Director's message archive
So what’s a population anyway?
Friends of PHS, and especially those of you who have used our data collections, appreciate the difference between a bona fide large population sample, and, say a group of patients with a particular disease, like cancer or IBD.
While such registries have been and remain invaluable for the study of clinical populations—especially when linked up with high-quality biologic and other markers such as images or tissue samples—but in general such registries cannot teach us much about the origins of disease, nor the relationship between what’s in our genome, the way we live our lives, and our risk for getting sick in the first place. We may assess those critical relationships only from the study of broader swathes of the population. Sometimes these “populations” are just large, assembled groups of volunteers, as we now see with our Baseline Study, our “Well” study and with national volunteer collections such as “All of Us” or the UK Biobank, in which genomic and other biologic data are linked to so-called “metadata,” typically a small slice of demographics and a clinical phenotype—viz. a 63 yo AA type II diabetic..
However, because such populations include people who join for their own unique (and unknowable!) reasons, and about whom we typically have very scant information, there are limits to what we can learn from such “populations.” Moreover, there may be substantial inferential pitfalls in trying to over-analyze as have been learned from the experiences with GWAS when people of different race and ethnicities were included without concern for how different they might be from others in the dataset. For example, there may be many reasons more Blacks than Whites get hypertension than gene mix. The bottom line is that the information about the people’s lives is often too skeletal in these types of “populations” for appropriate comparisons between and among them to be made.
For this reason, we are always excited when we gain access to populations which do share some critical common features such as geography or culture, groups of people for which we can reasonably infer that those things that cannot routinely measured are likely shared. So it is with great enthusiasm that we report some wonderful new developments this past month at PHS. Starting right here at home, the long-awaited data from the Health information Exchange in Solano County—flagship we hope to a Bay Area data lake some day—has begun to flow into a sandbox where we can for the first time explore both aggregate data on the almost ½ million residents and well as the more granular data from the 20% of the population who get their care from County operated clinics. Pilot studies to assess the quality and utility of these data have begun, and we are hopeful to be able to accept project proposals by Fall.
Looking a bit further away, our research partner in Bradford (UK), host to the remarkable Born in Bradford cohort of 12,000 children born 11 or so years ago has just hit the double-jackpot: a 10 million pound grant from the UK to launch “ActEarly,” a broad platform to study how community, school or health based interventions can impact health outcomes, and the buy-in from the local government to make all of the available administrative data sets from the city of 700,000 people available to the researchers. As long-standing collaborators and advisors, we have the broadest possible invitation to propose studies and avail this remarkable data model, which is available directly through our PHS portal.
And in case you were worried about the flow of future opportunities, we were invited this month to assist the Indian government in developing a national data eco-system, built on the model which has made PHS a hit at Stanford. As the world’s largest democracy becomes increasingly tractable for both observational and experimental study, is there any question left we will not be able to answer?
—Mark Cullen, MD, PHS Director
FEBRUARY 28, 2019
As winter quarter grinds to a soggy end, spring is bursting at PHS as many of seeds planted in months—even years—past begin to sprout. Three of the most exciting developments are highlighted in this month’s issue, but many more are about to bloom.
Air quality and health is at the front of our minds, and, after a recent convening this past quarter, the working group has seen a significant boost in momentum. Spurred by the success of two PHS Faculty Fellows—Eran Bendavid and Marshall Burke, who published an award-winning Nature article demonstrating the impact of particulate matter in driving infant mortality in Africa, we are excited to see what new research directions, resources and impact this group will cultivate. Additionally, generous support from Kari Nadeau’s lab and the Bravo Family Foundation has resulted in a post-doctoral fellowship to leverage the a new, rich aerial surveillance method to quantify local air quality as a means of assessing the health impact of California’s fire crisis. Soon, we hope to be able to make the data widely available to our PHS users for similar studies of environmental impact on a vast array of health outcomes spanning the life course. The working group meets next on Thursday, March 7, 2019. Email Rita Lonhart for details.
