Structure measures assess the capacity of a provider, clinic, or health care organization to deliver evidence-based care, and they may address such topics as staffing, data systems, and treatment procedures (Donabedian 1988). Structure measures are most often used in accreditation or certification programs that set minimal requirements for health care organizations (AHRQ 2013a) and may be helpful as a roadmap for organizations to follow when they assume responsibility for new activities or new populations (Teague et al. 2012; NCQA 2013).
With current federal efforts to spur delivery system reform, structure measures are being used to qualify organizations for participation in programs or eligibility for new incentives. For example, consistent with the ACA's incentives for developing new models of care and adoption of health information technology, structure measures have been introduced for behavioral health homes that take on responsibility for integrated mental and physical health care (SAMHSA-HRSA Center for Integrated Health Solutions 2012). To receive financial incentives for meaningful use of EHRs, providers are required to meet objectives for collecting and using particular types of data. Structure measures may also be used to certify providers, thereby making them eligible for new payment arrangements. For example, in some states, Medicaid providers must demonstrate that they meet certain standards before they can bill for Multisystemic Therapy (North Carolina MST Funding and Medicaid Standards n.d.).
In mental health care, structure measures are among the diverse means for gauging the fidelity of treatment (Teague et al. 2012). Fidelity can be conceptualized as a combination of provider competence in delivering treatment, provider adherence to treatment, and the extent to which the treatment differentiates itself from other treatments (Schoenwald and Garland 2013; Schoenwald et al. 2011a, 2011b). Because treatments are adapted to fit into service settings, structure measures can assess whether organizations have the capacity to deliver evidence-based psychotherapy (sufficient numbers of trained staff and procedures for supervision, for example) while process measures indicate something about the extent to which the adaptations result in the delivery of treatment that still contains the core components of the evidence-based model. Pairing structure and process measures may help to provide a more comprehensive understanding of the quality of care. For example, a study of mental health care in the U.S. Department of Veterans Affairs (VA) used a facility survey to assess the availability of evidence-based practices such as CBT as well as process measures of the receipt of evidence-based practices (the percentage of patients with PTSD who had any CBT visit during the study period) (Watkins et al. 2010, 2011).
Structure measures to assess the capacity of clinics or organizations to deliver evidence-based psychotherapy could address such topics as:
Availability of staff who are trained in evidence-based psychotherapy.
Adoption and implementation of protocols or guidelines for the use of evidence-based psychotherapy.
Availability of tools that support a consumer's engagement in therapy (web-based tools, workbooks, or homework materials).
Availability and use of tools for assessing symptoms and functioning, and the use of those tools to monitor outcomes.
Likely data sources for structure measures include data reported by an organization or clinic through surveys, reports, or documentation. Audits or reviews of documentation may be warranted to ensure the validity of these reports when payment is at stake.
Assesses the availability of services. Structure measures can provide critical information about the adequacy of the network of providers within a health system, health plan, ACO, or state in terms of their training and capacity to offer evidence-based psychotherapy. Such information is helpful for assessing gaps in accessibility of care, especially in rural or underserved areas.
Provide guidance on infrastructure development and best practices. By defining the infrastructure needed to provide evidence-based psychotherapy, structure measures could help clinics and organizations understand and implement new approaches to care. As described in a later chapter, these new approaches to care could include the use of repeated outcome assessments to report symptoms and functioning to guide care. These measures could also have the flexibility to address different types of psychotherapies that would be relevant to different populations or provider settings. Structure measures could build on existing approaches to fidelity training and supervision used in research and developed for specific psychotherapies. Similarly, structure measures could lay out expectations for data systems and protocols for collecting information on symptoms and functioning and how to adjust treatment when there is no improvement. Such an approach could offer flexibility to select tools that are relevant to different populations or conditions.
