How much after-hours charting is normal?

There is no clean normal, only distributions. In 2024, 24.5% of physicians reported 0 to 2 hours a week, 13.6% reported 6 to 8, and 22.5% more than 8 [1]. The classic event-log study of 142 family physicians found 1.4 hours a day outside clinic hours inside a 5.9-hour EHR day [14]. Compare yourself with peers in your specialty and FTE, not a national average.

Do AI scribes reduce pajama time?

It depends on who holds the stopwatch. EHR-log studies find note time improvements with no significant after-hours change [8][10][11]. Self-report studies find after-hours documentation falling about 0.90 hours [15], and a randomized trial found a 0.50 hour per day drop in work outside work that did not survive removing the top 3% of observations [16]. The well-being benefit is consistent; the reclaimed-evening benefit is not.

Is it better to chart during the visit or catch up at the end of the day?

For straightforward visits, closing the note in the room wins, because recall decay makes the same note longer later. For complex visits, a short batched block between sessions is usually more accurate. The behavior that reliably produces pajama time is deferring everything to a block that does not exist in your schedule.

Will a shorter note actually save me time?

Not automatically. Physicians who shortened notes cut copy/paste and templates; those who cut time in notes reduced manual typing and imported more team content [7]. If your goal is minutes rather than readability, target manual text and delegation first.

Should I ask for protected EHR time or a scribe?

Ask for protected time first. No vendor, no contract, no learning curve, and the published trade (about 13 RVUs a week initially, recovering over time) is a cleaner conversation than a per-clinician software license [4].

Back to blogs

How to Reduce Pajama Time: A Physician's Playbook

How to reduce pajama time, with cited evidence on what actually cuts after-hours charting, what does not, and a 30-day plan to finish notes on time.

Reviewed by O. Daniel Odutola, MD, MBA, MPHNovember 9, 202616 min read

The fastest way to reduce pajama time is to stop treating it as a charting problem. After-hours EHR work is mostly the overflow from three separate pipes (notes, inbox, and orders/results), and each one responds to a different fix. The interventions with the strongest published evidence are protected in-clinic time for asynchronous work, high-intensity team documentation, and inbox rerouting away from your personal queue. Ambient AI scribes help with note time, and the evidence that they shorten your evening is weaker than the marketing suggests: the objective EHR-log studies find no after-hours change, while the improvement shows up mainly in what clinicians report on surveys.

That last sentence is what vendor blogs leave out, so start there.

What "pajama time" actually measures, and why your dashboard is probably lying to you

"Pajama time" is EHR activity logged outside your scheduled clinical hours. The AMA survey version is weekday use between 5:30 p.m. and 7 a.m. In the 2024 AMA Organizational Biopsy (nearly 18,000 physicians, 43 states, more than 100 organizations), 22.5% reported more than 8 hours a week outside normal hours, up from 20.9% the year before, and another 13.6% reported 6 to 8 hours [1].

Read that again. Average workweek hours fell to 57.8 from 59.0 over the same period, and burnout has declined three years running to 41.9% in 2025 [1][2]. Pajama time went the other way. Whatever the wellness committees accomplished, it did not get you out of the chart at night.

The measurement is shakier than most dashboards admit. Arndt, Sinsky, and colleagues showed in Annals of Family Medicine that Epic's "Time Outside Scheduled Hours" excludes the 30 minutes before and after each clinic session, so two physicians with identical 0.7 FTE schedules and identical behavior can post after-hours numbers two full hours apart, purely from how their sessions are blocked [3]. They proposed WOW8 (work outside 8 scheduled patient hours), which counts everything before, between, and after appointments.

Practical consequence: when your medical director shows you a Signal report, ask what denominator it uses before you accept it as a measure of your life. Track total EHR minutes per clinic day alongside the after-hours figure, because the after-hours number can improve while your day gets worse.

For how documentation time is distributed across a clinical day, our clinical documentation guide walks through the full anatomy of EHR time.