From the outset PHS has had multiple global partners—notably in Denmark (Aarhus University and the Statens Serum Institut), the United Kingdom (Born in Bradford), Israel (Clalit), the Czech Republic (Recetox) and Taiwan—but our imminent launch in India represents a sea-change. Recently announced, PHS is developing an Indian satellite program, with a footprint in Mumbai. From that Center we will be replicating the Data Core we have built on campus, amassing data sets and collaborating with partners including the Bill and Melinda Gates Foundation, Google, Indian Institute of Technology (IIT) bombay, PATH, Tata Trusts, WISH Foundation and Wadhwani AI, and both State and National governmental agencies. The resource will be widely available to faculty and trainees for studies of health and development, and we will soon have the ability to provide logistic support, data services and office space for Stanford researchers working in-country. The initial focus will be on maternal and child health, sanitation, health service and use of AI for social good, however we expect to facilitate data access in support of studies focusing on a wide range of social science, environmental and biomedical issues.
Another data asset available in increasingly robust and user-friendly form is the American Manufacturing Cohort. This database is comprised of multiple, linked datasets that collectively provide an invaluable resource for understanding work-life exposures, health, and economic outcomes in a cohort of over a quarter of a million U.S. light and specialty metals workers and their dependents. Extensive investigations of longitudinal occupational, environmental, social, economic and health exposures and outcomes over the life course have already generated nearly a hundred papers. Now, with extensive efforts in the conversion to a common data model, data cleaning and upgrades to supporting metadata, these data will become more available to the researchers across campus for appropriate studies. A new and exciting venture for PHS, a dedicated website is being developed (expected to go live in March 2019) to detail the history of the data, body of research to date, current collaborators, and facilitate access to data and documentation to a broader audience of investigators across disciplines. We hope this will serve as a model for things to come. For more information about accessing these data, please contact Erika Tribett or Emma Hallgren.
Which reminds me — in so many ways it’s all about the data. I am usually loathe to boast, but I could not resist sharing this email from a colleague recruited to UCLA:
With access to the PHS data these past few years, I feel I have been able to make meaningful contributions to the field of GI. My group has had multiple (I have lost count) oral and poster presentations at national/international conferences, been featured in specialty newsletters, and won a “Best Clinical Poster Award” at a national conference. I first-authored 2 publications in Gut (2nd highest impact factor for GI) and co-authored 3 publications in high-impact/well-known GI journals. All of this would not have been possible without the PHS data. So, I personally thank you and your wonderful staff for the incredible opportunity. Of note, my upcoming chief in GI and co-director at UCLA are well aware that this will be a resource I will greatly miss. There is no resource like Stanford PHS there or elsewhere.
But our data team never rests on its laurels. In the coming months we will be both expanding the range of data (stay tuned!) and undertaking projects to improve these data for researchers across campus. This includes further legal research to simplify the dreadful DUA process for new datasets; continued efforts to tame the complex legal and ethical landscapes complicating research projects; developing new, standard ontologies to allow harmonization of data of novel kinds; and, in a collaboration with Google of potentially game-changing scope, a project to allow linkage in common data models for our (mostly) “high-risk” data with the literally millions of low risk data sets through machine-learned common data models. And because we live in a world of collaborations, we are exploring with many of our peer institutions the possibility of a shared data platform in the months to come. So keep our links handy: phs.stanford.edu and phsdata.stanford.edu!
—Mark Cullen, MD, PHS Director
OCTOBER 22, 2018
As we all embark on the new academic year, the Center for Population Health Sciences (PHS) would like to share several new initiatives and invite you to fill your calendar with upcoming events and reach out to get involved.
For the next several months, much energy will focus on large-scale projects enjoying ongoing support. This includes the public availability of our American Manufacturing Cohort (AMC) data--longitudinal administrative and health data for over 230,000 U.S. manufacturing workers and their spouses and dependents to investigate occupation as a critical source of exposures impacting economic and health outcomes over the life course. We are also focusing on the development of a health information exchange in Solano County that the we expect will expand to all counties in the Bay Area. This initiative--BRIDGES (Bay Area Research Innovation Discovery: Governance, Evaluation and Sustainability)--is a collaboration between public health departments and their constituents, community clinics, health systems, and academic researchers to develop infrastructure and processes for scalable aggregation, interoperability and secure sharing of a variety of data including EHR structured and unstructured data to address research questions of public health importance (for more information, please contact the project's PI, Dr. Lorene Nelson). We also continue to work with the Bill and Melinda Gates Foundation to take a deep dive into understanding the role of gender norms in sex-specific morbidity and mortality, and a Lancet paper series will soon present preliminary results. We continue collaborative research and accepting proposals for data available through our partnerships with Clalit, Born in Bradford and Denmark.
