CPT coding by interventional radiologists: a multi-institutional evaluation of accuracy and its economic implications

Automated vs. manual coding of neuroimaging reports via natural language processing, using the international classification of diseases, tenth revision

Objective

Natural language processing (NLP) can generate diagnoses codes from imaging reports. Meanwhile, the International Classification of Diseases (ICD-10) codes are the United States' standard for billing/coding, which enable tracking disease burden and outcomes. This cross-sectional study aimed to test feasibility of an NLP algorithm's performance and comparison to radiologists' and physicians' manual coding.

Methods

Three neuroradiologists and one non-radiologist physician reviewers manually coded a randomly-selected pool of 200 craniospinal CT and MRI reports from a pool of >10,000. The NLP algorithm (Radnosis, VEEV, Inc., Minneapolis, MN) subdivided each report's Impression into "phrases", with multiple ICD-10 matches for each phrase. Only viewing the Impression, the physician reviewers selected the single best ICD-10 code for each phrase. Codes selected by the physicians and algorithm were compared for agreement.

Results

The algorithm extracted the reports' Impressions into 645 phrases, each having ranked ICD-10 matches. Regarding the reviewers' selected codes, pairwise agreement was unreliable (Krippendorff α = 0.39-0.63). Using unanimous reviewer agreement as "ground truth", the algorithm's sensitivity/specificity/F2 for top 5 codes was 0.88/0.80/0.83, and for the single best code was 0.67/0.82/0.67. The engine tabulated "pertinent negatives" as negative codes for stated findings (e.g. "no intracranial hemorrhage"). The engine's matching was more specific for shorter than full-length ICD-10 codes (p = 0.00582x10-3).

Conclusions

Manual coding by physician reviewers has significant variability and is time-consuming, while the NLP algorithm's top 5 diagnosis codes are relatively accurate. This preliminary work demonstrates the feasibility and potential for generating codes with reliability and consistency. Future works may include correlating diagnosis codes with clinical encounter codes to evaluate imaging's impact on, and relevance to care.

Society of Urologic Prosthetic Surgeons' Coding and Billing Position Statement on Ancillary/Adjunct Penile Prosthesis Surgical Procedures: Part I

OBJECTIVE

To create a society position statement on common adjunct penile prosthesis (PP) procedures. While the Medicare Current Procedural Terminology code book lists descriptions of procedures, it is very brief and lacks detail in the small subspecialty of prosthetic urology. At educational/research meetings, wide variation was found in how experts in prosthetic urology code the same procedures, and need for a standardized format in billing common ancillary surgery was voiced.

METHODS

A subcommittee within the Society of Urologic Prosthetic Surgeons developed a survey assessing coding options for several procedures commonly adjunct to PP placement, which was distributed in the fall of 2022. The results of the survey were used to develop consensus statements on coding adjunct PP procedures; statements were distributed among society membership and meetings for approval.

RESULTS

Thirty members replied to the survey; demographics were obtained as follows: 73% were trained in a fellowship, 50% identified as university/academic practitioners, and 50% in community/private practice; and 63% respondents place more than 50 implants annually. Only 1 of the 30 respondents stated confidence in coding for these ancillary procedures. Specifically, differences in how to code curvature correction procedures were observed throughout the survey results.

CONCLUSION

Only 1 in 30 prosthetic urologists expressed confidence in coding and billing of adjunct PP procedures, further confirming the need for a society position statement. Therefore, we generated a consensus society position statement on common surgeries that are adjunct to PP placement.

Artificially Intelligent Billing in Spine Surgery: An Analysis of a Large Language Model

STUDY DESIGN

Retrospective cohort study.

OBJECTIVES

This study assessed the effectiveness of a popular large language model, ChatGPT-4, in predicting Current Procedural Terminology (CPT) codes from surgical operative notes. By employing a combination of prompt engineering, natural language processing (NLP), and machine learning techniques on standard operative notes, the study sought to enhance billing efficiency, optimize revenue collection, and reduce coding errors.

