Judge says $400 million algorithm unfairly denies Medicaid benefits to thousands
Thousands of Tennessee residents were illegally denied benefits, including Medicaid, due to data and programming errors in an algorithm system used by the state to determine eligibility for low-income residents and people with disabilities, as ruled by a District Court judge this week.
Can algorithms be trusted?
The TennCare Connect system – built by Deloitte and other contractors for over – was supposed to analyze income and health information to automatically determine eligibility for benefit programs. However, in practice, the system often fails to load the correct data, assigns beneficiaries to the wrong households, and determines eligibility incorrectly, according to Judge Waverly Crenshaw Jr. of the Middle District of Tennessee.
“When an applicant is entitled to Medicaid, administered by the state, luck, perseverance, or legal action should not be required for them to receive that health coverage,” wrote Crenshaw in his opinion. The decision resulted from a class action lawsuit filed in 2020, on behalf of 35 adults and children who were denied benefits.
Automatically denied
The Tenn Care Connect system was launched in 2019 as a result of several years of state efforts to modernize its Medicaid system and adhere to new eligibility criteria and legal requirements. Under the, states must provide a single application process that collects residents’ information and determines their eligibility for health and disability benefit programs. Crenshaw found that applicants’ eligibility for all available programs was not considered before removing their coverage. Deloitte greatly benefited from the modernization effort nationwide, and won contracts to build automated eligibility systems in more than 20 states, including Tennessee and Texas. Now, the Federal Trade Commission to investigate Deloitte’s practices in Texas, where thousands of residents are also said to have been irregularly denied benefits due to the company’s faulty systems.
