As the healthcare industry moves into a changing world of reform, the importance of reliable, accurate, and complete data has never been more critical. Historically, the transactional data we rely on to understand the burden or illness of the population and the risk, complexity, and severity of health conditions has been limited.
In a recent article for ICD-10monitor (ICD-10 Myths Part 2: Coding Specificity),I reported on an analysis of three years of claims data representing more than 15 million professional claims. According to this analysis, the primary diagnosis for 54 percent of all involved claims was “unspecified,” “other,” or just “symptoms” or “signs.” In looking at claims for cardiac rhythm disorders, for example, in well over 50 percent of claims there was no way to determine the nature of the cardiac rhythm problem. In this example, if we try to examine this data to better understand the cost of care or the number of cases of atrial fibrillation treated, we could be off by more than half.
There are few other industry segments for which transactional data about what was done and why is as non-specific as the data we rely on to manage the healthcare of a population.
Will ICD-10 Solve This Problem?
While it would be great to believe that the implementation of ICD-10 will result in better data throughout the industry, providers reporting claim transactional data can be just as vague in ICD-10 as they were in ICD-9. In many instances there are opportunities to be even less specific. ICD-10 provides an opportunity to be more specific, but this does not ensure that everyone will take advantage of that opportunity. In another recent ICD-10monitor article (ICD-10 Myths Part 2: Coding Specificity) a more detailed discussion demonstrates some of the challenges in this area. It was pointed out that in ICD-10, there are two valid codes for “injury unspecified” and “illness unspecified” that in theory might be used to describe virtually any condition. Although this is a rather extreme example, the reality is that there are many ICD-10 codes that are very vague. While any provider treating a patient should document which side of the body is being treated, ICD-10 allows for the coding of “unspecified side.” Likewise, ICD-10 has a code for “respiratory failure, unspecified,” whereas in ICD-9 only “acute” and “chronic” respiratory failure are allowed. There is no doubt that ICD-10 has many more specific codes, but if these codes are not used appropriately, there will be no advantage to migration to ICD-10.
In many instances we have become so focused on the codes that we have lost sight of the fact that codes are just a vehicle for capturing important medical concepts. Without this connection, the codes are just a set of meaningless characters, regardless of the code type.
The Role of the Clinician
The clinician is the source of all data about the clinical condition as it is represented in ICD-9 and ICD-10 codes. The data represented in these codes is entirely constrained by the observations and documentation of the clinician performed in the course of assessing the patient. Only the clinician is licensed to diagnose and treat patients. Unfortunately, many clinicians do not see the value proposition for them in capturing more accurate data from each patient encounter. Some feel that they know what they need to know to treat the patient, viewing additional documentation as a burden that takes away from good patient care. On the other side, however, good documentation was an important part of what we were taught in medical school and residency. There is no doubt that poor documentation of the key parameters of patient conditions compromises healthcare quality as well as the continuity of care, as patients often are treated by many different providers.
While I’m sure that many pilots feel that the required documentation and data capture for flying is burdensome and in many instances unnecessary, it is a key part of the aviation business and clearly an important part of aviation safety. Most law enforcement professionals will tell you that they hate the paperwork and feel that it takes away from their job. But without that documentation, we could not convict a single criminal. Like it or not, being specific about the important medical concepts of the medical condition is what we signed up for as healthcare professionals.
The Role of the Payers
Payers in theory are responsible to ensure that limited financial resources are used appropriately to create quality of services, broad access to needed services, patient safety, and affordable healthcare coverage. To accomplish this goal, payers and other managers of healthcare populations must have accurate, reliable data. Payerss also are charged to ensure that payment is fair, and commensurate with the severity and complexity of each service. Unfortunately, since the traditional focus has been more on service payment than the management of health conditions, there has been little focus on the quality of diagnosis-based codes from the payor perspective. In some instances, payers have encouraged providers to submit less-than-specific codes. Some payers historically have accepted “short codes” (which are not really valid codes) because their processing systems can only look at the first three characters of the codes. Some payers only process based on the primary codes and do not consider other codes that further describe the patient condition. Some payers require the submission of a vague code for a range of conditions because it’s easier to process, and because their system rules and logic are not sophisticated enough to utilize the greater level of detail. Payerss should never tell providers which diagnosis codes to use. They are not licensed to diagnose and treat patients. Without an assessment of the patient how can you know what code best represents the patient condition?
What’s the Solution?
There is little doubt that unless we understand the financial implications of clinical conditions and the clinical implications of financial decisions, healthcare delivery will not progress. Technology will not solve this problem. That’s like thinking that the best word processing system will make you a great author. Healthcare facts are technology-independent, and there is no amount of technology that can create facts that haven’t been captured. Big data sounds great, but more garbage in just means more garbage out. It all comes down to incentives. How can we align incentives to make payers want better data regarding patient conditions? How do we make providers sees the value proposition to them and their patients in transactional data? While there is no magic bullet, here are some key considerations:
- Payment should be aligned with the management of the patient condition, rather than the count of services delivered.
- Providers should have open access to data that defines what they do and how they compare with their peers.
- We must apply judgment carefully in order to avoid assuming that variation equates with poor performance. The best performers in healthcare are outliers just about as much as the worst performers. Expecting the norm may result in aspiring to mediocrity. Most data analysis raises more questions and seldom does any single report provide a final answer. A healthy dose of intellectual honesty must be applied to all data submitters and data consumers. We should not attempt to make data say something it cannot.
- Payers and providers alike should be held accountable for the quality and completeness of data reflecting what was done for the patient and why.
- Electronic health records should be focused on collecting details important for patient care and less focused on driving clinicians to “cut and paste” to meet some pre-defined model that may not be relevant to the patient condition.
- Systems to capture data and find the appropriate codes should be as simple as most Internet searches.
- EMR systems, cheat sheets, problem lists, and superbills must avoid creating the temptation to crosswalk from a short list of vague ICD-9 codes to an equally vague short set of codes in ICD-10.
- “Decomplexification” is an important goal. There is lots of complexity in how we do things that is artificial and adds no value. To quote Albert Einstein: “Things should be as simple as possible, but no simpler.”
The goal of moving to a new standard for the definition of patient conditions is clearly important, but it will not happen automatically. The good news is that rapidly evolving payment models are leading to a greater focus on the details of the patient condition, particularly on the hospital side. Without clinicians fully understanding the value proposition for better documentation and better data, however, and without payers providing incentives for good data and disincentives for bad data, we will not reach this goal.
About the Author
Dr. Nichols is a board certified orthopedic surgeon with a long history in health information technology. He has a wide range of experiences in healthcare information technology on the provider, payer, government and vendor side of healthcare business. He has served in positions in executive management, system design, logical database architecture, product management, consulting and healthcare value measurement for the last 15 of his 35 years in the healthcare industry. He has given over 100 presentations nationally related to ICD-10 over the past three years on behalf of payers, providers, integrated delivery systems, consulting groups, CMS, universities, government entities, vendors and trade associations. He co-chairs the WEDI (Workgroup on Electronic Data Interchange) translation and coding sub-workgroup and has received WEDI merit awards three years in succession. He is also an AHIMA approved ICD-10 coding trainer. He is currently providing consulting services as the president of Health Data Consulting Inc.
Contact the Author
Comment on this article
Editor for icd10monitor.com