Today we have a press release explaining that CDKL5 Deficiency Disorder has been granted an ICD-10 diagnostic code, and that it will become effective October 1 2020.

This was quite a major achievement for us. Let me explain you why it matters and how we did it.

WHAT ARE THE ICD-10 CODES

Imagine I give you a large box, and you cannot see what’s inside. I only give you a list of descriptive cues:

- Furniture

- Wood

- Four legs

- No seatback

With the last cue we can now rule out a chair, but it could still a table or a seat stool. This is not the Taboo game where your partner has to guess a word and you have to describe it using other words. This is rather a game of “you can only use words from the dictionary” and a dictionary that has no word for table.

Believe it or not this is the daily life of clinicians when they see patients with rare diseases. While rare diseases have names, these are often not included in the International Classification of Diseases, and without a code for them, clinicians need to use other codes to describe their patients, and it ends up looking as if they are just giving cues.

The International Classification of Diseases is the medical classification list from the WHO, that uses alphanumeric codes to ensure we can call the same disease the same way, no matter the country or the language.

But a lot of the rare diseases are missing in that classification.

Until now, the international codes (ICD-10 codes) for a child with CDKL5 deficiency disorder might look like this:

- F88 Global developmental delay

- G40.411 Epilepsy, refractory, with generalized seizures

- H47.61 Cortical visual impairment

- R28.2 Apraxia (when you have a motor impairment due to a brain – not muscular- problem)

- Q99.8 Genetic abnormality

- and so on

And this is happening exactly like this for a lot of rare diseases: clinicians have to almost use cues, using the symptoms and diagnoses for which there are codes, because the name of the disease is not in the dictionary and they have to stick to words in the dictionary.

Now imagine that there are over one thousand cases with CDKL5 deficiency (CDD) around the word, all coded with some variations of descriptive codes as above. And now I sit you in front of the medical database and ask you the following questions:

How many patients are out there with CDD? You don’t know, because you cannot perform a search since there is no common code
How many children are born per year and what is their life expectancy? Same, no idea because we cannot track them
Should this particular child have access to occupational therapy? You might not know if the physician didn’t use the right codes that give access to it, but if we had a code for CDD the answer could only be “yes, absolutely”
And should this child have access to ganaxolone once approved? Not sure, because while regulatory agencies recognize CDD as a stand alone disease, the international medical classification does not.

As you can see this is one problem with a lot of consequences.

I am telling you all this because we have changed it. Starting October 1 2020, when the next revision of the ICD-10 dictionary becomes effective, there will be a unique code to designate CDD and that is G40.42.

From October we won’t need to use a list of cues to tiptoe around the name-less disorder, we will be able to code into the medical record the disease for what it is. We will know how many people are out there because it will be possible to search how many people have that code in their medical records, and where they are, and make sure they have access to the services and therapies that they are entitled to.

This was a big big milestone, and if your rare disease has no ICD-10 code and you want to understand how we did it, keep reading below.

THE APPLICATION PROCESS

The country where these ICD-10 codes are most important is the US because of the insurance and reimbursement system, which requires the use of specific codes. There it is called the ICD-10-CM, for Clinical Modification. The revision of the dictionary is done once a year, every October 1, and it is managed by a division of the National Center for Health Statistics (under the CDC).

There are two reviews a year, and in order to get into the agenda you have to submit an application 3 months before. For example the next review for new code proposals is in March 2020, and the deadline was December 6 2019.

We applied in June 2019, and got accepted to be included in the review that took place in September 10-11. This was followed by an open consultation period of two more months, during which you are encouraged to secure support letters, and then we got the news 2+ months after that. Since we started preparing our application around March (2 months before the application deadline), the entire process took 10 months and remember that the effective date is October of the following year, so this is not something you can do in a rush.

STEP 1 – PROPOSAL

The proposal itself is not hard. It was about 2 pages. I wrote it following very useful advice from others who had gone through the process before (see section below), and it was submitted as a joint application from us (Loulou Foundation) and the IFCR, with the support of the US CDKL5 Centers of Excellence. This helped convey the message that having a unique ICD-10 for the disease was of critical importance for the clinicians treating these patients.

The 2 pages essentially argued that CDD is a unique disorder (not under another disease code), with unique medical needs (not identical to other diseases), and that not having a code was introducing variability in coding between patients that was hurting patient care and access to therapies.

The difficult part is what comes after those arguments, and that is what code you are proposing and why. Because you just don’t say “hey we need you to create a code” but a proper “we need a code for the disease, and this is how the new code should look because it belongs to this X section of the coding dictionary”.

That was our headache for those months. We are talking about a developmental and epileptic encephalopathy with a monogenetic cause. So it fits in three places in the classification:

- It could be under neurological conditions

- It could be under conditions with epilepsy

- It could be under genetic conditions

At the research Foundations we are not experts in medical coding, and clinical practitioners are not either, they are users of the medical coding. So we had to talk to people at the leading medical organizations (like the American Epilepsy Society and the American Academy of Neurology) who specialize in medical coding to help us see which code to propose. The review team at the National Center for Health Statistics also helped, by reaching out to WHO to check what might be the best placement.

