receive the best quality product for their money. But in health care, the stakes are much higher: A malfunctioning toaster can result in burned toast, but poor medical service can result in permanent and serious damage.

The approaches to assuring quality vary. They include: (1) performance report cards; (2) Total Quality Management; (3) practice guidelines; and (4) peer review. Performance report cards have been used to show how well a health care plan performed. [FN]237*1113 These report cards generally define the quality measures in terms of preventive care, prenatal care, mental health care, chronic disease management, access to care, and member satisfaction. [FN]238 Performance report cards, however, present a number of problems. First, the science of quality management is still too new to be applied on a national or even a state basis. [FN]239 Second, in an era where the primary concern is cost containment, the costs of implementing report card proposals might be prohibitive. [FN]240 Third, most health care institutions are not equipped to produce the kinds of data required for a report card. [FN]241 Finally, even if the other problems are overcome, there is still a significant likelihood that data will be “de-emphasized, overemphasized, or ignored.” [FN]242

Total Quality Management (TQM) [FN]243 has frequently been used in manufacturing facilities. [FN]244 One basic step in the process of implementing TQM is an agreed-upon set of principles and measures. [FN]245 Health care is an art, not a manufacturing facility. As a result, providers, policy makers, and quality experts have a difficult time agreeing on principles and a core set of measures. One problem is assuring that measures used are not solely geared to the white, middle-class, insured urban populace. Another problem *1114 may be the inability or unwillingness of providers to conceive of the poor and minorities as customers to whom they are accountable. [FN]246

Clinical practice guidelines were developed in an effort to standardize some aspects of health care practice. [FN]247 Practice guidelines were developed as an educational resource to support physicians in clinical patient care management [FN]248 and thus improve quality, assure appropriate utilization, and reduce costs. [FN]249 Use of practice guidelines present several problems. First, developers have failed to use a “coherent, standardized vocabulary” in creating practice guidelines. This failure results in uncertainty and ambiguity when applying the parameters to specific clinical problems. [FN]250 Another significant issue has been the development of practice guidelines in a chaotic way. This has included using clinical research subject to statistical manipulation, doing superficial analysis of the literature, and injudiciously using anecdotal information. [FN]251

Furthermore, practice guidelines may be flawed because they are subject to bias. In fact, publication bias, bias in the selection of study patients, the continuous evolution of medical knowledge, and methodological inadequacies undermine the validity of studies on which many practice guidelines are based. [FN]252 When considering guidelines, it is important to take into consideration the economic self-interest that may have impacted the development of certain practice guidelines.

The traditional view of quality is on peer review with a focus *1115 on physician mistakes or “errors.” [FN]253 The benefit of such a focus has been a clear definition of “bad medicine.” [FN]254 The problem with such a focus is that it overlooks too many other causes of poor quality health care besides individual responsibility. [FN]255 For instance, assuring quality care will require identifying and correcting systemic failures, poor administrative design for review of health care, inadequacies in training of health care providers, and inappropriate practice incentives. [FN]256

Each model has acknowledged strengths and weaknesses, but managed care must establish at the outset what the measure of quality shall and will be so that the people receiving the care can share in their assessment of the service.

B. Section 1115 Medicaid Waivers and Managed Care

Utilization review and financial risk-shift guidelines, which delineate quality, are incorporated into managed care plans, but so are financial incentives. Utilization review involves the use of an independent reviewer to evaluate the physicians' treatment decisions to determine if the treatment is necessary and if it will be delivered in the most cost-effective manner. [FN]257 These two forces may work in such a way that a patient may be unaware that care decisions are given or withheld based on price rather than the welfare of the patient. [FN]258 In this situation, the need for quality assurance is apparent. Furthermore, for disenfranchised patients such as the poor and minorities, it is essential. [FN]259

To obtain quality health care through the legal system, a patient must become a “complaining patient.” [FN]260 Many disenfranchised people have learned to be silent before administrative power. When quality guidelines are vapid, when patients are ill, or *1116 when patients are accustomed to being burdened or bewildered by the bureaucracy, there is a greater hesitation to complain.

