Lessons from Crowdsourcing Technology for Guiding a Patient-Centered Research Agenda
|
If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL). |
Abstract
Background
Severe atopic dermatitis (AD) is poorly understood and treatment options are limited. Wide variation in practice among clinicians suggests limited agreement on standards of care. Further, patients report poor quality of life associated with the condition. In the absence of proven treatment options, patients with AD seek answers online from patient communities and social media platforms. Information posted in these fora provide insight into the real-world treatment and management issues faced by patients with AD, and provide a guidepost for patient priorities for future research.
Objective
Using natural language processing (NLP) techniques to aggregate patient-generated data from social media sites, we sought to understand/characterize the universe of treatments used by AD patients, their experience with these treatments, and their perceived gaps in treatment and management options as well as desires/needs for future research.
Methods
Using NLP and machine learning techniques, we collected and analyzed informal patient conversations conducted on publicly accessible online social networks, blogs, and forums. Employing targeted healthcare NLP technology, dominant themes were extracted automatically from free form online patient communication. Using a computational framework that combines supervised and unsupervised machine learning techniques to enhance language processing and account for idiosyncracies in language, we identified salient themes related to patients’ experience with AD, the breadth of treatments described, and comments on treatment options. We cross–referenced this data with the findings of a literature review published articles describing treatments for AD and quantified the overlap between data points gathered through a natural language processing/aggregation approach and treatments identified in the published literature, noting commonalities and differences and highlighting promising areas for future research.
Preliminary Results
Patient comments from social media venues express significant distress and frustration with managing AD, and wide variation in treatment/management practices. Patients report a need for help with discomfort, cosmetic/appearance issues, and social support. These needs are underserved by current research efforts.
Conclusions
Natural language processing techniques may provide insight into patients real-world experience with managing a difficult-to-treat dermatology condition and offer direction for future patient-centered research.
Severe atopic dermatitis (AD) is poorly understood and treatment options are limited. Wide variation in practice among clinicians suggests limited agreement on standards of care. Further, patients report poor quality of life associated with the condition. In the absence of proven treatment options, patients with AD seek answers online from patient communities and social media platforms. Information posted in these fora provide insight into the real-world treatment and management issues faced by patients with AD, and provide a guidepost for patient priorities for future research.
Objective
Using natural language processing (NLP) techniques to aggregate patient-generated data from social media sites, we sought to understand/characterize the universe of treatments used by AD patients, their experience with these treatments, and their perceived gaps in treatment and management options as well as desires/needs for future research.
Methods
Using NLP and machine learning techniques, we collected and analyzed informal patient conversations conducted on publicly accessible online social networks, blogs, and forums. Employing targeted healthcare NLP technology, dominant themes were extracted automatically from free form online patient communication. Using a computational framework that combines supervised and unsupervised machine learning techniques to enhance language processing and account for idiosyncracies in language, we identified salient themes related to patients’ experience with AD, the breadth of treatments described, and comments on treatment options. We cross–referenced this data with the findings of a literature review published articles describing treatments for AD and quantified the overlap between data points gathered through a natural language processing/aggregation approach and treatments identified in the published literature, noting commonalities and differences and highlighting promising areas for future research.
Preliminary Results
Patient comments from social media venues express significant distress and frustration with managing AD, and wide variation in treatment/management practices. Patients report a need for help with discomfort, cosmetic/appearance issues, and social support. These needs are underserved by current research efforts.
Conclusions
Natural language processing techniques may provide insight into patients real-world experience with managing a difficult-to-treat dermatology condition and offer direction for future patient-centered research.
Medicine 2.0® is happy to support and promote other conferences and workshops in this area. Contact us to produce, disseminate and promote your conference or workshop under this label and in this event series. In addition, we are always looking for hosts of future World Congresses. Medicine 2.0® is a registered trademark of JMIR Publications Inc., the leading academic ehealth publisher.
This work is licensed under a Creative Commons Attribution 3.0 License.