Employees of Private Companies

Medium-sized and large companies (i.e., more than 100 employees) have increased their explicit coverage for drug treatment significantly since 1983. In 1988, 74 percent of employees in such companies had coverage, an increase from 66 percent in 1986 and 43 percent in 1983 (Morrisey and Jensen, 1988; Bureau of Labor Statistics, 1989a). The BLS 1988

Employee Benefit Survey (EBS) included much more detailed questions than any previous survey about the character of such coverage. Only sketchy statistical summaries of the responses to these items are available as yet, but these summaries are indicative of the direction of this coverage.

The 1988 EBS survey indicated that 28 million of the 31 million employees of firms sampled by the survey had employer-sponsored health insurance. Of the 31 million, 20.6 million were covered by plans that had an explicit provision for drug treatment of said that as a matter of course they would provide reimbursement for detoxification or rehabilitation charges. For the other 10 million employees of medium-sized and large firms, drug treatment episodes were excluded from their health insurance coverage.

Of the 21 million employees with drug treatment coverage, nearly all (96 percent) would be reimbursed for residential or inpatient drug detoxification—which is not drug treatment per se (referred to in this connection as ''rehabilitation"), although it is certainly indicative of a drug-related diagnosis. Inpatient or residential treatment was covered for 16 million employees, and outpatient treatment was covered for 17 million. There were limitations on this coverage, however, that differed from the standard limitations in the applicable health plans. ¹ For the most part, the limitations involved a differential cap on dollars or on number of days or visits, rather than different copayments, deductibles, or maximum out-of-pocket costs (Table 8-1). The most frequently imposed inpatient limit was 30 days per year; the most frequent outpatient limit was 20 or 30 visits per year. The typical inpatient limitation was based on the average chemical dependency inpatient treatment plan.

TABLE 8-1

Details of Drug Dependence/Abuse Benefits (percentage) for the Covered Employees of Medium-sized and Large Firms.

State and Local Government Employees

Insurance coverage of public employees and their dependents is relatively better documented than insurance arranged through private employers. Within the public sector, coverage for drug treatment is virtually universal for federal employees and nearly so for state and local employees. But the types of benefits are highly variable across the different plans of the thousands of state and local government entities. An estimate of this coverage is available from a BLS survey (Bureau of Labor Statistics, 1988) conducted in 1987 of benefits provided to employees of state and local governments.

Health insurance coverage for drug treatment in 1987 was more widespread among publicly employed workers than in the private sector. Among the 10.3 million full-time employees of state and local governments in 1987, the BLS study estimated that 94 percent had health care coverage, and of these, 94 percent had coverage for some type of inpatient hospital treatment for drug abuse; it is uncertain how much of this coverage applied only to detoxification.

Special limitations were usually imposed on the amount of coverage for drug treatment. About 71 percent of the 94 percent with impatient coverage were subject to special limitations on care that were different from those for other health care procedures. The most common limitation (38 percent) was a cap on payment for inpatients days of mental health, drug, and alcohol treatment. Another 22 percent of covered employees were limited in the number of days that would be covered just for treatment of drug abuse. The most common limitation (15 percent) was a maximum of 30 inpatient days; 6 percent had higher limits, and 2 percent had lower limits.

Coverage for outpatient services was more restrictive. Some form of outpatient coverage was available to at least 81 percent of employees participating in health insurance plans. Yet for only 16 percent of these was the coverage equivalent to that for other health problems. Charging benefits against mental health limits was most common—affecting 35 percent of the 81 percent with outpatient coverage. Limits on annual visits applied to 13 percent of the covered group (9 percent with 30 or fewer visits, 2 percent with 50 or more visits). There were coverage limitations on maximum dollars, or different coinsurance rates of copayments, for 18 percent of the provisions.

