Sentara Health deploys AI case management tool co-developed with Regard

Not-for-profit Sentara Health has deployed an AI case management tool it co-developed with clinical decision support company Regard, Fierce Healthcare reports. Case managers can query charts in plain language to surface patient history, status changes and other details, and can view what changed in the last 24 hours. Sentara, which operates 12 hospitals and more than 400 care sites in Virginia and Northeast North Carolina, plans to expand the tool across acute care and into ambulatory care, per the same report.

The economic case is large on paper. Sentara executives say the tool could save nearly 90,000 hours, which they equate to more than 40 full-time employees and nearly $5 million in labor costs, according to Fierce Healthcare. For nursing and operations leaders, the target matters as much as the total. Chief health information officer Rollin Reeder, M.D., described how much time teams used to spend piecing together a patient's story before discharge planning could begin. That reconstruction work sits upstream of throughput, length of stay and the daily workload that drives case manager burnout.

The evidence is thinner than the headline. The 90,000-hour figure is a projection from executives, and no independent measurement has been published. Regard's CEO cites more than 20 million physician-accepted diagnoses, a company statement about its clinical product rather than a measured result for case management. The useful signal is directional: a large system is moving decision support from physicians to operational teams, and it is doing so alongside other AI work, including a pilot of GW RhythmX's precision care platform announced in mid-September.

The move for CNOs and COOs is to build your own baseline before any vendor conversation. This week, measure how many hours your case managers spend on chart review and documentation, and how long it takes from admission to the start of discharge planning. Those two numbers let you judge any tool on returned time rather than projected time, and let you decide where recovered hours go: more patient and family contact, safer discharges, or relief for a stretched team. Because these tools read across the full chart, include your security and compliance leads early on access controls and audit logging, so wider visibility for case managers does not become wider exposure for patients.

You are an operations advisor to the COO and CNO of [your organization], a health system with [number of case managers] case managers working in [your EHR].

Context: Fierce Healthcare reports that Sentara Health has deployed an AI case management tool it co-developed with Regard, a clinical decision support company. Case managers can query charts in plain language to surface patient history and status changes, and can view what changed in the last 24 hours. Sentara plans to expand the tool across acute care and into ambulatory care. Sentara executives say the tool could save nearly 90,000 hours. That figure is a projection from executives, and no independent measurement has been published.

Task: Design a one-week baseline study we can start this week. It should measure how many hours our case managers spend on chart review and documentation, and how long it takes from admission to the start of discharge planning. The goal is to judge any AI tool on returned time, not projected time.

Output: A draft workflow for the executive team with these parts: steps for the week, an owner for each step, how time is captured without adding burden, checks for data quality, how results feed a decision on where recovered hours would go, and when security and compliance leads ([names of your security and compliance leads]) review access controls and audit logging. Use [the units included in the study] and compare against [your current estimate of weekly chart review hours per case manager], if known.

Guardrails: Use only the facts above and the slots I fill. Cite Fierce Healthcare by name for any Sentara detail. Flag anything uncertain or that needs verifying, including the projected 90,000 hours. Do not request or include patient data; measure time and workflow only. This is a draft for our team to check, and nothing here is medical advice.