METHODS

The model was given 3 different types of prompts for 50 surgical operative notes from 2 spine surgeons. The first trial was simply asking the model to generate CPT codes for a given OP note. The second trial included 3 OP notes and associated CPT codes to, and the third trial included a list of every possible CPT code in the dataset to prime the model. CPT codes generated by the model were compared to those generated by the billing department. Model evaluation was performed in the form of calculating the area under the ROC (AUROC), and area under precision-recall curves (AUPRC).

RESULTS

The trial that involved priming ChatGPT with a list of every possible CPT code performed the best, with an AUROC of .87 and an AUPRC of .67, and an AUROC of .81 and AUPRC of .76 when examining only the most common CPT codes.

CONCLUSIONS

ChatGPT-4 can aid in automating CPT billing from orthopedic surgery operative notes, driving down healthcare expenditures and enhancing billing code precision as the model evolves and fine-tuning becomes available.

Pediatric Limb Lengthening and Reconstruction Surgical Coding Survey Results

BACKGROUND

In surgical specialties like orthopaedics, documenting the surgery performed involves applying the appropriate current procedural terminology (CPT) code(s). For limb reconstruction surgeons, the wide-ranging types of surgeries and rapid evolution of the field create a variety of factors making it difficult to code the procedures. We sought to (1) assess whether appropriate limb reconstruction codes currently exist and (2) determine whether there is agreement among experienced pediatric orthopaedic surgeons when applying these codes to similar cases.

METHODS

A REDCAP survey comprised of 10 common pediatric limb reconstruction cases was sent to experienced pediatric limb reconstruction surgeons in the United States. Based on the description of each case, the surgeons were asked to code the cases as they usually would in their practice. There were no limitations regarding the number or the types of codes each surgeon could choose to apply to the case. Nine additional demographic and general coding questions were asked to gauge the responding surgeon's coding experience.

RESULTS

Survey participants used various codes for each case, ranging from only 1 code to a maximum of 9 codes to describe a single case. The average number of codes per case ranged from 1.2 to 3.6, with an average of 2.5 among all 10 cases. The total number of unique codes provided by the respondents for each case ranged from 5 to 20. Only 3 of the 10 cases had an agreement >75% for any single code, and only 2 of the 10 cases had >50% agreement on any combination of 2 codes.

CONCLUSIONS

There are dramatic variations in coding methods among pediatric orthopaedic limb reconstruction surgeons. This information highlights the need to improve the current CPT coding landscape. Possible solutions include developing new codes that better represent the work done, developing standardized guidelines with the existing codes to decrease variation, and improving CPT coding education by developing limb reconstruction coding "champions."

LEVEL OF EVIDENCE

Level V.

Use of intraoperative MRI for resection of intracranial tumors: A nationwide analysis of short-term outcomes

OBJECTIVE

Recent evidence supports the use of intraoperative MRI (iMRI) during resection of intracranial tumors due to its demonstrated efficacy and clinical benefit. Though many single-center investigations have been conducted, larger nationwide outcomes have yet to be characterized.

METHODS

We used the American College of Surgeons National Surgical Quality Improvement Program database to examine baseline characteristics and 30-day postoperative outcomes among patients undergoing craniotomy for tumor resection with and without iMRI. Comparisons between outcomes were accomplished after propensity matching using chi-square tests for categorical variables and Welch two-sample t-tests for continuous variables.

RESULTS

A total of 38,003 patients met inclusion criteria. Of this population, 54 (0.1%) received iMRI, while 37,949 (99.9%) did not receive iMRI. After propensity score matching, the resulting groups consisted of an iMRI group (n = 54) and a matched non-iMRI group (n = 54). Procedures involving iMRI were associated with significantly increased operation length compared to those without (p < 0.01). Length of hospital stay was higher in patients without iMRI, with this difference trending towards significance (p = 0.05) in the unmatched comparison. Patients undergoing craniotomy without iMRI had a higher rate of readmission (p = 0.04). There was no significant difference in occurrence of other adverse events between the two patient groups.