Federal Employees

The federal system had nearly 4 million health insurance policies in force in March 1988, covering close to 10 million current employees, retirees, and dependents.² The specifics of federal drug treatment benefits were closely examined by the Office of Personnel Management (OPM) in a document that outlined the pertinent benefits of all offering within the Federal Employee Benefits Health Plan (U.S. Office of the Personnel Management 1988). Every plan was required to offer substance abuse treatment benefits; however, their were no specific coverage standards. and the nature of coverage varied widely. The common characteristics of all plans was to make no distinctions between drug and alcohol treatment benefits in addition, their monetary values, as calculated by OPM, were all heavily weighted toward inpatient treatment. In this sense the federal plans seemed more or less to endorse chemical dependency treatment concepts, by and large tending to focus benefits on hospital-based treatment to the exclusion of nonhospital residential programs and more importantly, to provide only minimal coverage for outpatient services.

Among the 23 fee-for-service plans available, the most common coverage package was judged to include $4,000 to $6,000 per year in potential drug treatment benefits, with significant special deductibles and copayments There was much variation around this coverage: 8 plans had total annual coverage of from #2,800 to $4,000, 10 were in the $5,000 to $10,000 range and 5 were worth $18,000 or more. In 15 policies, more than 90 percent of the value of these benefits was specifically designated for inpatient treatment in hospital-based facilities. Five fee-for-service plans offered no coverage for outpatient services, and 7 others limited such services to $250 to $400 per year. Benefits of $750 to $1,000 per year were provided by 6 plans, whereas 3 offered benefits worth $1,500 to $2,500.³

Health maintenance organizations (HMOs) had benefits similar in many ways to fee-for-service plans, although the major HMOs seemed to impose fewer constraints and limitations with regard to inpatient care and the same or fewer limitations with respect to covering outpatient care. Nearly all of the largest HMOs covered inpatient treatment up to 30 days with negligible or modest copayments. Outpatient treatment was covered by all HMOs, generally to a maximum of 20 annual reimbursed visits, which is close to, although somewhat short of, the average outpatient nonmethadone treatment plan. A significant number of plans stipulated copayments of $20 (or more) per outpatient visit, whereas about half the regional plans under one large HMO covered "all necessary outpatient counseling" at minimal copayment rates.

Employers and Coverage Decisions

Although the public sector has made a limited amount of treatment available for the past 20 years (primarily directed toward criminally active drug abusers), until recently there has been little recognition of the drug problem in the work force. Private insurance policies gave little explicit recognition to the need for this type of treatment. Drug treatment, if delivered, was reported under medical diagnoses. As recently as 1983, only 43 percent of workers in medium-sized and large private companies had explicit coverage for any kind of treatment (Morrisey and Jensen 1988).

The reasons for the lack of coverage are many and varied, as are the reasons coverage has dramatically increased over the past 15 years. Not the least of the problem has been the lack of recognition or actual denial among employers that there were many or any drug-abusing and dependent individuals in their work force. Furthermore, like alcohol problems, drug problems have at best been viewed as a character flaw or personal weakness and at worst as "willful misconduct".

Another problem has been uncertainty on the part of insurers. There is uncertainty about the extent to which the benefit will be used and how much to pay for these services. It is unclear what kind or kinds of treatment should be covered—what works and what the outcomes are. This uncertainty makes it difficult for insurers to price the benefit reasonably without leaving themselves (or the self-insured entity) exposed to large potential losses if usage or cost per treatment is greater than expected. This certainty can motivate overpricing of the benefit until sufficient time as the benefit may be rated based on experience. Inflated pricing for a benefit may discourage employers (or individuals) from purchasing the benefit.

Implicit in the rationale for the addition of coverage for drug treatment is that drug treatment may pay for itself, either through improved worker productivity or through a "health cost offset" effect. There has been no rigorous analysis of the productivity-improving effects of chemical dependency drug treatment. However, a large and growing literature (Holder and Blose, 1986; Holder and Hallan, 1986) suggests that the cost of treating alcoholics is recovered subsequent to treatment by reducing their insurance claims for health services. The conclusions of this literature, although subject to methodological weaknesses, have by inference been applied to justify drug treatment, even though there are no studies of cost offsets with clients with primary drug abuse problems.