Traditional quality assurance focuses on what a well-trained and well-regarded health care professional customarily does in treating patients. [FN]261 Now, however, the meaning has been clouded by cost containment. Under managed care what is measured first is the cost of the care given within the parameters of the managed care plan and only then is the quality of care assessed. Even then the quality of care given by a well-trained and well-regarded physician is not assessed, but instead the amount of care allowed is measured. Thus, the physicians are required to not only be caregivers, but also cost managers. [FN]262 Under some of the Medicaid managed care systems, there is an incentive for the physician to deliver less care than may be necessary. [FN]263

Medicaid was established to prevent a dual-track health care system. [FN]264 Health care in the United States was not intended to be two systems: one with good quality and access for those who can pay and another for the poor with low quality and little access. [FN]265 State-supported managed care plans arguably provide lower quality of care. With the newly eligible patients getting minimum care, the current Medicaid recipients are forced to bare the loss of benefits while the rest of the community is insulated from sharing in this cost. [FN]266 The rationing that has been established by some plans will greatly affect the quality of health care for those recipients in the region. The incentive of managed care is to provide less care to the recipient in order for the HMOs to retain, as profits, much of the capitation rate. Commentators argue that this will more directly affect the Medicaid recipients because they statistically fall in categories which are currently under served and already have higher health risks than the general population. [FN]267

Implementing a managed care system for Medicaid recipients through section 1115 waivers creates many quality health care issues. Many of these waivers, including Hawaii's, allow for health care plans comprised solely of Medicaid patients. Will this plan be as quality oriented as a plan which services the private sector? Under managed care, a patient loses the right, to some extent, to *1117 choose his or her own doctor and to switch doctors at will. Patients are forced to go through a gatekeeper doctor before they can see specialists. Due to the capitation payment system, it is in the managed care organizer's best interest to limit access to specialists. These are just a few of many issues facing Medicaid patients in a managed care organization.

Because Medicaid's fiscal and administrative structure contains no checks on state incentives to compromise the federal interest in cost-efficient quality health care, the HFCA has in the past employed stringent waiver review procedures as a method to ensure against the risk of poor quality care to minorities and the poor. [FN]268 The increased use of Section 1115 Waivers allows State Medicaid programs to shift widespread and exclusive use of managed care may have a significant impact on the delivery of quality care to poor and minority communities.

The Medicaid population includes a disproportionate number of vulnerable individuals. There has been a longstanding problem with the delivery of quality care to minority communities. Furthermore, many of the state programs are based on structurally unproven managed care designs. [FN]269 Although research has demonstrated that quality in Medicaid managed care can be equal to or better than fee-for service, the performance is uneven, and significant problems arise when the plans lack regulatory oversight. [FN]270 Providing quality assurance is essential. Universal access to health care means little if it is not quality health care. Ensuring both quality and access while implementing cost containing measures is a necessity.

C. Critiquing State Waiver Applications

Section 1115 waivers that have the potential of assuring quality are waiver applications which (1) set out sufficient quality goals, (2) provide for essential data collection,(3) have sufficient mechanisms to monitor and enforce the goals, (4) allow for mainstreaming*1118 and (5) promotes consumers' role in oversight. [FN]271


  1. Does the Waiver Set Out Sufficient Quality Goals?

In terms of assuring quality, data collection is meaningless without some clear goals by which managed care plans performance will be measured. [FN]272 Two states, Tennessee and Oregon, outlined clear health status goals. [FN]273 For example, Tennessee has an extensive list of goals either in the plan itself or in the handbook that is distributed by the MCOs. [FN]274 These goals range from targeting issues of low birth weight, treatment of otitis media, childhood asthma, and breast cancer to the percentage of children completely immunized at age two, number of emergency room visits, number of drugs dispensed, and mortality. The clinical areas of concern range from childhood immunization, pregnancy, pap smears and lead toxicity to hip fractures and dental screenings. [FN]275 Even though Tennessee outlines quality goals, they remain general, nonspecific, and are not quantified. [FN]276 Oregon's goals, although specific, are also nonquantifiable. [FN]277

On the other hand, Missouri did not set out quality goals. [FN]278*1119 Florida's waiver application does not have easily discernible quality goals or standards on which managed care plans will be evaluated. [FN]279 Hawaii structures its goals in the form of hypotheses which are nonspecific and vague. [FN]280 The New York waiver does not state clear quality goals by which the managed care plans will be measured. [FN]281

In addition to outlining clear health status goals, waivers need to outline reductions in the disparities in the health status between minority vs. nonminority. [FN]282 No waiver application addressed this issue. [FN]283

Does the waiver set out sufficient quality goals? Missouri was the only state that did not address this issue. [FN]284 Primarily because they did not address the health status difference between minorities and nonminorities as a goal, the other states (Florida, Hawaii, *1120 Illinois, New York, Oregon, and Tennessee) were ambiguous, at best, in addressing this issue. [FN]285







Does the Waiver Provide for Essential Data Collection?

