A Michigan Public Health Deep Dive: What State Public Health Leaders Are Asking About Data (and What Metopio Makes Possible)
If you've ever felt like your community health assessment was the end of the process rather than the beginning, you're not alone. The tension between producing a report and actually driving change was at the heart of a recent webinar Metopio hosted in partnership with the Michigan Association for Local Public Health.
During the session, Metopio CEO Will Snyder and VP of Data Heather Blonsky walked Michigan public health leaders through the full Metopio platform and dug into real Michigan data across poverty, mortality, and access to care. Here's what stood out.
AI in Public Health: Cautiously Optimistic
Drawing on surveys and focus groups conducted in partnership with NACCHO, Snyder began by describing the current landscape of AI use as one of cautious optimism. Public health teams are seeing real value in AI for workflow improvement (moving faster from question to insight, sorting through large datasets, drafting reports), but the field has learned some hard lessons about what happens when AI is expected to do more than it should.
The consensus from epidemiologists and program managers across the country is clear: AI should help surface and organize data, but interpretation belongs to the human expert. Drawing conclusions — especially on sensitive public health topics — isn't something AI should be doing on its own.
That feedback has driven Metopio's product decisions directly. PHI and sensitive data live in a completely separate environment, with zero AI access, period. When AI is used, it surfaces and presents data without interpreting it — using a method called text-to-SQL that converts plain language questions into database queries, so the AI is a translator, not a decision-maker. Every AI action is logged and auditable. And if AI isn't permitted in your jurisdiction, it can simply be turned off.
A Platform Built for the Full Workflow, Not Just the Assessment
One of the clearest themes of the session was the importance of continuity — making sure the work doesn't stop when the CHA does.
Metopio is built around that complete workflow: from data collection and automated CHA drafting, straight into CHIP planning and program management, through to real-time dashboards that let teams track whether their interventions are actually working. For teams pursuing PHAB accreditation, that seamless path from assessment to improvement plan is particularly valuable.
The platform's expansive program library gives health departments a starting point, while also allowing teams to design and build their own programs from scratch. Shared metrics across multiple programs mean you can track something like A1C reduction across five different interventions in a single combined view, with dashboards that update in real time as data comes in.
Michigan Data Visualized
The highlight of the session for many attendees was Blonsky’s live walkthrough of a Michigan-specific Health Atlas she built ahead of the webinar — and shared publicly so participants could explore on their own.
On poverty, she went beyond the standard federal poverty level to show why a single measure never tells the whole story. Severe housing cost burden — which captures where the cost of living has outpaced incomes — paints a different picture than flat poverty rates, tending to surface more urban pockets in cities like Detroit, Flint, and Lansing. Income share for the lowest quintile of earners adds yet another layer, showing not just that poverty exists but how concentrated it is relative to everyone else in the county.
On mortality, she walked through years of potential life lost, which captures premature deaths by tracking how many years people miss before age 75, broken down by cause. Cardiovascular disease topped the list across Michigan counties, followed by cancer and accidents, with COVID reshuffling the rankings in notable ways when filtered separately.
On access to care, two datasets stood out. Drive time to hospitals with obstetric care revealed stark rural gaps — some communities in the Upper Peninsula face two-hour drives to deliver a baby. And broadband download speeds showed meaningful gaps between the Upper and Lower Peninsulas, with pockets of poor connectivity even in urban areas, a direct obstacle to telehealth access for the communities that need it most.
Bring Your Own Data: Combining What You Collect with What Metopio Curates
Metopio closed the platform portion with a look at their “Bring Your Own Data” feature — the ability to bring primary or proprietary data collected by a health department directly into the platform and combine it with Metopio's curated dataset library.
One example showed sudden unexpected infant death data, analyzed in combination with neighborhood-level poverty and household crowding indicators. Rather than reporting small case counts by zip code, the BYOD approach allowed the team to characterize the neighborhoods most affected, identify where interventions should be targeted, and build a framework for measuring whether those interventions moved the needle over time. It's the kind of analysis that turns a data point into a plan.
What This Means for Michigan Public Health Teams
The through-line of this session was a simple but important idea: data should be the beginning of the work, not the end of it. Michigan public health leaders are navigating real resource constraints, evolving AI policies, and growing pressure to show that community health investments are paying off. The tools to do all of that — in one place, built for public health — exist now.