DPC Directory

Does direct primary care lower ER visits and costs? What the studies show

DPC Directory Editorial

The short answer: one well-built study found fewer emergency room visits among direct primary care (DPC) members, but it covered one employer and two practices, and it did not find that DPC saved that employer money. Nothing we found shows that DPC members have fewer hospital stays or better health. The larger savings figures quoted elsewhere come from reports that, in the cases we read, did not describe adjusting for members' health.

Here is each study we found, what it measured, how it was designed and who paid for it (where the source says), followed by what that means if you are deciding whether to join a practice.

The Society of Actuaries study (Milliman, 2020)

This is the most careful study of DPC and utilization we found. The Society of Actuaries commissioned the consulting firm Milliman to write Direct Primary Care: Evaluating a New Model of Delivery and Financing (May 2020; authors Fritz Busch, Dustin Grzeskowiak and Erik Huth). Its sponsor was the Society's research pool, and the report says its opinions are the authors' own. We read the full report.

The design. The authors studied one midsized, self-insured employer with about 1,000 employees and 1,000 dependents. The employer added a DPC option in July 2015 alongside a traditional preferred-provider (PPO) plan and paid the DPC membership fee for people who chose it. About half of the people covered chose DPC. The authors compared the two years after launch for 912 DPC members and 1,074 traditional-plan members, using claims data. They used Milliman's risk-scoring tool to adjust for age, sex and diagnoses. Without that adjustment the comparison would have been unfair: the DPC group scored 8.3 percent lower on expected claims, meaning it was healthier to start with. Milliman's summary of the study puts it this way: "Controlling for patient selection in DPC studies is imperative."

What it found:

  • Emergency room visits. Visits per 1,000 members were 36.39 percent lower for DPC members before risk adjustment and 40.51 percent lower after it. The report calls this statistically significant, with a 90 percent confidence interval of about 28 to 53 percent lower. This is notable because, based on diagnoses, the DPC group's expected ER claim costs were 5.6 percent higher than the traditional group's.
  • Total claims. Allowed claims per member per month were 19.90 percent lower before adjustment and 12.64 percent lower after it, so about a third of the raw gap came from the DPC group being healthier. The remaining difference was still statistically significant. The authors treat claims as a stand-in for how much care people used, with the DPC practices' services priced the same way as everyone else's.
  • Hospital stays. The DPC group had 40 admissions and the traditional group 63. The risk-adjusted difference was 19.90 percent lower, but the confidence interval ran from 53.83 percent lower to 14.03 percent higher. The authors conclude there was "insufficient data to determine whether the DPC option affected the inpatient hospital admission rate." (The 19.90 percent figure appears twice in the report, for two different measures. It is easy to mix up.)
  • The employer's total cost. After adding the DPC membership fee, the authors estimated the DPC option raised the employer's non-administrative plan costs by 1.3 percent. Their range ran from a 7.8 percent increase to a 5.2 percent saving. The employer did not share its actual membership fees, so the authors assumed fees from their own market survey (the range reflects statistical uncertainty in the claims difference, not uncertainty about the fee), and they left out administrative costs. Milliman's summary also notes the DPC option waived the medical deductible, so members would be expected to use somewhat more care from plan design alone; the authors did not adjust for that, which they say would have made the reduction in demand slightly larger.

What the authors say it cannot show, and one limit of our own:

  • It covers one employer and two DPC practices over two years; the authors expect results to vary by practice and employer.
  • The data could not assess quality of care, health outcomes or patient satisfaction.
  • Members who pay the fee themselves may differ from members whose employer pays it, so results may not carry over to people who buy DPC on their own.
  • Our own note, not the authors': members chose DPC, so risk adjustment on claims data can reduce the effect of selection but may not remove all of it.

Union County, North Carolina

Union County, NC added a DPC option from the provider Paladina Health in April 2015. The best-known account is a John Locke Foundation report ("Direct Primary Care for Local Governments," Spotlight #483, by Katherine Restrepo and Julie Tisdale). The Foundation is a North Carolina policy think tank; the report does not say who funded it, and it is advocacy-style: its subtitle is "Helping Union County Save $1.28 Million in Health Care Claims."

What it measured. For 2015–2016, it compared the average monthly medical and prescription claims per employee for staff who used only the county's high-deductible-style consumer-driven plan against staff who chose DPC (44 percent of employees). It reports that DPC participants incurred 23 percent less in medical expenses and 36 percent less in prescription expenses, an average difference of more than $260 per employee per month, or $1.28 million.

What it did not do. It reports no emergency room or hospital figures. It does not describe any adjustment for health status, so people who chose DPC may simply have been cheaper to cover. The report says the county reached budget neutrality by redirecting $750 it had been putting into a health reimbursement account to pay for DPC memberships. The report compares claims; it does not set the cost of the memberships against them. Its reported health improvements ("by a nearly 3 to 1 margin") are self-reports.

