UPDATE (Feb. 2, 2010): Based on some feedback from none other than John Mack, I’ve changed a couple of pieces of info from my original post, as well as some insights from iGuard about how they conduct their surveys. This can be found at the bottom of this post.]

On a couple of occasions, I’ve talked about the potential impact of having product reviews for pharma products (see “Why Pharma Needs Product Reviews” and “One more Reason Pharma Needs Product Reviews“). You can check out those posts to see my rationale for why pharma might benefit from having product reviews, but one important point is that product reviews are available for prescription pharma products right now.

If you’re like most people, when you shop online, you check out product reviews and use them to determine which product you ultimately purchase. Whether you’re on Amazon.com or Wal-Mart.com, you’re going to find customer reviews. Even more interesting, you can just go to Buzzillions, which aggregates reviews from a bunch of different sites. So, instead of a handful of reviews, you might have hundreds. When you search for a product category, you quickly see the top-rated products.

That makes shopping pretty easy.

Question is: would reviews from other patients help you select which drug you ultimately take? Data from Pew Internet says that 6 out of 10 online adults reported that user-generated content affects their treatment decisions. This content includes everything from blogs to newsgroups and discussion boards. It also includes physician and patient product reviews.

If you consider that a lot of the social media “advice” people get online comes in the form of subjective information from a single person or just a handful of people, you’re missing out on one of the most powerful aspects of social media: the wisdom of crowds. That’s the rationale behind Wikipedia and why it remains so accurate despite the fact that anyone can edit anything (almost). When the crowd is in charge, you tend to get more accurate information. However, if you’re looking for information about a pharma product online, you might find a discussion thread with only a few people debating the merits of the product.

I know that they say “three’s a crowd,” but it’s not when it comes to accurate information. You need more than three or four people for crowd-sourcing to work. It turns out that there are already some places to find this type of information online. You  can now see patient ratings for any product compared  to others in the class. It takes a little work and it isn’t nearly as convenient as Buzzillions, but it’s out there.

When it comes to pharma product reviews, I’ve been tracking four different sites (eHealthMe, eDrugSearch, iGuard.org, and PrescriptionDrug-Info.com) on the Pharma and Healthcare Social Media Wiki.

The largest and most well known of these is iGuard. iGuard now has almost 2 million members with a pretty wide variety of conditions. Here are the patient counts for some select conditions (graphic provided by iGuard).

For common diseases, these are pretty big numbers and certainly qualifies as a crowd by my standards. The way  iGuard works is by asking patients to score their treatments in both satisfaction and efficacy. However, not every user is allowed to score their treatments. To ensure there is no impropriety and that a vocal minority doesn’t impact the results, iGuard randomly selects a sample of patients to actually rate the treatment. That is, while 10,000 people might report taking a particular drug, only a select group of these actually are asked to review the product.  The value of this is that it keeps those who are most happy and most disappointed with the treatment from being the only ones who go through the trouble of rating a specific product. It’s a smart wrinkle in the system that makes it more reliable in my opinion.

When you visit iGuard, you can search for a specific product and see all of the relevant safety information, treatments for which the product is most commonly used, most reported side effects, and objective scores in both overall satisfaction and effectiveness. In addition to some numbers, there are also comments…lots of comments. These are similar to the verbatim reviews that you see in product reviews for other products. Some people are ecstatic with their treatments and some are very disappointed. And a  large percentage have questions about their treatments.

What’s lacking from iGuard is a simple list showing the highest-rated treatments in a given drug class (similar to how Buzzillions works). But, I thought I’d do the work for you. Since you can look up the ratings for any product, I just looked up the products and put them together in an easy to read format.

So, which are the top-rated products? Forget about all those head-to-head trials that payors want, but most companies are hesitant to conduct (for many reasons). If you want to know which treatment is best, why not check out its ratings? How far away is a future where patients select which products they want to take by using reviews such as those found on iGuard? I’m sure some of you are scoffing at this idea because you think physicians should be recommending treatments, not iGuard. Two questions for those of you thinking this: aren’t objective ratings guiding treatment requests better than DTC TV ads that also aim to get people to ask for a specific treatment? And if these ratings are available, why would physicians ignore them? How long before they too use these types of reviews to decide which treatments to prescribe?

The ratings today aren’t “clean” enough to replace clinical trials, as they include confounding factors such as multiple indications and dosages grouped together. For example, when it comes to the ratings for atypical anti-psychotics, these data might have the ratings for patients with bipolar mania and schizophrenia lumped together regardless of condition or dosage. But, it’s certainly an interesting start and something to which pharma companies need to start paying attention, as it foreshadows a future when they have even less control over how their brands are seen in the market.

