BRENNEN

Industries / Health / Healthcare / Outpatient surgery centers

Scheduling

This is the work of fitting cases into operating rooms and staff hours, predicting how long each case will take, and getting each patient to arrive ready.

7

Milestones achieved, from 1996 to 2022.

2

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1996

    Surgeons beat the software

    At the University of Washington Medical Center, 26 of 43 surgeons estimated their operating times significantly better than the hospital's commercial scheduling system, and none did worse. Errors were still large, at 34% of a 157-minute average case, and a regression model that combined surgeons' estimates with case details improved accuracy by 11.5% to 18.2%.

    Lessonthe authors expected the next gains to come from giving surgeons timely feedback from historical data, not from better software alone.

    pubmed.ncbi.nlm.nih.gov

    Credit

    I. H. Wright, C. Kooperberg, B. A. Bonar and G. Bashein

  2. 1999

    Fitting add-on cases in

    An add-on case is one booked into time left open after the day's schedule is set. Franklin Dexter, A. Macario and R. D. Traub treated this as bin packing, the problem of fitting items of different sizes into containers, and simulated 10 published algorithms on data from two surgical suites. The best, Best Fit Descending with fuzzy constraints, gave operating room utilization 4% higher than the worst.

    pubmed.ncbi.nlm.nih.gov

    Credit

    F. Dexter, A. Macario and R. D. Traub

  3. 2000

    Case times are log-normal

    In 40,076 cases at a large teaching hospital, the times for each combination of procedure and anesthesia type fitted a log-normal curve, one with a long tail of slow cases, better than the bell-shaped normal curve. The authors recommended log-normal models for predicting the time of each procedure type.

    pubmed.ncbi.nlm.nih.gov

    Credit

    D. P. Strum, J. H. May and L. G. Vargas

  4. 2003

    Faster turnovers save little

    Turnover time is the gap between one patient leaving a room and the next one arriving. Using data from the operating room information systems of four academic hospitals, Dexter and colleagues showed that cutting average turnover by 3 to 9 minutes would cut staffing costs by 0.8% to 1.8%, and cutting it by 10 to 19 minutes by 2.5% to 4.0%, mostly by allocating less operating room time. They wrote that the other arguments for shorter turnovers, such as fitting in more cases, had been shown to be unsound.

    Lessonthe number an operating room committee watches most closely is not always the one that moves cost.

    pubmed.ncbi.nlm.nih.gov

    Credit

    F. Dexter, A. E. Abouleish, R. H. Epstein, C. W. Whitten and D. A. Lubarsky

  5. 2005

    Parallel rooms cut idle time

    A team at Harvard Medical School in Boston built a three-room suite: an operating room, an induction room where anesthesia was started, and an early recovery area, with extra anesthesia and nursing staff so that steps usually done one after another ran at the same time. For the same surgeons and case mix, time per case outside the operation itself fell from 67 to 38 minutes and more cases were done each day. Costs per case rose, but the extra cases paid for them.

    pubmed.ncbi.nlm.nih.gov

    Credit

    W. S. Sandberg and colleagues

  6. 2021

    Predicted times, tested at random

    At Memorial Sloan Kettering Cancer Center in New York, 683 colorectal and gynecology cases in 2018 were randomly assigned either to machine-learning predictions of case length or to the usual estimates. The mean absolute error fell from 59.3 to 49.5 minutes, and the wait before a delayed start fell from 49.4 to 16.3 minutes per patient, without longer gaps between cases. The authors wrote that the real-world effect of such a model had apparently not been studied before.

    pubmed.ncbi.nlm.nih.gov

    Credit

    C. T. Strömblad and colleagues, Memorial Sloan Kettering (JAMA Surgery)

  7. 2022

    Pit-stop operating lists

    Imran Ahmad, a consultant anaesthetist at Guy's and St Thomas' NHS Foundation Trust in London, says he came up with high-intensity theatre (HIT) lists in February 2022, modeled on Formula 1 pit crews. "Theatre" is the British word for operating room. A HIT list takes one kind of procedure, two theatres, extra staff and several planning meetings, so the surgeon moves straight to the next prepared patient. Ahmad says a surgeon on a normal list may operate for as little as 30% of the day, and for 90% on a HIT list. Other NHS trusts have since published their own, among them Portsmouth, which ran up to 10 day-case robotic gallbladder removals in a working day with one team and one theatre.

    emjreviews.comnature.compubmed.ncbi.nlm.nih.gov

    Credit

    Imran Ahmad and colleagues at Guy's and St Thomas'; J. Gray, G. I. van Boxel and colleagues at Portsmouth

Watch for

  1. Case building for surgical clinics

    LeanTaaS's iQueue for Surgical Clinics builds cases, tracks whether each patient is ready and contacts patients before surgery. LeanTaaS lists results of 10% total case growth, up to 72% of previously lost surgeries recovered and more than 250 surgeon hours reclaimed a year, and says surgical clinics lose up to 35% of potential cases each year to breakdowns in coordination, prior authorization and patient readiness, at a cost to hospitals and ambulatory surgery centers.

