What's Well & Good in Policy

Transcript 44 lines

Cold Open Stats Overview Paper Walkthrough free_promo

Cold Open

Jenny When a government or workplace announces a new wellbeing program, what makes you believe it will actually change anything?

Davis I wanna see a loop, not a slogan, like who checks if it worked next week and what they do when it didn’t.

Jenny Yeah, because “we care” is free, but measurement costs time, and if nobody’s tracking outcomes, it’s basically vibes.

Davis And coordination matters too, because if the nurse, the benefits office, and the community group all act like separate islands, the person in the middle does all the work.

Jenny Exactly, and this week we’ve got a perfect example of good intentions without follow-through, so… welcome to Wellbeing Policy on paperboy.fm.

Stats Overview

Davis Quick map of the week: we screened about fourteen sixty papers and qualified one hundred twenty-four for the show. That’s work from about four hundred twenty unique authors across fifty-one countries, so the net’s wide even when the pile’s smaller.

Jenny Smaller is right—qualified papers fell from one forty-seven to one twenty-four, down twenty-three, about sixteen percent. Are we seeing fewer strong policy evaluations, or did the mix tilt toward qualitative work—thirty-four qualitative versus twenty-eight surveys—which can be rich but harder to compare head-to-head?

Davis And the top-of-funnel dropped too: total hits went from one six five three to one four six five, down one eighty-eight, about eleven percent. When the feed thins like that, I expect fewer COVID-19 leftovers and more steady-state stuff like mental health and public health, which are our top two themes at twelve and eight papers.

Jenny But here’s the twist: unique countries jumped from forty to fifty-one, up eleven, about twenty-eight percent. So we’re seeing more geographic spread even as total volume falls—are we pulling in more regional journals, or are policy analysis and sustainability papers just coming from more places this week?

Davis Author-wise, it’s a lot of new blood: about one hundred first-time authors—meaning their first-ever paper in the metadata—plus about two hundred emerging researchers, and one twenty-five experienced. That lines up with the through-line for this episode: systems that measure and follow through beat big announcements, and early-career teams often publish the on-the-ground surveys and case studies that show whether coordination actually happened.

Paper Walkthrough

Paper 1 Patterns and predictors of outcome monitoring amongst link workers: Learnings from the National Social Prescribing Link Worker Survey 2025

Jenny Alright, let’s get into the papers. Paper one is called “Patterns and predictors of outcome monitoring amongst link workers: Learnings from the National Social Prescribing Link Worker Survey 2025,” and it’s basically asking a simple thing: when link workers send you to community support, are they actually tracking whether you got better?

Jenny They surveyed four hundred nine Social Prescribing Link Workers in England who had active caseloads, and only fifty-seven point five percent said they record patient outcomes “often” or “very often.”

Jenny And it’s even lower for service-level outcomes—forty-five point five percent—so the system can’t reliably answer “is this service working,” even though national guidance says outcomes should be reported via the Social Prescribing Information Standard, which is just the shared rulebook for what data to collect and how.

Davis If training and seniority don’t predict reporting, what actually changes behavior day to day—like, what makes a link worker bother to measure outcomes when they’re slammed?

Jenny They ran descriptive stats and then logistic regression—so, a model that estimates how much a factor changes the odds of reporting—adjusting step by step for other variables, and education, training, job seniority, and most demographics didn’t move the needle.

Jenny But awareness mattered: people who knew about the Social Prescribing Information Standard were about two point three times as likely to report outcomes, with an odds ratio of two point two nine and a ninety-five percent confidence interval from one point five three to three point four two, and people who shared outcomes with managers or stakeholders were over six times as likely, odds ratio six point zero three, confidence interval two point nine three to twelve point four three.

Jenny One catch though: it’s self-reported behavior from one workforce group in England, so it can’t prove what causes better monitoring, it can only show what travels together.

Davis That “sharing outcomes” result feels like the whole episode’s theme—implementation beats intention—because the credential stuff doesn’t matter if nobody asks for the numbers and uses them.

Davis Also, it’s wild that ninety-one percent say they capture qualitative notes, but only forty-eight point seven percent use a quantitative measure—meaning a standardized score you can compare across people—and even then most are using ONS4, which is that four-question life satisfaction set.

