Introduction

Monitoring the successes and challenges of accessing treatment, especially for substance use disorder (SUD), requires accurate, comprehensive and timely information that can support quality improvement efforts and effective coordination of care [1,2,3]. There are several public sources of data that are maintained by the Substance Abuse and Mental Health Services Administration (SAMHSA) which provide information regarding SUD treatment services and patient utilization [4]. These public databases have varying purposes and are used for both research and clinical decision-making.

These database systems are embedded in health information systems to capture patient-related information. The Treatment Episode Data Set (TEDS) serves as a source for analyzing utilization of treatment in relation to population demographics [5]. The Drug Abuse Warning Network (DAWN) has been used for assessing SUD prevalence and potential demand for treatment [6], though there is potential for underestimation [7, 8]. Information on the availability of treatment services, which is found in the National Survey for Substance abuse Treatment Services (N-SSATS), is used for making referrals, as well as assessments of treatment capacity for SUD in context of policies and population needs [9, 10]. There is rich information in these sources, but differing reporting requirements introduces the potential for variability in estimates and administrative strain on the reporting facilities [11,12,13].

Variability in metrics may not be the only concern for these public databases. Typically, these sources can have delays of up to a year before publishing reports, increasing the likelihood that data will not accurately represent behavioral health facility characteristics by the time of public consumption [13, 14]. Incorrect or potentially outdated information would undoubtedly have a negative impact on the ability of providers and patients to find appropriate SUD treatment. Additionally, current data does not capture information provider shortages and wait times for intake, two important factors which impact treatment accessibility and utilization [15]. Considering these limitations, the authors of this study developed and administered a survey for the purpose of collecting these additional data elements. The survey provides a potential framework for capturing these and other data elements used to assess facility and workforce capacity.

Methods

Data collection

Development of the study survey first began with inclusion of questions from the N-SSATS [16] which would be used for comparison purposes. Additional questions regarding facility workforce capacity, hours of operation and intake procedures were designed and included. The design of these questions was determined by the study research team, which is comprised of clinicians, academic researchers and statisticians. To demonstrate the supplemental information collected, a comparative survey inventory for both the N-SSATS and the study survey is provided in Table 1. The final study survey consisted of five sections: 1. Practice hours and insurance; 2. New Patient intake procedures; 3. Treatment services; 4. Licensed behavioral health professionals; 5. Unlicensed behavioral health professionals. After finalization, the survey was converted into an electronic form using the Research Electronic Data Capture (REDCap™), an online data collection and data management application.

Table 1 Comparative survey inventory
Full size table

N-SSATS data collected in 2018 for behavioral health facilities in Indiana were downloaded from the SAMHSA Behavioral Health Treatment Services Locator, which can be accessed at https://findtreatment.samhsa.gov/. Aside from information on facility name, address and contact information, N-SSATS service data are collected primarily through yes/no or checkbox questions. Therefore, each item is formatted as indicator variables so that if a respondent replied “Yes” or checked the item, a “1” is placed in the data field. These formatted data were uploaded to the electronic survey in REDCap™ to serve as baseline data.

The telephone study survey was administered by trained research staff, who adhered to a survey script and data entry instructions. Though SAMHSA surveys are completed through both paper survey and electronic survey, this study only used the telephone survey to ensure consistency in the method for data collection and entry. Facilities that declined to respond to the survey or did not answer after three phone calls were considered non-respondents. Facilities which answered questions to at least one section of the survey were considered respondents. Survey administration began in August 2018 and concluded in November 2018.

Descriptive analysis

All facility data gathered in REDCap™ were exported to Microsoft Excel. Differences between respondents and non-respondents with regards in questions asked in the N-SSATS were analyzed using the chi-square analysis to test for non-response bias. A second chi-square analysis was conducted to determine differences in responses to the N-SSATS and the study survey among respondents. Descriptive summary was produced for additional quantitative data collected regarding wait times and provider capacity. Statistical analysis was conducted in SAS 9.4 with significance at α = 0.05. Geographic maps were developed to display the distribution of behavioral health facilities in Indiana using ArcGIS 10.8. The geolocation of behavioral health facilities was based on the geocoordinates provided in the N-SSATS.

Qualitative analysis

The study survey included three open-ended questions regarding referral, intake and treatment: “What is the intake process for referrals?”; “How do patients access services?”;” Are there requirements for patients to engage in treatment?”. Answers to these questions were summarized into themes after two phases of analysis using the code-to-theory method [17]. In the first phase of analysis, a small team of researchers assigned one or more categories to responses based on level of detail given. Themes were then created by grouping related categories. After presenting the initial themes to the full research team, recommendations were provided for refining categories and themes. Themes and categories were then finalized in the second phase.

