Discovering new diseases with the internet: How to find a matching patient

[article index] [] [@mattmight] [+mattmight] [rss]

Genome and exome sequencing are the greatest diagnostic breakthroughs in the history of rare disease.

When sequencing identifies a genotype already associated with human disease, it can short-circuit years of costly and painful one-off disease tests.

But, if sequencing turns up “variants/mutations of uncertain clinical significance,” then a new kind of diagnostic odyssey unfolds.

Narrowing down which variant is responsible for a disorder may require “functional studies”: going to the lab to study cells or genetically modifying organisms in an attempt to link the mutations to the presentation of the disorder.

(Functional studies are not and are unlikely to ever be covered by insurance.)

Alternatively – and preferrably – you can find a second patient to confirm discovery of the disorder.

This article below describes how to use the internet to find a second case for a previously unknown genetic disorder.

If you find success with this approach, please email me to let me know how it worked out for you.

Can you find patients online?

In the case of my son, sequencing revealed several variants of interest. After considering each variant, mutations in his NGLY1 gene were deemed most likely responsible for his condition.

There was just one problem: no other patient had ever been identified with a disease caused by this gene.

We were fortunate: researchers working with us performed basic functional studies on cells that raised our confidence to “highly likely” that this gene was the cause of my son’s disorder.

My wife and I then had to grapple with a question facing many these days: “What do you do when you are the first patient? What do you do you’re the only patient?”

To confirm the diagnosis and build a community, I created a blog post designed to find other patients like him.

It worked.

In just two months, two more confirmed patients popped up in Turkey.

In a little under three years, 35 patients with my son’s disorder were found, even though we estimate roughly 500 to 2000 living patients worldwide based on the frequency of deleterious mutations in the NGLY1 gene.

Internet-driven patient-finding

The core idea behind internet-driven patient-finding is to create a “reverse dragnet” to catch all of the other patients with the same disorder when they search for information about the disorder online.

You want to make it easy for other patients to find you.

Setting up a patient-finding web site

Creating a web site is the first step in finding other patients.

The web site need not reveal personally identifiable information as long as there is a way to contact you anonymously.

A blog hosted on something like Google’s blogspot is a perfectly reasonable and simple platform for hosting a patient-finding site.

Even a facebook page may be sufficient to find more patients, although you must ensure that the page is public and indexable by Google.

Creating content for the patient-finding web site

The web site should be a complete description of the patient, including all terms that researchers or other patients might type in while searching for information online:

  • List all gene variants identified in a sequencing report.

    • For each variant, list both the the gene name and the mutation. The mutation is going to be in a notation that looks something like c.NumberLetter>Letter or p.LetterNumberLetter. List both c. and p. notations if available. (Ask if the c. and p. notations are with respect to the full protein rather than an splicing variant that may have skipped exons.)

    • For each gene name, list any synonyms for the gene name as well.

  • List all symptoms, past and present.

  • Use both medical terminology (human phenotype ontology (HPO)) to describe symptoms as well as corresponding terms that parents and patients are likely to use instead.

  • Describe the history of the condition from birth, including all suggested diagnoses, and all attempted and current treatments, procedures and medications.

  • List the outcome/effectiveness of each treatment/procedure/medication.

  • Include both abbrevations and expansions of medical terms; for example: “elevated aspartate transaminase (AST).”

  • Make it easy to get in contact: Listing an email address and a phone number and including a web contact form at the top and bottom of the page will help.

  • Include the names of suspected genes in the title, in the header and in the body of the web site. Gene names are the most important for Internet-based matchmaking.

  • Including pictures, especially if there are dysmorphic features, will assist in confirming matches.

Example sites

Here are a handful of examples of sites that have connected patients online:

Editing Wikipedia

Wikipedia regularly ranks near the top of search results for gene names.

Inserting an entry about the possible clinical signifance of the gene that references your site will increase the chance that a patient, parent or researcher searching for the same gene will find your site.

Use the following form to create a template for inserting text into wikipedia:

Gene name
Variant / mutation(s) e.g. p.R401X, c.1202A>T

Obviously, you can change the text here to suit your needs.

User Ruraldreams created a guide to posting about rare disease on wikipedia.

Remember to update Wikipedia when you confirm the disease: update the correct page to confirm the disorder, and remove the information from the genes that did not match.

Setting Google alerts

Google alerts provide a way to be continuously searching the interet for a given term.

When a search term pops up somewhere on the web for the first time, Google will email you.

Set a Google alert for all gene names of interest, the specific variants of interest and all rare symptoms.

Buying Google AdWords

Buying Google AdWords on the name of the gene and especially rare symptoms is a reasonable strategy for patient-finding, especially given that gene names and rare symptoms should be cheap AdWords.

Using gene names for domain names

Google gives increased weight to a site if the search term appears in the domain name.

Buying a domain name for the gene name, e.g., is an inexpensive way to rank more highly for the gene.

In the long run, a gene-based site name can serve as an online hub for the community.

Minor search engine optimization

There are a few options avaliable to anyone for providing a slight boost to search engine rankings:

  • Enable SSL / https to secure the site.

  • Include the gene name/rare symptoms in the <title> and <h1> tags.

  • Include the gene name/rare symptoms in a URL (if not in the domain).

  • Make the page “mobile-friendly” since Google now weights sites higher if they have a mobile-friendly version.

Using Google analytics and webmaster tools to find “cloaked” patients

Using Google Analytics and Google Webmaster Tools allows you to understand the traffic coming to your site.

Most critically, you can examine the search terms used and the location of the visitors.

If you see a surge in traffic from a particular country, that could be a sign that researchers there have discovered a patient.

It’s possible, however, that the researchers won’t reach out.

Consider doing a short, one paragraph translation into that language and posting it on your site, in case patients follow up with a search themselves.

Standard registries

In addition to the internet, there are a number of registries available for searching and reporting cases in the hopes of finding a match, including: