Firm quantifies extent of addiction workforce shortage

In an attempt to quantify the well-recognized gap between the size of the behavioral health workforce and the demand for treatment, a consulting firm has developed a proprietary tool called the Provider Availability Index (PAI). Advocates for Human Potential (AHP), based in Sudbury, Mass., and having recently finished a workforce project for the Substance Abuse and Mental Health Services Administration (SAMHSA), created the PAI using labor force databases and prevalence estimates of mental illness and substance use disorders (SUDs), nationally and by state.

The PAI quantifies the number of providers available per 1,000 individuals in need of treatment. In principle, it is similar to the Health Resources and Services Administration’s (HRSA’s) professional shortage area designation for mental health. But the PAI is broader, since HRSA doesn’t include SUDs and bases its shortage designations solely on the availability of psychiatrists in a region, according to AHP.

The PAI therefore constitutes a new standardized measure for comparing labor force availability by state, separately for any mental illness, serious mental illness, and substance abuse. Here we are focusing only on SUDs—which is somewhat problematic in itself, in that AHP admits that the databases it relied on are not specific for SUDs.

Nationally, the PAI for SUDs is 32.1, meaning that there are 32.1 providers for every 1,000 individuals needing substance use disorder treatment. However, in some states the number is much lower. AHP divided the states into quartiles, relative to the national average. The lowest quartile states are Nevada, where the PAI is 10.8, Georgia (17.3), Texas (17.7), Indiana (18.3), Florida (20.0), Arizona (20.1), Hawaii (21.6), Louisiana (22.3), South Carolina (22.9), Michigan (22.9), Arkansas (24.4) and Alabama (25.5). Kentucky and Utah also are in the lowest quartile, but the exact PAI was not available for them.

“Our motivation for this initiative stems from widespread consensus that overall and geographically based shortages of behavioral providers exist, and will be exacerbated by health reform, without reliable estimates of the inadequacy,” says Jeff Zornitsky, director of strategic initiatives at AHP. The measures derived for the analysis will be of use to Single State Authorities (the state agency directors in charge of the Substance Abuse Prevention and Treatment Block Grant) and treatment providers, among others, Zornitsky says.

The AHP data can be used to direct more resources to the neediest states, a priority for SAMHSA, according to Zornitsky. “By identifying individual states, we could tell a state you have a problem, and suggest to them what they should consider doing,” he says.

AHP hopes to help states perform better in terms of workforce, because the gaps between provider supply and service demand are expected only to get bigger with Medicaid expansion.

Asked what the lower quartile states have in common that could be responsible for their bigger gap, Richard Landis, senior director of strategic planning and workforce development at AHP, says, “I think it has to do with how the occupation is viewed, whether people are choosing to go into the occupation in that state, or to stay in that state to work.”

The real problem “comes down to salary,” says Landis. Colleges, in particular community colleges, are producing people who are trained to be SUD providers but who can’t find work that pays well enough. Landis cites a dilemma seen in Genesee County, Mich., where AHP did a study that explains the workforce problems “on the ground.”

“They were having an awful time, even though they had schools that could provide the labor force,” he says. Graduates went to other states to work, because salaries in Michigan were too low. “We need to work with these programs so they can pay more,” Landis says. Paying salaries that are too low creates a death spiral. “When they have job openings that aren’t filled, they treat fewer people, and they get even paid less,” he explains.

Behavioral health salaries are lower, even for the same job categories, than in general medicine, says Zornitsky. “People with comparable levels of education—for example, social workers—get paid much less at a behavioral health site,” he says. Therefore, the costs of college, two years of training, and getting licensed don’t appear to be worth it.

The workforce gap isn’t new, so how is it that SUD treatment has survived so far? Mainly because so many people who go into the field are “mission-oriented,” says Zornitsky. “They know they can make more money in regular healthcare, but they care more about behavioral,” he says. “The vast majority of people in this work have been personally or family-affected” by SUDs, he says. “They’re more passionate about it.”