How to reduce pajama time: five levers, ranked by published effect size

In the order the evidence supports, not the order vendors sell them.

Get protected time inside the clinic day for asynchronous work. At Yale, 41 general internists got one appointment slot per half day for messages, prior authorizations, and refills, against 89 controls, from November 2021 through June 2024. After-hours EHR use fell roughly 25 minutes a week and nonworkday use about 29 minutes, at a transient cost of about 13 RVUs a week [4]. This is the most actionable ask you can bring to a scheduling conversation: it converts invisible unpaid evening work into visible schedule capacity.
Push team documentation past 40% of note text, or do not bother. In a national cohort of 18,265 ambulatory physicians, adopters of coauthored documentation saw visits rise 6.0% and documentation time drop 9.1%. The stratified result is the one that matters: only adopters with more than 40% of note text authored by someone else actually reduced documentation time, by 53.9 minutes a week, rising to 72.2 after a 20-week learning period. Low-intensity adopters got the extra visits and none of the time back [5]. A scribe used for two patients a day is a productivity intervention dressed up as a wellness intervention.
Reroute the inbox before you optimize it. Confluence Health in north central Washington cut after-hours EHR work 67% (166 to 54 minutes per physician) and total EHR time 17% (391 to 326 minutes per 8-hour clinic day), with family physician burnout falling from 41% to 26%. The tactics were unglamorous: route messages to staff pools rather than individuals, auto-close resolved threads, filter "thank you" replies, expand self-scheduling, and change the portal language from "message your doctor" to "message your physician's office" [6].
Cut copy-forward and template text, not your typing speed. Across 130,079 ambulatory Epic users, physicians who shortened notes did it mainly by cutting copy/paste text (down 18.9%) and templated text (down 17.2%), while those who cut time in notes wrote less manual text (down 27.3%) and imported more from other team members (up 21.1%) [7]. Shorter notes and faster notes are two different projects. Our piece on note bloat covers which sections to cut first.
Add ambient AI, with calibrated expectations. See the next section.

If you do one thing this quarter, do number one. It is the only item here that does not depend on you working differently while tired.

Cut after-hours charting with Physicians' Copilot.

See how your offer compares

Compare real pay data before you sign anything

Compare pay

What each intervention actually bought: a benchmark table

Intervention	Evidence base	Effect on after-hours EHR time	Effect on documentation or note time	Productivity effect
Protected in-clinic slot for async work	41 internists vs 89 controls, 2021 to 2024 [4]	About 25 fewer min/wk; about 29 fewer min/wk on nonworkdays	Not the primary outcome	Transient dip of about 13 RVUs/wk, then recovered
Team documentation, high intensity (>40% of note text by others)	18,265 physicians, national cohort [5]	Reported as total EHR time and time outside clinic hours	53.9 fewer min/wk; 72.2 after a 20-week learning period	Visits up 6.0% (2.5/wk)
Team documentation, low intensity	Same cohort [5]	No meaningful change	No meaningful change	Visits up a comparable amount
Ambient AI scribe (real-world adoption)	8,581 clinicians, 5 health systems [8]	No significant change	16.0 fewer min per 8 patient-care hours; 27.3 for power users	0.49 more visits/wk (1.7%)
Ambient AI scribe (randomized)	238 physicians, 14 specialties [10]	Not reported as primary	Nabla 9.5% less time-in-note; DAX no significant change	Not reported
Ambient AI scribe (randomized, pediatric subspecialty)	23 physicians [11]	No significant change (-9.4 min/scheduled day)	No significant change	No significant change in wRVUs/encounter
Ambient AI scribe (self-reported, uncontrolled)	263 clinicians, 6 health systems, 30 days [15]	Self-reported after-hours documentation down 0.90 hours	Cognitive task load down 2.64 of 10	Not reported
Ambient AI scribe (stepped-wedge RCT)	66 practitioners, 2 states, 24 weeks [16]	Work outside work down 0.50 h/day, not robust to removing top 3% of observations	Note time down 0.36 h/day	Billing codes improved
Organizational inbox and workflow bundle	Confluence Health program [6]	67% reduction (166 to 54 min per clinic day)	Total EHR time down 17% (391 to 326 min)	Not reported

Every row that objectively moved after-hours time required someone other than you to change something. Every row that only changed your tooling moved note time instead. Keep that in mind when a demo promises your evenings back.