CONCLUSION

Despite increasing operative length, iMRI is not associated with higher infection rate and may have a clinical benefit associated with reducing readmissions and a trend towards reducing inpatient length of stay. Additional nationwide analyses including more iMRI patients would provide further insight into the strength of these findings.

Neural Machine Translation-Based Automated Current Procedural Terminology Classification System Using Procedure Text: Development and Validation Study

BACKGROUND

Administrative costs for billing and insurance-related activities in the United States are substantial. One critical cause of the high overhead of administrative costs is medical billing errors. With advanced deep learning techniques, developing advanced models to predict hospital and professional billing codes has become feasible. These models can be used for administrative cost reduction and billing process improvements.

OBJECTIVE

In this study, we aim to develop an automated anesthesiology current procedural terminology (CPT) prediction system that translates manually entered surgical procedure text into standard forms using neural machine translation (NMT) techniques. The standard forms are calculated using similarity scores to predict the most appropriate CPT codes. Although this system aims to enhance medical billing coding accuracy to reduce administrative costs, we compare its performance with that of previously developed machine learning algorithms.

METHODS

We collected and analyzed all operative procedures performed at Michigan Medicine between January 2017 and June 2019 (2.5 years). The first 2 years of data were used to train and validate the existing models and compare the results from the NMT-based model. Data from 2019 (6-month follow-up period) were then used to measure the accuracy of the CPT code prediction. Three experimental settings were designed with different data types to evaluate the models. Experiment 1 used the surgical procedure text entered manually in the electronic health record. Experiment 2 used preprocessing of the procedure text. Experiment 3 used preprocessing of the combined procedure text and preoperative diagnoses. The NMT-based model was compared with the support vector machine (SVM) and long short-term memory (LSTM) models.

RESULTS

The NMT model yielded the highest top-1 accuracy in experiments 1 and 2 at 81.64% and 81.71% compared with the SVM model (81.19% and 81.27%, respectively) and the LSTM model (80.96% and 81.07%, respectively). The SVM model yielded the highest top-1 accuracy of 84.30% in experiment 3, followed by the LSTM model (83.70%) and the NMT model (82.80%). In experiment 3, the addition of preoperative diagnoses showed 3.7%, 3.2%, and 1.3% increases in the SVM, LSTM, and NMT models in top-1 accuracy over those in experiment 2, respectively. For top-3 accuracy, the SVM, LSTM, and NMT models achieved 95.64%, 95.72%, and 95.60% for experiment 1, 95.75%, 95.67%, and 95.69% for experiment 2, and 95.88%, 95.93%, and 95.06% for experiment 3, respectively.

CONCLUSIONS

This study demonstrates the feasibility of creating an automated anesthesiology CPT classification system based on NMT techniques using surgical procedure text and preoperative diagnosis. Our results show that the performance of the NMT-based CPT prediction system is equivalent to that of the SVM and LSTM prediction models. Importantly, we found that including preoperative diagnoses improved the accuracy of using the procedure text alone.

Arteriovenous Vascular Access-Related Procedural Burden Among Incident Hemodialysis Patients in the United States

RATIONALE & OBJECTIVE

As the proportion of arteriovenous fistulas (AVFs) compared with arteriovenous grafts (AVGs) in the United States has increased, there has been a concurrent increase in interventions. We explored AVF and AVG maturation and maintenance procedural burden in the first year of hemodialysis.

STUDY DESIGN

Observational cohort study.

SETTING & PARTICIPANTS

Patients initiating hemodialysis from July 1, 2012, to December 31, 2014, and having a first-time AVF or AVG placement between dialysis initiation and 1 year (N = 73,027), identified using the US Renal Data System (USRDS).

PREDICTORS

Patient characteristics.