Individual claims data is essential to being able to assure uniform data. Individual claims data allows a better assessment of health plan risk and health plan behavior including discriminatory practices. [FN]286 New York's waiver does not address data collection. [FN]287 The other six states (Florida, Hawaii, Illinois, Missouri, Oregon, and Tennessee) have ambiguous approaches regarding individual data collection. [FN]288

For example, notwithstanding the importance of claims data, Florida will not collect 100% encounter level data. [FN]289 It argues that collecting patient encounter data would be too costly. [FN]290 Both Florida and Hawaii propose to monitor quality by obtaining information from new and existing databases, [FN]291 surveys, [FN]292 medical *1121 records analysis, [FN]293 interviews, [FN]294 and case studies. [FN]295 However, while Hawaii indicates that it will be collecting encounter data, it is unclear whether that data will be collected based on each individual encounter or claim. [FN]296 New York appears to have established data collection systems that closely resemble the collection of individual claims data. [FN]297

Oregon proposes to collect not only medical claims data but *1122 dummy claims data as a method of evaluating quality. [FN]298 Finally, Tennessee requires each managed care organization to provide detailed information on provider and recipient activity, including encounter data, type of care provided, levels of care provided, outcomes of care, and use of preventive services. [FN]299 Furthermore, Tennessee asserts that it will employ an external contractor for monitoring who will use the Quality Assurance Reform Initiative prepared by the HCFA and the Kaiser Family Foundation as the standards for monitoring. [FN]300 In fact, Tennessee devotes eighty-five pages of its waiver to quality control monitoring. [FN]301

However, the primary problem with the waiver applications is the near total failure to require the monitoring of quality of care with regard to race. Florida is the only state that even mentions monitoring quality goals based on demographics. For example, Florida's data will specifically focus on vulnerable groups. [FN]302 Because it fails to define vulnerable groups, however, it cannot be assumed that they will target the collection of data based on race. [FN]303 Hawaii does not require that data be broken down by race, *1123 income level, or gender, but it does require comparisons between providers and counties. [FN]304 New York suggests that the data will be broken down by a number of categories, but it will not be collecting data based on race or ethnic background. [FN]305 Finally, Oregon's proposal does not specifically identify the data by race or ethnicity. The forms used for the survey are very basic and do not ask for identifying data. [FN]306

Does the waiver provide for essential data collection? New York's waiver does not address data collection. [FN]307 The other six states (Florida, Hawaii, Illinois, Missouri, Oregon, and Tennessee) have ambiguous approaches regarding individual data collection [FN]308 and do not require the collection of data based on race.




  1. Does the Waiver Application Outline Sufficient Mechanisms to Monitor Managed Care Plans and Enforce the State's Quality Assurance Standards?

Managed care organizations have a profit incentive. They are motivated to increase their profit margin over the profit margin of the previous year. This incentive is a “moral hazard.” It may operate to cause a managed care organization to deny care and to restrict access so as to increase profit. This power must be checked in order to assure access to quality care. [FN]309 Thus, any waiver must include sufficient resources for the state to monitor and enforce their plans' contractual obligation to provide quality care. [FN]310

  • 1124 Five states (Hawaii, Illinois, New York, Oregon, and Tennessee) had standards in their waiver application which addressed resource allocation to monitor and enforce quality assurance goals. [FN]311 For example, Oregon's requirements include written policy and procedures for an internal quality assurance plan, a medical record-keeping system that conforms with professional medical practice, and written procedures for handling client complaints and grievances.

In order for a waiver to have sufficient mechanisms to monitor and enforce quality goals, the waiver should require that managed care organizations have corrective action plans which would be enforceable by the state or the recipients. Oregon is the only state that had such standards in its waivers. [FN]312

Finally, the waiver application should include penalties that can be used when enforcing the requirements of the waiver. These penalties should include a range of graduated enforcement tools, such as monetary fines, enrollment restrictions, and revocation of contract. [FN]313 Without such graduated tools, states are likely to be reluctant to enforce the contract since the only option would be revocation. [FN]314 Only two state waivers (Florida and Oregon) had penalties in the waiver that could be used when enforcing the requirements of the waiver. [FN]315 For example, Florida provides for the imposition of significant financial penalties to the participating health plans for failure to perform the functions specified in the contract.

Did the waiver application outline sufficient mechanisms to monitor managed care plans and enforce its quality assurance standards? The answer to that question for Oregon and Illinois is in the affirmative. [FN]316 Only one state (Florida) did not address monitoring or enforcement issues. [FN]317 However, the other states were ambiguous in their approach to this issue. [FN]318




  • 1125 4. Does the Waiver Application Provide for the Recipients to Have the Choice to Participate in a Managed Care Plan That is Mainstreaming?