Why ambient AI scribes cut note time but not charting after hours

The largest real-world analysis to date, published in JAMA on April 1, 2026, followed 8,581 ambulatory clinicians across five academic health systems (Mass General Brigham, Emory, UCSF, Yale New Haven, and UC Davis), of whom 1,809 adopted an ambient scribe across Ambience, Nuance DAX Copilot, and Abridge. Adopters spent 13.4 fewer minutes in the EHR and 16.0 fewer minutes on documentation per 8 hours of patient care, and 21.3 and 27.3 fewer minutes among power users at 50% utilization or more. Visit volume rose 0.49 a week (1.7%). After-hours time showed no significant change [8][9].

Randomized data points the same way. In a three-group pragmatic trial of 238 outpatient physicians across 14 specialties in NEJM AI, Nabla cut time-in-note 9.5% (95% CI, -17.2% to -1.8%; P=0.02) while DAX Copilot showed no significant change (-1.7%; P=0.66). Both arms improved on Mini-Z and task-load scores, and clinicians reported clinically significant inaccuracies "occasionally" (2.7 and 2.8 on five-point scales) [10]. In a randomized quality improvement trial among 23 pediatric subspecialists, DAX improved experience and self-reported burnout while pajama time did not budge (-9.4 minutes per scheduled day; 95% CI, -41.2 to +22.4) [11].

Across an observational cohort of 8,581, a randomized trial of 238, and a randomized trial of 23, the log-based pattern is consistent: ambient AI makes the note cheaper and does not by itself get you out of the chart at 9 p.m. Utilization is the other reason. In the NEJM AI trial, DAX was used in 33.5% of 24,696 eligible visits and Nabla in 29.5% of 23,653 [10]. In the JAMA cohort, only 32% of adopters reached high-frequency use [8]. A tool used on a third of your patients cannot clear a third of your evening, and the visits clinicians skip tend to be the complicated ones that generate the longest notes.

There is a sharper way to read this literature, and it is the one to carry into a purchasing conversation. Split the studies by how they measured. Where the outcome came from EHR logs, after-hours time did not move [8][11]. Where it came from clinicians reporting on themselves, it did. A 30-day study across six health systems found self-reported after-hours documentation fell 0.90 hours (SE 0.19) and burnout dropped from 51.9% to 38.8% (odds ratio 0.26; 95% CI, 0.13 to 0.54) among 263 ambulatory clinicians [15]. Check the design before quoting it: pre and post surveys, no control arm, voluntary enrollment, one coauthor employed by the scribe vendor.

The randomized evidence lands in between and is candid about it. In a 24-week stepped-wedge trial of 66 practitioners across two states, work outside work fell 0.50 hours a day (95% CI, -0.90 to -0.09), but that finding lost significance once the top 3% of daily observations were removed [16]. Work exhaustion and disengagement improved robustly (-0.44 points; 95% CI, -0.62 to -0.25; P<0.001), and note time fell 0.36 hours a day.

The well-being signal reproduces across designs. The minutes-at-home signal does not. Buying an ambient scribe to feel less shattered at the end of clinic is well supported. Buying it to reclaim a set number of evening hours means asking the vendor which studies measured logs and which measured feelings.

Use ambient AI. Buy it for the reason it works, a shorter and less miserable note at the moment of care, and pair it with a fix aimed at the inbox.

