In what other profession can one be as ill-prepared to understand the complexity of appropriately billing for services as in medicine? After years of intense study of the clinical minutia of nephrology and passing multiple examinations culminating in board certification, nephrologists enter practice generally naive of how to appropriately document, code, and bill–which are the life blood of operating a successful practice and earning an income. Add to this the complex and growing body of regulations, an increase in the required coding specificity and payer requirements, and we have a formula for physician frustration, burn-out, and failure.

Further, since physicians do not readily admit to failure, many often simply relegate coding and billing as “unimportant” and focus on caring for patients.

Lean on (others)

After completing fellowship training, nephrologists typically “learn” piecemeal about coding and billing from colleagues or maybe, if ambitious, go to a training seminar. Nephrologists frequently depend on their practice’s “experts” who assist with billing, but often they too are unable to keep up with the increasing level of complexity. Try as they may, even the best practice staff are often not able to provide the level of expertise necessary in today’s changing and challenging practice environment.

Nephrologists can also be intimated by the alphabet soup of payment-related regulations: ICD-10, PQRS, MU, MACRA, MIPS, and APM.

All this confusion can also put a nephrologist in legal danger. The Centers for Medicare & Medicaid Services places a significant focus on fraud and abuse regulations concerning submission of physician billing, coding and documentation. Over-documentation as well as under-documentation are equally considered fraudulent and may increase your practice’s risk. Getting your coding and billing “right” the first time can save your practice some time, resources, and money.

A test of knowledge

With this in mind, we analyzed the coding accuracy of nephrologists that had signed up for education classes 1  from May 1, 2014  to Feb. 28, 2017. Physician vintage (years of experience), coding accuracy (based on physician documentation) and risk were all reviewed. “At risk” codes were defined as evaluation/management (E/M) levels that were deemed to be at audit risk due to frequency of use by the nephrologist, i.e., if the nephrologist always billed a certain code such as 99214 for an established office visit.

The analysis focused on comparing how each nephrologist coded prior to receiving education and their coding accuracy following continuous education.


We evaluated data from 97 physicians that work in general nephrology practices across the United States. These nephrologists had different years of practice experience (see Figure 1).

Nephrology coding and billing data was analyzed by years of experience (vintage), and grouped into categories for nephrologists with 1-5 years, 6-9 years, and 10+ years of experience respectively. We reviewed the most common e/m codes used for nephrology (see Table 1) and areas that auditors were currently monitoring.

Education was provided to nephrologists via a web-ex and phone conferences, reviewing both areas where coding was accurate as well as areas where incorrect coding occurred based on the nephrologist’s documentation.

During education sessions, the nephrologist is shown specifically where the errors occurred either in their documentation and/or in their coding of an appropriate E/M level.

Depending on how the nephrologist scored on their initial coding review, a follow-up coding review is performed either at 30, 60, 90, or 120 days later (with more frequent coding reviews if the nephrologist had a lower coding accuracy percentage). After each coding review, education is provided again, focusing on where each nephrologist’s errors occurred and what was needed to document and code correctly.


Continuous education following coding reviews improved the accuracy of coding by nephrologists. Nephrologists with 1-5 years of experience had the smallest difference between their initial coding accuracy pre-education and following continuous education, but even this group demonstrated an improvement of 6%. Nephrologists with 6-9 years of experience showed the greatest improvement, moving from a 67% initial coding accuracy to 85% following education—a 27% improvement. Nephrologists with 10+ years of experience had an improvement of 12% in their coding accuracy following education (see Figure 2).

Details of the types of coding errors varied and general findings are summarized below in Table 2.

Overall across all levels of experience, the initial accuracy on coding review averaged 73% vs. a post-education average of 85%, resulting in an improvement, on average, of 16%.

In terms of reviewing coding accuracy by type of code, specific discrepancies were found in codes with high medical decision-making requirements (e.g. codes 99205, 99215, 99223, 99233, and 99291).

Improvements in coding accuracy were noted for all physician experience cohorts and for all codes. While such improvements were relatively consistent for four months, they deteriorated by month 6 (see Figure 3). Thus, as the period extends between coding reviews, nephrologists may revert to older patterns of less focus on coding accuracy and appropriate documentation.


We found improvement in all nephrologist vintage/experience cohorts following coding reviews accompanied by focused education. Nephrologists with 1-5 years’ experience showed the best initial coding review results, while nephrologists with 6-9 years’ experience having the worst initial coding review results and those nephrologists with >10 years’ experience had initial coding review results in between them.

Given that nephrologists with the least experience had the most accurate initial results, one wonders if their typically lower clinical burden at the earlier stage in their practice allows them to be more thoughtful when documenting and coding or if this group’s proximity to completion of their fellowship produces a generally higher level of documentation. These conclusions are supported by our data that this cohort is most often over-documenting as we reviewed their detailed coding performance.

Nephrologists with 6-9 years’ experience are typically at the peak of their practice productivity, excessively busy and/or distracted, and may have formed bad habits and therefore may score lower on initial coding review accuracy. Our data that this cohort is generally under-documented/over coded as we reviewed their specific performance supports this theory.

Finally, nephrologists with 10+ years’ experience have initial coding review accuracy falling between the other cohorts, which may suggest some “wisdom” of seniority. This cohort may have a combination of the luxury of being somewhat less busy with clinical activities, have more overall experience and understanding of appropriate documentation, coding and billing, and value its importance to the business success of their practice. This experienced group, similar to the cohort with 6-9 years’ experience, typically had errors resulting from under documentation and/or overcoding services, suggesting that continued monitoring also yields improvement in coding accuracy even among the most experienced nephrologists.


Coding reviews can help nephrologists assess the accuracy and completeness of the documentation of services provided as well as to determine if that documentation supports the claims ultimately submitted for payment. It is critical to the success of a nephrology practice to be accurately reimbursed for services provided.

Our data suggests that when presented with coding review results and applicable education, all nephrologist cohorts showed improvement in their coding accuracy. But this improvement declined after six months (if coding reviews and education did not continue).

Nephrologists who receive continuous reviews, monitoring and education on their coding accuracy and appropriate documentation maintain a higher coding accuracy profile than those without such monitoring.

Given changing reimbursement and regulations as well as the increased scrutiny by Medicare and other payers, a structured coding review appears an appropriate process improvement measure for nephrology practices and may minimize a practice’s audit risk. Further, given the complexity of such regulations and the day-to-day demands of running a practice, nephrologists may elect to retain an independent expert to provide an objective review of performance and expertise that an individual practice may not have available. Finally, such a structured and on-going coding review process may also help with a practice’s effective OIG mandated compliance program.


  1. Nephrology Practice Solutions (