OUTCOME

Successful AVF/AVG use and intervention procedure burden.

ANALYTICAL APPROACH

For each group, we analyzed interventional procedure rates during maturation maintenance phases using Poisson regression. We used proportional rate modeling for covariate-adjusted analysis of interventional procedure rates during the maintenance phase.

RESULTS

During the maturation phase, 13,989 of 57,275 patients (24.4%) in the AVF group required intervention, with therapeutic interventional requirements of 0.36 per person. In the AVG group 2,904 of 15,572 patients (18.4%) required intervention during maturation, with therapeutic interventional requirements of 0.28 per person. During the maintenance phase, in the AVF group 12,732 of 32,115 patients (39.6%) required intervention, with a therapeutic intervention rate of 0.93 per person-year. During maintenance phase, in the AVG group 5,928 of 10,271 patients (57.7%) required intervention, with a therapeutic intervention rate of 1.87 per person-year. For both phases, the intervention rates for AVF tended to be higher on the East Coast while those for AVG were more uniform geographically.

LIMITATIONS

This study relies on administrative data, with monthly recording of access use.

CONCLUSIONS

During maturation, interventions for both AVFs and AVGs were relatively common. Once successfully matured, AVFs had lower maintenance interventional requirements. During the maturation and maintenance phases, there were geographic variations in AVF intervention rates that warrant additional study.

Reimbursement for Hip Fractures: The Impact of Varied Current Procedural Terminology Coding Using Relative Value Units

BACKGROUND

Many orthopedic practices routinely code hip fracture hemiarthroplasty as Current Procedural Terminology (CPT) 27125 even though 27236 is the correct CPT code. Our objective is to determine the financial impact this simple mistake has on surgeon reimbursement.

METHODS

Our data comprised cases assigned International Classification of Diseases, Tenth Revision code S72.001A through S72.035A and CPT code 27125 or 27236 within the American College of Surgeons National Surgical Quality Improvement Program 2016-2017 database. Relative value units (RVUs) per CPT code and the Centers for Medicare and Medicaid Services reported that RVU conversion factor of $36.0896 per 1 RVU was used to calculate reimbursement per case. The dollar difference and percent difference per case was then calculated between cases assigned CPT code 27125 and those assigned 27236.

RESULTS

Our total sample consisted of 12,287 National Surgical Quality Improvement Program cases. Of those, 4185 (34%) were cases of a hip fracture treated with hemiarthroplasty that were incorrectly coded as CPT code 27125. That error in coding results in a decrease in reimbursement of $35.01 per case, a 5.51% difference.

CONCLUSION

Since the current healthcare reimbursement model relies solely on CPT codes to determine RVUs, it is imperative that orthopedic surgeons understand the financial impact of incorrect coding. Although correct coding of hemiarthroplasty procedures for hip fractures is an easy task to fix in the future, we hope that through this study a greater emphasis is placed on coding in orthopedic surgery.

A Descriptive Revenue Analysis of a Wound-Center IR Collaboration to Treat Lower Extremity Venous Ulcers

PURPOSE

To describe the revenue from a collaboration between a dedicated wound care center and an interventional radiology (IR) practice for venous leg ulcer (VLU) management at a tertiary care center.

MATERIALS AND METHODS

This retrospective study included 36 patients with VLU referred from a wound care center to an IR division during the 10-month active study period (April 2017 to January 2018) with a 6-month surveillance period (January 2018 to June 2018). A total of 15 patients underwent endovascular therapy (intervention group), whereas 21 patients did not (nonintervention group). Work relative value units (wRVUs) and dollar revenue were calculated using the Centers for Medicare and Medicaid Services Physician Fee Schedule.

