Medicare Loses Billions to Billing Errors

Champagnie, Samantha J.

doi:10.2139/ssrn.3454108

Cited by 4 publications

(12 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, several studies suggest that undercoding is more prevalent in visits for established patients. 8 Third, patients with two or more chronic conditions and a greater level of medical complexity are more likely to be seen for Level 3 or Level 4 encounters.…”

Section: Methodsmentioning

confidence: 99%

“…17 Finally, the assumption that one-third of office visits are potentially undercoded is based on the work of Holt and colleagues (2010), but is supported by multiple studies. 8 , 11 Due to the use of publicly available, de-identified primary care data, Institutional Review Board or Ethics Board approval was not required for this study.…”

Section: Methodsmentioning

confidence: 99%

“…Undercoding, overcoding, and miscoding is far more rampant than most healthcare professionals, including pharmacists, may realize. 8 Undercoding or “downcoding” occurs when a billing provider intentionally codes a visit as less serious or extensive than it actually was. Conversely, overcoding or “upcoding” involves submitting codes for services that were never received or reporting a visit as more intensive than it actually was.…”

Section: Introductionmentioning

confidence: 99%

“…In our experience, physicians and health system coders undercode to avoid insurance rejections or audits or even to make visits more affordable for patients. 8 , 9 , 12 Though studies have found overcoding to be as commonplace as undercoding, we will focus on undercoding in the predictive model that follows since it correlates most directly with sacrificed revenue that could be used to support clinical pharmacy services. 8 …”

Section: Introductionmentioning

confidence: 99%

“… 8 , 9 , 12 Though studies have found overcoding to be as commonplace as undercoding, we will focus on undercoding in the predictive model that follows since it correlates most directly with sacrificed revenue that could be used to support clinical pharmacy services. 8 …”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

The Fermi problem: Estimation of potential Billing losses due to Undercoding of Florida Medicare data

Tenpas¹,

Dietrich²

2023

Exploratory Research in Clinical and Social Pharmacy

View full text Add to dashboard Cite

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

The Fermi problem: Estimation of potential Billing losses due to Undercoding of Florida Medicare data

Tenpas¹,

Dietrich²

2023

Exploratory Research in Clinical and Social Pharmacy

View full text Add to dashboard Cite

Using natural language processing for automated classification of disease and to identify misclassified ICD codes in cardiac disease

Falter,

Godderis,

Scherrenberg

et al. 2024

European Heart Journal - Digital Health

View full text Add to dashboard Cite

Introduction ICD-codes are used for classification of hospitalisations. The codes are used for administrative, financial and research purposes. It is known however that errors occur. Natural language processing (NLP) offers promising solutions for optimising the process. Objectives To investigate methods for automatic classification of disease in unstructured medical records using NLP and to compare these to conventional ICD coding. Methods Two datasets were used: the open-source MIMIC-III dataset (n = 55.177) and a dataset from a hospital in Belgium (n = 12.706). Automated searches using NLP algorithms were performed for the diagnoses “atrial fibrillation” and “heart failure”. Four methods were used: rule-based search, logistic regression, term frequency-inverse document frequency (TF-IDF), XGBoost and BioBERT. All algorithms were developed on the MIMIC-III dataset. The best performing algorithm was then deployed on the Belgian dataset. Results After pre-processing a total of 1.438 reports was retained in the Belgian dataset. XGBoost on TF-IDF matrix resulted in an accuracy of 0.94 and 0.92 for AF and HF respectively. There were 211 mismatches between algorithm and ICD codes. 103 were due to a difference in data availability or differing definitions. In the remaining 108 mismatches, 70% were due to incorrect labelling by the algorithm and 30% were due to erroneous ICD-coding (2% of total hospitalisations). Discussion and conclusion A newly developed NLP algorithm attained a high accuracy for classifying disease in medical records. XGBoost outperformed the deep learning technique BioBERT. NLP algorithms could be used to identify ICD-coding errors and optimise and support the ICD-coding process.

show abstract

Medical Diagnosis Coding Automation: Similarity Search vs. Generative AI

Klotzman

2024

Preprint

View full text Add to dashboard Cite

Objective: This study aims to predict ICD-10-CM codes for medical diagnoses from short diagnosis descriptions and compare two distinct approaches: similarity search and using a generative model with few-shot learning. Materials and Methods: The text-embedding-ada-002 model was used to embed textual descriptions of 2023 ICD-10-CM diagnosis codes, provided by the Cen-ters provided for Medicare & Medicaid Services. GPT-4 used few-shot learning. Both models underwent performance testing on 666 data points from the eICU Collaborative Research Database. Results: The text-embedding-ada-002 model successfully identified the relevant code from a set of similar codes 80 percent of the time, while GPT-4 achieved a 50 percent accuracy in predicting the correct code. Discussion: The work implies that text-embedding-ada-002 could automate medical coding better than GPT-4, highlighting potential limitations of generative language models for complicated tasks like this. Conclusion: The research shows that text-embedding-ada-002 outperforms GPT- 4 in medical coding, highlighting embedding models usefulness in the domain of medical coding.

show abstract

Medicare Loses Billions to Billing Errors

Cited by 4 publications

References 0 publications

The Fermi problem: Estimation of potential Billing losses due to Undercoding of Florida Medicare data

The Fermi problem: Estimation of potential Billing losses due to Undercoding of Florida Medicare data

Using natural language processing for automated classification of disease and to identify misclassified ICD codes in cardiac disease

Medical Diagnosis Coding Automation: Similarity Search vs. Generative AI

Contact Info

Product

Resources

About