CPT to RVU conversion improves model performance in the prediction of surgical case length

Garside, Nicholas; Zaribafzadeh, Hamed; Henao, Ricardo; Chung, Royce; Buckland, Daniel M.

doi:10.1038/s41598-021-93573-2

Cited by 8 publications

(7 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, total wRVU has been demonstrated to correlate well with case complexity in multiple surgical specialties. 16,25,26 Additionally, the mFI-5 has been validated as a tool in predicting complications after free flap reconstruction. 27 We uniquely utilized a combination of these factors in an attempt to control for operative complexity and patient preoperative health.…”

Section: Discussionmentioning

confidence: 99%

Effect of 2 Teams and Operative Time on Complications After Oral Cavity Free Flap Reconstruction

Shaikh

Noor

Jafary

et al. 2023

Ann Otol Rhinol Laryngol

View full text Add to dashboard Cite

Objective: Evaluate the effects of operative time and 2 team approach on complications after soft tissue free flap reconstruction for oral tongue cancer. Methods: Patients with oncologic glossectomy with myocutaneous or fasciocutaneous free flap reconstruction were included from the 2015 to 2018 American College of Surgery National Surgical Quality Improvement Program. The primary predictive variables assessed were operative time and 2 team approach; control variables included age, sex, body mass index (BMI), 5-question-modified frailty index (mFI-5), American Society of Anesthesiologists (ASA) class, and total work relative value units (wRVU). Outcomes assessed included 30-day mortality, 30-day reoperation, hospital length of stay beyond 30 days, readmission, medical and surgical complications, and non-home discharge. Multivariable logistic/linear regression models were used to predict surgical outcomes. Results: Microvascular soft tissue free flap reconstruction of the oral cavity after glossectomy was performed on 839 patients. Operative time was independently associated with readmission, prolonged length of stay, surgical complications, medical complications, and non-home discharge. A 2-team approach was independently associated with prolonged length of stay and medical complications. The mean operative time of the 1-team and 2-team approach was 8.73 and 9.13 hours. The 1-team approach did not significantly increase operative time ( P = .16). Conclusions: In the largest study to date of operative time on post-surgical outcomes after glossectomy and soft tissue free flap reconstruction, we found longer operative times increased rates of postoperative complications and non-home discharge. The 1-team approach is non-inferior to the 2-team approach with respect to operating time and complications.

show abstract

Section: Discussionmentioning

confidence: 99%

Effect of 2 Teams and Operative Time on Complications After Oral Cavity Free Flap Reconstruction

Shaikh

Noor

Jafary

et al. 2023

Ann Otol Rhinol Laryngol

View full text Add to dashboard Cite

show abstract

“…This may, however, reflect the design of the experiment and the limited set of models assessed. For example, we would expect sign-out time and CPT code usage to partially reflect case complexity and uncertainty 40 , where more challenging cases could impact the measurement of relative value units (RVUs). However, sign-out time data is only a one-sided representation of the complexity of the text and may not represent true complexity in several cases.…”

Section: Discussionmentioning

confidence: 99%

Assessing the Impact of Pretraining Domain Relevance on Large Language Models Across Various Pathology Reporting Tasks

Lu,

Srinivasan,

Preum

et al. 2023

Preprint

View full text Add to dashboard Cite

Deep learning (DL) algorithms continue to develop at a rapid pace, providing researchers access to a set of tools capable of solving a wide array of biomedical challenges. While this progress is promising, it also leads to confusion regarding task-specific model choices, where deeper investigation is necessary to determine the optimal model configuration. Natural language processing (NLP) has the unique ability to accurately and efficiently capture a patient’s narrative, which can improve the operational efficiency of modern pathology laboratories through advanced computational solutions that can facilitate rapid access to and reporting of histological and molecular findings. In this study, we use pathology reports from a large academic medical system to assess the generalizability and potential real-world applicability of various deep learning-based NLP models on reports with highly specialized vocabulary and complex reporting structures. The performance of each NLP model examined was compared across four distinct tasks: 1) current procedural terminology (CPT) code classification, 2) pathologist classification, 3) report sign-out time regression, and 4) report text generation, under the hypothesis that models initialized on domain-relevant medical text would perform better than models not attuned to this prior knowledge. Our study highlights that the performance of deep learning-based NLP models can vary meaningfully across pathology-related tasks. Models pretrained on medical data outperform other models where medical domain knowledge is crucial, e.g., current procedural terminology (CPT) code classification. However, where interpretation is more subjective (i.e., teasing apart pathologist-specific lexicon and variable sign-out times), models with medical pretraining do not consistently outperform the other approaches. Instead, fine-tuning models pretrained on general or unrelated text sources achieved comparable or better results. Overall, our findings underscore the importance of considering the nature of the task at hand when selecting a pretraining strategy for NLP models in pathology. The optimal approach may vary depending on the specific requirements and nuances of the task, and related text sources can offer valuable insights and improve performance in certain cases, contradicting established notions about domain adaptation. This research contributes to our understanding of pretraining strategies for large language models and further informs the development and deployment of these models in pathology-related applications.

show abstract

“…The categorical variables comprise sex, patient class, service, primary physician, primary CPT, primary anesthesia type, location, and laterality. We reported that all the CPTs could be converted to relative value units (RVU) in a case length prediction model 14 . The RVU consists of 3 categories: physician work, practice expense, and professional liability.…”

Section: Methodsmentioning

confidence: 99%

Development, Deployment, and Implementation of a Machine Learning Surgical Case Length Prediction Model and Prospective Evaluation

Zaribafzadeh,

Webster,

Vail

et al. 2023

Annals of Surgery

Self Cite

View full text Add to dashboard Cite

Objective: To implement a machine learning model using only the restricted data available at case creation time to predict surgical case length for multiple services at different locations. Background: The operating room is one of the most expensive resources in a health system, estimated to cost $22 to $133 per minute and generate about 40% of hospital revenue. Accurate prediction of surgical case length is necessary for efficient scheduling and cost-effective utilization of the operating room and other resources. Methods: We introduced a similarity cascade to capture the complexity of cases and surgeon influence on the case length and incorporated that into a gradient-boosting machine learning model. The model loss function was customized to improve the balance between over-and under-prediction of the case length. A production pipeline was created to seamlessly deploy and implement the model across our institution. Results: The prospective analysis showed that the model output was gradually adopted by the schedulers and outperformed the schedulerpredicted case length from August to December 2022. In 33,815 surgical cases across outpatient and inpatient platforms, the operational implementation predicted 11.2% fewer underpredicted cases and 5.9% more cases within 20% of the actual case length compared with the schedulers and only overpredicted 5.3% more. The model assisted schedulers to predict 3.4% more cases within 20% of the actual case length and 4.3% fewer underpredicted cases. Conclusions: We created a unique framework that is being leveraged every day to predict surgical case length more accurately at case posting time and could be potentially utilized to deploy future machine learning models.

show abstract

CPT to RVU conversion improves model performance in the prediction of surgical case length

Cited by 8 publications

References 11 publications

Effect of 2 Teams and Operative Time on Complications After Oral Cavity Free Flap Reconstruction

Effect of 2 Teams and Operative Time on Complications After Oral Cavity Free Flap Reconstruction

Assessing the Impact of Pretraining Domain Relevance on Large Language Models Across Various Pathology Reporting Tasks

Development, Deployment, and Implementation of a Machine Learning Surgical Case Length Prediction Model and Prospective Evaluation

Contact Info

Product

Resources

About