Employer reports from DPC companies

In October 2026, Medical Economics reported results released by SALTA Direct Primary Care, a membership practice group in southeast Michigan, and the benefits administrator Ameriflex: ED visits down 31 percent, inpatient admissions down 71 percent, and cost growth of 0.9 percent a year against an expected 6.2 percent. The article says the release does not name an independent analyst, the number of employers or employees, the years covered, or a comparison method. We treat reports like this as claims from the seller, not findings. They are the kind of result the Society of Actuaries authors warned about when they said selection has to be controlled.

Peer-reviewed research

We did not find a peer-reviewed study of DPC that compares matched groups on emergency room use, hospital stays and total cost. Our search was not exhaustive, and we would add one if we found it. A 2019 review by the University of Wisconsin's Health Policy Programs Group, written before the Society of Actuaries study, described the published literature on DPC as mostly descriptive and survey information that "generally lacks rigorous studies on cost, quality, and outcomes." It also noted that a published study of MDVIP (which the Society of Actuaries report classes as concierge care), reported large savings from lower hospital use but did not adjust for members' baseline health, so it "does not allow conclusion about the impact of the delivery model." Concierge medicine and DPC are not the same thing; see our DPC vs concierge FAQ.

What holds up and what doesn't

Claim Where it stands
DPC members visit the ER less Supported in one risk-adjusted employer study. Not shown for people who buy DPC themselves.
DPC members have fewer hospital stays Not shown. The one adjusted study had too few admissions to tell.
DPC lowers total claims Supported in the same study (about 13 percent lower after adjustment), before the membership fee.
DPC saves the employer money Not shown. The same study estimated a 1.3 percent increase, with a wide range either way.
DPC improves health or quality Not studied in anything we read.

Two patterns explain most of the gap between the headlines and the evidence. First, utilization is not the same as cost: fewer ER visits and lower claims can coexist with higher total spending once the membership fee is added. Second, employer plans are not individual memberships: every study above looked at people whose employer chose, arranged or paid for DPC, and the Society of Actuaries authors say people paying their own way may differ.

What this means if you are choosing a practice

Does direct primary care reduce ER visits for you? The research cannot say. It cannot tell you that DPC will lower your bills or keep you out of the hospital either. What it supports is narrower: in one employer group, DPC members used the ER noticeably less than comparable colleagues did, which fits what DPC practices say they offer, more access to a clinician who knows you. The Union County report also describes same-day appointments and longer visits, but those come from the report, not from measured data.

So treat the decision as one about access, time and relationship, which you can check yourself, instead of a bet on proven savings:

  • Ask how you reach the doctor after hours, and how same-day visits work. Our FAQ on same-day urgent visits explains what to expect.
  • Ask what the membership does not cover. DPC does not replace insurance for an ER visit, a hospital stay or specialist care; see what happens when you need the ER or hospital.
  • Compare the monthly fee with what you spend on primary care now. Cost is a personal calculation, not a study result.
  • If an employer offers a DPC plan, read what to check before you enroll.

For the system-wide side of the question, see the panel-size critique of DPC.

Browse DPC practices by state and compare what each one publishes.

This is general information about health research, not medical, legal or financial advice; talk to your doctor about your care and a financial or benefits adviser about costs. The Society of Actuaries report, the John Locke Foundation report and the University of Wisconsin brief were read in full; the Medical Economics article was read, but we did not see the SALTA release itself. Sources were read on November 15, 2026.

Questions

Does direct primary care reduce emergency room visits?

One study found so. A 2020 Society of Actuaries report by Milliman compared about 900 DPC members with about 1,100 members of a traditional plan at one employer and found 40.51 percent fewer ER visits after adjusting for health status. It covered one employer and two practices over two years. We did not find a matched peer-reviewed study showing the same for people who buy DPC on their own.

Does direct primary care lower total health care costs?

Not shown. The same Society of Actuaries study found claims about 12.64 percent lower after risk adjustment, but once the DPC membership fee was added it estimated the employer's total plan cost rose 1.3 percent, within a range from a 7.8 percent increase to a 5.2 percent saving. Larger savings figures come from reports, such as a think-tank report on Union County, NC, that did not describe adjusting for members' health.

Does direct primary care mean fewer hospital stays?

We could not find evidence either way. In the Society of Actuaries study, DPC members had 40 hospital admissions and the traditional group 63, and the risk-adjusted difference was not statistically significant: the authors said there was insufficient data to tell. A company-reported figure of 71 percent fewer admissions came from a release that did not describe its comparison method.

Has any study shown that direct primary care improves health outcomes?

Not in anything we found. The Society of Actuaries authors said their data could not assess quality of care, health status or patient satisfaction, and a 2019 University of Wisconsin review said the DPC literature generally lacks rigorous studies of cost, quality and outcomes. DPC's access, visit time and relationship with your doctor are things you can check with a practice yourself.