No AI tool replaces clinical judgment. Verify all outputs against primary sources and your own assessment.

The part nobody schedules: after hours documentation is increasingly messages, not notes

Note-writing time per note has been falling. Epic Research analyzed 1.7 billion notes from 166,318 outpatient providers across 384 organizations from May 2020 to April 2023: average writing time dropped 11.1% (5.4 to 4.8 minutes per note) even as note length grew 8.1% (4,628 to 5,002 characters) [12].

Meanwhile the inbox exploded. A 2026 JAMA analysis using Epic Cosmos, covering 140 million patient records across 2,067 hospitals and 47,100 clinics, found portal messages rose 153% between 2020 and 2025 [13].

Put those two findings together. Your notes got faster. Your evenings did not. The growth is in asynchronous care nobody added to your template.

None of this is new. The 2017 event-log study that put the phrase on the map already found inbox management eating 85 minutes a day, 23.7% of total EHR time, second only to clerical work at 157 minutes [14]. We have had nine years to reroute that queue.

Three things that help, in descending order of how much control you have:

Insist on pool-based routing for results, refills, and forms. Confluence Health's 67% reduction came substantially from this [6].
Batch the inbox twice a day at fixed times. Continuous triage means you finish the day with the inbox cleared and the notes open, which is exactly the trade that produces pajama time.
Bill for asynchronous work where it qualifies. Nothing focuses an administrator on inbox volume like a line item.

Because after-hours EHR time and burnout travel together, the case for these changes belongs in the same conversation as your physician burnout metrics, not a separate wellness track.

A 30-day plan to finish notes faster

Thirty days is enough to get a baseline, run one change, and see whether it held.

Days 1 to 7: measure honestly. Pull your EHR efficiency report and write down three numbers: total EHR minutes per scheduled 8 hours, minutes after 5:30 p.m. on weekdays, and minutes on unscheduled days. Ask which denominator it uses, given the TOSH problem above [3].

Days 8 to 14: split that time into notes, inbox, and orders/results. Most people are surprised. If more than half is inbox, a scribe will not fix your evenings.

Days 15 to 21: change exactly one thing. Pick the lever matching your biggest bucket. Notes: close the note in the room for your five simplest visit types and strip one copy-forward block permanently [7]. Inbox: move one message category to a pool and batch the rest twice daily [6]. Orders: build two order sets for your highest-volume complaints.

Days 22 to 30: remeasure, then ask for the schedule change. Bring the Yale protected-time data to your division chief with your own numbers attached [4]. The ask is one asynchronous slot per half day: a transient dip of roughly 13 RVUs a week, for about 25 fewer after-hours minutes a week plus 29 fewer on days off.

One caution. Only high-intensity team documentation and the Confluence Health bundle produced an hour a week alone, and the latter was an organizational program, not a habit change. Stacking two or three is how the numbers get meaningful. Anyone promising a single tool that ends after-hours charting is selling, not citing.

Cut after-hours charting with Physicians' Copilot.

Key takeaways

Pajama time is rising even as workweeks shorten and burnout falls: 22.5% of physicians reported 8+ hours a week of after-hours EHR use in 2024, up from 20.9% in 2023 [1].
Ambient AI splits by measurement: EHR-log studies show 13.4 fewer EHR minutes per 8 patient-care hours and no after-hours change [8], while self-report and survey studies show after-hours documentation falling 0.90 hours and burnout dropping from 51.9% to 38.8% [15].
Team documentation bought time back only above 40% of note text authored by others: 53.9 fewer documentation minutes a week, rising to 72.2 after a 20-week learning curve [5].
Protected in-clinic time for asynchronous work cut after-hours EHR use about 25 minutes a week and nonworkday use about 29 minutes, at a transient productivity cost [4].
Vendor after-hours metrics undercount by excluding 30 minutes around each session, so verify the denominator before trusting your dashboard [3].