In the committee's view, the justification for insurance coverage for drug treatment does not and should not rest on insurance cost offsets. Most health care services are covered whether or not the treatment renders cost offsets. Many terminal or chronic illnesses might not be treated if the criteria of cost-effectiveness were applied. Advanced-stage cancer, stroke, and heart disease are primarily incident in older persons who have relatively short life expectancies even without the specific disease; they often have poor prognoses, and aggressive treatment tends to be very expensive (Hartunian et al., 1980). Similarly, organ transplants involve high costs and are undertaken with the expectation of modestly increasing life expectancy or quality of life but not necessarily of saving costs for the insurance plan. In the sense that drug treatment has no proven expectation for immediate reduction of health care expenditures and can be expensive, it is analogous to coverage of treatment for many terminal or life-threatening illnesses. There are, however, valid concerns about directing patients to the least expensive of equally effective treatments or providers. These concerns have been the most important recent trend influencing the extent of private coverage and are discussed in the next section.

Access to Coverage

The first issue about the relevance of state mandates for coverage of drug treatment is whether they in fact make coverage more available to beneficiaries. As of this writing, 10 states plus the District of Columbia mandate the inclusion of drug treatment benefits in group policies. Another 8 states mandate that insurers at least offer this benefit as an optional addition to basic coverage. Each state has a similar or identical mandate for coverage of alcohol treatment. ⁵

The extent of coverage (discussed earlier) for the 31 million employees (plus dependents) of medium-sized and large corporations and for 13 million public employees is much greater than would be indicated by the mandates enacted by state legislatures. States with mandates to cover or offer to cover drug treatment were home to 11.9 and 16.6 percent of the U.S. population, respectively. But in 1988, 74 percent of private employees in medium-sized and large firms that had company-sponsored health insurance had some kind of coverage for drug treatment. Among public employees the coverage rate in 1987 was 94 percent. Thus, the extent of insurance coverage for drug treatment is greater than would be indicated simply by state mandates.

A crucial issue with state insurance mandates is that private corporations that self-insure under federal ERISA (Employee Retirement Income Security Act) statutes effectively evade any insurance coverage mandates that are legislated by states. State coverage mandates are not likely to be a necessary or sufficient cause for any company to self-insure, but there is a clear tendency for self-insured companies to be less likely to cover drug treatment than companies with Blue Cross/Blue Shield coverage or employees covered under HMOs. Morrisey and Jensen (1988) found that employees of self-insured companies were much less likely to be explicitly covered for drug treatment (56 percent of employees were covered) than employees insured by a Blue Cross/Blue Shield carrier (76 percent) or an HMO⁶ (88 percent). Policies with commercial insurers, however, were the least likely to offer drug treatment coverage (50 percent). A further analysis by Jensen (1988) indicates that state mandates are not significant predictors of whether a company self-insures when other characteristics of the company are examined. The important predictors of self-insurance were the size of the state tax on health insurance premiums, the nature of the industry, and the characteristics of workers of the company. Self-insured companies do so for more economically compelling reasons than to avoid coverage mandates for drug or alcohol treatment. On the other hand, an accumulation of several relatively inexpensive mandates may be expensive enough for a company to opt for self-insurance.

Adequacy of Coverage

The adequacy of mandated coverage for drug treatment is highly problematical because coverage for drug abuse is for all practical purposes an afterthought to coverage for alcohol treatment; where coverage for drug treatment is mandated, it is virtually identical to that for alcohol treatment. Only in Maryland are there different limits on coverage for drug and alcohol abuse, and in that case drug treatment has a lower minimum coverage than alcohol treatment.