Results

Descriptive analysis

In the 2018 N-SSATS data, there were 287 SAMHSA-certified behavioral health treatment facilities located in Indiana and included in this study. Among these, 63 (22.0%) behavioral health facilities distributed across 33 Indiana counties responded to the survey. Table 2 provides a summary of the chi-square test for non-response bias based on responses to the 2018 N-SSATS data. Statistically significant differences were found with regards to providing residential services (p = 0.009), offering medication-assisted treatment (MAT) health screening (p = 0.0173) and regional location (p = 0.0064). Geographically, the highest proportion of respondents were in the Evansville, Lafayette, and Indianapolis regions (Fig. 1).

Table 2 Difference in N-SSATS data based on study response status
Full size table
Fig. 1
figure 1

Geographic distribution of respondent and non-respondent behavioral health treatment facilities in Indiana. a displays reported number of non-prescribing providers, and b displays the reported number of prescribing providers

Full size image

A second chi-square analysis tested the difference in source data among survey respondents (Table 3). As compared to responses to the N-SSATS data, respondents to the study survey were more likely to report prescribing buprenorphine (17.5% vs. 4.8%; p = 0.0233), offering health education (74.6% bs. 19.1%; p < 0.0001) and offering a sliding fee schedule (82.5% vs 42.9%; p < 0.0001). Facilities were also less likely to report offering group counseling (47.6% vs 95.2%; p < 0.0001), offering health screening (38.1% vs. 79.2%; p < 0.0001), and accepting self-pay (74.6% vs 96.8%; p = 0.0006).

Table 3 Chi-square test for differences in responses among study survey respodents
Full size table

Additional information collected from the study survey is also summarized in Table 4. Over 40% of respondents use a wait list for intake of new patients for SUD treatment, and nearly two thirds (65.4%) of these respondents reported having a wait time greater than 2 weeks. Just under half of respondents (49.2%) reported having at least one provider at the facility with a scope that includes prescribing medications, such as psychiatrists, psychologists, physician assistants and psychiatric advanced practiced registered nurses. Nearly all (90.5%) of respondents reported their location having at least one non-prescribing provider, such as registered nurses and behavioral health counselors and social workers [18]. The geographic distribution of the reported number of providers is presented in Fig. 1. The largest proportion of respondents (30.2%) reported that their clinical professionals work an average of 25–32 h per week, or the equivalent of 3 to 4 days.

Table 4 Workforce Capacity Questions
Full size table

Qualitative analysis

Of the 63 survey respondents, 54 (79.4%) answered at least one of the three open-ended questions regarding intake and requirements for SUD treatment (see Tables 5, 6 and 7).

Table 5 Themes and categories identified for the question “What is the intake process for referrals?”
Full size table
Table 6 Themes and categories identified for the question “How do patients access services?”
Full size table
Table 7 Themes and categories for the question “Are there requirements for patients to engage in treatment?”
Full size table

What is the intake process for patients referred to your facility?

Theme 1: overall intake process

Intake process typically involves three phases: 1) coordinating initial appointment/admissions; 2) completing assessments; and 3) determining treatment plans at one location. Responses revealed variation in the way in which the intake process is implemented. There were 16 facilities which indicated following the typical intake process of completing both the intake assessment and treatment plan at the initial appointment. Two facilities indicated continuation of care as their primary service, meaning that completion of an assessment and treatment plan establishment are completed by a referring agency. One facility indicated that their intake process involved completing an intake assessment at the initial appointment, followed by referral to a clinician for treatment planning as part of a separate appointment.

Theme 2: intake assessment

A total of 43 respondents indicated that some form of assessment was administered at intake in order to determine the appropriate treatment for the new patient. The most common responses referred to administering a non-specific intake assessment (n = 27) (example: “An assessment is completed”). Among those that reported using a formal assessment tool, the American Society of Addiction Medicine (ASAM) tool or the Substance Abuse Subtle Screening Inventory (SASSI) tool were most frequently cited.

Theme 3: patient engagement

Six facilities indicated that patients are engaged in shared decision-making regarding treatment during the intake process.

Theme 4: personnel involved

Seven facilities identified the clinical personnel that were involved in the intake and assessment process. The specific types of clinical personnel reported to be involved in these processes include clinical counselors, general clinicians, or substance use coordinators.

How do patients access services?

Theme 1: referral pattern

All 54 facilities indicated the method by which a referral to their outpatient clinical treatment can be made. The most common method was a self-referral by the patient (n = 32). However, 20 additional facilities indicated that both a patient and a referral site can initiate contact with a treatment facility. For the remaining facilities, referrals were accepted as mandated by a state agency or through continuation of care.

Theme 2: facility restrictions

A small number of facilities (n = 11) indicated restrictions to referrals. For instance, four facilities indicated that walk-ins were not accepted. The remaining facilities indicated limited availability for walk-ins, such as having designated walk-in days.

Theme 3: referral source

Thirteen facilities indicated the specific source from which their referrals are typically accepted. For instance, four identified a specific health care facility as the sources for their referrals, and eight indicated accepting referrals from government agencies, such as the Department of Child Services or the Indiana Recovery Works residential program. One facility reported receiving referrals from non-traditional sites, such as employers or educational institutions.