    Achieved whenan outside party publishes before-and-after case counts for named clinics or centers that use it.

    leantaas.com

    Claimed by

    LeanTaaS, June 2025

  2. Cameras that time each step

    Apella's system uses wall-mounted cameras and computer vision to record events such as a patient entering the room or being draped, in place of times typed into the health record. Apella says its event detection scored above 0.98 on F1, a combined measure of missed and false detections where 1 is perfect, across more than 100,000 cases, and that its use was associated with two more cases per operating room per month. It calls its peer-reviewed evaluation independent, but the paper on its system found here, on 2,502 knee replacements, was written by three Apella employees and a surgeon who consults for Apella.

    Achieved whena group with no ties to Apella measures the detection accuracy or the case gains at sites it chooses.

    apella.iopubmed.ncbi.nlm.nih.gov

    Claimed by

    Apella, February 2026

Open

  1. 01

    A planner for pit-stop lists

    Software that takes a list of cases of one procedure type, two operating rooms, a staff roster and the usual length of each step, and lays out a HIT list: when each patient is prepared in the anesthetic room, when each room turns over, and when the surgeon moves between rooms. Check it by replaying lists that hospitals have published, using only the case counts and timings given in their papers.

    The brief
    Credit · unclaimed
  2. 02

    A register of efficiency claims

    A public register of operating room efficiency claims made by scheduling and tracking software vendors such as LeanTaaS, Qventus and Apella. For each claim, record the site, the dates, the measure and how it is defined, the baseline, the result, who measured it, and whether anyone outside the vendor and its customer checked it.

    The brief
    Credit · unclaimed

Open 01 · the brief

A planner for pit-stop lists

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

Software that takes a list of cases of one procedure type, two operating rooms, a staff roster and the usual length of each step, and lays out a HIT list: when each patient is prepared in the anesthetic room, when each room turns over, and when the surgeon moves between rooms. Check it by replaying lists that hospitals have published, using only the case counts and timings given in their papers.

Why it is still open

HIT lists are planned by people in several planning meetings that bring every member of the patient's pathway together, Ahmad says. The published lists report their own timings, such as a thoracic team's 10 lung resections in 5 hours 45 minutes with 2 to 3 minutes between cases, and none of the papers read for this file describes planning software. A GitHub search for theatre list planners found one unrelated waiting-list model.

emjreviews.compubmed.ncbi.nlm.nih.gov

Built with what exists

Google OR-Tools, whose CP-SAT solver builds timetables under constraints (Apache-2.0 license); SimPy, a Python library for simulating queues and flows, to test each plan against random case times drawn from log-normal curves; Streamlit for a web page a theatre manager can use.

github.compypi.org

  1. First step

    Copy the case counts, start times, case times and turnover times from three published HIT-list papers into one table.

  2. Done looks like

    For at least 5 published HIT lists, the planner's predicted finish time is within 15 minutes of the reported finish, given only the inputs each paper reports, with the plans and code public.

Open 02 · the brief

A register of efficiency claims

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public register of operating room efficiency claims made by scheduling and tracking software vendors such as LeanTaaS, Qventus and Apella. For each claim, record the site, the dates, the measure and how it is defined, the baseline, the result, who measured it, and whether anyone outside the vendor and its customer checked it.

Why it is still open

Results reach buyers as vendor releases and case studies, each in its own measure: LeanTaaS reports case growth and surgeon hours, and Apella reports detection scores and cases per room per month. The one peer-reviewed paper on Apella's system found here was written mostly by its employees, and the most detailed comparison of the rival tools found here is a blog post by Apella, a competitor.

leantaas.compubmed.ncbi.nlm.nih.govapella.io

Built with what exists

vendors' press pages and case studies, trade press, searches of PubMed and Europe PMC, a spreadsheet or small database, and a language model such as Claude Fable 5.2 (see the Note "Three working devices") to pull the fields out of each source for a person to check.

  1. First step

    Collect every claim posted on the press pages of LeanTaaS, Qventus and Apella from 2024 to 2026 into the table.

  2. Done looks like

    At least 50 claims, each linked to its source and graded by evidence (vendor statement, customer statement, peer-reviewed with vendor authors, independent), with two people agreeing on the grade for at least 90% of entries.