EDITOR’S NOTE: This is the first installment of a four-part series that explores the past, present and future of clinical documentation integrity (CDI).

To understand where we are as a clinical documentation integrity (CDI) profession, it is important to understand how and why the profession began.

In the following months, I will be examining many of the metrics currently used to evaluate the “success” of a CDI professional and CDI department, through a series of articles. As we examine the origins and original intent of these metrics, it may help those who lead CDI departments to better understand the need to rethink how we measure success – and if it is really necessary to continue to measure a financial “return on investment.” The CDI profession has come a long way from where we started, so it amazes me that most CDI departments are still referred to as CDI “programs,” and continue to use the same metrics as when the program was launched. Our metrics must evolve as the role of the CDI professional evolves, and our metrics must accurately reflect the impact of our work by capturing the complexity of healthcare reimbursement as a reflection of the acuity of our patient population.

The origins of the CDI profession can be traced back to the implementation of the Centers for Medicare & Medicaid Services (CMS) Diagnostic-Related Group (DRG) system back in 1983. It was during this time that CMS drastically changed its reimbursement process, creating a relationship between coded claims data and reimbursement. The DRG Prospective Payment System (PPS) was crafted in response to a congressional mandate to control medical costs associated with caring for the elderly and disabled. Prior to implementation of the DRG system, Medicare made interim payments to hospitals throughout the year, and reconciled those payments with the hospital’s “allowable costs,” as detailed on a cost report filed each fiscal year. Implementation of the DRG system resulted in a per-case reimbursement mechanism wherein Medicare paid a flat rate for each inpatient hospital claim, in hopes of promoting efficiency, thereby keeping healthcare costs down.

The DRG PPS classified “all human disease according to affected organ system, surgical procedures performed on patients, morbidity, and sex of the patient … accounting for up to eight diagnoses in addition to the primary diagnosis.” It is a classification scheme composed of classes of clinically similar patients who medically would be expected to consistently use a similar amount of hospital resources. DRGs were arranged by physician panels into Major Diagnostic Categories (MDCs) to support clinical coherence. Each MDC was constructed to correspond to a major organ system, when possible. Medical classes usually include a class for neoplasms, symptoms, specific conditions relating to the organ system involved, and “other medical disease.”

Because surgical procedures that require the use of an operating room have a significant effect on hospital resource consumption, the MDCs were further subdivided into medical and surgical groups. To address the possibility of a patient having multiple procedures related to their principal diagnosis, and because a patient can only be assigned to one surgical class, the surgical classes in each MDC were defined in a hierarchical order. Patients are assigned to the highest surgical class in an MDC hierarchy. Like within the medical class, there is a “other surgical procedures” class to account for infrequent procedures that are reasonably expected to be performed for a patient in a particular MDC. The goal of the MDC structure is to clinically align the principal diagnosis and procedure, but that is not always possible, so the DRG system also includes a surgical class referred to as “unrelated operating room procedures.”

Most coding departments have a dedicated work queue to monitor cases that result in each DRG, to validate the assigned codes and their sequencing. It is important to note that not all “procedures” result in a surgical DRG. Procedures that are diagnostic in nature will not result in a surgical DRG. For example, a cardiac catheterization only results in a DRG change when related to the affected organ system, e.g., the cardiovascular system, but it remains a medical DRG.

Both medical and surgical DRGs were further subdivided based on the presence of a complication and comorbidity (CC) or pediatric age (0-17). There were as many as 115 pairs of DRGs that could be impacted by the presence of a CC. In addition the principal diagnosis, up to eight additional diagnoses could be reported to reflect “other factors affecting the patient’s care or treatment.” However, it only takes one diagnosis classified as a CC to “move” the DRG for maximum reimbursement under the DRG reimbursement mechanism. The DRG system created a way for hospitals to differentiate themselves from each other by demonstrating that they are caring for patients who require more resources than the “average” patient. The CC list remained virtually unchanged for almost 24 years, except for the addition of new diagnosis codes like HIV. The stability of the CC list made it easy to incorporate identifying these conditions into the coding workflow. In fact, by the mid 2000s, nearly 80 percent of patients had CCs, in part due to better coding of secondary diagnoses, reducing the power of the DRG system to discriminate hospital resource use.

CMS is required under the Social Security Act to adjust the DRG classifications, a process referred to as reclassification, and adjust relative weights, referred to as recalibration, annually. Changes go into effect each Oct. 1, the start of the federal fiscal year. These annual adjustments are necessary to adequately compensate for costs under the PPS. CMS has refined the DRG system in response to issues like overpayments for patient transfers, improper payments to hospitals for nonphysician outpatient services, and monitoring of DRG upcoding. Examples of how these issues were addressed include the 72-hour rule, whereby related outpatient services delivered in the three days before admission are included in the payment for the inpatient stay, and transfer DRGs result in reduced payment for shortened length of stay due to transfer of the patient to post-acute care.

I started my career as a CDI professional in 2008, which is when the CDI profession really started to gain traction within the healthcare industry. I entered the field as the manager of a CDI program at an academic medical center, with the understanding that there was no guarantee that CDI was here to stay. In fact, consider the use of the word “program,” emphasizing that we were not a department, like coding or quality, because if we did not show a return on investment (ROI), then the “experimental” program would be abolished. I’ll speak more about the concept of ROI in a later article, when I discuss CDI metrics related to performance, but first, let’s continue examining how the CDI profession was created in response to the DRG payment mechanism.