We understood that the placement doesn’t matter so much to us as patients because all you need is a code and what that code will give you access to (and easy of tracking). But we have seen for other diseases how disagreements over the coding section meant they had to resubmit, and then resubmit again, and that means 6 more months each time since there are only two review opportunities a year, so it is important to get the proposed placement right.

We ended up applying for a code within the overall epilepsies classifications following the WHO recommendation, although CDD could fit equally well in multiple places. During all of this feedback collection process I kept getting back to the medical societies to make sure they agreed with the proposed code and were fully informed about why one code proposal and not other.

STEP 2 – FACE TO FACE REVIEW

After our application we got the news that we had been invited to the review, so we were into the agenda. That was already good news.

We then had to send a clinician to the public review at the CDC, which took place September 11, to explain in a couple of minutes why we needed a code and why THAT code in particular (these are professional coders, remember).

We sent them a very charming young neurologist who runs one of the CDKL5 Centers of Excellence. It is hard to say no to Dr Eric Marsh, and he did a great job at answering the questions after his presentation (starting by “why an epilepsy code and not a neurological or genetic code?”).

There is an online link to follow the reviews remotely. I watched it over dinner (European time) with my family and I felt we were watching the Academy Awards.

Except you still don’t know if you got the award that night.

STEP 3 – SUPPORT LETTERS

After the face to face review there is an open consultation review of about 2 months, during which the code users (for example hospitals) can review all what was presented during the September review and reach out to the National Center for Health Statistics agreeing or disagreeing with the proposals. We were encouraged to request support letters to be sent to the review team during this period.

We got letters from some of the major medical societies (others don’t produce letters but call and talk to the review team, so we still wanted to make sure they would be supportive), from relevant umbrella patient organizations like the Epilepsy Foundation and NORD, and from all of the pharmaceutical companies working on CDD.

The message was always the same: “CDD is a unique disorder with unique medical needs that therefore needs a unique code, and we support the proposed code”.

We had already told all of them since before the June application deadline that we will be applying, discussed the specific code with the medical ones, sent the face to face agenda and on-line link to all of them for them to be aware of the September 11 review, and then finally collected their letters. Between Heidi Grabenstatter from IFCR and I we sent and received over 100 emails back and forth with all these organizations for the ICD-10 proposal.

Two months after closing that consultation period, last week we got the email we were waiting for: code accepted, will be included in the next revision of October 1st.

We have now prepared a press release and will make sure we notify all families and their clinicians so they can start incorporating the new code into their medical records as soon as it is possible.

WHAT WAS KEY

I think we had two key arguments that helped us get a code created for CDD. This means that if your rare disease is at an earlier stage of medical literature and therapy development you might not get it.

1 – There were some very strong medical literature articles stating in very conclusive terms that CDD is a unique disorder, not under Lennox-Gastaut syndrome (which had a code) and not under Rett syndrome (also with a code) and not under any other disease, just a standalone disease.

2 – We have 4 drugs in clinical trials. The best way to justify that you not only have a unique disorder, but that is also has unique medical needs, is to show that there are drugs in development specifically for CDKL5 deficiency disorder. This means that companies and regulatory agencies agree you need unique drugs! Quite a strong argument. Here is where I think many other rare diseases might face difficulties.

THE SUPPORT

I believe there are very few people in patient communities who will go through the process of a requesting an ICD-10 code creation more than once. I have only had to do it for CDD, and same for my counterpart Heidi at IFCR. So we relied on a lot of advice.

We needed people who has done it before not only for specific feedback on our applications materials, but even to have a mental picture of the whole process. How longs will it take? Is that hard? Who should we talk to? Is it normal is the application team doesn’t reply to emails?

Luckily for us two friend patient groups (Angelman and Dravet) had recently gone through the process, so we had extremely useful advice from Terry Jo, Nicole and Mary Anne. The team at Ovid Therapeutics also gave us advice, since they had supported those two previous applications.

And not just advice, but we got to see their application letters. That was fantastic.

We also had the support from the US network of CDKL5 Centers of Excellence, all of which not only reviewed by also co-signed the application. When you have some many hospitals backing up a proposal, you know the reviewers have to take it seriously. And Dr Marsh was willing to free up a whole day with very short notice (think 4 business days before the face to face review in Maryland). He and I worked through the night to put together the presentation, which has specific requirements down to how to name a file without using spaces…. so believe me that was quite some challenge.

And last, the support letters. I feel fortunate that so many medical organizations, large patient organizations and pharm companies would be so willing to support us in our request for an ICD-10 code. This includes people who we didn’t know before we started this process, and who would make sure the Board of their society will have the perfect letter for them to sign before deadlines. Huge thank you to all of you.

Because we had so much help and support I would like to offer the same to those of you starting this journey: if you are a rare disease patient group and are considering applying for an ICD-10 code for your disease, reach out to me and I will be happy to be for you what all of these people were for us.

I hope this entry was informative and useful. Let me know if you have questions!

Ana Mingorance, PhD