Most of the state legislatures have virtually mandated only one modality, chemical dependency treatment, and made barely enough provision for a typical course of outpatient nonmethadone treatment. Of nine state drug abuse mandates that specify minimum days of inpatient coverage, six call for minimum annual coverage of 28 or 30 days; the other three call for minima of 21,45, and 60 days.⁷

Three other state mandates cover minimum annual dollar limits for inpatient reimbursement, with values of $3,000 (per 30-day period), $4,500, and $9,000, respectively. The $9,000 coverage is for hospital-based inpatient rehabilitation treatment and is marginally or less than adequate for a 28 to 30-day stay. The lower dollar limits clearly preclude the use of most chemical dependency treatment programs at the rates typically charged. There is a great deal of evidence, however, that these rates can be drastically reduced without cutting into patient care costs by simply reducing the extraordinary rates of return that characterized these programs during the 1980s (Health Care Advisory Board, 1988). Another state mandates coverage for residential treatment ''pursuant to a treatment plan" with no minimum specified for days of care or dollars. Three states mandate $1,500 to $2,000 annual coverage for outpatient treatment of drug abuse but specify no minimum coverage for inpatient services. Another three states simply require policies to offer optional coverage of an unspecified nature.

Cost Containment

State mandates recognize several mechanisms for containing the costs of drug treatment. The primary method allowed for this purpose is the use of less expensive competitive facilities for delivery of residential treatment. Alternative treatment facilities are recognized by 34 of the 35 states that have drug or alcohol abuse mandates, usually under the proviso that the facility be licensed by the state substance abuse authority or accredited by the voluntary Joint Commission on Accreditation of Healthcare Organizations (JCAHO) or the Commission on Accreditation of Rehabilitation Facilities (CARF).

Many nonhospital residential facilities have lower cost structures than hospital-based programs and charge appreciably less per day of treatment. They do not have the continuing onsite medical facilities, equipment, and personnel required for hospital licensure, but then again, these capacities are not needed for most drug treatment clients. Insurance plans thus are often given the option of covering drug treatment in lower cost facilities. A frequent criticism of health insurance plans by nonhospital treatment providers, however, is that many insurers and third-party administrators do not in fact cover treatment in nonhospital facilities, even though these facilities are licensed by the state and accredited by JCAHO or CARF for drug and/or alcohol treatment. Although it may be in the financial interest of insures to cover treatment in these facilities, insurance plans reportedly have been reticent to do so because of uncertainty about the quality of care delivered in nonhospital-based programs.

Two state drug coverage mandates, those of North Dakota and Arkansas, specify flexibility for the policy beneficiary. In North Dakota the basic mandate is for a minimum of 60 days in a hospital plus 120 days of partial hospitalization and 20 outpatient visits. Part of the inpatient care may be exchanged for partial hospitalization care on a two-for-one basis. Arkansas mandates a minimum total value of services of $6,000 per year, delivered in hospital or nonhospital freestanding facilities or by outpatient providers. Alabama in its alcohol treatment mandate allows a trade-off of inpatient (hospital) care for treatment in a state-licensed, short-term residential alcohol treatment facility or a three-for-one exchange for outpatient treatment.

The 15 jurisdictions that mandate minimum levels of outpatient benefits range in value from $900 to $2,500 per year, or 20 to 45 visits (hours) per year. These benefits tend to have maximum copayment rates of about 20 percent.

The Value of Additional Mandates

The committee has reservations about the value of additional state mandates for drug treatment coverage. First, coverage for drug treatment is more widespread than the extensiveness of state mandates would indicate. There are clearly reasons other than mandate enactment for the spread of coverage —perhaps the increasing realization by employers that drug treatment is a valuable benefit for their employees and for the company. Second, state mandates do not apply to the growing number of companies that self-insure under the federal ERISA statutes, especially companies with more than 500 employees, of which the percentage self-insuring is now at least 60 percent. ERISA does not deal with the coverage of drug treatment services or other matters that states have attempted to address with mandates. Third, the nature of coverage mandated by many states is too much captive to the chemical dependency model, which is not the only available modality of drug treatment.

Finally, state benefits mandates are quite rigid in their structure. Only a few states permit flexibility or the trading-off of benefits of different kinds to encourage treatment purchasers and providers to seek the most cost-effective treatment choices. Typically, states mandate a minimum benefit for inpatient treatment and minimum benefit for outpatient treatment, with no opportunities to substitute less inpatient for more outpatient or greater amounts of less expensive treatments for smaller amounts of more expensive ones. Only one state, Oregon, mandates what seems, in the committee's view, to be the most sensible option: a simple minimum dollar value of insured drug treatment coverage. If that dollar value is realistic in terms of competitive prices, it enables companies and individual beneficiaries to seek the best treatment values while using managed care strategies to guard against inappropriate use and to collect useful information about provider characteristics and performances.