RESULTS

Three sources of revenue were identified: evaluation and management (E&M), diagnostic imaging, and procedures. The pathway generated 518.15 wRVUs, translating to $37,522. Procedures contributed the most revenue (342.27 wRVUs, $18,042), followed by E&M (124.23 wRVUs, $8,881), and diagnostic imaging (51.65 wRVUs, $10,599). Intervention patients accounted for 86.7% of wRVUs (449.48) and 80.0% of the revenue ($30,010). An average of 33 minutes (38.3 hours total) and 2.06 hours (36.8 hours total) were spent on E&M visits and procedures, respectively.

CONCLUSIONS

In this collaboration between the wound center and IR undertaken to treat VLU, IR and E&M visits generated revenue and enabled procedural and downstream imaging revenue.

From imaging to reimbursement: what the pediatric radiologist needs to know about health care payers, documentation, coding and billing

Medical coding and billing processes in the United States are complex, cumbersome and poorly understood by radiologists. Despite the direct implications of radiology documentation on reimbursement, trainees and practicing radiologists typically receive limited relevant training. This article summarizes the payer structure including the state-based Children's Health Insurance Programs, discusses the essential processes by which radiologists request and receive reimbursement, details the mechanisms of coding diagnoses using International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes and imaging services using Current Procedural Terminology (CPT) and Healthcare Common Procedure Coding System (HCPCS) codes, and explores reimbursement and coding-related issues specific to pediatric radiology. Appropriate documentation, informed by knowledge of coding, billing and reimbursement fundamentals, facilitates appropriate payment for clinically relevant services provided by pediatric radiologists.

Arteriovenous Fistula Maturation in Prevalent Hemodialysis Patients in the United States: A National Study

BACKGROUND

Arteriovenous fistulas (AVFs) are the preferred form of hemodialysis vascular access, but maturation failures occur frequently, often resulting in prolonged catheter use. We sought to characterize AVF maturation in a national sample of prevalent hemodialysis patients in the United States.

STUDY DESIGN

Nonconcurrent observational cohort study.

SETTING & PARTICIPANTS

Prevalent hemodialysis patients having had at least 1 new AVF placed during 2013, as identified using Medicare claims data in the US Renal Data System.

PREDICTORS

Demographics, geographic location, dialysis vintage, comorbid conditions.

OUTCOMES

Successful maturation following placement defined by subsequent use identified using monthly CROWNWeb data.

MEASUREMENTS

AVF maturation rates were compared across strata of predictors. Patients were followed up until the earliest evidence of death, AVF maturation, or the end of 2014.

RESULTS

In the study period, 45,087 new AVFs were placed in 39,820 prevalent hemodialysis patients. No evidence of use was identified for 36.2% of AVFs. Only 54.7% of AVFs were used within 4 months of placement, with maturation rates varying considerably across end-stage renal disease (ESRD) networks. Older age was associated with lower AVF maturation rates. Female sex, black race, some comorbid conditions (cardiovascular disease, peripheral artery disease, diabetes, needing assistance, or institutionalized status), dialysis vintage longer than 1 year, and catheter or arteriovenous graft use at ESRD incidence were also associated with lower rates of successful AVF maturation. In contrast, hypertension and prior AVF placement at ESRD incidence were associated with higher rates of successful AVF maturation.

LIMITATIONS

This study relies on administrative data, with monthly recording of access use.

CONCLUSIONS

We identified numerous associations between AVF maturation and patient-level factors in a recent national sample of US hemodialysis patients. After accounting for these patient factors, we observed substantial differences in AVF maturation across some ESRD networks, indicating a need for additional study of the provider, practice, and regional factors that explain AVF maturation.

Variation in National ACGME Case Log Data for Pediatric Orthopaedic Fellowships: Are Fellow Coding Practices Responsible?

BACKGROUND

The introduction of the 80-hour work week for Accreditation Council for Graduate Medical Education (ACGME) accredited fellowship programs initiated many efforts to optimize surgical training. One particular area of interest is on recording and tracking surgical experiences. The current standard is logging cases based on Current Procedural Terminology codes, which are primarily designed for billing. Proposed guidelines from the ACGME regarding logging exist, but their implementation is unknown, as is the variation in case volume across fellowship programs. The purpose of this study was to investigate variability in the national case log data, and explore potential sources of variation using fellow surveys.