REFERENCES

Berg S. "Doctors work fewer hours, but the EHR still follows them home." American Medical Association. Published August 19, 2025. Reports 2024 AMA Organizational Biopsy data (nearly 18,000 physicians, 43 states, 100+ organizations). https://www.ama-assn.org/practice-management/physician-health/doctors-work-fewer-hours-ehr-still-follows-them-home. Accessed September 29, 2026.
"Physician burnout rate continues to decline, falling to nearly 42%." American Medical Association. Published April 16, 2026. Reports 2025 AMA Organizational Biopsy and National Physician Comparison Report data (nearly 19,000 responses, 106 organizations). https://www.ama-assn.org/practice-management/physician-health/physician-burnout-rate-continues-decline-falling-nearly-42. Accessed September 29, 2026.
Arndt BG, Micek MA, Rule A, Shafer CM, Baltus JJ, Sinsky CA. "Refining Vendor-Defined Measures to Accurately Quantify EHR Workload Outside Time Scheduled With Patients." Annals of Family Medicine. May 1, 2023;21(3):264-268. doi:10.1370/afm.2960. https://www.annfammed.org/content/21/3/264. Accessed September 29, 2026.
Kanaparthy NS, Holmgren AJ, Sun Y, Paek H, Williams B, McLean R, Doolan R, Goodman KE, Rotenstein L, Melnick ER. "Protected Time for Electronic Health Record Work and Physician Productivity." JAMA Network Open. December 1, 2025;8(12):e2546550. doi:10.1001/jamanetworkopen.2025.46550. Effect sizes as reported in Medical Economics coverage of the same study, December 2025: https://www.medicaleconomics.com/view/protected-ehr-time-cut-after-hours-work-with-little-productivity-loss-study-finds. Citation retrieved via PubMed (PMID 41329489). https://doi.org/10.1001/jamanetworkopen.2025.46550. Accessed September 29, 2026.
Apathy NC, Holmgren AJ, Cross DA. "Physician EHR Time and Visit Volume Following Adoption of Team-Based Documentation Support." JAMA Internal Medicine. October 1, 2024;184(10):1212-1221. doi:10.1001/jamainternmed.2024.4123. Retrieved via PubMed (PMID 39186284). https://doi.org/10.1001/jamainternmed.2024.4123. Accessed September 29, 2026.
Berg S. "Small steps add up to much less 'pajama time' for physicians." American Medical Association. Published June 26, 2024. Confluence Health results. https://www.ama-assn.org/practice-management/physician-health/small-steps-add-much-less-pajama-time-physicians. Accessed September 29, 2026.
Apathy NC, Hare AJ, Fendrich S, Cross DA. "I had not time to make it shorter: an exploratory analysis of how physicians reduce note length and time in notes." Journal of the American Medical Informatics Association. January 18, 2023;30(2):355-360. doi:10.1093/jamia/ocac211. Retrieved via PubMed (PMID 36323282). https://academic.oup.com/jamia/article-abstract/30/2/355/6794246. Accessed September 29, 2026.
"JAMA Study: AI Scribes Deliver Modest EHR Time Savings Across 5 Major Health Systems." HIT Consultant. Published April 1, 2026. Detailed reporting of the JAMA multisite study (8,581 ambulatory clinicians, 1,809 adopters, five health systems). https://hitconsultant.net/2026/04/01/jama-ai-scribe-study-ehr-time-savings-burnout-reality-check/. Accessed September 29, 2026.
Aguilar M. "Large AI scribe study finds modest time savings, inconsistent use." STAT News. Published April 1, 2026. https://www.statnews.com/2026/04/01/ai-ambient-scribes-modest-time-savings-clinical-documentation/. Primary study abstract: JAMA, April 2026. https://jamanetwork.com/journals/jama/article-abstract/2847319. Accessed September 29, 2026.