While it’s good to be passionate, this also means that employers have gotten away with paying their workers too little. The average salary for a social worker in behavioral health was $38,600 in 2010, compared to $47,200 for social workers employed in medical settings, says Zornitsky.

Medicaid expansion

The workforce and SUD prevalence data used for the PAI analysis came mainly from the Bureau of Labor Statistics (BLS) and the 2010 National Survey on Drug Use and Health, respectively. The PAI for SUDs in Medicaid expansion states is 33, similar to the national average of 32.1. This is worrisome, Zornitsky says, because “you know there’s going to be increased demand, but they don’t have any more availability than the national average.”

There are job openings in the behavioral health field, but half of those openings occur because someone retired or left the field—what AHP calls “replacement demand.” What is needed are job openings that occur by new job growth, according to AHP. Replacement demand exceeds 50% in seven of the bottom quartile states, and four of these states are Medicaid expansion states.

Medicaid expansion states that already are in the lowest quartile have the greatest difficulty providing services. Some of these states also have a relatively low score in terms of readiness to adapt to a changing healthcare business environment, as measured in another recent AHP analysis.

BLS job categories

Unlike a 2012 workforce report from the Addiction Technology Transfer Center (ATTC) Network, which was specific to SUDs, the PAI is based not on a provider survey but on the labor force. The occupational categories cover people who provide SUD services, but not exclusively.

Data sources for workforce information were based on employment occupation statistics and projections from BLS. The only available occupational categories used by BLS for behavioral health, and therefore used by AHP, are psychologists, psychiatrists, counselors, social workers and psychiatric technicians and aides. There are significant omissions in the BLS classifications, which do not accurately represent the specific workers in the addiction field and tend to rely more on the idea that SUDs are included in the mental health categories.

In addition to BLS classifications, AHP used three worker categories based on the federal Digest of Education Statistics for 2009-2010: mental health counselors, SUD/addiction counselors, and rehabilitation counselors. The various classifications used don’t include some of the emerging occupations in the addiction field, such as certified peer specialists or recovery coaches.

For estimates of SUD prevalence, AHP used SAMHSA estimates for three categories: any mental illness (which includes all mental illnesses but not SUDs), SUD, and serious mental illness (SMI) for individuals 18 to 64. This omits the prevalence data for the youngest and the oldest segments of the population, which is one reason that the prevalence estimates are low.

Defining the SUD workforce using BLS categories is impossible, says Landis. BLS has combined mental health and SUDs, “and there’s no way to de-combine them,” he says. “Everyone recognizes that it’s difficult to find out the scope of the problem because the data isn’t there.”

And despite the present widespread attention to opioid addiction, the most vital providers in this area in the SUD workforce are left out: opioid treatment programs that provide methadone maintenance, and physicians who prescribe buprenorphine or naltrexone. “They’re a very small proportion of the workforce,” says Landis.

However, AHP did ask the 900 SUD providers in the recent readiness to change survey whether they provide medication-assisted treatment. Forty-four percent said they don’t, says Zornitsky. One reason was that medication-assisted treatment wasn’t part of their “treatment philosophy,” he says. But also cited were workforce problems—having difficulty recruiting a prescriber, or not being trained to administer medication-assisted treatment.

Possible improvements

As possible solutions, particularly for states with the biggest gap, AHP recommends improving “people management practices,” because the high replacement demand specifically means that staff retention remains a problem.

Another recommendation from AHP: Improve the data collection system of the behavioral health labor force. Until more is known about what it looks like, not much can be done in the SUD sector. Last year SAMHSA submitted a report to Congress that was supposed to be about the SUD workforce, but was actually about the broader behavioral health workforce. It was a reflection of the poor information on job categories, but the ATTC report, while more useful in terms of SUDs, was based on a survey of providers.

Also, AHP recommends that new low-level occupations be introduced into the field, such as certified peer specialists, recovery coaches and care coordination specialists. However, these entry-level positions must include “career ladders that lead to sustainable family wages,” says Zornitsky. This will “reduce the workload of highly trained professionals and create family supporting jobs for all behavioral health professionals,” he says.