Okay, so you probably just want to see the rankings now, well, here it comes. Each includes the drug name, patient satisfaction score, patient effectiveness score, average score (average of the previous two scores), total number of user comments, and the total number of individual patients on iGuard taking the treatment who supplied a rating [UPDATE: The number taking the survey is not available, but iGuard suggests that approximately 10-30% of people take the survey for each product. See more details in the update at the end of this post]. All but the “average score” come from iGuard; I added that as a way to stack rank all the treatments. It’s the average of the patient satisfaction score and patient effectiveness score. A few things you should know about this data: both brand and generics (where available) are included together under the brand name (e.g., Prozac and fluoxetine) and all dosages and indications are included together. This is how iGuard supplies the data. For each drug class, the drugs are listed in order by “average score” (high to low). The scores are 1 to 10 (10 is the highest).

So, without further ado, here are the ratings for some of the top classes of drugs available today.

If you want to check out the raw data and play with it yourself, you can get it here in the form of  a Google Doc spreadsheet and, of course, go to iGuard and look up different conditions for yourself.

A few things to note that I found interesting in compiling this information. First, the number of patients for each drug was impressive to me in most cases. This is a lot of data that could potentially tell you a lot about how a brand is performing in the “real world.” Second, there are a large number of comments for most treatments. These hold a wealth of information that can tell you exactly what patients like, don’t like, need information about, and what they say when you’re not listening. One bit of warning: if you work for one of the companies that doesn’t want you looking around online lest you accidentally see an adverse event, don’t go to iGuard. You’ll see them. However, most of the events are well within the expected events for most products.

As I mentioned in my testimony at the FDA hearings on social media, there are people here that want and need answers to their questions. Without them, in many cases, they’ll likely stop their treatment. The example I gave in my testimony was that of a woman experiencing  joint pain while taking Arimidex (a product I used to work on at AZ).

This is unfortunately a common side effect with this class of drugs, but the patient doesn’t know this nor does she know how she might treat these symptoms. She notes that she has a visit with her oncologist (Arimidex is for breast cancer) in one week, yet she’s still on iGuard asking anyone for help with this side effect. To me, that shows how desperate she is for information. Do you think that without an answer she’ll continue her treatment? Without her treatment she’ll reduce her mortality by as much as 40%. But, there’s no one to answer her question.

This is perhaps the biggest flaw with iGuard’s model. In some cases, an iGuard representative does answer some questions, but generally they refer people to their physician (a reasonable response in most cases). However, they can’t get to every question and might not have the in-depth knowledge and experience with some of these treatments like manufacturers might. My question to all of you: assuming iGuard’s system allowed you to answer questions, like the Arimidex one above, would you? You probably say “no,” as it would likely be a “letter of the law” violation of DDMAC rules since you couldn’t include all the necessary fair balance. But what if you did answer the question in a fair, non-promotional,and objective manner? Say, like this:

I think the FDA would have a pretty tough time explaining why they punished a company for helping someone stay compliant with a treatment their doctor prescribed; one that will likely save her life. I’m not naive enough to think they wouldn’t issue a warning letter for this, but my argument would be that if the FDA is there to protect the public health, does banning these types of interaction help or hurt public health? If this woman simply stops her treatment because she got no answer to her question, then that hurt public health. Period.

So, who’s going to step up and take the risk? If every company did this all at once, think of the impact that would have in the public eye and at FDA. A stretch, I know, but who said doing the right thing was easy?

UPDATE DETAILS (Feb. 2, 2010): I asked iGuard a bit more about how they conduct their surveys and if the number of patients they supply on the site (and seen in the table) is the same as the number of patients who have completed the survey. It turns out that they are not the same number. About “10-30%” of patients taking a specific treatment complete the survey, according to iGuard. As I mentioned already, the surveys are not sent to everyone, but instead to a random sample of patients to avoid “deck stacking.” iGuard added a few other points about the surveys to help show the rigor of the survey tools and analysis they use.  These are direct quotes from an iGuard representative (in italics):

1)  At no point in time is a patient asked to simply rate the “effectiveness” or “satisfaction” of their medication. These are derived scores calculated from questions in TSQM as is specified in the published TSQM methodology. [Note: more on TSQM can be found here.]

2) There is no connection between our qualitative patient comments published on our website and our TSQM statistics. The statistics are derived from our TSQM surveying. Our patient postings are purely spontaneous interactions between patients and our site.

3) Personally, I find the side-effect frequency data much more compelling than the effectiveness / satisfaction data – and its actually this side-effect information that we get much more patient feedback on. Rather that going to their doctors with satisfaction stats, patients tend to go to their doctors with side-effect stats citing a symptom that they had been having for some time and never connected with their medication. A good example of this, believe it or not, is muscle pain and statins!

4) Most importantly, though, iGuard’s primary purpose is as a medication monitoring system – providing patients with drug safety reports, alerts and recalls. We run the TSQM program so patients become comfortable with participating in research (our business model), and publish some of these statistics only as a benefit of general user participation. It is for this reason that we don’t publish the ratings data head-to-head on our site, nor do we make a business of selling the TSQM data that we collect.