(a) Identification. Prognostic assessment software of mild cognitive impairment is a prescription-only device that combines information from multiple clinical data sources to produce an output indicative of the likelihood that a patient diagnosed with mild cognitive impairment will progress to more severe forms of cognitive impairment, or will develop dementia, such as Alzheimer's disease dementia, within a specified period of time. This device is intended for adjunctive use and not intended as a stand-alone diagnostic or prognostic tool.
(b) Classification. Class II (special controls). The special controls for this device are:
(1) Data obtained from premarket clinical performance validation testing and postmarket surveillance acquired under anticipated conditions of use must demonstrate that the device performs as intended when used to analyze data from the intended patient population, unless FDA determines, based on the totality of the information provided for premarket review, that data from postmarket surveillance is not required.
(i) Data provided from paragraph (b)(1) of this section must demonstrate the accuracy, precision, and reproducibility of the device for predicting progression of patients with mild cognitive impairment to more severe forms of cognitive impairment or to dementia, such as Alzheimer's disease dementia, based on a clinically relevant reference standard for the diagnosis of such disease states or conditions. The test data set must include data acquired from a patient population that is representative of the intended patient population normally seen in clinical practice. The test data set must be independent from data used in training/development of the device.
(ii) Objective performance measures (e.g., sensitivity, specificity, positive predictive value, negative predictive value) must be reported with relevant descriptive or developmental performance measures.
(2) Software verification, validation, and hazard analysis must be provided. Software documentation must include a detailed, technical description of the model/algorithm(s) and algorithm inputs and outputs.
(3) Human factors/usability assessment must demonstrate that the intended user(s) in the intended use environment can correctly use the device and interpret the device output(s).
(4) Physician labeling must include:
(i) A detailed summary of the clinical performance testing methods, including results of the performance testing for tested performance measures/metrics, selection criteria, and the patient demographics;
(ii) A description of the patient population that was used in development or training of the device algorithm/model;
(iii) Device limitations or subpopulations for which the device may not perform as expected or for whom the device has not been validated;
(iv) A statement that the device is not a stand-alone diagnostic or prognostic tool and that the device output should only be interpreted in the context of all available clinical information, including patient history, medications, and comorbidities; and
(v) Information for interpretation of the device outputs detailing the risks associated with misinterpretation of the device outputs.
(5) Patient labeling must include information explaining the device output(s) and the associated risks if the device output is incorrect.
[91 FR 69692, Nov. 5, 2026]