Support credentialing and payment. Structure measures could support or complement existing credentialing processes, thereby allowing purchasers and health plans to select clinics or provider organizations that are equipped to provide evidence-based psychotherapy services and outcomes monitoring. Alternatively, performance on structure measures could be used as the basis for differential payment and incorporated into new models, such as the behavioral health home.
Support consumer choice. Health plans or other entities could make information available to consumers about the capability of providers to allow them to select providers that have expertise in evidence-based psychotherapy, including treatments specific to their conditions. This information could be provided in health plan provider directories and also made available to other providers to facilitate referrals to care consistent with consumers' desired treatment.
Lack of clarity on necessary structures. Defining structure measures that are clear but also flexible enough for implementation in different settings can be challenging. Although there is evidence to support the use of certain psychotherapies, research has not identified the structures that contribute to psychotherapy outcomes (for example, addressing the matter of whether there is an evidence base for the amount of training or supervision needed to implement the treatment with fidelity in typical treatment settings).
Limited resources for investing in training and infrastructure development. Building the capacity to deliver evidence-based psychotherapy may require significant investments in staff time as well as outlays for expert training, consultation, and supervision as therapists learn and implement new skills. This is especially a concern for proprietary therapies. Building data systems to support outcomes monitoring may also represent new expenses for mental health providers, particularly given the lower rates of EHR adoption in these settings (Druss and Dimitropoulos 2013).
Documentation burden. Documenting the steps involved in implementing evidence-based psychotherapies or use of structured processes for monitoring outcomes may impose an additional burden on clinics or organizations, and reviewing this information with an external auditor adds to costs. Managed care organizations and states now credential providers based on licensure, and they typically obtain self-reported information on expertise and training. New efforts to certify health homes offer an avenue for implementation, but auditing and documenting structures described here would be a new activity. Because of restrictions on access to behavioral health records in some locations, auditors may have to make special arrangements with provider organizations, health plans, or state agencies to access documentation for external review. It is also unclear whether the proprietary nature of some psychotherapies would limit the kinds of standards set for supervision or training.
Although audits could focus on protocols and reports, they might also include using independent raters to assess whether providers demonstrate competency in delivering evidence-based psychotherapy. Many studies have used trained, independent raters who either directly observe the delivery of therapy or review video or audio of therapy. Independent raters are considered the most valid reporters of fidelity to psychotherapy because they have been specifically trained to recognize the delivery of the therapeutic elements and, because their rating has no personal bearing for them, their responses are the least likely of all raters to be biased (Barber et al. 2007; Schoenwald et al. 2011a). Clinics or organizations could contract with independent raters as part of a system of evaluating the competency of their providers (although clinics would need to take special care to protect privacy of consumers).
Credentialing on the basis of reports from independent raters could be burdensome and costly. Nonetheless, some states are doing this. For example, beginning in fiscal year 2014, the Texas Department of State Health Services will require providers who offer psychotherapy to adults with depression to demonstrate competency in CBT, which will be determined by an independent rater who reviews and scores a videotaped therapy session (Robinson 2012). Providers who score below an established threshold will be prohibited from providing psychotherapy to adults with depression until their competence has been established. The state has given local mental health authorities a list of organizations qualified to provide independent competency ratings, which costs $150-$300 per videotape reviewed. As of this writing, we do not know how often providers must demonstrate competency, nor do we know the total cost to the Texas health system to obtain independent ratings.
Structure measures are appealing as an initial step for setting expectations related to how providers should implement evidence-based psychotherapies and monitor consumer symptoms and functioning over time. This is particularly useful given the very limited information on the use of evidence-based therapies in routine settings and concerns that where the therapies are implemented they are not being done with sufficient adherence to protocols demonstrated to be efficacious in research trials. Furthermore, as described in a later chapter, processes for outcomes monitoring are critical because even in randomized trials showing the benefit of evidence-based treatment, many participants do not benefit and efforts to adjust treatment are needed. Structure measures are likely to serve best as building blocks to guide the delivery of care and as complements to other measures.