Needs Assessment Methodologies in Determining Treatment Capacity for Substance Use Disorders: Environmental Scan Final Report. RESULTS II: NEEDS ASSESSMENTS FROM OTHER PROVIDER TYPES

09/16/2019

This section provides an introduction to the research question that asks, in part, "Are there needs assessments from other provider types (for example, primary care physicians) that could be applied to the SUD treatment workforce?" Two models are discussed: general health care and the Medicaid 1115 Waiver Delivery System Reform Incentive Program (DSRIP). The final report will address the additional components of the question that call for recommendations: "What are the differences between other workforces and the SUD workforce that might make such application difficult, and how could these difficulties be overcome?".

General Health Care

The most obvious place to identify approaches in other provider types of systems that might serve as a model for SUD needs assessment is in the general health care field. Two types of health care needs assessment that are particularly promising methodological models are the HRSA's designation of HPSA and the needs assessments conducted in preparation for CMS DSRIP applications. While these are conducted at a scale and complexity of methodology that is beyond the scope of most locally conducted SUD needs assessments, they merit review as models of best practices.

Health Professional Shortage Areas (HPSA)

HPSA assessments, which are conducted by primary care clinics applying for HRSA funding, address provider shortages in primary care, dental care, and mental health. Determination of eligibility for funding is based on a scoring algorithm that is specific to each of the three provider types. Several features of the HPSA determination process are relevant to SUD needs assessment, particularly in methods for addressing the complexities of assessing capacity as discussed above. Briefly, these include:

  • Determinations made on the basis of three types of shortage: geographic (shortages affecting the entire population of a specified area), population (affecting specific subgroups such as low income), or facilities (specific types of health care facilities).

  • Scoring based on weights for a set of several components: for mental health these are population-to-provider ratio, percent poverty, ratio of elderly and youth to the general population, prevalence of alcohol and substance abuse, and travel time to care facility.

  • Detailed definitions for components, for example how to calculate and count full time equivalent (FTE) psychiatrists.

These methods provide for quantitative comparison of needs across geographical areas and across different types of needs within an area, thus serving the function of needs assessment in providing guidance for priority-setting and decision making--in this case, for allocation of resources to areas with the greatest need.

As models for SUD treatment capacity needs assessments, the HPSA approach has one primary limitation: it is a method for relative allocation of resources across areas--directing resources to localities where the shortages are greatest--but is not a method for planning at the system level. That is, it does provide an indication of how allocations should be made within areas: how many psychiatrists, nurses, social workers etc. should be provided in a particular area. In this respect, the HPSA methodology is similar to the SAMHSA block grant formulas, which also use social indicators to determine the size of the allocation to individual states but not how the state's allocation should be used. To adapt this methodology for SUD treatment system capacity needs assessment, it would be necessary to combine it with the "what should be" component to determine how the workforce should be distributed in a balanced system of care. An example is McAuliffe's Rhode Island needs assessment, which determined the appropriate location of treatment facilities based on local area social indicators. The IMU on which the Health Provider Shortage Area designation is based would need to be quite different for SUD. The IMU score is based on four weighted factors: providers-population ratio, percent at 100% of the FPL, percent aged 65 and over, and the infant mortality rate. The indicators suggested by Green et al. (2016) would be better suited to SUD. Finally, the designation of medically underserved area is determined by a cut-off IMU score of 62 or below (out of a possible 100). An appropriate cut-off score for SUD would need to be developed.

Delivery System Reform Incentive Program (DSRIP)

DSRIP is a type of CMS Section 1115 Waiver created by the ACA that allows Medicaid funding to be used to create incentives for providers to pursue a wide range of system reforms. (The Kaiser Family Foundation website provides an overview of DSRIP waivers.[7]) States typically require eligible entities to submit a plan for approval that outlines the specific projects and metrics they intend to implement, features of which could provide models for smaller-scale SUD needs assessments. An example of a DSRIP report is the Capital Region, Mohawk and Hudson Valley DSRIP Community Needs Assessment (NY CNA) conducted by the Albany Medical Center and Ellis Performing Provider Systems (PPS) (Albany Medical Center PPS & Ellis PPS, 2014).

The NY CNA presents a huge amount of information about health care resources and utilization, as well as the health needs for the general population, Medicaid insured, and uninsured populations for 11 counties. Some features of this report that offer models for SUD treatment needs assessments include:

  • Engagement with a diverse group of organizations in developing the CNA including community health centers; health plans; local business; philanthropy; colleges/universities; school districts; faith-based; housing; mental/behavioral health; transportation; social services; and media.

  • Diverse methods of data collection designed to fill gaps in any one source. These included a provider survey, a community-based organization survey, a consumer survey, focus groups with Medicaid members and the uninsured concentrated on three groups (chronic disease, mental health and substance abuse), and listening sessions held with providers, including physicians, members and the general community.

  • Detailed information about providers from the state's Provider Network Data System, which the authors considered, based on research reports, to be more reliable than the widely used Area Health Resources Files.

  • The distribution of behavioral health providers in local areas (square miles by neighborhood) in relationship to the density of beneficiaries with behavioral health conditions.

  • A summary of systems assets and resources identified and quantified through the collaborative process of developing the CNA.

Locally conducted SUD treatment needs assessments are unlikely to have the resources that were available for the NY CNA, and this group also benefitted from New York's exceptionally rich health information systems. However, the practices identified in the points above could be readily adapted for use on a smaller-scale.