Can You Crowdsource a Cure? Using Collective Intelligence to Guide Off-Label Treatment Decisions

Healthcare providers are increasingly being encouraged, and even financially incented, to follow evidence-based treatment guidelines for their patients.
However, what if guidelines don’t exist? Or, what if certain, potentially effective, treatments may help, but evidence to support their use is limited? What is a physician to do?
Charles Rocamboli, DO, an emergency physician at Hi-Desert Medical Center in Joshua Tree, Calif., who previously worked in biotech venture capital development, experienced the problem first hand: Patients were asking about treatments, sometimes not FDA-approved, sometimes falling under the “alternative medicine” umbrella, that they had heard of working for friends. Often, Dr. Rocamboli had heard the anecdotal evidence too, but had difficulty supporting patients’ use of it without more extensive data.
“There’s huge dissection between the info we’re fed as physicians and the anecdotal info we hear from patients and other doctors,” he told me in an interview.
In 2011, he had an idea for a way to help mitigate this mixed information: He would create a search engine-like database for medical conditions, both common and rare, where patients could enter data on how well a certain treatment (approved, off-label or alternative) worked for them. The engine would analyze the data and present it to other users to help them inform their own medical decisions.
After three years of development, Dr. Rocamboli launched CureCrowd in March of this year, at SXSW in Austin, Texas. The site is in beta, but is actively seeking patient data on treatment effectiveness for four common conditions: depression, insomnia, back pain and Celiac disease. CureCrowd will expands its database over time, hoping eventually to offer comparative, crowd-sourced data on a variety of treatments for various chronic and urgent medical conditions.
All data analyzed by the site is user-generated, and users can only create one data entry for each potential treatment. All data entered into the site is reviewed by a physician to ensure it’s “clean” — in a form that can be meaningful analyzed. Physicians review entries and categorize and group treatments. For example, a physician might help guide the system to recognize that Ibuprofen and Motrin are the same drug, so that data on both the generic and name-brand can be combined for analysis.  
The site doesn’t want to be a self-diagnostic site, but instead, aspires to be a registry of sorts to assess the effectiveness of off-label and alternative treatments, compared to more traditional ones, for a number of diseases and conditions. The data can be accessed by physicians, or by patients directly and then shared with physicians, to expand discussions around treatment options.