3 Fundamental Challenges With Medical Coding That We Must Face Today

Nicola Sahar
3 min readNov 13, 2019

When I started medical work, I believed that the biggest challenges that we faced in healthcare could be solved in the clinic. I wanted to increase access to care for all patients to make sure that no patient was slipping between the cracks. As I started rotating in the clinics, however, I saw many pervasive issues in healthcare that impacted patient care well before and after they arrived in my clinic. I began to understand that truly caring for my patients is not just about caring for them in the clinic. One of the most subtle but insidious challenges, for example, is how fragmented and unusable our health data is, which causes fundamental issues across the whole system.

The best example of this is medical coding, where we have in place today complicated documentation and coding practices to ensure accurate coding and billing. Hospitals spend over $3B on medical coding to code medical notes into narrowly defined insurance codes, yet, still face an estimated $31B in missed payments due to a lack of support to coding teams and errors in documentation. It is difficult to quantify the negative impact that the undercoding gap can have on the healthcare system. To bring visibility to this problem, I built an undercoding calculator that quantifies the losses brought about by these challenges for hospitals.

Today, there are 3 fundamental challenges with medical coding that we must face today:

  1. Medical coders miss clinical context in up to 50% of cases which can reduce code value: Coders spend 90% of their time searching charts for clinical context that can increase the value of codes, but they don’t have the time or clinical expertise to find all of this context.
  2. Coders are 8x less productive than employees in other sectors: From the EHR to computer-assisted coding and abstraction software, coders are using several disconnected tools to complete a chart, which can reduce efficiency and accuracy.
  3. Providers do not document up to 40% of clinical context and active chronic conditions: Providers do not have the time or coding expertise to capture all clinical information required for accurate coding of every active chronic condition, as they are trained to focus on the chief complaint.

After we estimate the size of the undercoding gap, we can begin to work towards its solution. We need to start thinking about the use of innovative technologies to create intelligent solutions that can make the lives of medical coders easier and more efficient. If built in the right way, this next generation of medical coding solutions will help us unlock the true value of our health data to truly transform patient care.

I believe this vision is possible today, and am leaving clinical practice to make it a reality. My first plan of action is to help hospitals and medical coders estimate the amount of undercoding that occurs. As I mentioned above, I built an undercoding calculator to help you with this — I would love it if you could check it out and let me know if this was useful for you.

--

--

Nicola Sahar

Medical doctor and machine learning researcher passionate about using technology to transform patient care.