A $600 million system was supposed to streamline aid for people with disabilities.
Instead, it's ‘a giant mess’
Six years after promising to modernize Minnesota’s biggest assistance program for people with disabilities, the state is struggling to fix a system racked by cost overruns, computer breakdowns, delays and staff turnover.
The system, known as MnChoices, was supposed to streamline and simplify a program that disburses more than $3 billion a year to Minnesotans with physical and developmental disabilities.
Instead, it has produced wildly inconsistent results and a chorus of complaints from the county officials and front-line employees who use it.
The state has already spent $600 million on MnChoices, and its annual operating expenses are forecast to surpass $170 million — more than 10 times what officials predicted. Nearly one-fourth of the 2,300 workers who were hired and trained to use MnChoices have quit.
“MnChoices is a runaway spending train that needs to be stopped,” said Sen. Jim Abeler, R-Anoka, chairman of the Human Services Reform Finance and Policy Committee. “It was supposed to standardize the evaluation process, but it does not appear to have standardized anything.”
By sending trained county assessors equipped with laptops and a computerized questionnaire into families’ homes, MnChoices aimed to bring rigor and consistency to the process of evaluating tens of thousands of people who apply for state assistance.
But the program crashes so often that the assessors have developed elaborate workarounds and sometimes simply resort to pen and paper. A long-awaited software upgrade to address the problems, promised for this year, has been postponed until 2021.
The technical problems are more than a nuisance: They disrupt the delivery of vital services to tens of thousands of Minnesotans with physical and developmental disabilities. County supervisors say MnChoices is so unstable that a single errant keystroke can determine whether a medically fragile child gets round-the-clock care at home — or almost no support at all.
Frustrated families are challenging the results of their MnChoices assessment in rising numbers. Last year 32% of eligibility assessments in the most common category of aid were reversed on appeal, according to documents obtained by the Star Tribune.
An outside analysis for the Minnesota Department of Human Services found the MnChoices software had not been tested to assure it produced consistent results for aid recipients.
“Most of us became social workers because we really wanted to help people, and that became difficult with MnChoices,” said Jennifer Bagne-Walsh, who supervises assessors in Olmsted County in southern Minnesota. “It’s a giant, inefficient and mostly impersonal tool.”
Officials at the Minnesota Department of Human Services, which oversees Medicaid, defended MnChoices, saying it is more consistent than its predecessor and is still being enhanced. The agency has simplified the assessment document by removing about 100 questions and reorganizing it to make interviews more “person-centered” and conversational. The state also is developing new training for county assessors on how to improve their interviewing techniques.
Stacy Twite, assistant DHS commissioner for community supports, added that the agency will be making changes to the computer system to reduce the number of crashes.
“I think the MnChoices system works better than it did back in 2013 when it first launched, and we continue to make improvements,” Twite said.
Patty Pflepsen, a retired Dakota County assessor, said “MnChoices was so big and so cumbersome that it created one more barrier between you and the person seeking help.”
Conceived more than a decade ago as a way to modernize a program that helps thousands of Minnesotans with severe disabilities such as cerebral palsy, traumatic brain injuries and developmental delays. Today, some 90,000 Minnesotans receive this coveted form of assistance, known as a Medicaid “waiver.” The sums, which typically range from $30,000 to $100,000 per recipient per year, can be used for personal care aides, physical therapy, home medical equipment and other supports that allow recipients to remain at home rather than moving into nursing homes or other costly institutions.
The new computerized questionnaire, designed with the help of outside contractors, combined several screening tools into a single assessment for people of all ages and disabilities. The overall system was designed to remove human subjectivity by requiring county workers to be certified and trained on the same tool with the same questions.
Technical problems dogged the system from the start, according to internal memos obtained by the Star Tribune and interviews with county officials.
Over budget every year
Since its inception in 2014, MnChoices’ total spending has exceeded projections.
Source: Minnesota Department of Human Services
The online platform failed regularly, often while assessors were interviewing people in their homes. Even now, MnChoices is plagued with regular “blackouts,” in which the system goes dark for hours or even days at a time while the state makes upgrades and attempts to fix glitches, county officials say.
“It was a giant mess,” said Sheana Schlichter, a former assessor from Scott County who now works as a consultant for families seeking waivers. “Our biggest problem is the system kept crashing.”
In interviews, more than a dozen current and former county assessors described their frustrations.
“We kept hoping that the technology would improve and the process would be more streamlined, but it just kept getting worse,” said Bagne-Walsh, the supervisor from Olmsted County. “The bigger the MnChoices tool gets, the more bogged down it gets.”
Some counties estimate that it can take an assessor 12 to 16 hours to complete a MnChoices evaluation — several times longer than the old, paper-based screening system. To speed up the process, assessors said they sometimes skip entire sections of the evaluation, which has reduced the program’s reliability.
