Correspondence


Nature Biotechnology 27, 888 - 890 (2009)
doi:10.1038/nbt1009-888

The power of social networking in medicine

Catherine A Brownstein1, John S Brownstein2, David S Williams III1, Paul Wicks1 & James A Heywood1

  1. PatientsLikeMe, Cambridge, Massachusetts, USA.
  2. Children's Hospital Boston, Harvard Medical School, Harvard-MIT Division of Health Sciences and Technology, Informatics, Boston, Massachusetts, USA.

Correspondence to: Catherine A Brownstein1 e-mail: cbrownstein@patientslikeme.com

To the Editor:

Your editorial in last year's September issue1 outlines some of the ways in which PatientsLikeMe (http://www.patientslikeme.com), an Internet-based social networking site launched in March 2006 to capture patient-reported data for people with life-changing diseases, is opening up new ways of testing treatments and speeding patient recruitment into clinical trials. We would like to provide readers with more detail concerning the utility of the system in alerting patients to new off-label uses of existing approved drugs as well as identifying potential new safety issues.

Currently, PatientsLikeMe has 16 disease communities, which in turn represent information from over 40,000 patients, who can participate 24 hours a day. The US Health Insurance Portability and Accountability Act (HIPAA) regulations do not apply as the information is directly entered by users. These users can anonymously share treatment, symptom, progression and outcome data with the entire community, facilitating knowledge management and transfer for all to benefit (Fig. 1). As the decision to participate is entirely voluntary and the patients control the entry of their data, PatientsLikeMe embodies the sprit of HIPAA.


The platform is freely accessible on the Internet without subscription fees, and is free of traditional advertising, such as banner ads or popups. Revenue is derived through the sale of aggregated, anonymized data, clinical trial awareness programs and market research surveys for the pharmaceutical and biotech industries, health insurers and research institutions. Ultimately, as the ability to more and more effectively measure outcomes evolves, the goal is to directly partner with industry in the discovery process to accelerate the development of biomarkers and new treatments.

Unlike most electronic medical records, which exist to store all information about the patient and serve the practitioner and the needs of their institution for liability reduction and billing, the PatientsLikeMe framework exists specifically to address the integrated health and disease needs of each population it serves.

At the core of the tool is the patient's primary question: “Given my current situation, what is the best outcome I can expect to achieve and how do I get there?” The site breaks “my current situation” down into constituent parts, such as the patient's genetic background, diagnosis history, estimated pathology and level of functioning. By collecting data in a systematic framework, aggregating them and then presenting compact visualizations of the course of a disease, PatientsLikeMe gives patients the tools to answer the question, “What is the best outcome I can expect to achieve?” Finally, the site answers the patient question “How do I get there?” in a variety of ways, whether it is symptom reports, treatment reports or interactions with other peers.

The framework requires patient-reported outcome measures that are psychometrically robust, correlate with the latent construct of disease pathology (where known) and are meaningful to the patient. Where available, the preferred instruments used in clinical trials for drug approval are used (e.g., the amyotrophic lateral sclerosis (ALS) functional rating scale (ALSFRS-R) in ALS2 and the unified Parkinson's disease rating scale (UPDRS-III) in Parkinson's disease3). In some cases, such as multiple sclerosis (MS), new instruments have been developed to overcome shortcomings of existing measures.

Rather than disseminating medical advice, PatientsLikeMe serves as a platform for peers to interact with one another in a data-driven context. Patients are able to speak with other 'expert' patients that have already been through the decisions they are wrangling with and have advice to share. Most of this discourse is publicly available on users' personal pages or on the forum, ensuring it is archived for others to benefit from the learning (Table 1).


One exciting prospect is the opportunity for patients, especially those who do not have years to wait for a clinical trial, to monitor and participate in real-world natural experiments. An example is the use of lithium as a treatment in ALS; when a small study and a number of media reports suggested that lithium carbonate could halt the progression of ALS, many patients immediately obtained the drug off-label4. In response to patient demand, PatientsLikeMe built a live study tool to allow patients with ALS to record their experiences on lithium, including side effects, perceived efficacy, lithium blood levels and any improvements they noted. Despite numerous obvious limitations of the system and the clear understanding that it is no surrogate for a properly designed clinical trial, the tool has offered preliminary assessment on how the drug performed in the 'real world'; in fact, the system worked so well, data were available for hundreds of ALS patients before researchers were able to begin enrolling for a follow-up study.