Extent, Costs, and Trends of Coverage

The private tier of drug treatment providers is largely oriented toward treating the employed population and their family members. The majority of this population, about 140 million individuals, have specifically defined coverage for drug treatment in their health insurance plans. About 48 million others who are privately insured do not have specifically defined coverage for drug treatment, although coverage may occur de facto under general medical or psychiatric provisions. As of 1988, the health plans of about 67 percent of full-time employees of firms with 100 or more employees offered specifically defined coverage for some types of drug treatment, although the actual extent of benefits under these defined coverage provisions is uncertain.

Actuarial studies of claims experience yield rather modest estimates for the overall cost of covering drug treatment. Data about drug treatment expenditures tend to be buried under more inclusive headings and behind "horror stories" involving troubled adolescents with multiple diagnoses spending months in psychiatric facilities. Nevertheless, the committee estimates that a health plan with typical coverage now spends 1 percent or less of its total outlays for explicit drug treatment, most of it for hospital inpatient charges—with a large fraction of that cost devoted to detoxification. However, there has been a substantial apparent growth in the rate of drug treatment claims in recent years, particularly for insured adolescents. It is difficult to know how much of this increase is actually due to the replacement of psychiatric or medical diagnoses with more revealing or accurate drug problem diagnoses versus an increased demand for drug treatment in the insured population. Possibly, both processes are occurring.

Although this growth is disturbing to the degree it increases the aggregate cost of health insurance premiums, it is desirable if it means that an increased number of those who need treatment are seeking and receiving it, particularly if the treatment delivered is appropriate, effective, and reasonable in cost. Some payers, however, reacting in part to the high costs in a small number of cases and the high incidence of recidivism, have strongly questioned the value of drug treatment episodes. There is a movement at least rhetorically to view drug treatment as part of the non-medical/surgical fringe of health coverage that may be differentially limited (rather than cut back evenly with other benefits across the board) to trim increasing overall costs.

Mandating Drug Treatment Coverage

There are legislative mandates in 18 states plus the District of Columbia that require certain categories of employer-supplied group health plans to specifically cover—or offer optional coverage for—drug and alcohol treatment. (Another 19 states require some degree of coverage for alcohol treatment only.) In the committee's judgment, private coverage of drug treatment is beneficial to individuals and employers and should be included in every health package; however, legislative mandates at the state level have not necessarily proved to be an effective way, and are clearly not the only way, to induced adequate coverage. Most of those in the insured population whose plans include explicitly defined coverage for drug treatment reside in states that do not have legislative mandates for such coverage. Moreover, the political process has often produced less-than-optimal mandatory provisions that are difficult to adjust and overly rigid and that pay too much attention to limits on the length of stay and the number of visits rather than to the cost and effectiveness of treatment. Most mandatory provisions have the constraining effect of funneling people toward one particular modality of treatment by favoring inpatient stays of prespecified lengths.

The committee believes that the development of soundly derived standards for admission, care, and program performance will do more at this time to generate appropriate coverage than a further set of mandates. If mandates are to be used, efficiency and fairness dictate that they be applied to all competing insurers. Yet if the private market leaves large numbers of the insured population without coverage for drug treatment, it may be necessary for government to intervene. Such action could involve subsidies for drug treatment coverage, tax preferences for certain kinds of coverage, or mandates, with the choice dependent on judgments about the incidence, efficiency, and equity of alternative ways of financing coverage.

Optimal Coverage Provisions

Private insurance provisions (including most legislatively mandated benefits) often include financial incentives for beneficiaries to seek more expensive hospital or residential treatment. Insurance coverage until very recently has heavily favored hospital-based inpatient stays over outpatient visits and continues to encourage the "gold standard" medical model rather than more explicitly psychological or socially oriented treatment. Although residential drug treatment, including hospital treatment, often serves clinically important functions such as permitting intensive therapy, isolating the patient from an adverse environment, or treating concurrent psychiatric or medical complications, the hospital-specific components of such programs (e.g., 24-hour onsite medical coverage) do not seem to be the therapeutically important elements in the drug treatment programs that are sited there, even though the availability of these components is used to justify charging acute care hospital rates for all clients.