Are there requirements for patients to engage in treatment?

Theme 1: level of requirement enforcement

When asked whether there were any patient requirements for treatment, 4 requirement levels were identified. These ranged from no patient requirements (n = 7) to having individualized requirements (n = 15). The predominant response (n = 26) was that facilities had requirements specific to the treatment they offered, such as attendance policies for counseling or group therapy.

Theme 2: counseling services requirements

Many facilities offered different forms of counseling (n = 41) and had requirements based upon the type of counseling offered. For instance, 25 facilities had requirements for attending group therapy or support groups. Another 13 indicated the specific frequency of weekly counseling attendance that was required.

Theme 3: additional requirements

Beyond counseling, facilities also reported having other adherence requirements for patients. These included medication management, sobriety and curfew. Other requirements depended on the type of treatment program assigned to the patient.

Discussion

Real-time behavioral health facility data were collected by this study’s telephone survey to address factors impacting access to care. The N-SSATS collects information that is valuable to informing referrals for behavioral health treatment. This study survey, however, went beyond N-SSATS and gathered information which may have implications for coordination of care and can also help inform potential gaps in accessibility. For instance, results demonstrate that there may be a prevalence of wait lists for intake appointments and limited acceptance for walk-ins among treatment facilities in Indiana. These findings are important when considering that increased delays to treatment initiation are associated with continued substance use and lower likelihood of completing treatment programs [19, 20]. Patients may also experience discouragement with multiple attempts to schedule the first appointment [21].

Delays in the time to initiating treatment have been a persistent issue which disproportionately affects racial minorities, low-income populations, those in the criminal justice system, and those seeking methadone treatment [19, 22]. Barriers to timely admission to a treatment program also impact health outcomes and the quality of services received [23]. Of course, there are many factors which can influence this phase of intake, such as the patient’s insurance status, the type of treatment that is being sought and the referral source, but the impact of these barriers is not independent from each other [23]. For instance, expansion of Medicaid through the Patient Protection and Affordable Care Act has had no significant impact on behavioral health treatment utilization, indicating that coverage alone will not increase access without successful coordination of care [24, 25]. The implementation of quality improvement strategies, such as using electronic health records and national databases to measure wait times and personnel capacity, may assist behavioral health facilities in responding to increasing population demand [3].

In addition to addressing the referral phase of the intake process, results from the qualitative analysis suggest that the use of assessment tools for diagnosing patients and determining treatment plans lacks consistency. The type of assessment tools used ranged from standard national tools to general psychometric tools. Though the type of tool used is at the discretion of the facility, there are concerns with making informed decisions given the differing psychometrics, focus, and target population of many assessment tools [26]. There have been recommendations promoting the use of standard assessment tools such as the SASSI and ASAM [27, 28], but these tools have not proven to have comparable effectiveness [29,30,31,32,33,34]. Given this evidence, it may be worth examining alternative standard guidelines for using assessment tools which may aid in ensuring patients receive equitable and timely services [35].

Finally, the number of respondents unable to answer or provide a complete assessment of the licensed professionals at their facility is concerning for understanding facility capacity. As with behavioral health treatment and utilization data, there are public sources of workforce data which could be used for verification of workforce capacity. However, many sources lack validity, reliability, and alignment with a minimum data set (MDS) structure [36]. This MDS survey tool design has been used for developments of a state license survey in Indiana which collects supplemental information on professional practice information, capacity, and services provided by each professional [37,38,39,40]. Thus, state level workforce data could serve as a supplement to calculating behavioral health service capacity within treatment facilities.

Limitations

There are notable limitations to this study. First, this study relied on data collected from a survey which may introduce the potential for response bias. This may be reflected by the fact that there was no control as to who at the behavioral health care facility answered the phone and responded to the survey. However, because a publicly available phone number provided in the N-SSATS was used, responses most likely mirror the responses that potential patients or providers would receive when seeking specialized treatment for SUD. Second, the questions designed for collecting additional data from behavioral health facilities were not validated and could have impacted the types of answers provided by behavioral health facilities.

The chi-square test results suggest the possibility of non-response bias based on certain facility characteristics. Additionally, because of the relatively small sample of behavioral health facilities in the state that responded to the survey, the generalizability of the results is limited. Finally, results from the qualitative analyses may be subject to preconceptions; however, the potential for this risk was minimized through the implementation of a standard methodology for qualitative research.

Conclusions

The results of this study highlight the need for comprehensive and timely information about treatment facilities which can be used for clinical decision-making, facilitating care coordination and assessing organizational effectiveness. Data not typically captured by public sources, such as wait times, intake procedures and current capacity, were helpful for identifying factors which can inform referrals to behavioral health treatment. These factors can also aid in the examination of disparities in the quality and accessibility of treatment services. The data that can be collected from a survey such as this study survey, or from existing data sources, could also be used to support development of referral networks. However, additional research is needed to determine the minimum dataset that is needed to address these issues related to accessibility.