At the same time a myriad of new projects are taking form:
- Assessment of the relative health impacts of controlled fires--the best known prevention for the current fire crisis--versus the respiratory and cardiac risks of the wildfires which result in the absence of such burns. A new fellowship in environmental health will commence with generous support from Kari Nadeau’s lab.
- We will collaborate with the Alzheimer’s Disease Center to study new population aspects of the epidemic, including the role that work factors play in long term risk for dementia, and the disparities in social supports and care among different ethnicities. Two recent grant supplements from the National Institute on Aging will support this work.
- Our Office of Community Engagement has embarked on several major cancer-related initiatives with support from the Cancer Center and PCORI. These include the Peer Navigator program to raise awareness of breast cancer among African American women in the East Bay and the war on melanoma; a proposal to address the potential benefits of targeted “Precision Health” efforts among Latinos in Salinas is under evaluation as this multi-faceted effort expands exponentially.
- Efforts to better understand the epidemic of opioid abuse, suicide and other “diseases of despair” are accelerating in collaboration with the Veterans Administration and with colleagues at UC Berkeley. Several projects are burgeoning, boosted by the recent visit of Edmon Begoli from the Oak Ridge National Laboratory, which has taken the national lead in data management for governmental agencies.
Behind each of these successes lies the continued growth and success of our data core. We have been flattered in recent days by requests to share our technology with many peer institutions, the National Bureau of Economic Research, and a host of domestic and foreign NGOs. Indeed, the data core and its infrastructure and data governance has resulted in the enthusiastic invitation to launch a satellite data center in India, likely to occur early in this academic year. This will undoubtedly be the subject of a forthcoming symposium in which we look forward to your participation. Stay tuned!
—Mark Cullen, MD, PHS Director
JUNE 7, 2017
End of Year Wrap-Up
By any standard it has been a very satisfying year for the team at PHS, despite—or perhaps in part because of—the extraordinary events we did not see coming. Over 730 of our Stanford colleagues and trainees have become members; more than 1000 came to one or more of our events. And what events! Drawing on ideas from each of our working groups, our seminar series brought together many of the threads that connect our work: Dalton Conley (Princeton) and Steve Cole (UCLA) explored various mechanisms by which our lives get under our skin—and hence the origin of most health disparities—including neuro-humeral and epigenetic pathways among others. Nicole Bush (UCSF) and Ursula Staudinger (Columbia) presented evidence that social conditions are strong determinants, respectively, of mental development and cognitive decline at the extremes of the life course. Peter Muennig (Columbia) and Sandro Galea (BU) presented a compelling case for public health activism.
If these invited guests roused interest with their presentations, response was no less overwhelming for our global data partners who came to present collaborative research opportunities. In January, we welcomed our colleagues from Clalit, the largest healthcare payer/provider system in Israel. Their visits attracted an outpouring of interest, and research projects have already been launched, the potential limited only by resources and distance until the day we can streamline access to the servers where data reside. In February, we hosted our colleagues from Born in Bradford, a UK-based prospective pregnancy and birth cohort established to examine how genetic, nutritional, environmental, behavioral and social factors affect health and development during childhood in low-income communities.
Speaking of servers, over 100 “users” are clocking time on our data portal, engaged in some 200 projects. And we haven’t even moved to the Google Cloud, a transition we expect (in the Fall!) will make the experience immeasurably more satisfying.
All of this spurred our now 14 working groups to “go the next mile”, exploiting the administrative, data and community engagement support we were able to offer. Four resulted in plans for large jointly sponsored symposia. In May, we hosted a timely and important “Child Health and Immigration” conference. Events on sex and gender, aging and gene-environment interaction are in the offing for the Fall. Our emerging Pacific Rim Alliance for Population Health (PRAPH) will bring together colleagues from 8 countries in Taipei at the end of September to share research opportunities for our members and some of the most advanced universities in Australia and Asia. The four newest working groups—adolescence; behavior and health; food and nutrition and the environment-- appear each to have struck a chord, as we begin programming for the next academic year. Interested? Let us know!
