MODELS FOR FUNDING TREATMENT SERVICES
A US view: the Cobb-C4 models
Both fee-for-service and capitated managed-care systems for buying healthcare services have intractable financial conflicts of interest inherent in their designs.
In the first, the incentive is to provide more treatment than is necessary. In the latter, the opposite—financial incentives for undertreating. Usually, neither model of purchasing or funding treatment was used with logical, objective means of monitoring accountability.
To date, most monitoring has been crude (eg: mortality rates), inconsequential (eg: what % of incoming calls are answered by the third ring), or of questionable validity (eg: patient satisfaction questionnaires and complaint rates). More recent efforts attempted to address quality of care for addictions treatment in terms of more relevant measures such as monitoring rates of engagement in treatment once contact is made and retention in terms of completing the recommended care once it has been initiated.
Still, such monitoring does not address the question of whether treatment services are effective, much less, cost-effective. Also, such monitoring is not linked to the design of treatment services so as to produce optimal or acceptable objective results, nor to address questions about whether services are of adequate structure, intensity or duration to yield expected outcomes.
Accomplishing this goal requires information and innovation. Information was derived from reviewing the literature on the variables and using available data to explore factors which influence treatment outcomes. Factors influencing outcome expectations include patient characteristics and the relationships with the intensity and duration of treatment services. Development of models also included consideration of contracting arrangements that might be suitable from the perspectives of payers, providers and patients.
This work has progressed to the point that two basic models can be proposed. One is more recognisable as applicable to public-sector service and the other to private-sector contracting, but in point of fact, either can be used for public or private sector contracting.
The distinguishing features of the models are that they can tie financial incentives and penalties to treatment results. In many cases, such results are not simply measures of abstinence from substance use, which can be difficult to determine with confidence. Rather, the outcome measures can be objective and measurable — eg: what proportion of treated employees are still employed; a year or two after starting treatment; what proportion of welfare recipients continue to manifest substance use- related problems that precluded employment; what proportion of substance related offenders were rearrested or reincarcerated for a substance-related offence in a specified time.
The following is a brief description of each of the COBB/C4 funding models. Each has certain advantages and disadvantages relative to the other. It should be pointed out that the COBB/C4 models consider substance dependence as a chronic condition to be treated as any other chronic medical condition. That means that treatment is not to be considered a time -imited event, but rather will involve long-term management with indefinite periodic delivery of services as needed. In contrast, substance abuse may be more of an acute or time delimited problem that can actually be resolved with brief interventions or time-limited services. In either case, the goal is to provide results as defined in the contracting vehicles.
INDEPENDENT CASE MANAGEMENT
The Independent Case Management (ICM) model is less radical in its departure from more commonly used managed-care models. It is distinct from the traditional capitated models in that the fee, or cost, is derived from a funding stream or contract independent from the funding for the actual treatment services. Treatment funds not expended revert back to the payer and are not retained by the management organisation.
There are significant differences in the COBB-C4 model from most of the current arrangements.
The first difference is that the care-management organisation must be a clinical organisation, but not a provider. That is, the care-management organisation must employ professionals of sufficient skills and credentials so as to do the actual face-to-face assessments needed to make proper diagnoses and design a treatment plan. Case management by remote control via formulas implemented by clerks and phone authorisations without any patient contact are not acceptable. This means that the care management organisation must be local and have means of being accessible at the point where people are likely to seek care.
The care management entity must be independent of the care providers, in reality as well as name. This means that shell corporate structures in which a provider group owns the care-management organisation, the care management organisation owns the provider programmes, or both are owned by a holding company is unacceptable for the ICM model.
Another major distinction is that with the ICM model, the care-management entity is accountable for treatment results. Cost is not the only criteria. Do the services authorised by the care managers address problems and result in outcomes that are related to tangible results? The results to be achieved may be abstinence, but in most cases, it will involve other issues related to recovery and of interest to society an those funding treatment services. These issues include employability, commission of crimes, ability to parent, use of healthcare resources, etc.
The ICM model assumes that other services beyond basic treatment services will be required. These so-called “wraparound” services might include factors to enable treatment (eg: transportation expense, child care) or to enhance recovery efforts (eg: vocational counseling, social skills training). Investments in the wraparound services can not only be cost-effective, but essential in long-term gains.
Financial bonuses for exceeding expectations in terms of outcomes given service expenses and financial penalties for failure to perform can be built into the contracting framework. Over time, expectations can be adjusted based on history within a location and on experiences in other geographical areas.
ICM models have been in operation in the US for years, with the delivery of services to recipients of TANF (Temporary Assistance to Needy Families) funds. The results have been that more people needing services were identified and more services were provided for the same level of funding. In England, similar models are in pilot stages of development.
ACCOUNTABILITY BASED CONTRACTING
Accountability Based Contracting (ABC) models provide a strategy by which either a care-management organiSation or provider network go at varying levels of risk for provision of all services from case finding to primary treatment to maintenance care. One way of conceptualising the ABC model is as an augmented case-rate system. The provider/management organisation contracts with the payer to provide those components of the continuum of care needed to achieve a desired result. An example might be the easiest way to appreciate this model.
An employer with 10,000 workers in a specified geographic area is served by a provider or provider network. Of this workforce, epidemiological data suggests that about 600 will have a substance-use disorder. Actuarial data also suggests that only 30 of those 600 are likely to enter treatment in any given year and perhaps as few as 15 of the 30 will be able to remain employed by the company. The others will have relapsed and been found positive in random testing or otherwise will have been terminated or quit working.
In this example, the goal could be to more precisely estimate the prevalence of substance use disorders, increase penetration of services to 10% of the afflicted workforce (60 people), and achieve a retained employability rate of at least 65% (40 workers). The provider/case-management organisation is to provide a bid for doing so. Such a bid would involve employee assistance services if the company does not have an internal EAP or coordinating with the established EAP for case finding strategies, provide primary treatment services, and maintenance care. A variable in the contract can be a case-rate fee for each case identified so that the provider is not penalised for being more successful than anticipated in identifying and treating afflicted employees.
Employees who leave the company for reasons other than cause should not be part of the equation since they might leave for better jobs elsewhere on achieving recovery. A bonus might be agreed on if at least 75% of employees are retainable post-treatment. In such a scenario, the provider/manager must be able to deliver effective, efficient services to be competitive. Also, ongoing care would need to be accessible over the period of the contract for employees diagnosed as dependent, because they are likely to require periodic services to avoid relapses.
Competitive bidding in the ABC models will require knowing the characteristics of the workforce, the implications of those characteristics for treatment outcome expectations, and the likely results given various levels and intensity of services. At a general level, these are obtainable from the scientific literature and data available to COBB-C4. With a given workforce, these parameters can be refined over time.
A public-sector ABC model might involve a similar bidding situation for treating people arrested for driving under the influence or for inmates with addiction problems released from state penal institutions. The objective would be to reduce the anticipated substance-related rearrest rates. In this scenario, the bidding entity would again need to evaluate the characteristic of the target population, determine likely service needs, and costs related to services likely to achieve the desired targets.