Another application of the site is to study side effects or adverse events. For widely prescribed drugs, these may not emerge until some time after launch. In 2007, the US Congress passed legislation requiring drug companies to study the safety of medicines after they have been approved for use5. PatientsLikeMe provides services to the pharmaceutical industry that permit them to 'listen' to patients on their medications in a real-world setting while also allowing submission of adverse event data to the US Food and Drug Administration.

In addition to identifying potential signals, the platform can also be used to identify shifts in patient perceptions and behaviors in response to a safety event. The patient sentiment towards the multiple sclerosis drug Tysabri (natalizumab), which was linked with three cases of progressive multifocal leukoencephalopathy (PML) in July and October of 2008, was able to be accurately assessed by context analysis of the PatientsLikeMe MS forum.

An analysis of 141 posts in the discussion forum based on the keyword 'Tysabri', comparing the period from January 1, 2008, to July 31, 2008, to the three-week period after the July PML announcement, found that overall opinions toward Tysabri remained positive. Although patients were discoursing more on the topic of product safety, announcing the PML cases did not adversely affect the overall positive sentiment. This was confirmed by the number of patients taking Tysabri, which increased almost 12% between June 2008 and September 2008 according to Biogen-Idec (Cambridge, MA, USA), the drug's manufacturer.

ADVERTISEMENT

Patients also eagerly evaluate the medications they are taking for their conditions. This provides an opportunity to compare real-world experiences on a medication with those reported by the clinical trials. For example, the main side effect reported by PatientsLikeMe users on selective serotonin reuptake inhibitor (SSRI) Lexapro (escitalopram) was “Decreased sex drive (libido),” at 24% (n = 149), whereas the clinical trial data on Lexapro report 3% (n = 715)6. This discrepancy of sexual side effect reporting in SSRIs has recently been investigated by Landen et al.7. The results of PatientsLikeMe drug evaluations and Landen et al. are in accordance: the actual rate of sexual side effects from SSRI use may be significantly higher than clinical trial reported results. One reason for this discrepancy may be that the open-ended questions used by clinical trials result in an underestimation of sexual side effects6. It is thus possible that the PatientsLikeMe system may be able to detect drug or treatment problems that have not shown up in trials.

Although these findings are promising, challenges in the development of this tool remain. Future work should focus on accounting for reporting bias, misinformation, properly interpreting unstructured information and an in-depth evaluation of the continuously shifting catchment population. Furthermore, with a medical framework now firmly established, next steps for PatientsLikeMe include validation studies to establish that participation on the site improves real-world outcomes, enhanced tools for pharmaceutical companies to help improve their products and services, more powerful tools to empower patients to improve their outcomes, and ultimately changing the pace and quality of evidence-based medicine.



Competing interests statement

The authors declare  competing financial interests.

Top

References

  1. Anonymous. Nat. Biotechnol. 26, 953 (2008). | Article
  2. Cedarbaum, J.M. et al. J. Neurol. Sci 169, 13–21 (1999). | Article | PubMed | ChemPort |
  3. Fahn, S., Elton, R. & Members of the UPDRS Development Committee. in Recent Developments in Parkinson's Disease, vol. 2 (eds. Fahn, S. Marsden, C. Calne, D. & Goldstein, M.) 153–163, 293–304 (Macmillan Health Care Information, Florham Park, NJ, 1987).
  4. Fornai, F. et al. Proc. Natl. Acad. Sci. USA 105, 2052–2057 (2008). | Article | PubMed
  5. US Congress. Food and Drug Amendments Act of 2007, 110th Cong., 1st sess. (2007).
  6. Lexapro [package insert]. (Forest Pharmaceuticals, Inc., St Louis, 2006).
  7. Landen, M., Hogberg, P. & Thase, M.E. J. Clin. Psychiatry 66, 100–106 (2005). | PubMed | ChemPort |

MORE ARTICLES LIKE THIS

These links to content published by NPG are automatically generated.

REVIEWS

Retinoid signalling in the development of the central nervous system

Nature Reviews Neuroscience Review (01 Nov 2002)