The research, led by Matthew J. Mulligan and colleagues at the University of Utah’s Spencer Fox Eccles School of Medicine, addressed what has come to be known in health services circles as the inbox problem. Patient portal messaging exploded during the COVID-19 pandemic, when in-person visits were curtailed and patients turned to electronic communication for everything from routine questions to urgent concerns. Studies cited by the Utah team document the consequences: more time tethered to the EHR, longer workdays that extend into evenings and weekends, and measurable links between after-hours inbox work and clinician exhaustion. Prior analyses have associated high EHR time with degraded quality-of-care metrics, and surveys of physicians consistently rank in-basket management among the most demoralizing parts of modern practice.

The logic of the intervention was straightforward. When a patient sends a medical advice request through the portal, that message does not have to be answered by the physician alone. Medical assistants can handle many routine requests using standing protocols, such as confirming that a stable patient’s medication refill matches an existing care plan, or relaying normal lab results with prepared explanations. Nurses can triage symptom-related messages using established decision rules, escalating only those that genuinely require physician judgment. The University of Utah team embedded standardized protocols directly into the EHR, integrating them into the existing workflows of primary care clinical staff rather than creating parallel systems. Because the initiative used existing personnel and existing technology, it required no new funding streams, a design choice the investigators describe as cost-neutral.

To test whether this approach actually moved the needle, the researchers conducted a quasi-experimental pre-intervention and post-intervention study across eleven primary care clinics within a large academic health system. The primary outcome was the percentage of patient medical advice requests, abbreviated PMARs, that were ultimately received by a physician. This metric captures the fraction of message traffic that escaped the care team’s front line and reached the doctor’s inbox. The researchers used multivariable mixed-effect regression analysis, a statistical approach that controls for potential confounding factors and accounts for the clustering of messages within individual physicians, which is essential when some clinicians handle far higher volumes than others.

The results showed a statistically significant shift. Before the intervention, 61.6 percent of PMARs were received by a physician. After the standardized protocols were integrated into staff workflows, that figure fell to 57.6 percent, a difference with a p-value below 0.001, indicating that the change was very unlikely to be due to random variation. In the fully adjusted multivariable mixed-effect model, the odds that a physician received a given PMAR after the intervention were lower than before, with an odds ratio of 0.83 and a 95 percent confidence interval of 0.82 to 0.84. Translated into practical terms, the intervention corresponded to a 4.6 percent reduction in the adjusted mean percentage of patient messages reaching physicians.

Statisticians will note that an odds ratio of 0.83 describes a modest effect, and the authors are candid about this. A four-percentage-point drop in the share of messages reaching physicians will not, by itself, transform the working lives of primary care doctors drowning in portal traffic. Yet the study matters for several reasons. First, it demonstrates that a low-cost, workflow-embedded intervention can produce a measurable, system-wide change across a large network of clinics, not just a pilot ward or a single motivated practice. Second, it provides quantitative evidence for a strategy that national recommendations have long championed on theoretical grounds: leveraging the full care team rather than concentrating message work on the most expensive and most constrained resource in the clinic, the physician.

The broader context makes the modest effect size easier to interpret. The volume of electronic inbox messages has been rising for years, driven by portal adoption, patient expectations of rapid electronic communication, and algorithmically generated messages such as automated result notifications and care-gap reminders. Research published in Health Affairs in 2019 linked physicians’ well-being specifically to in-basket messages generated by algorithms, and studies of EHR use trends among academic primary care physicians from 2019 to 2023 found clinicians increasingly tethered to the record. Against this rising tide, even holding the line requires effort. A 4.6 percent relative reduction in physician-received messages represents swimming against a current that pushes in the opposite direction every year.

The Utah study also contributes to a growing literature on team-based inbox management. Previous work has explored using medical assistants to manage portal messages, nurse triage models for patient messages, dedicated in-basket managers sometimes dubbed inboxologists, and simple tools to reduce overall portal message volume. What distinguishes the new study is its systematic, multi-clinic evaluation with a clear primary outcome and rigorous adjustment for confounders. By embedding the protocols within the EHR itself, the team ensured that medical assistants and nurses encountered the decision support at the moment a message arrived, rather than relying on memory or separate reference documents. This design detail matters, because workflow integration is repeatedly identified as the difference between interventions that stick and those that fade after the initial enthusiasm.

Still, the headline finding cuts both ways. After the intervention, physicians received 57.6 percent of all patient medical advice requests, meaning the majority of electronic patient communication still demanded physician attention. The authors conclude that integrating standardized protocols into primary care staff workflows was associated with a modest reduction in physician workload, while emphasizing that PCPs continued to receive more than half of all patient messages. This residual burden has implications for clinician well-being and retention, both of which the study identifies as casualties of the inbox problem. Burnout research has repeatedly connected inbox volume and after-hours EHR time to exhaustion among primary care clinicians, and workforce analyses warn that the current design of primary care may be structurally unsustainable.