For social, racial, and economic reasons, attempts to mainstream has not been an overwhelming success. [FN]319 Nevertheless, mainstreaming is an important concept which is essential to quality control. Managed care plans comprised solely of Medicaid patients have the potential of becoming “Medicaid mills.” [FN]320 Because “mainstream” managed care plans are regulated by an agency other than the state Medicaid agency, the outside agency could provide a check on the state Medicaid agency, cutting cost from the managed care contracts. [FN]321 At a minimum, the state waiver application should require that recipients have the choice of at least one managed care plan that sees non-Medicaid recipients. Only three states (Missouri, New York, and Oregon) contained such provisions. [FN]322 The other states (Florida, Hawaii, Illinois, and Tennessee) had no such provisions. [FN]323

Recognizing that mainstreaming by itself is not enough, the managed care plans must be able to provide comprehensive services to the Medicaid population. Thus, the waiver should require that the states assess the ability and desire of mainstream plans to serve the client population comprehensively, including enabling services. Only three states' waivers (Florida, Oregon, and Tennessee) *1126 have such provisions. [FN]324 Finally, recognizing that ability and desire do not always translate into fact, waivers should provide that states will frequently assess client satisfaction. Five of the seven waivers (Florida, Illinois, New York, Oregon, and Tennessee) have such provisions. [FN]325

For example, Oregon keeps track of client satisfaction and complaints by requiring managed care providers to report written grievances from clients on a quarterly basis. Satisfaction surveys have been conducted since 1986. In addition, in 1990, Oregon established a survey to elicit responses from clients who disenrolled from a managed care plan. Also in 1991, a series of client hearings were sponsored throughout Oregon. It plans to follow up on these complaints and has considered establishing regional workshops, client councils, and focus groups.

Did the waiver application provide recipients with a choice to participate in a managed care plan that is mainstreaming? As to mainstreaming, all the states but one (Oregon) were ambiguous in their provision.




  1. Does the Waiver Assure Recipients' Ability to Impact Quality of Care Provided to Them Individually and Collectively?

An important aspect of quality assurance is assuring that patients have the ability to impact the kind of quality they receive. Three ways that waivers can provide for patients to impact quality is to (1) assure the patient's ability to “vote with their feet”; (2) require the establishment of formal consumer boards and require the inclusion of minority patients on the board; and (3) establish *1127 mechanisms for reporting of complaints and advocacy. [FN]326

If consumers have the freedom to change managed care plans when they are dissatisfied, then the plans are likely to respond to their needs. [FN]327 To be able to vote with their feet, patients must have some choice and mobility. Preferably, the waivers would not authorize mandatory enrollment unless there are at least three plans from which a person could make a choice. [FN]328 Furthermore, consumers should be allowed to disenroll from any plan for any reason, at least within an open enrollment period of every six months. [FN]329 Finally, consumers should have the right to change providers within any plan at any time. [FN]330 Only one state waiver (Oregon) required minimal patient choice as a component of their demonstration project. [FN]331 However, seven states (Florida, Hawaii, Illinois, Missouri, New York, Oregon, and Tennessee) provide for a disenrollment period. [FN]332

For example, after an enrollee selects a plan in New York, the enrollee has thirty days in which he or she may switch to another plan. Only Missouri provided recipients with the right to change providers within in the plan at anytime. [FN]333

Another way of assuring a consumer's ability to impact quality is to assure that they (or their advocates) have a role in the governing boards and advisory. [FN]334 In addition, there should be a formal consumer advisory board that provides input and support to the consumers' representative and to the governing board. [FN]335 Any formal board must include representation by minority consumers. Four states (Illinois, Missouri, New York, and Oregon) required a formal consumer advisory board. [FN]336 No state clearly provided for the inclusion of minority consumers on the board. However, just *1128 allowing consumer representation on boards will be insufficient if there is not the provision of resources necessary for consumers to participate in the boards. [FN]337 Those resources include transportation, educational materials and services, child care, and translation services.

Both the state and the plans have financial interests which may color the ability to accurately assess problems related to access and quality. Thus, the waiver should establish a mechanism (such as an Ombudsprogram) to monitor access and quality of care. [FN]338 In fact, an Ombudsprogram could provide direct assistance to recipients, legal advocacy, access to and review of financial and medical audits, and client education. [FN]339 States should also provides a toll-free enrollee hotline to provide immediate assistance to recipients and to channel concerns appropriately. [FN]340

Does the waiver assure recipients the ability to impact the quality of care provided to them individually and collectively? No state had an overall positive waiver application. Three states (Illinois, Missouri, and Oregon) were ambiguous in their provisions. [FN]341



D. Assuring Quality Health Care: Conclusion

Assuring quality health care is extremely important in the *1129 managed care environment. Managed care actually operates under inverse incentive, where the providers make more money and the third party payers spend less as they reduce services. This incentive may encourage plans to cut services, particularly to vulnerable populations, which are marginally necessary. It is not enough for the states to increase access to health care; they must also assure the quality of the care given.

Although no waiver completely ignored this area, none of them was positive overall in the treatment of this criteria. The most striking feature about all of the states was the lack of detail with which they addressed this issue. This is particularly true with regard to standards to assure quality of health care to minority patients.