Why did the growth of CDI occur in 2008? It’s because CMS updated the DRG payment mechanism to the Medicare Severity (MS)-DRG (version 25 DRGs) beginning with the 2008 fiscal year. First, MS-DRGs no longer included pediatric age distinctions. Secondly, the CC list was completely revised, with implementation of MS-DRGs to reflect significant acute disease, acute exacerbations of significant chronic disease, advanced or end stage chronic disease, and chronic disease associated with extensive disability. The revision of the CC list resulted in an estimated decrease in Medicare patients with CCs from 80 to 40 percent. Lastly, the MS-DRG system also classified diagnoses as major complications or comorbidities (MCCs) based on relative resource use. At the time, approximately 12 percent of all diagnosis codes were classified as a MCC, plus 24 percent as a CC and 65 percent as a non-CC; of course, this has changed over time, especially with implementation of ICD-10-CM.

Another important change that occurred in 2008 was specification of a secondary diagnosis as present on admission (POA) (or not). The reporting of the POA indicator allowed complications that occurred post-admission to be identified. It is important to note that the use of “complications” was to identify a condition that occurred during the admission, but did not imply wrongdoing on the part of the provider. CMS always referred to CCs as complications and comorbidities, but with implementation of POA, they were better able to differentiate a comorbidity condition that existed prior to admission from one that occurred during an admission. This change was required by the Deficit Reduction Act, which also mandated documentation of Hospital-Acquired Conditions (HACs). Initially, CMS designated 12 conditions as HACs (the list was later increased to 14 conditions, and has not changed since the 2014 fiscal year). If a condition classified as a HAC is present on admission, it continues to be classified as a CC or MCC, whereby it can potentially increase reimbursement. However, if the diagnosis was not POA, then it was no longer to be classified as a CC or MCC, potentially resulting in a lower MS-DRG assignment and reduced reimbursement.

So, where are we now? In 2019, there were 759 severity-adjusted MS-DRGs, based on the patient’s clinical condition and treatment strategies, as defined by discharge diagnosis and procedure codes. Unlike the DRG system, the MS-DRG system currently has 335 base DRGs, many of which are further divided into two or three DRGs, sometimes referred to as a pair or triplets, based on the presence of a secondary diagnosis defined by CMS as a CC or MCC. MS-DRG assignment can also be influenced by discharge destination and use of a specific drug.  Implementation of the MS-DRG system meant more opportunity to capture secondary diagnoses classified as CCs or MCCs, compared to the old DRG system, potentially resulting in higher revenue. It was around this time that the CDI profession also began separating from coding. That’s not to say that coders were not working as CDIs; rather, CDI was being seen as more than an “extension” of coding, allowing it to become its own entity. As hospital executives came to better understand this reimbursement mechanism, they saw the potential advantage of implementing a CDI program, to see if such a program could increase revenue.

Another significant change occurred beginning with the 2013 fiscal year. This change was not related to the structure of MS-DRGs, so it didn’t have a direct relationship with code assignment, but it did have an impact on reimbursement. Although the origins of CDI programs were aligned with implementation of the PPS and DRG reimbursement mechanism, many CDI departments struggled to continue that alignment as CMS moved to a quality-based payment structure. Beginning with 2013, CMS introduced mandatory quality programs, beginning with value-based incentive payment (better known as the Hospital Value Based Payment Program, or HVBP) and the readmission reduction policy (better known as the Hospital Readmission Reduction Program, or HRRP). These programs were followed by implementation of the hospital-acquired conditions penalty, better known as the HAC Reduction Program (HACRP). It is one thing to monitor cases for HACs as a program, which could result in the loss of a CC or MCC on a particular claim. It is another, much more complicated process to monitor the potential impact of each of these programs at a per-case level. I’ll discuss more about these programs and how CDI departments are incorporating a quality approach in a future article.

Although the initial CDI mission was to capture diagnoses classified as CCs and MCCs, many CDI programs have evolved beyond this focus and have tried to incorporate a “quality” approach. Ironically, the move to such an approach was not precipitated by CMS’s change in the Inpatient Prospective Payment System (IPPS) rules that incorporated quality as it was by the myth that focusing on revenue (e.g., the capture of CC and MCCs) was “leading” or “noncompliant.” Additionally, around this time there was increased visibility of the concept of mortality – and, luckily, the early mechanisms used to measure mortality using severity of illness (SOI) and risk of mortality (ROM) were closely aligned with reimbursement, so it was easy to “focus on quality” and “let reimbursement follow.” However, that relationship is not as clear as it once was. As a result of all these different influences, there is much inconsistency across CDI departments regarding how “quality” is defined and incorporated into workflow.

As you can see, there is a lot to examine when we look at where the CDI profession began, as well as where CMS Medicare reimbursement began, and how both have evolved through the years. We have evolved as a profession, so why haven’t our metrics? I hope you will join me as we explore this topic more in the coming months.


Cheryl Ericson, RN, MS, CCDS, CDIP

Cheryl is the Director of CDI and UM/CM with Brundage Group. She is an experienced revenue cycle expert and is known internationally for her work as a CDI professional. Cheryl has helped establish industry guidance through contributions to ACDIS white papers and several AHIMA Practice Briefs in the areas of CDI, Denials, Quality, Querying and HIM Technology.

You May Also Like

HCCs: The Role of CDI and Risk Scores

HCCs: The Role of CDI and Risk Scores

Predicting coding patterns using the HCC risk scores can be a valuable endeavor. EDITOR’S NOTE: Longtime RACmonitor contributing correspondent Frank Cohen, a senior healthcare analyst,

Read More

Leave a Reply

Your Name(Required)
Your Email(Required)