For health systems searching for actionable steps, the study offers a template. Standardized protocols can be written, embedded in the EHR, taught to medical assistants and nurses, and applied to the large category of routine patient messages that do not require a physician’s diagnostic reasoning. The intervention required only a small internal grant of ten thousand dollars from the University of Utah’s Division of General Internal Medicine, underscoring that the barrier is organizational will rather than capital. The remaining challenge, the data suggest, is larger: redesigning primary care so that the physician’s inbox contains only what genuinely requires a physician, while the rest of the care team handles the rest, safely and consistently. Until that redesign matures, the inbox problem will persist, but the Utah results show that even incremental, protocol-driven changes can shift the balance of work back toward the team.

Subject of Research: Impact of standardized care-team protocols on primary care physicians' electronic health record inbox workload

Article Title: The Inbox Problem: Standardized Protocols for Primary Care Teams to Address Patient Messages

Article References: Mulligan, M. J., Conroy, M. B., Codden, R. R., Wu, C., Flynn, M., Loucao, S., McCormack, E., & Johnson, S. A. (2026). The Inbox Problem: Standardized Protocols for Primary Care Teams to Address Patient Messages. Journal of General Internal Medicine. https://doi.org/10.1007/s11606-026-10643-9

Image Credits: AI Generated

DOI: 10.1007/s11606-026-10643-9

Keywords: primary care, EHR inbox, patient portal messaging, team-based care, standardized protocols, medical assistants, nurse triage, physician burnout, quality improvement, electronic health records, clinician workload, patient messages

A new perspective published in the Journal of General Internal Medicine by Farhad Mehrtash and JoAnn E. Manson of Harvard Medical School and Mass General Brigham proposes a practical way out of this impasse. Their answer is a team-based implementation of the 5A’s framework—Assess, Advise, Agree, Assist, Arrange—endorsed by the United States Preventive Services Task Force for behavior counseling. The core insight is a division of labor: clinicians handle only the brief, high-value steps that require their expertise, while the rest of the care team absorbs the time-intensive work of counseling, resource navigation, and follow-up. The model is designed to fit inside the constraints of real-world primary care rather than demanding an idealized clinic that does not exist.

The mechanics of the workflow begin before the clinician ever enters the exam room. Team members measure vital signs and patients complete lifestyle screening questionnaires, ideally through the patient portal at home or on a tablet at check-in. The screening covers six behaviors: tobacco use, body mass index, physical activity beyond regular work, dietary pattern, alcohol consumption, and sleep duration and quality of at least seven hours per night. Patients also complete the Rapid Eating Assessment for Participants, Shortened Version 2, a validated dietary screener recommended by the American Heart Association that also captures alcohol and physical activity. Crucially, patients also report social barriers such as food insecurity or lack of transportation, which can derail even the best-designed lifestyle plans. The team then summarizes everything for the clinician, who reviews the results to complete the Assess step.

What follows is deliberately compressed. During the Advise step, the clinician explains how the screening results relate to health risks—for example, telling a patient who smokes that their habit increases the risk of cancer, heart disease, and stroke. During the Agree step, clinician and patient use shared decision-making to select one achievable behavior change, such as taking 15-minute walks or cutting back two cigarettes per day. The entire clinician interaction is designed to take under five minutes. For patients facing social barriers, the team first prioritizes connecting them to existing resources such as the Supplemental Nutrition Assistance Program, local food banks, or Medicaid non-emergency medical transportation before targeting lifestyle change itself, an acknowledgment that advice about diet means little to someone who cannot reliably obtain food.

The Assist and Arrange steps belong to the team. Medical assistants, nurses, health coaches, or community health workers help patients translate goals into specific daily actions and track progress using tools matched to each behavior and to the patient’s digital literacy. Options range from paper-based food journals, weigh-ins, trigger logs, and sleep diaries to apps such as MyFitnessPal, step trackers, and Sleepio. When patients need structured programs beyond brief counseling, teams connect smokers to quit lines via electronic referral and enroll patients with prediabetes in the Diabetes Prevention Program. Clinicians refer patients with diabetes or dyslipidemia, or those needing weight, dietary, or alcohol support, to registered dietitians for individualized medical nutrition therapy, and refer persistent sleep problems to sleep medicine or behavioral sleep specialists for cognitive behavioral therapy for insomnia. In the Arrange step, teams schedule follow-up contacts by phone, portal, or visit to review progress, adjust goals, and track referral completion through the electronic health record.

Concerns that delegation would add to clinicians’ administrative burden appear to be unfounded. In a national study of more than 18,000 physicians, shifting documentation to the team lowered clinicians’ total electronic health record time rather than adding to it. The authors also lay out the financial plumbing that makes the model sustainable. Practices can bill for preventive counseling under codes 99401 through 99404, smoking cessation under 99406 and 99407, medical nutrition therapy under 97802 through 97804, obesity counseling under G0447, and physical activity and nutrition assessment under G0136. When social barriers impede treatment, Community Health Integration services under code G0019 reimburse team members for up to 60 minutes per month of resource navigation per patient. Coverage varies and payment for preventive counseling remains low, but value-based and capitated payment models, which reward preventing downstream chronic disease rather than paying only for counseling visits, make the framework more financially viable. Embedding referral and follow-up tracking in the electronic health record also helps programs persist after temporary grant funding ends, and lifestyle medicine assessment tools such as the American College of Lifestyle Medicine Short Assessment Form are now being integrated into platforms like Epic.

The evidence that each targeted intervention works is substantial. Delivering the Assist step for tobacco cessation increased the odds of quitting by 40 percent, and the Arrange step by 46 percent. Brief alcohol counseling in adults screening positive for unhealthy use reduced consumption by roughly 1.6 drinks per week and lowered the odds of exceeding recommended limits by 40 percent at six to twelve months. Medical nutrition therapy reduced total cholesterol by 4.64 to 20.84 milligrams per deciliter, LDL cholesterol by 1.55 to 11.56, and triglycerides by 15.9 to 32.55. The Diabetes Prevention Program’s lifestyle intervention, targeting seven percent weight loss, reduced diabetes incidence by 58 percent in adults with prediabetes over three years, an effect that persisted as a 27 percent reduction after fifteen years. Even small doses of activity matter: adding just five minutes of daily moderate-to-vigorous physical activity may prevent six to ten percent of all deaths, with the greatest benefit—roughly thirty percent lower mortality—among the most sedentary. And in a randomized trial of 642 adults across 35 primary care practices, four sessions of nurse-delivered sleep restriction therapy significantly decreased insomnia severity at six months.

The gap between this evidence and everyday practice is enormous. While 65 percent of smokers receive cessation advice, only 13.4 percent of visits include dietary counseling and just 5.3 percent include physical activity counseling. Only a quarter of patients with diabetes receive any lifestyle counseling, and fewer than 5 percent of adults with prediabetes are referred to the Diabetes Prevention Program. The authors attribute these failures to payment models that reward procedures and hospital-based care over prevention, combined with practices lacking the time, staffing, and payment support to implement evidence-based programs. Yet team-based referral systems demonstrably improve enrollment: a St. Louis program using community health workers enrolled 30 percent of referred patients in the Diabetes Prevention Program and a blood pressure self-monitoring program, and expanding medical assistants’ roles within team-based workflows raised smoking cessation counseling rates from 45 to 89 percent, sustained beyond one year.

A common objection is that most practices lack the staffing such models assume. Only 11.4 percent of practices reach the recommended two-to-one medical-assistant-to-clinician ratio, and more than half employ just one medical assistant per clinician. The framework’s answer is flexibility: because it does not depend on any specific role, it adapts to whatever staff a practice already has, whether nurses, health coaches, or community health workers. At minimum, a practice with a single medical assistant can deliver all five steps by splitting the workflow across the visit—screening at intake by the assistant, brief Advise and Agree during the visit by the clinician, and Assist and Arrange at checkout by the assistant, who connects patients to quit lines or prevention programs. The model is not merely theoretical for underserved settings: in one randomized trial, medical assistants trained in self-management support and follow-up raised the proportion of safety-net patients meeting at least one clinical goal for diabetes, blood pressure, or cholesterol to 46.4 percent, compared with 34.3 percent under usual care.

The broader significance of the proposal lies in its refusal to treat prevention as an ideal that must wait for a perfect health system. By limiting the clinician’s role to brief assessment and goal setting while leveraging team-delivered counseling, structured referrals, and existing billing mechanisms, the 5A’s framework turns what has long been framed as an impossible time problem into a solvable workflow problem. With up to 35 percent of adults lacking adequate sleep and chronic disease driving the bulk of morbidity and mortality, the authors argue that systematically addressing modifiable behaviors within routine care is one of the highest-yield opportunities in medicine—and one that, with the right division of labor, primary care practices can begin implementing now.

Subject of Research: Team-based implementation of the 5A's framework for lifestyle behavior counseling and chronic disease prevention in primary care

Article Title: Implementing the 5A’s for Lifestyle Behavior Counseling and Chronic Disease Prevention

Article References: Mehrtash, F., & Manson, J. E. (2026). Implementing the 5A’s for Lifestyle Behavior Counseling and Chronic Disease Prevention. Journal of General Internal Medicine. https://doi.org/10.1007/s11606-026-10860-2

Image Credits: AI Generated

DOI: 10.1007/s11606-026-10860-2

Keywords: 5A's framework, lifestyle medicine, chronic disease prevention, primary care, behavioral counseling, team-based care, cardiovascular disease, type 2 diabetes, Diabetes Prevention Program, preventive medicine, health coaching, medical assistants