We were fortunate to fund two rounds of Spectrum PHS pilot grants this year, awarding 30 transdisciplinary projects close to a million dollars. These awards continue to drive our repertoire of projects, with several already producing manuscripts and attracting outside funds. Notably, many will engage with our expanding catalogue of datasets, including a handful slated to use the Health Information Exchange in Solano County, and the EMRs of the local Federally Qualified Health Centers (FQHC) as proof of the concept to study whole populations right here in California. A recent meeting with health officers from all nine Bay Area counties signaled some considerable interest in expanding this approach over the coming years, a prospect made even more appealing by the evolving negotiations with state agencies for a wider swath of California administrative data.
Finally, we are very excited to announce our first cohort of faculty fellows! We are honored to have such an outstanding cohort of scholars committed to advancing PHS’ mission:
· Sanjay Basu, MD, PhD, assistant professor of medicine. Professor Basu has significantly expanded PHS’ health disparities research agenda.
· Eran Bendavid, MD, MS, assistant professor of medicine. Professor Bendavid has been working closely with the School of Earth System Science to advance PHS’ work related to the environment and health.
· Suzan Carmichael, PhD, professor (research) of pediatrics. Professor Carmichael continues to contribute to PHS’ research on the first 1,000 days of life.
· Lisa Chamberlain, MD, MPH, assistant professor of pediatrics. Professor Chamberlain is supporting PHS’ efforts to strengthen its advocacy and community engagement efforts.
· Liran Einav, PhD, professor of economics. Professor Einav has been instrumental in helping PHS create a university-wide administrative data center.
· David Grusky, PhD, professor of sociology. Professor Grusky, bringing the perspective of a different discipline, is also actively engaged in this project (above).
· Latha Palaniappan, MD, MS, professor of medicine. Professor Palaniappan will be teaching the Practice in Medicine course this fall, integrating key PHS concepts and principles into the curriculum.
· David Rehkopf, ScD, MPH, assistant professor of medicine. Professor Rehkopf helped to successfully launch PHS’ gene-environment working group, engaging faculty from across the university.
· Nigam Shah, MBBS, PhD, associate professor of medicine and of biomedical data science. With advice and guidance from Professor Shah, PHS has made considerable progress toward the creation of an integrated data commons (datasets will be transformed into a common model and sit on a cloud-based platform that includes innovative visualization and analytic tools that make it easy to access, use, and link very large datasets).
Our fellows will serve as our “inner circle,” providing a sounding board for our plans and ideas. In addition, they will assume the role of PHS “ambassadors,” representing us at internal and external events, meetings, and research activities. As our circle of faculty fellows expands, as we continue to add representatives from across the university, we hope to have them serve as focal points for interest at key schools and departments (recalling that PHS, as part of Spectrum, is administered by Ann Arvin and the Provost as an independent center).
For those of you who will be leaving campus for the summer, we wish you a great break and look forward to reconnecting in the fall. For those of you who will spend at least part of the summer here, keep checking our event’s page for summer activities. In addition, there will be no better time to drop by and visit 1070 Arastradero, meet our Community Engagement and Data Core teams, and learn more about the opportunities for you and your colleagues.
—Mark Cullen, MD, PHS Director
MAY 1, 2017
A Nuanced Perspective
Well we are now three months into the new political reality, and mercifully most of us have moved on from the incredulity and grief stage into confronting the only issue that really matters: How shall we proceed with our work to improve the present and future health of the US and world’s population?
At first blush the answer appears simple: we should proceed with our core missions of advancing the scientific knowledge base, and the training of those who continue carrying the torch, even if for now policy advocacy may appear frustrated. In fact, it is probably wise—maybe even critical-- to direct our energies towards defending the resources necessary to do both, resources that many across the political spectrum respect and even support. Serious things, like the budgets of our science agencies and the funds needed to guarantee our schools and universities are not eviscerated.
