Karl Y. Bilimoria
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Karl Y. Bilimoria has a total of 99 co-authors. Here are the top co-authors by number of shared publications:
- Yue-Yung Hu
Surgical Outcomes and Quality Improvement Center (SOQIC), Department of Surgery, Indiana University School of Medicine, Indianapolis, IN
5 shared publications - Anthony D. Yang
Division of Surgical Oncology, Indiana University School of Medicine
3 shared publications - Elaine O. Cheung
Department of Medical Social Sciences, Northwestern University Feinberg School of Medicine, Chicago, Illinois
3 shared publications - Jeanette W. Chung
Jeanette W. Chung, PhD, is Senior Research Professor of Surgery, SOQIC, Indiana University School of Medicine, Indianapolis, Indiana, USA; and
3 shared publications - Ryan P. Merkow
Department of Surgery, University of Chicago, Chicago, IL
3 shared publications - Anne Madeleine Stey
Department of Surgery, Feinberg School of Medicine, Northwestern University, Chicago, USA
2 shared publications
Latest Publications
- Open Access
Large language model-augmented implicit surgical video review
Surgical video interpretation is a promising medical artificial intelligence application. However, no existing video annotation method preserves the spatiotemporal complexity of surgeon reasoning. Here we show that verbal reasoning and visual attention can be converted into structured, machine-actionable records of intraoperative behaviours. Our method decomposes transcribed verbal commentary into video-anchored semantic feedback chunks, which are classified via a large language model, with spatial grounding to surgical scenes via eyegaze or cursor tracking. We demonstrate method validity and scalability on structured and unstructured annotation tasks. For quality feedback on full-length colorectal procedures, the method reached near-human fidelity for chunking (cosine similarity: 0.95±0.01) and semantic classification across observations (Cohen’s κ: 0.71±0.07) and evaluative triggers (Cohen’s κ: 0.67±0.14), with excellent usability ratings. For structured critical view of safety assessment in laparoscopic cholecystectomy, implicit annotation yielded excellent agreement with explicit reviewer ratings (Cohen’s κ: 0.83, 0.49 and 0.81 across three criteria). We anticipate this method will advance surgical data science by enabling scalable construction of meaningfully annotated surgical video datasets.
Open AccessLarge language model-augmented implicit surgical video review
Surgical video interpretation is a promising medical artificial intelligence application. However, no existing video annotation method preserves the spatiotemporal complexity of surgeon reasoning. Here we show that verbal reasoning and visual attention can be converted into structured, machine-actionable records of intraoperative behaviours. Our method decomposes transcribed verbal commentary into video-anchored semantic feedback chunks, which are classified via a large language model, with spatial grounding to surgical scenes via eyegaze or cursor tracking. We demonstrate method validity and scalability on structured and unstructured annotation tasks. For quality feedback on full-length colorectal procedures, the method reached near-human fidelity for chunking (cosine similarity: 0.95±0.01) and semantic classification across observations (Cohen’s κ: 0.71±0.07) and evaluative triggers (Cohen’s κ: 0.67±0.14), with excellent usability ratings. For structured critical view of safety assessment in laparoscopic cholecystectomy, implicit annotation yielded excellent agreement with explicit reviewer ratings (Cohen’s κ: 0.83, 0.49 and 0.81 across three criteria). We anticipate this method will advance surgical data science by enabling scalable construction of meaningfully annotated surgical video datasets.