Lukac PJ, Turner W, Vangala S, Chin AT, Khalili J, Shih YT, Sarkisian C, Cheng EM, Mafi JN. "Ambient AI Scribes in Clinical Practice: A Randomized Trial." NEJM AI. November 26, 2025;2(12). doi:10.1056/aioa2501000. Retrieved via PubMed (PMID 41497288). https://doi.org/10.1056/aioa2501000. Accessed September 29, 2026.
Shin HS, Williams H, Braykov N, Jahan A, Meller J, Orenstein EW. "The Influence of Artificial Intelligence Scribes on Clinician Experience and Efficiency among Pediatric Subspecialists: A Rapid, Randomized Quality Improvement Trial." Applied Clinical Informatics. July 17, 2025;16(4):1041-1052. doi:10.1055/a-2657-8087. Retrieved via PubMed (PMID 40675605). https://doi.org/10.1055/a-2657-8087. Accessed September 29, 2026.
"Two Years After Coding Changes Sought to Decrease Documentation, Notes Remain 'Bloated'." Epic Research. Published July 6, 2023. Analysis of 1.7 billion notes, 166,318 outpatient providers, 384 organizations, May 2020 to April 2023. https://www.epicresearch.org/articles/two-years-after-coding-changes-sought-to-decrease-documentation-notes-remain-bloated/. Accessed September 29, 2026.
"Portal messages in Epic have more than doubled since 2020." Healthcare IT News. Published June 2026. Reports NYU Langone Health analysis published in JAMA, June 22, 2026, using Epic Cosmos (140 million patient records, 2,067 hospitals, 47,100 clinics). https://www.healthcareitnews.com/news/portal-messages-epic-have-more-doubled-2020. Accessed September 29, 2026.
Arndt BG, Beasley JW, Watkinson MD, Temte JL, Tuan WJ, Sinsky CA, Gilchrist VJ. "Tethered to the EHR: Primary Care Physician Workload Assessment Using EHR Event Log Data and Time-Motion Observations." Annals of Family Medicine. September 2017;15(5):419-426. doi:10.1370/afm.2121. Clinic hours defined as 8:00 a.m. to 6:00 p.m., Monday through Friday. Retrieved via PubMed (PMID 28893811). https://www.annfammed.org/content/15/5/419. Accessed September 29, 2026.
Olson KD, Meeker D, Troup M, Barker TD, Nguyen VH, Manders JB, Stults CD, Jones VG, Shah SD, Shah T, Schwamm LH. "Use of Ambient AI Scribes to Reduce Administrative Burden and Professional Burnout." JAMA Network Open. October 1, 2025;8(10):e2534976. doi:10.1001/jamanetworkopen.2025.34976. Pre/post quality improvement study, 263 ambulatory clinicians, 6 health systems, February to October 2024; no control arm, and coauthor M. Troup is affiliated with Abridge AI, Inc. Retrieved via PubMed (PMID 41037268). https://doi.org/10.1001/jamanetworkopen.2025.34976. Accessed September 29, 2026.
Afshar M, Baumann MR, Resnik F, Hintzke J, Sullivan AG, Wills G, et al. "A Pragmatic Randomized Controlled Trial of Ambient Artificial Intelligence to Improve Health Practitioner Well-Being." NEJM AI. November 26, 2025;2(12). doi:10.1056/aioa2500945. 24-week stepped-wedge individually randomized trial, 66 practitioners, 71,487 notes. ClinicalTrials.gov NCT06517082. Retrieved via PubMed (PMID 41625485). https://doi.org/10.1056/aioa2500945. Accessed September 29, 2026.

Research support: article metadata for references 4, 5, 7, 10, 11, 14, 15, and 16 was retrieved from PubMed.

FAQs

It is EHR activity logged outside your scheduled clinical hours. The AMA survey definition is weekday use between 5:30 p.m. and 7 a.m. [1]. Epic's Time Outside Scheduled Hours is narrower, excluding 30 minutes before and after each session, which is why the same physician looks different on two reports [3].