Opinion · Duty hours # The ACGME says its work-hour rules are too hard to schedule. That isn’t a reason to cut them. The ACGME’s proposal leans on one argument: today’s rules make resident schedules too hard to build and to monitor. We build those schedules for a living. That problem is solvable, and solving it is the programs’ job, not something residents should pay for in rest. Jackson Cantrell · Updated October 12, 2026 · 9 min read **Short answer** The ACGME has proposed cutting rest after 24-hour call from 14 to 12 hours and removing the every-third-night call cap, the eight hours between shifts, night float limits and specialty limits. Its stated reason, again and again, is the burden of building and monitoring schedules. That burden is real, but it is a tooling problem: software can build a schedule inside every one of these rules and check it instantly. Burden is an administrative reason, not a safety one, and it shouldn’t be paid for with residents’ rest. Comments close October 22, 2026; the changes would take effect July 1, 2028. ## What the ACGME proposed On September 8, 2026 the Accreditation Council for Graduate Medical Education (ACGME), which accredits every US residency program, posted a proposed rewrite of its Common Program Requirements, the rules every residency must meet. It says the package cuts the number of requirements by 43 percent. The work-hour section changes most. ### Resident work-hour rules, today and as proposed From the ACGME’s draft requirements and rationale statement, September 2026. | Rule | Today (2026) | Proposed (July 2028) | |---|---|---| | **80 hours a week** | Averaged over up to four weeks | Kept; the average must also fall within a single rotation | | **Rest after 24 hours of in-house call** | At least 14 hours | At least 12 hours | | **Time off between shifts** | Should have 8 hours (not citable) | Removed | | **One day off in seven** | Averaged over four weeks | Kept; 24 consecutive hours, averaged within a single rotation | | **Longest shift** | 24 hours plus up to 4 for handoffs | Unchanged | | **In-house call frequency** | No more often than every third night, averaged | Removed | | **Night float** | Counts toward the limits; specialties may cap it | Separate rule removed; specialties can no longer cap it | | **Staying late by choice** | Allowed for a sick patient, a family or a unique teaching event | Removed | | **88-hour exception** | On the books; enforcement suspended in February 2026 | Removed | | **Specialty-specific limits** | Specialties may add stricter limits | Removed; existing ones deleted | | **At-home call** | Counts toward 80 hours; never on a day off | Programs must also define which home-call work counts and monitor its intensity | For a rule-by-rule walk through what each change means on a calendar, see our [guide to the proposed 2028 rules](https://jacks.med/notes/acgme-proposed-2028-work-hour-rules). ## The ACGME’s case, in its own words The rationale statement explains each change. Building and monitoring schedules is the thread that runs through it: - **Shorter rest after call:** “to reduce burden related to developing resident schedules, while retaining sufficient time for rest.” - **Specialty limits:** “variation in work hour requirements among specialties within an institution creates significant burden for scheduling and monitoring.” - **Night float:** removing specialties’ caps “will allow programs more flexibility to develop resident schedules that best support education and patient care.” - **Every third night:** the other limits already “make it nearly impossible for a program to assign a resident in-house call more frequently than once every third night on average,” so monitoring the cap “adds burden that is likely unnecessary.” An ACGME spokesperson told MedPage Today that “the intent is not to reduce protections for residents.” But in the rationale the reason given for most of these changes is not that a protection was found unnecessary. It is that keeping it is work. ## Why “too hard to schedule” doesn’t hold up We build scheduling software for residency programs and attending groups, so this is the part of the argument we know best. Here is what we see. **Every one of these rules is arithmetic on a calendar.** Hours worked in a rotation, hours between the end of a call and the next shift, full days off, how often call comes around. None of it needs judgment to check. Software checks all of it on every draft, instantly, and can build the schedule inside the rules instead of checking it afterwards. Our [free duty-hours checker](https://jacks.med/notes/acgme-duty-hours-call-schedule) tests a resident’s month against every rule in a browser, in less time than it takes to read this sentence. **We have tested it on a real program.** We took one OB-GYN program’s actual year of weekend and holiday call: 37 residents, three sites, primary and backup, with the program’s own rules for rest, rotations, PGY level, weekend caps and yearly targets. Our solver filled every seat, from four weeks up to the full year of 1,323 seats, with zero rule breaks. ([The full test](https://jacks.med/notes/claude-residency-call-schedule) is written up, including how a general-purpose chatbot did on the same year.) Compliant schedules are not too hard to build. They are too hard to build by hand. **The ACGME’s own words make the point.** It says the every-third-night cap is one the other rules make “nearly impossible” to break. A rule that almost never binds costs software nothing to keep checking. The “burden for scheduling and monitoring” created by different rules in different specialties is exactly what scheduling software is for: each program keeps its own rule set, and every schedule is checked against it. **The ACGME tightens where it thinks it matters.** The same proposal makes the 80-hour average harder to meet, requiring it to hold within a single rotation. That makes schedules harder to build, and the ACGME accepts it. Scheduling effort is a cost it is willing to impose when it believes a rule protects people. The question for every other rule should be the same: does it protect patients and residents? Not: is it work to schedule around? **The strongest counterpoint, and our answer.