We identified 123 postings by 95 users that met the criteria based on the key phrases. Among these comments, more referred to treatments (29/123, 23%) than to symptoms or outcomes (9/123, 7%). Almost half of the comments (56/123, 45.5%) included at least one question, and half of these questions were explicit requests for advice (34/123, 28%).
The following are typical examples of comments in three major categories: (1) targeted questions to others with relevant experience, (2) advice and recommendations, and (3) forming and solidifying relationships based on similarity. We also estimated how many of these comments led to ongoing discussion among users. Names of users have been changed.
Targeted Questions to Others With a Shared Experience
When considering a new treatment, one user observed what another member was using and stated:
I notice you are using ginger root and you believe it is slowing your progression. I'm very interested in this. Can you tell me more about how it's working for you?
Another user, also curious about a nutraceutical, conducted a more complete inquiry. He sent almost identical versions of an in-depth request for information rather than addressing a specific comment to each individual on the treatment:
I see you are using Glyconutrients. What are the exact ones that you're using, how long have you been using them for, and what benefits if any have you seen. I have heard a lot of encouraging things about them, but I have yet to hear anything about their use by ALS patients. Are they helping with a particular symptom? Please let me know what you have learned by taking these supplements. Blessings to you and your family.
In such comments, users with a particular treatment question often addressed their question to other members already using that treatment. For the above two cases, the questions were about nutraceuticals and their perceived efficacy. In other comments, users asked about prescription pharmaceuticals, dosage levels, or experience using a piece of equipment. In all of these scenarios, one user with a question apparently identified another user with relevant experience and then asked about his or her perception of the treatment’s efficacy.
The graphic depiction of the length of time on a treatment (in the Gantt charts) pointed one user to identify another as an appropriate recipient for his question. Since this man was considering a feeding tube, he asked a woman on the site about her experience:
Peter: [Jen], I'm a new member of PLM like yourself. I notice you have had a tube for about 8 months. I'm having difficulty eating, so the neurologist suggested I look into getting one. My meeting with the gastroenterologist did not leave me with desire to get one. It would help me if you would send me a message about your experience, pro and con, with your feeding tube.
In this case, Peter explicitly referenced the amount of time Jen had been using an assistive technology as evidence of her value as an advisor. Although he had consulted with health care providers, he sought out another patient’s opinion with the implication that it would contribute to his own decision.
In the above cases, users identified a single feature of a profile then asked an appropriate question. Other users made more sophisticated observations based on multiple charts and data types. For example, Adam, an ALS patient considering the use of a breathing assistance device—bilevel positive airway pressure (BiPAP)—asked the following:
Adam: Hi [D] I am [Adam] in the PLM web site. My als was like yours breathing onset. I see your FRS improved a bit after you went onto BIPAP in april 06. Did it in fact make that much difference.??
To ask this question, Adam apparently cross-referenced two charts in the profile to see the relationship between beginning to use various interventions, including a BiPAP, which is displayed in one type of chart, and experiencing improved function, which is displayed in another. In this case, the relationship looked clear (). In fact, the displayed clarity of the relationship appeared to give Adam pause since he asked for confirmation.
Patient profile of PatientsLikeMe member “D” (with added explanatory remark "Based on the charts..", which is not part of the original screenshot)
In the preceding examples, users with a specific question identified another member and addressed their question to him or her. The criterion leading to that identification appears to be simply taking the medication or using the treatment. In one case, a user referenced the amount of time a member had been using a technology as a factor in identifying an individual as a credible resource. Using other members as a resource to inform treatment decisions emerged as a reoccurring use of comments within the site (29/123, 24%).
Advice and Recommendations
Browsing the site, users frequently posted their remarks on one another’s profiles, in some cases sharing their own relevant experience. One man observed another’s symptom:
I see you note emotional lability. I had that very bad, but now I take a compound of dextromethorphan and quinidine that controls it beautifully.
In this case, a user offered personally acquired knowledge to another member listing a shared symptom (depicted within a modified Gantt chart). This was not an isolated instance: in five of the comments users provided similar recommendations specifically around observed symptoms, including bed sores and cramping. In each case, the comment offered advice based on a positive personal experience and included a treatment recommendation and a method for administering that treatment.
Users’ advice went beyond sharing personal treatment and symptom experiences. Within the “about me” section of the profile, many users provided their city and state of residence. Some members reading the profile referenced this information to make geographically appropriate suggestions. For example, one patient wrote to a caregiver on the site about a local support group:
Hi [Bill]. ... There will be another ALS support group starting up next Tuesday Feb.20th in Holt. Just checking to see if your parents were interested...
In similar references to location, four users either mentioned or explicitly invited others to support groups.
In one case, a user noted the individual’s region and type of onset to suggest a research study on a new technology designed to address her specific situation:
George: [Joanne], I see you are legs onset; have you heard about the new diaphragm implants they are doing at Case-Western and Johns Hopkins? It means you don't have to vent to breathe.
George had a piece of information about a location- and topic-specific resource available to patients. He referenced Joanne’s diagnosis history (onset type) to make a recommendation. Using posted data, he was able to connect the individual to the resource.
In these cases, users offered advice and recommendations to others. In most cases, these recommendations stemmed from personal experience with taking a drug or using a device, but they also stemmed from personal research, as in the last example where a member offered knowledge of a research study to another user. In other online applications, individuals may share personal experience through messages broadcasted to a large audience. In this context, users delivered targeted messages to particular users they think may benefit from them.
Forming and Solidifying Relationships Based on Similarity
Comments also function as a mechanism for creating and maintaining relationships, particularly around points of similarity: 25% (31/123) of the comments we analyzed identified a shared attribute, hobby, or concern within a broader comment or question. Locating a similar patient, one member quoted what they had in common medically as a basis to invite further contact:
Hi [Michael], I see we are pretty similar. I am 62 dx 11/06 with leg onset. I need a walker to help me walk. I move slower and have had a few recent falls due to my leg dragging. I would like to be available if you want to compare progress. I started noticing symptoms a year ago, but just dx this month.
Referencing diagnosis history, this user made a connection with another member in the community. For patients in unusual situations, the site allows for finding a similar individual even when there are only a few. In the following case, another user expressed her pleasure in finding others with a shared but atypical disease progression:
hi [Rachel]. yes same boat indeed. i am so glad to find this site because i see there are many of us with slower progression than stereotypical. the support groups locally really focus of immediate need patients and us long timers are not so immediate except we still have concerns and fears, etc so it has been so great to see how long timers cope with losing our function slowly and wondering which part is going to fail next. hers my personal email; … id love to talk more.
In these cases, the first patient had explicitly invited further contact, and the second suggested a willingness to share data even beyond the anonymous structure of the site by giving her personal email address.
In addition, there were examples of patients seeking out others based on non-medical criteria: 18 of the 31 comments on similarity were based on non-medical criteria including location, employment history, astrological sign, and shared interests. As with the medically based examples, the site facilitated meeting of people with shared concerns—people who probably would not have met offline.
Initiating Ongoing Discussion
All 123 comments were analyzed as single units, but the reality is that comments may occur within ongoing exchanges. Without looking at the contents of private messages, we examined the full exchange of comments and messages between the sender and the recipient of all the comments studied. We found that these comments served to both continue an exchange between the two users (59/123, 48%) and initiate new exchanges (64/123, 52%). In the initiate cases, 56% (36/64) of the comments received at least one reply. In more than half the replies (20/36, 55%), the recipient continued the exchange in the “public” sphere of the site either through comments only (12/36, 33%) or through a combination of private messages and comments (8/36, 22%). On the other hand, among the cases where a comment emerged in an ongoing exchange, 57 of 59 comments were responded to, with comments being used in 68% of the exchanges.