METHODS

National ACGME case log data for pediatric orthopaedic fellowships from 2012 to 2015 were reviewed, with particular attention to the domains of spine, pelvis/hip, arthroscopy, trauma, and other (which includes clubfoot casting). To explore potential sources of case log variability, a survey on case logging behavior was distributed to all pediatric orthopaedic fellows for the academic year 2015 to 2016.

RESULTS

Reported experiences based on ACGME case logs varied widely between fellows with percentage difference of up to 100% in all areas. Similarly, wide variability is present in coding practices of pediatric orthopaedic fellows, who often lack formal education on the topic of appropriate coding/logging. In the survey, hypothetical case scenarios had an absolute difference in recorded codes of up to 13 and a percentage difference of up to 100%.

CONCLUSIONS

ACGME case log data for pediatric orthopaedic fellowships demonstrates wide variability in reported surgical experiences. This variability may be due, in part, to differences in logging practices by individual fellows. This observation makes meaningful interpretation of national data on surgical volume challenging. Proposed surgical experience minimums should be interpreted in light of these data, and may not be advisable unless accompanied by standardized and specific guidelines for case log entry. Efforts to optimize training in the post 80-hour era will require accurate data to serve as a starting point for future educational efforts.

Medicare Part B Intensity and Volume Offset

Under Medicare Part B, adjustments to the fee schedule are made under the assumption that physicians and hospitals make up for fee reductions through increased service provision called 'volume offsetting'. While historically, researchers have found evidence of volume offsetting, more recent studies have called into question its magnitude and existence. This study is the first to propose and empirically evaluate an alternative hypothesis of offsetting, namely the alteration of billed or provided services as a means of 'intensity offsetting'. Evaluating both forms of offsetting, it finds strong evidence of intensity offsetting and little to no evidence of volume offsetting. Simulating a 10% reduction in the Medicare fee schedule, this study estimates that across different procedures between 22% and 59% of a fee reduction will be offset through alterations in service intensity.

Discordance in current procedural terminology coding for foot and ankle procedures between residents and attending surgeons

OBJECTIVE

Because of the importance of current procedural terminology (CPT) coding in both resident education evaluation and practice management, this study was undertaken to evaluate the correlation and interrater reliability between residents and attending physicians in CPT coding for orthopedic foot and ankle surgeries as well as to determine attending surgeons' and residents' familiarity with and confidence in the coding process.

METHODS

CPT codes from resident case logs were compared with those submitted by attending surgeons, and Pearson's correlation coefficient and interrater reliability were calculated to examine coding congruency. An online survey was also used to examine attending surgeon and resident perceptions and habits regarding CPT codes and the coding process.

RESULTS

CPT codes recorded by 20 residents (1164) were compared with those recorded by 3 attending foot and ankle surgeons (1259). Correlation between attending and resident codes was poor (r = -0.015). Interrater reliability demonstrated a kappa value of 0.04, indicating poor agreement. Compared with attending CPT coding, residents concordantly coded 42% of the time, with an individual resident range from 2% to 65%. Additionally, 43% of residents reported being uncomfortable about foot and ankle CPT coding, and they reported rarely or never discussing CPT codes with attending surgeons in the perioperative period.

CONCLUSIONS

Resident and attending surgeon concordance in CPT coding for foot and ankle procedures is poor, and residents have a low level of confidence in logging CPT codes, possibly because of a lack of training and preparation in coding. Because CPT coding is used not only for practice management but also has implications for evaluating institutions by accreditation bodies, educational initiatives to improve resident confidence and accuracy with CPT coding may be warranted.

