Length of Stay False Diagnosis Cut Off Insurer's Payment for Treatment of Elderly Woman

Health insurance companies, particularly through Medicare Advantage plans, have deployed AI algorithms to predict when they can cut off payment for elderly patients' medical care. The investigation focused on NaviHealth's nH Predict algorithm, which uses patient data to generate precise predictions about recovery timelines and discharge dates. The system was used by major insurers including UnitedHealthcare, Humana, and Blue Cross Blue Shield plans to manage over 31 million Medicare Advantage beneficiaries. Specific cases included Frances Walter, an 85-year-old with a shattered shoulder whose algorithm predicted 16.6 days of care, leading to payment denial on day 17 despite her inability to dress herself or walk. Another case involved Dolores Millam, whose algorithm predicted 15 days of nursing home care after leg surgery, with payment terminated despite her inability to bear weight and requiring 24-hour assistance. The denials forced patients to spend life savings or go without care, with appeals processes lasting up to 2.5 years. Federal inspectors found insurers were using internal criteria beyond Medicare's rules, and appeals showed most denials were eventually overturned, indicating systematic inappropriate use of AI predictions.