- 1 October 2025
- Published by Journal of Graduate Medical Education in Journal of Graduate Medical Education
- Vol. 17 (5) , 676-681
Accreditation Council for Graduate Medical Education Work Hour Requirements: A Review of Recent Evidence
- 1 October 2025
- Published by Journal of Graduate Medical Education in Journal of Graduate Medical Education
- Vol. 17 (5) , 676-681
Accreditation Council for Graduate Medical Education Work Hour Requirements: A Review of Recent Evidence
- 1 August 2025
- Published by Journal of Graduate Medical Education in Journal of Graduate Medical Education
- Vol. 17 (4) , 486-496
Association of the 2017 ACGME US Resident Duty Hour Policy Change With Hospital Quality and Patient Experience
Background In 2017, the Accreditation Council for Graduate Medical Education (ACGME) updated select US residency duty hour requirements to improve continuity of care and resident education. It is unknown if this policy change affected hospital quality of care and patient experience. Objective To evaluate the association of the 2017 duty hour policy change with hospital quality and patient experience in teaching vs nonteaching hospitals. Methods In this observational difference-in-differences (DiD) study, hospital quality (patient outcomes) and patient experience metrics were obtained from Centers for Medicare & Medicaid Services Care Compare data from before (July 1, 2014-June 30, 2016) and after (July 1, 2017-June 30, 2019) the 2017 policy change in teaching and nonteaching hospitals. Primary outcomes include hospital quality (five 30-day readmission indicator rates, five 30-day mortality rates, 8 patient safety indicators), and patient experience (5 measures from Hospital Consumer Assessment of Healthcare Providers and Systems survey). Results A total of 2935 hospitals (250 teaching; 2685 nonteaching) were included. When comparing before and after the 2017 policy change, teaching hospitals had greater reductions in hospital-wide readmission (DiD coefficient -0.26; 95% CI, -0.34 to -0.18; P <.001), heart failure readmission (DiD coefficient -0.48; 95% CI, -0.78 to -0.19; P <.002), and stroke mortality rates (DiD coefficient -0.56; 95% CI, -0.94 to -0.19; P <.01) than did nonteaching hospitals. There were no significant differences between teaching and nonteaching hospitals in other outcomes before vs after policy change. Conclusions After the 2017 duty hour policy change, there was no evidence of worsening of hospital quality or patient experience in teaching hospitals, compared to nonteaching hospitals.
- 1 August 2025
- Published by Journal of Graduate Medical Education in Journal of Graduate Medical Education
- Vol. 17 (4) , 486-496
Association of the 2017 ACGME US Resident Duty Hour Policy Change With Hospital Quality and Patient Experience
Background In 2017, the Accreditation Council for Graduate Medical Education (ACGME) updated select US residency duty hour requirements to improve continuity of care and resident education. It is unknown if this policy change affected hospital quality of care and patient experience. Objective To evaluate the association of the 2017 duty hour policy change with hospital quality and patient experience in teaching vs nonteaching hospitals. Methods In this observational difference-in-differences (DiD) study, hospital quality (patient outcomes) and patient experience metrics were obtained from Centers for Medicare & Medicaid Services Care Compare data from before (July 1, 2014-June 30, 2016) and after (July 1, 2017-June 30, 2019) the 2017 policy change in teaching and nonteaching hospitals. Primary outcomes include hospital quality (five 30-day readmission indicator rates, five 30-day mortality rates, 8 patient safety indicators), and patient experience (5 measures from Hospital Consumer Assessment of Healthcare Providers and Systems survey). Results A total of 2935 hospitals (250 teaching; 2685 nonteaching) were included. When comparing before and after the 2017 policy change, teaching hospitals had greater reductions in hospital-wide readmission (DiD coefficient -0.26; 95% CI, -0.34 to -0.18; P <.001), heart failure readmission (DiD coefficient -0.48; 95% CI, -0.78 to -0.19; P <.002), and stroke mortality rates (DiD coefficient -0.56; 95% CI, -0.94 to -0.19; P <.01) than did nonteaching hospitals. There were no significant differences between teaching and nonteaching hospitals in other outcomes before vs after policy change. Conclusions After the 2017 duty hour policy change, there was no evidence of worsening of hospital quality or patient experience in teaching hospitals, compared to nonteaching hospitals.
- 4 October 2022
- Published by Elsevier in Journal of Surgical Education
- Vol. 80 (1) , 72-80
National Comparison of Program Director Perceptions vs. Resident Reports of the Learning Environment and Well-Being
- 4 October 2022
- Published by Elsevier in Journal of Surgical Education
- Vol. 80 (1) , 72-80
National Comparison of Program Director Perceptions vs. Resident Reports of the Learning Environment and Well-Being
- 17 August 2022
- Published by Springer Nature in Archives of Dermatological Research
- Vol. 315 (3) , 629-632
Readability of online patient education materials for Merkel cell carcinoma
The Internet is a significant source of information for patients. According to the National Institutes of Health, patient education materials (PEMs) should be at or below an eighth-grade reading level. Merkel cell carcinoma (MCC) is a rare and aggressive skin cancer that affects patients over 50 with rising incidence. Unfortunately, US adults aged 65 + have the least proficiency in health literacy. This study assessed the readability of online PEMs and factors that contribute to readability. We retrieved 50 PEM websites and extracted primary content. A readability software package calculated six readability statistics and generated a consensus standard readability. Overall, only eight articles had a standard reading level of eighth-grade level or below (16%). The median standard reading level was at the 11th-grade level. We also examined MCC PEMs from cancer treatment institution websites (N = 20). We determined whether they contained institution-specific information, meaning they contained text information about the institution-specific expertise and specialist team. Websites containing this information (N = 13) had a significantly higher reading level than websites that did not (N = 7) in five of six readability metrics (p < 0.05). We concluded that MCC PEMs with institution-specific information led to significantly higher reading level scores. We propose that such information may increase cognitive load, as patients are learning about their disease and treatment and contending with the institution-specific information. The Cognitive Load Theory principles of intrinsic load (learning the material relevant to the disease and treatment) and extraneous load (institution-specific information and increased reading level) are constrained by limited working memory. Working memory decreases with age; hence, the patient demographic most sensitive to increased extraneous load tends to overlap with that of MCC. As patients typically read pages linked from their search engine, we suggest moving institution-specific information to another page, separate from the PEMs.