Researching the reliability of accreditation survey teams: lessons learnt when things went awry

Accreditation of health organisations, occurring in over 70 countries, is predicated upon the reliability of survey teams judgements, but we do not know the extent to which survey teams are reliable. To contribute evidence to this issue, we investigated the reliability of two survey teams simultaneously assessing an organisation. The setting was a large Australian teaching hospital, and data were derived from interviews, observations and survey documents. Participants were from four groups: hospital staff, accreditation agency personnel and surveyors, and research staff. Thematic analysis was employed to identify significant factors that influenced the study. The two survey teams ratings and recommendations demonstrated high levels of agreement. However, while a common understanding of the study existed, the research was compromised. There were difficulties enacting the study. Contrary to negotiated arrangements, the pressure of the study resulted in surveyors discussing evidence and their interpretation of standards. Uncontrollable circumstances (late changes of personnel), and unexpected events (a breakdown of working relationships), challenged the study. The twin lessons learnt are that a consistent survey outcome is likely to be reached when reliability of process and consistent application of standards are pursued, and research requires negotiating challenges and relationships.

Accuracy of administrative claims data for polypectomy

BACKGROUND

The frequency of polypectomy is an important indicator of quality assurance for population-based colorectal cancer screening programs. Although administrative databases of physician claims provide population-level data on the performance of polypectomy, the accuracy of the procedure codes has not been examined. We determined the level of agreement between physician claims for polypectomy and documentation of the procedure in endoscopy reports.

METHODS

We conducted a retrospective cohort study involving patients aged 50-80 years who underwent colonoscopy at seven study sites in Montréal, Que., between January and March 2007. We obtained data on physician claims for polypectomy from the Régie de l'Assurance Maladie du Québec (RAMQ) database. We evaluated the accuracy of the RAMQ data against information in the endoscopy reports.

RESULTS

We collected data on 689 patients who underwent colonoscopy during the study period. The sensitivity of physician claims for polypectomy in the administrative database was 84.7% (95% confidence interval [CI] 78.6%-89.4%), the specificity was 99.0% (95% CI 97.5%-99.6%), concordance was 95.1% (95% CI 93.1%-96.5%), and the kappa value was 0.87 (95% CI 0.83-0.91).

INTERPRETATION

Despite providing a reasonably accurate estimate of the frequency of polypectomy, physician claims underestimated the number of procedures performed by more than 15%. Such differences could affect conclusions regarding quality assurance if used to evaluate population-based screening programs for colorectal cancer. Even when a high level of accuracy is anticipated, validating physician claims data from administrative databases is recommended.

Evolving roles of radiologists, nephrologists, and surgeons in endovascular hemodialysis access maintenance procedures

PURPOSE

The aim of this study was to evaluate the changing relative roles of radiologists, nephrologists, and surgeons performing endovascular hemodialysis access maintenance procedures.

METHODS

Medicare Physician Supplier Procedure Summary Master Files from 2001 through 2008 were analyzed for procedure codes for hemodialysis access angiography, angioplasty, percutaneous thrombectomy, and open surgical interventions. Using physician specialty code data, component procedure volume for all 3 endovascular services was extracted for radiologists, nephrologists, and surgeons. Percentage changes were calculated for all groups. National trends in percutaneous and open interventions were compared.

RESULTS

Between 2001 and 2008, the total Medicare fee-for-service component procedure volume for dialysis access angiography, angioplasty, and percutaneous thrombectomy increased by 102%, 171%, and 52%, respectively. In 2008, radiologists performed 50% of angiography, 47% of angioplasty, and 46% of declotting procedures, down from 82%, 82%, and 84%, respectively, in 2001. In contrast, nephrologists increased from 4%, 5%, and 4% to 22%, 27%, and 21% of services, and surgeons increased from 7%, 5%, and 4% to 22%, 19%, and 16%. As percutaneous procedures increased in frequency, open surgical interventions declined by 43%.

CONCLUSION

Nationally, endovascular hemodialysis access maintenance procedures have increased as open surgical interventions have declined. Nephrologists and surgeons have both experienced marked relative increases in endovascular procedure volumes as radiologists, previously by far the predominant providers of these services, now only perform approximately half.