But under the surface there is a more nuanced perspective. For one thing, it is crucial that even as we fight for science (as many of us did last week, here and across the country), we must not be seduced into the view that scientific “facts” will assure that our perspectives on human health and environment will prevail if only we generate more data and more certainty. The truth is, as psychologists have known for over fifty years, that even for many scientists our “posteriors”—what we glean from the results of our research—are most often seen through the lens of our “priors”—our conviction before we did the research. For lay people the problem is a thousand times more intense: results are almost invariably viewed as “proof” of those earlier beliefs, no matter how irrational or unfathomable. If we are to change the dialogue and assure the success of our (scientifically defensible) agenda, we will need more than just science—we will need to understand how and why our adversaries have come to their convictions, starting from common ground. So if, for example, many may fear environmental controls as job-killing (like ACA, you may recall), perhaps we should engage on the extraordinary potential of environmental reclamation to spur economic growth rather than the impacts on health and environment. That needn’t mean we should stop studying the health and ecologic impacts, of course; in fact, the current policy environment is likely to spawn myriad natural experiments in which we can learn faster than we might prefer just how important pro-health policies really are. We need only the faith to believe that a better time to roll out these observations will come.
But even the notion that we should abandon advocating for better policy in this climate may be self-defeating. True, we will unlikely see major progressive change from Washington any time soon, and in that domain the best we can hope for is a good defense of past progress. But at the same time it is not at all clear how states, cities and yes, even large corporations may react this political climate. How many other governments will follow California’s lead on environment and immigration? And do we think that big companies will blithely revert to old (very very) bad habits of dumping and polluting? Will their Boards or consumers support such behavior? My own speculation is that we may find some very strange new bedfellows in the coming months; we should not abandon hope that once a few states and a few prominent companies defy the national race to the bottom the complexion of the populist revolution may start to change.
Bottom line: now is not the time to hide away and dream for a better day, but to double-down on the hard work we are doing, with the sure conviction that this will be rewarded, perhaps sooner than we think!
—Mark Cullen, MD, PHS Director
FEBRUARY 9, 2017
Translation Medicine: A New Paradigm Emerges
The holy grail for biomedicine is discovering life-changing drugs and devices and then bringing these discoveries to the bedside. No one is talking about it much, but there is a paradigm shift occurring in biomedicine’s sacred pathway from discovery to real-world applications that benefit society. This pathway has traditionally involved five phases:
· T0: Discovering a potential for a new treatment in a laboratory based on advances in scientific knowledge
· T1: Testing its safety on humans (i.e., is it safe for humans to use?)
· T2: Assessing the efficacy of the drug or device (i.e., how well does the drug or device work)
· T3: Testing its effectiveness compared to the best known treatments or placebos
· T4: Demonstrating its real world health impact on large populations (i.e., under real world conditions do people adopt the drug or device and does it improve health outcomes?)
There is no denying that a fabulous armamentarium of drugs and devices have traversed this path, and we are much the better for it. We can only be thankful for the myriad cures and palliatives that have contributed so vitally to our present ability to treat and modify once life-ending or life-destroying conditions.But, the truth is, that this schema is premised on two sets of assumptions that no longer apply: First, that the goal of all translation is treatment of disease; in fact, we now face the bolder societal challenge of learning enough to prevent disease (with the same precision with which we cure diseases). Second is the increasing recognition that there is more to enhancing health and preventing disease than drugs and devices. If social and behavioral determinants are as important as evidence now suggests, disease prevention and enhancement will demand far better appreciation of how we can effectively modify the ways we live as individuals and populations based on “diagnosis” of the most effective ways to achieve this. Even the delivery of healthcare itself must be researched and optimized, as we are fast learning, not something a drug or even a device will likely fix.
The troves of data newly available and linkable, along with the analytic and computational tools to exploit them will almost certainly change the translation paradigms; it seems likely much of the future discovery will occur in “dry” labs like PHS. And where the object of intervention is social, behavioral or environmental change, and the target is whole populations, it would appear equally likely that new techniques and methods will evolve to supplant traditional clinical trials for demonstrating effectiveness.
It is probably too early to draw the map of this brave new world, or even to predict how it will intersect with the time-honored translational “line” plotted above. But one thing I will boldly predict—as a biomedical research community we had better prepare to translate across two dimensions, and likely even more!
—Mark Cullen, MD, PHS Director
SEPTEMBER 26, 2017
A Welcome Message from the Director
On behalf of the faculty and staff here at the Center for Population Health Sciences, welcome to the new academic year. Hopefully you have all had a restful summer. We all did, but of course our Center never sleeps. In today’s bulletin, I invite you to explore the many new and exciting offerings on our ever-expanding website as well as our calendar of upcoming events.