There is currently a movement a foot to reduce hospitalizations, mainly as a result of cost-containment measures, especially precertification, utilization review, and negotiation of preferred provider arrangements. The committee's principal response to these developments is to favor them in general, but it specifically recommends that curbs on unit-of-service costs for inpatient care be strengthened and that payers insist on the generation of reliable performance/outcome data. There are two reasons why it would be unwise to institute blanket denials of coverage for hospital-based drug treatment. First, in some states and localities, hospital-based programs are the only sites providing residential treatment. Second, a certain proportion of the individuals who seek drug treatment also have problems for which a course of acute hospital care is appropriate, namely, complications or co-occuring medical or psychiatric disorders.

Altogether, such cases in which it may be justifiable and necessary to initiate drug treatment services in a hospital setting may total one-fourth of privately covered clients who seek drug treatment. This figure is only guesswork, however, pending the advent of objective diagnostic assessment, systematic follow-up data collection, and systematic services research and evaluation of private treatment programs. Whatever the numbers involved, the committee recommends that drug treatment services at hospital sites be reimbursed separately from other diagnoses or hospital services, as there appears to be no compelling reason why these services for most drug treatment patients should routinely command fees comparable to acute care rates rather than to reasonably competitive residential treatment rates.

Insurers and employers need to become better informed about drug treatment and to structure their benefits to support controlled access to a broad range of the most appropriate, effective, and efficiently priced treatments rather than to a narrow (and expensive) band of options that are similar in form to the treatment of acute medical conditions. Private insurance, health maintenance organizations, and other health financing plans should cover appropriate, adequate, cost-effective drug treatment and not reimburse the cost of excessive, inappropriate treatments or charges (Table 8–2).

TABLE 8–2

Preferred Sites of Types of Treatment for Selected Categories of Drug Treatment Clients.

The committee recommends that private risk bearers, in lieu of arbitrary payment caps or exclusions, institute rigorous, independent preadmission review (where possible) and concurrent review of all hospital and residential admissions as a way to control access and utilization, ensure appropriate placement, and manage costs. Preadmission review may not be necessary for outpatient admissions, but early concurrent utilization review is important for outpatient treatment to ensure that diagnostic criteria are observed and charges are reasonable. Employee assistance programs can serve as utilization managers in cases in which their personnel have appropriate training for matching patients to treatment. Hospital utilization should be managed under the same terms as recommended for public coverage (see the section on utilization management in Chapter 7). In general, utilization management and indicators of performance are needed to meet concerns about costs and inappropriate treatment. In this area, as in other dimensions of health care, the stress should be on efficient delivery of effective care, in which responsible clinical innovation is encouraged, tested, and used when its worth is demonstrated.

The committee recommends that private payers insist that providers participate in and agree to the publication of regular, independent follow-up up surveys to determine client outcomes, taking into account data on admission characteristics such as problem severity. Providers and payers should be able to compare treatment results with overall program norms to ensure that good performance is maintained and poor performance recognized when it occurs.

The committee recommends that the provisions of drug treatment benefits, including deductibles, copayments, stop-loss measures, and scheduled caps, be similar to provisions for treatment of other chronic, relapsing health problems. Except in terms of limitations on the length of stay and number of visits, such provisions are mostly the rule today. The committee believes that sound cost-containment and managed care arrangements and reliable performance and outcome measurements will in short order obviate the need for separate length-of-stay and dollar caps on coverage. Nonhospital residential and outpatient treatment delivered in state-certified treatment programs should be covered. Covered limitations, charge schedules, and cost-containment incentives. (e.g., copayment schedules) should be adjusted to reflect the findings of research on appropriate models, lengths, and costs of drug treatment—especially the recognition that longer residential and outpatient stays are strongly correlated with more beneficial results.