National fluid shifts: fifteen-year trends in paracentesis and thoracentesis procedures

PURPOSE

The aim of this study was to evaluate national trends in paracentesis and thoracentesis procedures and the relative roles of specialty groups providing these services.

METHODS

Medicare Physician Supplier Procedure Summary Master Files from 1993 to 2008 were analyzed for paracentesis and thoracentesis procedure codes. Using physician specialty identifier codes, procedure volumes were extracted for radiologists, primary care physicians, and surgeons for both procedures. Volume data were extracted for gastroenterologists and pulmonary and critical care medicine physicians, respectively, for paracentesis and thoracentesis. Frequency by site of service was similarly evaluated. Relative changes were calculated.

RESULTS

Between 1993 and 2008, paracentesis procedures on Medicare fee-for-service beneficiaries increased by 133% (from 64,371 to 149,699), and thoracentesis procedures decreased by 14% (from 147,363 to 127,444). Services by radiologists increased by 964% (from 10,456 to 111,275) and 358% (from 14,531 to 66,602), respectively, while all other targeted groups experienced declines. For paracentesis, radiologist and gastroenterologist procedure shares changed from 16% and 32%, respectively, in 1993 to 74% and 6% in 2008. For thoracentesis, radiologist and pulmonary and critical care medicine physician shares changed from 10% and 49% to 52% and 27%. Relative shifts in site of service to the hospital outpatient setting occurred for both procedures.

CONCLUSIONS

Since 1993, paracentesis procedures on Medicare beneficiaries have more than doubled, while thoracentesis volumes have declined slightly. Radiologists now far exceed gastroenterologists and pulmonary and critical care medicine physicians, respectively, as the predominant providers of these services. Those shifts are likely attributable to both the incremental safety of imaging guidance and also the unfavorable economics of these procedures.

Change in use of gadolinium-enhanced magnetic resonance studies in kidney disease patients after US Food and Drug Administration warnings: a cross-sectional study of Veterans Affairs Health Care System data from 2005-2008

BACKGROUND

Exposure to gadolinium in patients with kidney disease has been linked to risk of developing nephrogenic systemic fibrosis. The US Food and Drug Administration (FDA) has issued warnings against the use of gadolinium in this population. We studied the impact of these warnings on the use of gadolinium-enhanced magnetic resonance (GE-MR) studies in patients with decreased estimated glomerular filtration rate (eGFR) and the practice of measuring serum creatinine before gadolinium exposure.

STUDY DESIGN

Cross-sectional study of patients who had undergone MR studies from October 2002 to September 2008.

SETTING & PARTICIPANTS

Patients receiving medical care in the US Department of Veterans Affairs Health Care System.

PREDICTOR

Date of MR imaging, serum creatinine level, and eGFR using the 4-variable Modification of Diet in Renal Disease (MDRD) Study equation.

OUTCOMES & MEASUREMENTS

The rate of MR studies performed with and without gadolinium from July 2005 to September 2008 in patients with different stages of kidney disease, defined using eGFR. The proportion of GE-MR studies with a screening serum creatinine level.

RESULTS

There was a 71% decrease in the rate of GE-MR use in patients with GFR<30 mL/min/1.73 m2 2 years after the release of the first public health advisory, although studies continued to be performed in patients with stages 4 and 5 chronic kidney disease. The proportion of GE-MR studies with serum creatinine measured within 1 month before the study increased by 99%.

LIMITATIONS

Data available up to September 30, 2008. Indications for the GE-MR studies were not assessed. The accuracy of Current Procedural Terminology and International Classification of Diseases, Ninth Revision coding was not assessed.

CONCLUSION

There was a large decrease in the use of GE-MR studies in patients with GFR<30 mL/min/1.73 m2 and a large but not universal increase in the practice of measuring serum creatinine before GE-MR after the release of the FDA warnings.