5 posts in the category “Treat Us Right”

Treat Us Right: Mapping What Patients Think About Medications

Posted December 15th, 2010 by

One of the ways we can better understand whether you, as patients, are having a positive or negative treatment experience is to “listen” to the conversation you’re having in our forum.  By understanding whether you are having a positive, negative, or neutral experience with a particular treatment you are taking or are considering taking, we can measure the impact of different events on the overall community.

For example, in 2008 we measured the impact on our multiple sclerosis community of a corporate announcement by Biogen about a serious and sometimes fatal side effect of Tysabri (occurs in about 1 in 1000 patients).  The results revealed that patients were indeed frightened by the announcement, but these patients were also so positive about Tysabri’s benefits, that most planned to continue taking the medication regardless of the risk.

Visualizing Perception of Sentiment
We visualize movement in your sentiment via perceptual maps and longitudinal bar charts.  The perceptual map here shows how patient perception (indicated via forum conversations in one disease community) is moving regarding different medications over four periods of time. (Note: each color represents one medication;  the shading represents the change of perception over time with the darkest shade being most recent).  From period to period, it becomes clear which medications you perceive work the best (i.e., Medication D for efficacy) and those that have the most side effects (i.e., Medication A for safety).


A stacked bar chart graph is a way to further break down the sentiment.  For example, the chart below shows the volume of posts about Medication E’s perceived efficacy, whether positive, negative, or neutral by month over time.  This visual allows us to evaluate if certain events impact your perceived efficacy of a particular medication; to create this graph, we look both at volume of posts (spikes) as well as proportion of posts by sentiment (colors).


Why is that important?  Because studies have shown that people who stay on their medications long term get the best health outcomes.  By measuring patient sentiment of discussions, we can predict if patients may discontinue taking their medications and why.  Knowing that, along with the information you share as part of your profiles, helps in research of how outcomes change over time and the impact of peer influence.

These methods are also used in creating our PatientsLikeMeListenTM service for industry partners.  Their interest is in understanding aggregate perceptions and what influences patient behavior so that they can keep patients like you on medication.  As part of this service, we show them which types of patients are most likely to stay on medication appropriately and which ones might be better off changing medications.

Our goal in analyzing patient sentiment overall and providing the PatientsLikeMeListenTM service for industry partners is to amplify your voice to anyone listening:  Treat Us Right.

PatientsLikeMe member dwilliams

Treat Us Right: Comparing our Community to the General Population

Posted December 14th, 2010 by

Our recent series entitled Share and Compare focused on how patients like you can better answer the question, “How do I put my experience in context?” The answer, in part, comes from how much information you share to help create that context of real-world patient experiences.  Think of it this way – with every piece of information you share, you are contributing directly to research.

When we’re conducting research, one of the things we look at is how similar or different you are to the populations at large.  We even have minimum criteria for a person’s data to be usable.  For example, if you indicate whether you’re male or female, you make it that much easier in determining how you “fit in.”  That one piece of information helps us know if our population is in fact representative of a disease, or whether we’re only getting one specific type of patient (e.g., males with fibromyalgia who don’t have much pain).  If we do get more of one type of patient, it becomes more difficult to draw any conclusions from that population and apply them to the general public.

So, you may be wondering why we need to compare to the published literature/general public?  Why can’t we just say that our conclusions apply to our users and leave it at that? The answer to this question has many parts:

  1. We have the ability to positively impact everyone with disease, not just our current members.  Ideally, we will apply knowledge gained through research in our communities to all people living with diseases.
  2. From a research perspective, we have to know our biases, and how to correct for them if possible. For example, we tend to have more women than men in our populations.  By knowing that, we can “correct” for it in our analyses by making sure our proportions are correct when we look at a sample of users.
  3. We can know how our discoveries fit in with other information known about a disease.  For example, let’s say we figure out that patients who have had fibromyalgia for 15+ years improve their quality of life by doing Treatment X.  If we don’t know how many patients have had fibromyalgia for 15 years or how many do Treatment X and don’t improve, the discovery loses some of its power from lack of context.  Perhaps it isn’t a discovery at all!  However, if we have data from our community to answer those questions and can compare it to published literature, we can trust more in our discoveries.

Here’s a great example of what can happen with the data you share.  Recently, we evaluated our fibromyalgia community characteristics with the Demographics Survey sent out early in 2010.  For some of our communities, the survey had fantastic results.  We are now able to declare with confidence that our community very closely matches the fibromyalgia community at large (Table 1).



By maintaining your profiles and keeping accurate records of side effects, medications, background information, and outcomes (such as quality of life), you are participating in groundbreaking research that is already yielding fantastic results. Our research team has presented at prestigious conferences and written dozens of abstracts and papers. Working together, PatientsLikeMe has discovered new symptoms and compared treatment efficacy; we are also working towards creating an accurate picture of how medications work in the real world so you get the right treatment for you. This is just beginning.

PatientsLikeMe member cbrownstein