** The burden is real for the people carrying it. In many programs a chief resident builds the year in a spreadsheet on nights and weekends, with a one-year term and a full clinical load, and checks the rules by eye. We hear it from chiefs. But that is an argument for giving chiefs better tools and real support, not for giving residents less rest. ## What’s at stake for patients and residents **Less guaranteed rest after every 24-hour call.** Two hours shorter, on every call, for every resident. With the every-third-night cap gone, 24-hour calls could also come around more often, limited only by the 80-hour average and one day off in seven. **The evidence doesn’t cover this.** The best argument for loosening the rules is the FIRST and iCOMPARE trials, which found flexible schedules no worse for patient outcomes. Those trials loosened limits on shift length and time between shifts. In FIRST, both groups still kept the 80-hour week, one day off in seven and the every-third-night call cap. The proposal removes that cap. An ACGME-commissioned review of the evidence, published in 2025, concluded that consensus on how work hours affect resident well-being “continues to be elusive.” And in ROSTERS, residents on shorter shifts slept more but made more serious errors, a reminder that handoffs carry risk too. The evidence cuts in more than one direction. That is why rules like these should be set on safety, not on scheduling convenience. **The rules already bend.** A 2013 commentary in the Journal of Graduate Medical Education collected surveys in which more than 60% of surgery residents said they had falsely reported duty hours to appear compliant. Hours are mostly self-reported. Looser rules on top of weak monitoring compound each other. Monitoring that runs off the schedule itself, automatically, would do more for compliance than removing the rules being monitored. **The burden of proof is on removing a protection.** Taking two hours of rest away from every post-call resident should require evidence that it is safe, not evidence that keeping it is inconvenient. Burden is an administrative reason. Residents feel it as fatigue, and patients feel it through the residents caring for them. Emergency medicine residents have said much the same. The Emergency Medicine Residents’ Association said on October 7 that the changes “threaten trainees’ education and quality of life” and that “the protected time of rest should not be reduced.” Even the American Academy of Emergency Medicine, which told MedPage Today it “supports reducing unnecessary administrative requirements in graduate medical education,” added that “flexibility should not come at the expense of resident well-being, patient safety, or high-quality training.” ## What programs should do instead - **Build the schedule inside the rules.** Treat work-hour limits as the constraints a schedule is built from, not a check run after it is published. Software does this for every draft. - **Monitor from the schedule, not from memory.** Hours, rest and days off can be computed from the schedule and every swap automatically, instead of relying on residents to log them and someone to audit the logs. - **Give chiefs real tools and support.** The person building the year shouldn’t be doing it by hand on their nights off. That is where the burden actually lives. - **Comment by October 22.** Through the form on the ACGME’s review-and-comment page for the Common Program Requirements (Residency). If you agree that scheduling burden is solvable, say so, and ask the ACGME to keep 14 hours of rest after call and the every-third-night cap. This is the work we do. jacks is a scheduling platform for residency programs and attending groups: we set your program up with you, it builds the whole year with every work-hour rule checked while it builds, balances everyone against a fair target, and keeps the schedule inside the rules through every swap and call-out. Today’s rules are not too hard to keep. They just need the right tools. ## Questions people ask ### What is the ACGME proposing for resident work hours? A rewrite of the work-hour section of its Common Program Requirements, posted September 8, 2026. The 80-hour week and the 24-plus-4 shift limit stay, and the 80-hour average would have to fall within a single rotation. Rest after 24 hours of in-house call drops from 14 to 12 hours, and the eight hours between shifts, the every-third-night call cap, the night float rule, the 88-hour exception and specialty-specific limits would be removed. ### Why is the ACGME changing the rules? Its rationale statement frames the revision as reducing burden, a 43 percent cut in the number of requirements. For several changes it names scheduling directly: the shorter post-call rest is meant “to reduce burden related to developing resident schedules,” and specialty limits are said to create “significant burden for scheduling and monitoring.” ### Are the current work-hour rules really too hard to schedule? No. Every one of them is arithmetic on a calendar: hours worked, gaps between shifts, days off and how often call comes around. Scheduling software can build a schedule inside all of them and check every draft instantly. In our own test on a real 37-resident program’s full year of call, our solver filled every seat with zero rule breaks. The burden is real for chiefs working in spreadsheets, and the fix for that is better tools, not less rest. ### Is the 80-hour limit going away? No. The proposal keeps the 80-hour weekly limit averaged over up to four weeks and adds that the average must fall within a single rotation. What it removes are the narrower protections around it, most notably two hours of guaranteed rest after every 24-hour call and the every-third-night call cap. ### How can residents and programs comment? Through the comment form on the ACGME’s review-and-comment page for the Common Program Requirements (Residency), by October 22, 2026. The ACGME invites comments from its community of interest and does not reply to individual comments. A final draft goes to its board in June 2027, with a proposed effective date of July 1, 2028. --- Source: https://jacks.med/notes/acgme-work-hour-rules-controversy · jacks, scheduling for residency programs and attending groups · hello@jacks.med