Since our data portal and fabulous team of data brokers are always a hit, I start by highlighting our new datasets and tools. We now have over 40 datasets from around the world and are rapidly expanding our trove. Over the summer, we added new search tools that make it easier than ever to search, visualize and link these datasets (new homegrown tools are also on the way!). For those new to PHS, or who have been eager to use key datasets but need to learn more about them, we are continuing with our formal data training sessions this Fall. Possibly most exciting of all is the arrival to our team of Amy O’Hara, former senior economist from the US Census Bureau, as our new Associate Director, Data. She will be leading our efforts to host new datasets from across the campus and integrating our work with the Federal Statistical Research Data Center on campus.
Fresh from its stunning mini-symposiumhighlighting the accomplishments of our Stanford Precision Health project (and drawing an audience of 265 in August!) our Office of Community Engagement (OCE) has expanded its support for community-engaged research. With a new Associate Faculty Director, Community Engagement, Lisa Goldman-Rosas, and research teammate Kendra Smith, opportunities will include assistance with study design, community-engaged research strategies/methodologies, and evaluation. We encourage you to visit the site and see for yourself!
In addition to Amy and Lisa, we are pleased to welcome Latha Palaniappan, Professor of Medicine, as our new Associate Faculty Director, Education and Lisa Chamberlain, Associate Professor of Pediatrics, as our new Associate Faculty Director, Policy and Advocacy. Latha will be leading our expanded programs for medical student and post-graduate medical education while Lisa will bring to PHS an expanded repertoire of projects she has developed for trainees in her own department in this critical time for the future of health and healthcare in our country.
Not surprisingly with the influx of new data, faculty and scientific staff our repertoire of projects and collaborations is growing exponentially (far beyond the scope of this letter to even summarize). I will mention only a few “hot spots” from my own calendar. I just returned from the Research Center for Toxic Compounds in the Environment (RECETOX), located in the Czech Republic (Brno), that serves as the central “lab’ for EU environmental studies and hosts a now 25-year-old birth cohort (yes, fantastically, born the year of the Velvet Revolution!). We’re looking forward to collaborating with RECETOX as it launches a new birth cohort. This week I will be in Delhi to meet at the Indian Institute of Technology with leaders from the health and data science communities to discuss development of a Stanford-India “center” to host projects of joint interest spanning health development and environmental issues. After Delhi I’ll be off to attend the second Pacific Rim Alliance for Population Health (PRAPH) meeting in Taipei, joined by our counterparts from seven Asian countries. All of these emerging partnerships mean we’ll be able to offer our members access to exciting new global datasets. Stay tuned!
If all of this isn’t enough, we have plenty of exciting events on the horizon. We will be hosting our annual PHS symposium on November 6th, – so save the date! While you are at it, hold January 25th for our annual community engagement conference. Buoyed by its success last year, we’ve assembled an outstanding line up of PHS seminar speakers who will cover a diverse set of timely topics (e.g., population genetics, social media analytics, health economics, public health informatics).
Finally, don’t miss our Gender Matters: Why Sex in Science Isn’t Enough symposium on October 18th – we are very fortunate to have Janine Clayton and Hunt Willard as keynote speakers!
We hope to hear from you and see you all soon!
—Mark Cullen, MD, PHS Director
DECEMBER 31, 2016
A Hope in Uncertain Times
In light of concern about the future of public funding for our work in this political environment we are thrilled to report success in funding from two Foundations. The Sex and Gender working group -- and our new group on Adolescent Health -- got a huge shot in the arm from the Gates Foundation, receive two grants totaling $2.4 million. The first grant, led by Gary Darmstadt, will investigate available data from around the world to assess the relationship between local gender norms and the well-being of teen girls, with a strong focus on low and middle income countries. Darmstadt is also PI on a second large project to evaluate a maternal and child health intervention Gates is currently undertaking in Bihar, India, which will ultimately provide yet another database for the study of the First 1000 Days of Life working group.