The new system also broke the traditional relationship between counties and vulnerable individuals. In the past, people with disabilities who sought state assistance were typically evaluated by a county social worker — the same person might help them apply for aid, navigate the available services and develop a care plan.
Under the new system, people seeking waivers found themselves being evaluated by newly hired assessors who often had little experience in social services or medical care. Turnover was high, which meant that families often saw a different assessor each year.
Assessors also were so busy getting through the questionnaire that they had little time to view conditions in the home or engage in a broader conversation about the applicant’s goals, said Patty Pflepsen, a retired assessor from Dakota County.
“It is already incredibly hard for anyone with a disability to open up about their issues,” Pflepsen said. “MnChoices was so big and so cumbersome that it created one more barrier between you and the person seeking help.”
Sheana Schlichter, a former assessor from Scott County, now works as a consultant for families seeking waivers. She was on her way to meet with a mother in Northfield to help them get the services their 13-year-old son needs.
In the four decades since the federal government approved the idea of Medicaid waivers, no other state had succeeded in developing a comprehensive assessment tool for the full range of physical, cognitive and behavioral disabilities. Today most states still rely on a variety of screening forms and questionnaires that determine eligibility based on age, diagnosis or medical condition.
Minnesota’s project “was the holy grail,” said Mary Sowers, executive director of the National Association of State Directors of Developmental Disabilities Services. “No other state has figured this out.”
The state of Arkansas embraced the Minnesota technology and even kept the MnChoices name. But so far, the rollout has been rocky. Assessments have been so late that a private contractor was recently fined $200,000 by Arkansas state regulators. Patient advocacy groups note that denial rates have risen for people who had previously been eligible for Medicaid services.
Margaret Patterson, a public health nurse in Ramsey County, was an early supporter of MnChoices and even suggested some of the health-related queries on the assessment questionnaire. The idea of creating a single portal for all people, young and old, seeking in-home supports “made logical sense,” she said. And she still thinks the expansive survey was an effective way to gather data and track a person’s health condition and support needs over time.
But early on, Patterson said, it became apparent that MnChoices was too big and unwieldy to be effective. The system crashed so regularly that assessors had to develop shortcuts. Oftentimes, assessors took handwritten notes while sitting in people’s homes and then later entered the information into the MnChoices system.
“It was definitely not ready for prime time,” Patterson said.
Eventually, Patterson decided that the problems with MnChoices were beyond repair. After 40-plus years working as a public health nurse, she decided to retire in 2016 from her job as a supervisor of assessors in Ramsey County. “I could not, in good conscience, continue” working with the program, Patterson said in a recent interview.
Despite reports of dysfunction and the critical role that MnChoices plays for vulnerable people, the DHS has done little to ensure that the program achieves one of its primary goals, according to internal documents and interviews with county supervisors: to ensure consistency in reviewing applicants. State officials have never tested MnChoices for “inter-reliability” — whether different people using the same system will generate consistent results. Such reliability testing is critical for a system that touches tens of thousands of lives, technology analysts say.
A national consulting firm hired to evaluate MnChoices concluded that “such testing is imperative.” Abeler called the state’s failure to conduct reliability testing “startling and irresponsible.”
DHS officials said they are preparing a request for proposals to conduct inter-reliability testing.
Today, the problems with MnChoices are so persistent that some counties often revert to the legacy paper-based system, with its multitude of screening questionnaires and forms.
A Star Tribune review of state records found that counties are using paper screening tools for about a quarter of all assessments, even though they have had five years to adopt the new system. Last year in Hennepin County, nearly 40% of all assessments were completed with paper, while 33% of assessments in Ramsey County were paper-based, state records show.
County officials gave a simple reason for sticking with paper: Speed. The paper assessments are shorter and faster to complete, which has helped counties keep up with the staggering demand for Medicaid waivers — 110,000 assessments last year.
The widespread use of paper assessments, however, conflicts with official DHS policy, which says all initial assessments must be done with MnChoices.
DHS officials said use of MnChoices has expanded each year, with a majority of counties reporting in a recent survey that they used the program for all their assessments last year.
County officials and disability advocates fear this inconsistency will deepen the already wide disparities across Minnesota, where access to waivers is often based on where one lives rather than need.
“There is a high level of anxiety because families never know what they’re going to get from year to year,” said Barnett Rosenfield, a supervising attorney with the Minnesota Disability Law Center. “There is a desperate need for consistency in the Medicaid waiver system, and that can only come from the state.”
Reporting Chris Serres, Glenn Howatt
Photography Richard Tsong-Taatarii
Editing Dave Hage, Eric Wieffering, Deb Pastner, Catherine Preus
Graphics C.J. Sinner, Ray Grumney, Glenn Howatt, Anna Boone
Design Anna Boone, Josh Penrod, Jamie Hutt
Development Anna Boone, Jamie Hutt