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.
"We really are different than anything out there because, first off, we are in no way a diagnostic site. A lot of websites [prompt users] to put in symptoms and self-diagnose. That is a somewhat dangerous process," explains Dr. Rocamboli. "The ability to diagnosis is truly is physician skill set. What treatments work, you can look that up in a book."
But for those treatments that aren't yet in the book, CureCrowd could one day be the go-to resource for physicians and patients. While the treatment data will remain anecdotal, using larger-scale anecdotal evidence to inform treatment decisions is better than having patients make decisions based on a single tale from a neighbor or coworker.
Take for example, the site's data on depression treatments. Based on user-generated data, Celexa is the most effective drug to treat depression (see screenshot). Yet, running and yoga were rated as more effective overall than any drug, including Paxil, Prozac, Zoloft and Cymbalta.
"If one person says it, it's one thing, says Dr. Rocamboli. "If a 1,000 people say it, then that's another thing."
Such a database could be especially helpful to patients with rare conditions. More patients are diagnosed with rare diseases than AIDS and cancer combined — and less than 400 FDA approved drugs are available to treat more than 7,000 rare diseases.
"We lend value to off-label and alternative use," explains Dr. Rocamboli.
"Some very basic [naturally occurring] medicines like morphine and insulin would be not be developed by pharma today," since pharmaceutical manufacturers tend to invest in synthetic drug development, he explains. "Today, those would be pushed off as alternative medicine.
"We circumvent the issue by aggregating data an giving leverage and justification to use that doesn't have as much research supporting it," he adds.
For patients with rare conditions, or for whom traditional treatments haven't taken hold, a database like CureCrowd could be just what the doctor ordered.