Making home visits safer at Cityblock

Making home visits safer at Cityblock

Making home visits safer at Cityblock

Cityblock staff visit members at home, and some visits carry risk. Warnings about threats or weapons sat in free-text notes that staff could miss while preparing for a visit. After shadowing an EMT in the Bronx, I got the problem onto our roadmap and designed a way to log safety risks in Commons, our care management platform.

Role

Senior Product Designer

Senior Product Designer

Duration

4 weeks

4 weeks

Team

Product manager

1 Engineer

Care operations

Compliance

Product manager

1 Engineer

Care operations

Compliance

Skills
Skills

Field research

Product strategy

Information architecture

Usability testing

Field research

Product strategy

Information architecture

Usability testing

Results

Results

  • Rolled out to 200+ field staff across all six markets

  • Safety risks added for 80+ members at launch

  • Five visits changed in the first eight weeks after staff noticed risks

Rolled out to 200+ field staff across all six markets

Safety risks added for 80+ members at launch

Five visits changed in the first eight weeks after staff noticed risks

Warnings were written down in spots where they might be missed

Warnings were written down in spots where they might be missed

"Oh, this one... I might have to call them first. They get angry sometimes."

An EMT said this to me before a visit. He knew because he had been there before. He told me another EMT once saw a gun on a kitchen table during a visit and was told it was loaded.

I hadn't considered how dangerous these visits could be for Cityblock staff. It had been easier to think every member was grateful for our team's care. I asked him how he made sure others visiting the home would know.

He showed me how he wrote warnings into the Need to Know section or logged visit notes on the Timeline. But free text could not alert anyone. A visit could be booked without the mental health clinician when required, then cancelled at the door or carried out unsafely.

There was a separate system for logging risks, LogicGate. Staff used it for training once or twice a year, so it was often forgotten, and nothing logged there showed up in Commons.

Reflections
Reflections
Reflections

Safety was not on the roadmap

Safety was not on the roadmap

My team had been asked to focus on improving general scheduling metrics for the rest of the quarter. I met with my lead engineer to discuss what we could build, and worked through compliance questions with my design manager and product manager.

Then, I took my shadowing notes to product leadership. They agreed to prioritize this work, and gave our team four weeks.

To meet that deadline, we cut a two-way sync with LogicGate and kept the two things staff needed most: a way to log a risk as structured data, and a way to surface risks during visit prep.

Compliance's categories confused staff

Compliance's categories confused staff

Compliance sorted risks into three groups: to self, to others, and from the environment. In testing, staff could not decide where some risks belonged.

I rebuilt the list with staff in mind. Personal risks follow the member, so they include anything to be aware of whether we meet them inside or outside of their home. Environmental risks come from the home or neighborhood.

I also kept the full list visible instead of hiding it in a dropdown, so staff could see what could be logged as a risk.

Seeing a risk was not enough

Seeing a risk was not enough

I spoke with three EMTs, a community health worker, and a mental health clinician. Staff without clinical training told me they often did not know what to do when they saw a risk. They would message a manager and wait.

So I worked with care managers and Compliance to write guidance for each risk. It appears above the list once a risk is logged.

Where risks are logged and displayed

Where risks are logged and displayed

Most warnings were being typed into Need to Know, so I put the "Add safety risk" button right under that section to nudge staff toward this new flow.

Because members may see staff screens during their visit, the sidebar alert only shows a risk is logged, not what it is, and one click opens the details.

Any staff member can log a risk right away, but only managers can remove one. I decided it would be better for Commons to surface risks immediately, rather than waiting for a manager to approve them.

What happened after launch

What happened after launch

We searched old Need to Know notes for keywords and added risks for 80+ members. In later sprints, I built on it: assessment scores could log a risk, and scheduler guidance incorporated risk when booking.

Because we cut the sync, staff were asked to log risks in both Commons and LogicGate. Compliance could at least compare the two lists to see what was missing.

However, if I did this again, I would prioritize syncing between the two systems in the MVP. While adjusting the list's taxonomy was valuable, reducing the documentation burden was more valuable to tackle first.

There's more to this story…

This page just covers the highlights. For the nitty-gritty, like how I made the case to leadership, let's talk.

There's more to this story…

This page just covers the highlights. For the nitty-gritty, like how I made the case to leadership, let's find a time to talk.