- 17 August 2022
- Published by Springer Nature in Archives of Dermatological Research
- Vol. 315 (3) , 629-632
Readability of online patient education materials for Merkel cell carcinoma
The Internet is a significant source of information for patients. According to the National Institutes of Health, patient education materials (PEMs) should be at or below an eighth-grade reading level. Merkel cell carcinoma (MCC) is a rare and aggressive skin cancer that affects patients over 50 with rising incidence. Unfortunately, US adults aged 65 + have the least proficiency in health literacy. This study assessed the readability of online PEMs and factors that contribute to readability. We retrieved 50 PEM websites and extracted primary content. A readability software package calculated six readability statistics and generated a consensus standard readability. Overall, only eight articles had a standard reading level of eighth-grade level or below (16%). The median standard reading level was at the 11th-grade level. We also examined MCC PEMs from cancer treatment institution websites (N = 20). We determined whether they contained institution-specific information, meaning they contained text information about the institution-specific expertise and specialist team. Websites containing this information (N = 13) had a significantly higher reading level than websites that did not (N = 7) in five of six readability metrics (p < 0.05). We concluded that MCC PEMs with institution-specific information led to significantly higher reading level scores. We propose that such information may increase cognitive load, as patients are learning about their disease and treatment and contending with the institution-specific information. The Cognitive Load Theory principles of intrinsic load (learning the material relevant to the disease and treatment) and extraneous load (institution-specific information and increased reading level) are constrained by limited working memory. Working memory decreases with age; hence, the patient demographic most sensitive to increased extraneous load tends to overlap with that of MCC. As patients typically read pages linked from their search engine, we suggest moving institution-specific information to another page, separate from the PEMs.
- 15 June 2022
- Published by Elsevier in Preventive Medicine
- Vol. 161, 107110
The association between gun shows and firearm injuries: An analysis of 259 gun shows across 23 US cities
- 15 June 2022
- Published by Elsevier in Preventive Medicine
- Vol. 161, 107110
The association between gun shows and firearm injuries: An analysis of 259 gun shows across 23 US cities
- 15 February 2022
- Published by Springer Nature in Nature Reviews Clinical Oncology
- Vol. 19 (4) , 221-222
Risk-adjusting away volume as a quality metric for surgical oncology: a perspective worth re-visiting
- 15 February 2022
- Published by Springer Nature in Nature Reviews Clinical Oncology
- Vol. 19 (4) , 221-222
Risk-adjusting away volume as a quality metric for surgical oncology: a perspective worth re-visiting
- 1 November 2021
- Published by Wolters Kluwer Health in Journal of the American College of Surgeons
- Vol. 233 (5) , S142
Post-Discharge Deterioration after Pancreatic Surgery: Call for an Actionable Cancer Surgery Quality Measure
Clinical deterioration, defined as end-organ-dysfunction following surgery, is a devastating, yet potentially preventable set of complications, usually occurring after an inciting event. The extent to which deterioration occurs post-discharge after pancreatectomy is unknown. Our objectives were to (1) characterize post-discharge deterioration after pancreatectomy, and (2) identify factors associated with post-discharge deterioration.
- 1 November 2021
- Published by Wolters Kluwer Health in Journal of the American College of Surgeons
- Vol. 233 (5) , S142
Post-Discharge Deterioration after Pancreatic Surgery: Call for an Actionable Cancer Surgery Quality Measure
Clinical deterioration, defined as end-organ-dysfunction following surgery, is a devastating, yet potentially preventable set of complications, usually occurring after an inciting event. The extent to which deterioration occurs post-discharge after pancreatectomy is unknown. Our objectives were to (1) characterize post-discharge deterioration after pancreatectomy, and (2) identify factors associated with post-discharge deterioration.