When Nursing Homes Text Each Other, Three Very Different Communication Patterns Emerge — and One May Be Putting Residents at Risk

Columbia, Missouri — The way a nursing home’s care team communicates — not just how often, but who talks to whom — may have more to do with patient outcomes than most administrators realize.

A new study published in International Journal of Medical Informatics took a close look at how nursing home care teams use text messaging to coordinate care around hospital transfers, and the results reveal three starkly different communication patterns — each with its own implications for resident safety and team effectiveness.

Researchers from the University of Missouri applied social network analysis to more than 5,000 text messages linked to 585 nursing home-to-hospital resident transfers collected over five years. The goal was simple: map how communication actually flows, and what that structure says about how care gets delivered.

What they found was that nursing homes don’t communicate the same way at all. They fall into three distinct models.

Three Ways Nursing Home Teams Actually Communicate

The first model — the one researchers called “Integrated” — showed high message volume, broad participation across roles, and low hierarchy. In these facilities, information moved freely between team members, and text messaging had become a general collaboration tool rather than a niche workaround. These teams had what the researchers described as mature adoption, where no single person controlled the information flow.

The second model, “Hub-and-Spoke,” looked more like a traditional command structure. A central figure — often a charge nurse or senior clinician — acted as the communication switchboard, with most messages flowing through them. It’s a protocol-driven model that works, but one that concentrates risk in a single individual. If that person is off shift, communication can stall.

The third model, “Siloed,” was the most concerning. Despite generating high message volume, these facilities had low role diversity in their networks — nurses texting nurses, aides texting aides, but rarely across those lines. Information wasn’t getting to the people who needed it, even when plenty of messaging was happening.

Why This Matters for Hospitalizations

The study didn’t track clinical outcomes directly, but its implications point in one direction: communication structure shapes how quickly and accurately teams can respond to a resident who’s declining. When a facility operates under a siloed model, the right information might never reach the physician or nurse practitioner who could intervene before a hospital trip becomes unavoidable.

The researchers also noted that individuals who occupied “bridging” positions — those who connected otherwise separate clusters — functioned as informal opinion leaders, even without formal authority. These connectors, whether they knew it or not, were doing some of the most important work in the facility.

This finding builds on what some nursing home operators have already discovered: that informal communication networks often drive care quality more than organizational charts. As industry reports have noted, on-call coaching models that improved care coordination cut hospitalization rates significantly in facilities that invested in building those connections deliberately.

What Administrators Can Do With This

For nursing home leaders, the study offers a practical frame. Before investing in new technology or training, it’s worth asking what communication model your facility currently runs on — and whether that structure supports the kind of fast, cross-disciplinary information flow that prevents avoidable hospitalizations.

The researchers suggest social network analysis as a tool for evaluating those structures — a method that doesn’t require expensive software, just a willingness to look at who’s actually talking to whom, and who isn’t.

As text messaging becomes increasingly embedded in care team workflows, understanding the shape of those networks may be just as important as the content of the messages themselves.