The Virtual Choir

The Virtual Choir

Robert S. Miller, FASCO

Apr 06, 2012
In a previous life I used to be a church musician. I started piano in the 1st grade, and since I attended a pretty traditional parochial elementary school, it was natural to gravitate in this direction. Over the years I’ve been involved in church and school choirs as a pianist, organist, singer, arranger, composer, and conductor, and I did this in high school, college, med school, and beyond, as recently as about four years ago. Lest you confuse me with someone with real talent, let me say I was never classically trained and I would probably embarrass myself pretty badly today if I tried to perform anything complicated. But I’ve long maintained my love for classical choral music, both sacred and secular, and I’ve always greatly enjoyed listening to choral works on my iPod.

A few years ago, while browsing the iTunes store, I discovered a composer and conductor named Eric Whitacre, who has had a somewhat meteoric rise to fame and international acclaim in this genre. He has an interesting personal story. He grew up in rural Nevada in the ‘70s-‘80s and attended UNLV, not exactly known as a powerhouse for classical music. As he tells it, even though he could sing, he couldn’t read music when he started college. On a whim, he joined a college choir and almost immediately had a transformative experience with the music and the blending of voices that led to a BA in music, then a Masters at the Juilliard School, and over the next 20 years lavish praise, fame, and multiple awards as a prolific composer and conductor of choral and symphonic music. I’m sure it doesn’t hurt that, now at 42, he is described as “boyishly handsome” and is an articulate and passionate speaker. His music is known for its dense harmonies, dissonance, and unusual rhythms. I find it very beautiful and challenging. But what I really found fascinating was an experiment he started in 2009 with something called the Virtual Choir. It’s probably easier to watch this 11-minute “TED Talks” video, which went insanely viral last year, than read my summary, but I will try to be brief. It seems that a young girl sent him a fan video on YouTube of her singing the soprano part to one of his pieces called Sleep. He was struck by the tribute, and he recognized the purity and sweetness of her artistry. He then had an inspiration about how to harness the creativity of multiple other would-be performers who might be singing to their video cameras and uploading to YouTube, hoping to create something beautiful that would make them famous. So what he did was to put out a call, using social media, encouraging singers all over the world to video themselves singing one of the parts of his work Lux Arumque. He ended up receiving almost 200 hundred videos from a dozen countries. Then with a collaborator, he assembled his Virtual Choir into a single video, with himself conducting these virtual voices, and the results are, well, astounding. He went on to produce Virtual Choir 2.0, with over 2,000 separate singers from dozens of countries, and then just this week released the video of Virtual Choir 3.0, with 3,746 singers from 73 countries performing his work Water Night. I know you are busy people, but take 30 minutes out of your evening to view all three of these videos. Even if you are not a fanboy like me, the artistry is remarkable, and the music sublime and exhilarating.

So what in the world does this have to do with medicine, oncology, or ASCO Connection? As I followed this story, it struck me how there are some parallels with health care and the experience of illness. In the Virtual Choirs, Whitacre and thousands of others with the common interest of choral music were coming together to share an artistic vision and create a larger work. While each singer was recording an individual vocal track in the privacy of his or her own bedroom, there was an undeniable social connection. The need to connect with other people who share our interests and experiences is one of the strongest forces of our humanity. And I see this every day in the practice of oncology. Yes, this is the era of personalized medicine, but so many of our patients have a fundamental need to connect with other patients and families that are going through the same thing they are, a need that we as health professionals can’t possibly fill. People need to learn from each other in a social and collaborative fashion, and the wonders of the Internet and the explosion of so many social media channels today have enabled this to a degree never before possible. Much research has confirmed this, and while patients will usually come to their physicians for the authoritative voice we have, they still need to process this knowledge through the filter of other people. As oncologists, we would be foolish to deny this reality, and I would maintain that a much greater good could come from our efforts to facilitate it, if not harness it for a larger purpose.