Davis So my practical takeaway is boring but real: if you run social prescribing, you don’t start with another training day, you start by making the standard unavoidable and making the data flow upward, and this paper’s decent evidence—moderate, not definitive—says that’s where the leverage probably is.

Paper 2 US State Policy Index for Population Health Analyses

Davis That bit about “making the standard unavoidable” sticks with me, because this next paper is basically a standard for policy context.

Davis It’s called US State Policy Index for Population Health Analyses, and Montez, Gutin, and Monnat build a simple zero-to-one score for every state, every year, from nineteen eighty through twenty twenty-three.

Davis Plain version: states that stack more policies that help people stay financially stable and avoid harm tend to have fewer people dying in midlife.

Davis Their index runs from zero to one on a conservative-to-liberal continuum, and it’s built from eleven specific policies meant to strengthen economic security, expand safety nets, and discourage risky behaviors.

Jenny When you say it “predicts” working-age mortality, are we seeing policy effects, or just that states differ in a hundred other ways—industry, demographics, health care access—and the index is basically a vibe check?

Davis They try to show it’s not just a vibe by validating the tool first, then checking outcomes.

Davis On validation, their index correlates about zero point eight nine with two existing broad state policy indices, which says it’s tracking the same underlying policy orientation but with a cleaner, replicable recipe.

Davis Then they link higher scores over nineteen eighty to twenty twenty-three to lower working-age mortality rates, but the big limitation is baked in: it’s state-level correlation, so you can’t, from this alone, claim the policies caused the mortality changes.

Jenny I still like it as a tool, because it turns “policy environment” into a number you can actually update every year instead of cherry-picking one law.

Jenny And it fits our thread that wellbeing is system-shaped—just, like, with the warning label that this is strong pattern-finding, not a magic lever you pull and watch deaths drop next quarter.

Paper 3 Primary Health Care Services and Continuity of Care Are Associated With Better Health Outcomes in the Older Population

Jenny You just said “state-level correlation,” and I keep thinking about what that looks like at the person level when the system actually follows through.

Jenny So here’s a 2026 Journal of the American Geriatrics Society paper called Primary Health Care Services and Continuity of Care Are Associated With Better Health Outcomes in the Older Population, looking at older adults on long-term care in the community from twenty sixteen to twenty nineteen.

Jenny Plain version first: older people did better when they kept seeing the same primary care doctor and when their visits skewed preventive, not just lots of visits.

Jenny They had one hundred twenty thousand five hundred twenty-two people age sixty-five plus, and compared folks seeing a known primary care physician versus a new one.

Jenny Seeing a known doctor was linked to fewer hospitalizations across nine outcomes, like fractures with a subdistribution hazard ratio of zero point seven two, meaning about a twenty-eight percent lower risk, with a ninety-five percent confidence interval from zero point six six to zero point seven eight.

Jenny And the “preventive-heavy” care pattern, about thirty-four thousand people, had lower pressure-injury hospitalizations than the “high overall use” pattern, with an sHR of zero point six four, confidence interval zero point five two to zero point eight zero.

Davis Okay, but how did they decide “known” versus “new” doctor, and how sure are we this isn’t just healthier, more organized patients being the ones who can keep continuity?

Jenny They ran it as a retrospective cohort, so they’re looking back at administrative care patterns, then using propensity scores to balance groups on measured differences, and survival analysis to track time to hospitalization or death.

Jenny “Known” versus “new” is basically continuity of primary care, meaning your visits concentrate with the same physician rather than bouncing around, and the care patterns split into things like high preventive care versus high total volume.

Jenny But you’re right on the core limitation: because it’s retrospective, they can’t fully rule out unmeasured stuff, like family support or baseline frailty, that both helps you keep the same doctor and lowers your hospitalization risk.

Davis Still, the practical read is kind of blunt: don’t design long-term care so people are constantly re-introducing themselves to a new GP, because that churn shows up as more fractures and pressure-injury admissions.

Davis And I like that it fits our “wellbeing is system-shaped” thread, because “continuity” isn’t a vibe, it’s scheduling rules, staffing stability, and whether the system rewards preventive visits instead of just counting visits.

Davis It’s big and pretty careful for an observational study, but it’s not a randomized trial, so I’d treat it as a strong signal for how to organize care, not a guarantee that switching doctors will magically change outcomes next month.