The goal of this pragmatic cluster randomized clinical trial is to compare management of suspected infection in nursing home residents with dementia The main questions it aims to answer whether residents with dementia in nursing homes randomized to use a multicomponent intervention to optimize suspected infection management ( versus usual care) use less antibiotics and fewer burdensome interventions.

This 52-month study (8 months preparation; 34 months trial conduct; 10 months data analyses and manuscript preparation)is a parallel cluster ePCT of an intervention to improve infection management for suspected infections among residents with moderate to advanced dementia (N=600; N=300/arm) living in NHs (N=50; N=25/arm). Facilities are members of a provider managed care network, Iowa Health Care Quality Network ('Network'). The intervention will be similar to the original TRAIN-AD program, but adapted for moderate to advanced dementia and rolled out in the Network. At each intervention NH, the intervention will be implemented for 24-months. Control NHs will employ usual care. In all NHs, there will be a 2-month startup period, 12-month resident enrollment period, and 24-months data collection period (Figure 2). Eligible residents with dementia will be identified using the EHR and Minimum DataSet (MDS) and followed up to 12 months. Outcome data will be ascertained from the EHR. Randomization and program roll out will be at the facility level. Analyses at the resident level. Clinical outcomes compared between arms at 12 months will be: 1. Antimicrobial courses/person-year in residents with moderate to advanced dementia (primary outcome) and all residents with dementia (secondary outcome) (Aim 1); and 2. Number of burdensome procedures/person-year used to manage suspected infections among residents with moderate to advanced dementia and those at all stages of dementia including: hospital transfers, urine specimens, chest x-rays, and blood cultures (secondary outcomes) (Aim 2). A process evaluation of implementation will be conducted in the intervention arm based on the RE-AIM framework1 (Aim 3) using quantitative and qualitative data (stakeholder interviews).

Inclusion Criteria:

1. Age > 60
2. A diagnosis of dementia (any type)
3. Cognitive Functional Scale (CFS) > 1
4. NH length of stay >90 days

The CFS score categorizes cognitive impairment status based on data in the electronic health record into: 1. None, 2. Mild, 3. Moderate, and 4. Severe (advanced). For the primary outcome, the analysis will be restricted to residents with a CFS score of 3 or 4.

Exclusion Criteria:

1. Less than 60 years of age
2. Living in nursing home for less than 90 days
3. Does not have diagnosis of dementia
4. Does not meet CFS >1 score