But for us as physicians, it’s more than simply allowing patients to share their war stories and their tips on dealing with treatment, with some type of blind but vaguely disapproving eye. I think this social urge is a greatly underutilized resource for discovery. Think of the whole concept of patient-reported outcomes. As physicians, we know or should know that we generally do a lousy job in assessing the patient experience of illness. In our field this very pointedly is the patient experience of treatment toxicity, fear of the disease process, and loss of control. The data are so much better when we let patients tell their stories directly. And think how easy it would be to enable this with all of the simple and widely accessible technologies available today, like YouTube and cell phone texting. As investigators, we could create our own Virtual Choir of our patients, individually and collectively, contributing content that we can amass and analyze for trends and causality. Think how simply elegant an experimental design could be. Say we wanted to know the temporal trend of patient-rated bone pain following a Neulasta injection. Sure, we think we know this from existing clinical trial data, but mostly it’s what we’ve asked our patients to tell us. What if we told every breast cancer patient getting dose-dense AC starting on day 2 of each cycle to text an integer value from 1-10 corresponding to their level of bone pain to 11435? And what if the data were displayed real time on a website that refreshed every ten seconds with all other participants’ results, so that people feeling lousy in their homes that night could see other peoples' scores and know they weren’t alone? I would predict that this collaborative content creation would enhance study compliance like no $25 Starbucks gift card ever could. And if instead of a Virtual Choir of 2,000 or 4,000, we had tens of thousands of patients doing the same thing? Talk about Big Data! Yes, our current broken clinical trial system can’t support something like this. But I believe we can figure out a better way to harness the power of social media, collaborative data creation, and the patient voice to overcome these barriers and transcend our current maddening limitations.

Whitacre had a vision of musical creation that transcended a single concert hall or venue, and the results were breathtaking. In our field, I believe we are no less creative and passionate. What are the provocative questions we can answer by being as visionary and bold?


The ideas and opinions expressed on the ASCO Connection Blogs do not necessarily reflect those of ASCO. None of the information posted on is intended as medical, legal, or business advice, or advice about reimbursement for health care services. The mention of any product, service, company, therapy or physician practice on does not constitute an endorsement of any kind by ASCO. ASCO assumes no responsibility for any injury or damage to persons or property arising out of or related to any use of the material contained in, posted on, or linked to this site, or any errors or omissions.


Howard Jack West, MD

Apr, 06 2012 7:16 PM

  I've been thinking about many of the same issues, and like you, I've been intrigued by these collaborative projects, including Eric Whitacre's Virtual Choir along with more lowbrow but functionally similar efforts like the Johnny Cash Project and the Star Wars Uncut: Director's Cut project.  It's all part of the concept that people from all over can contribute a little to produce something much bigger.     

   While I agree that pooling patient data the way that PatientsLikeMe is striving to do can add to our understanding, I think that there's a lot of potential in developing a "poor man's cooperative group" by focusing on situations that are uncommon to rare, like adenoid cystic carcinoma or large cell neuroendocrine NSCLC, solitary and potentially "precocious" brain met, etc., and creating some kind of very basic template of saying "let's just plan to treat these people in a uniform way, off protocol, then report how people did."  This is not research in the same sense that it would replace a prospective randomized trial, but in situations where we might see one patient every year or two or three with a situation, we can either treat them in a solitary way, never have anyone learn anything from it, or we can aggregate the data from people everywhere and create a "virtual collective case series" that would be far more powerful and larger than any single institution experience of a dozen patients over a decade or more.  If you or others are interested, I wrote about this in a little more detail in my own blog on Medscape.   

   One point worth making is that as oncology becomes more molecular, more granular, we're breaking up big categories into much smaller subgroups that aren't amenable to the same "big trial at 400 centers" strategies that we used a decade ago.  You can't do 1000-patient trials on 2% subpopulations; these patients are scattered all over and aren't seen at most centers in a high enough concentration to justify opening many of these clinical trials.  We need to develop new ways to develop clinical trials to address these newly defined, geographically dispersed, small subgroups.

-Jack West

Back to Top