At the same time, the Sloan Foundation has given us a grant of $1.8 million to solve a series of problems related to our emerging data platform. A collaboration between faculty at the Law School, Graduate School of Business, School of Medicine, and School of Humanities and Sciences, will see funds directed at the study of the contracting process itself. The process remains daunting and often rate-limiting to our ambition to provide an open, transparent yet secure platform for even the most sensitive of clinical, biologic and administrative data, including those with critical identifiers available (very securely!) to allow linkage among records. A second focus of the grant will be the crucial linkage to federal administrative and statistical data such as NHANES, Census, and Labor (and possibly IRS if we’re successful in our efforts!). While the Census Bureau hosts a Facility on campus (up at IRiSS), access for investigators has been both technically challenging and encumbered -- issues the funds will be used to directly remedy. The third focus will be to further develop the technology for the platform to ultimately make our data a true “public good.”
During these uncertain times know that we at PHS remain steadfast in our mission to discover the causes of health and disease over the life course, to impact our population and our patients by what we learn, and to teach the next generation how to carry on this work.
—Mark Cullen, MD, PHS Director
OCTOBER 1, 2016
A Welcome Message from the PHS Director
Whether you’re returning to campus after a great summer break, coming up after a grueling summer of work, or new to the Farm altogether, it’s my pleasure to welcome you to the Stanford Center for Population Health Sciences, and share a few of the exciting developments that have happened while you weren’t looking.
First and foremost are our data assets. As you will see flipping to our new Data Portal, many of your most frequent requests —for example commercial claims data and Medicare—are moving from their vendors to our Data Core, and we are moving along in our development of tools to make these datasets more easily accessible for you. At the same time, we are identifying and documenting additional datasets that we hope will be of interest to schools and departments across the university. We also have even more data that you will be able to access through liaison with one of our partners. To help you navigate through this treasure trove, our rapidly expanding “concierge” data broker team is at your disposal. In addition, our team of engineers and designers are working furiously to expand our data offerings and their ease of accessibility and us.
As fast as the dataset pool is growing, the scope and depth of active initiatives—many already funded—expands weekly. Organized around our working groups, many of these initiatives offer excellent opportunities for research trainees and new collaborators. Our team can help you broker these openings by emailing us at: PHS Research. And of course you can add to this growing body of work by applying for PHS pilot funding before October 7.
Speaking of working groups, there are some new ideas swirling for areas that are reaching cross-campus salience and dovetail with new data assets. Come to one of our upcoming brown bag lunches to learn more about these ideas, and see if one may excite your iterest, such as “food and nutrition”, or “adolescence,” or “behavior change” or “the environment,” all ideas raised by members in search of new opportunities, colleagues and data.
Of course it takes more than data and even analytic support to make for a great project, especially if it involves human subjects or community resources. That’s why we strategically merged with the Office of Community Engagement (formerly Office of Community Health) into a more integrated team, all housed at 1070 Arastradero. This team plays a critical liaison role between our local community and our academic institution and serves as a valuable resource to any investigator or research group. As part of our concierge service, we can now provide you with a variety of consultation services ranging from a single question about community partnerships and engagement strategies to long-term advising on collaborative community research projects. It’s never too early in the research planning process to engage the experts and rally community support.
Finally, we have begun to explore opportunities for inculcating PHS into the educational and training fabric of Stanford. In addition to our natural role coordinating existing courses and syllabi for coherence and accessibility to prospective students, we are introducing this fall a research trainee interest group, providing a social and instructional home to clinical fellows moving into dry lab research, doctoral and post-doctoral trainees in relevant disciplines, and other junior research trainees from across the campus. Not only do we hope to raise interest in population health, but we are confident that bringing these scholars together will have rewards for the future of team science that we cannot even imagine! Imagine!
Thanks and looking forward to seeing you in person on October 10th!
—Mark Cullen, MD, PHS Director
SEPTEMBER 1, 2016
Treating a Rare Tumor
Over 7,000 diseases classified as rare and affect more than 25 million Americans. Definition of what is classified as a ‘rare disease’ varies; the US Orphan Drug Act of 1983 defines it as a disease that affects less than 200,000 people. A report of rare cancer incidence using the CINA (Cancer in North America) dataset applied the published definition of fewer than 15 cases per 100,000 per year; by this estimate, 60 of 71 cancer types classified in this report were considered rare, accounting for 25% of adult tumors. In Europe, rare cancers account for 22% of all cancer diagnoses with 541,000 new diagnoses every year. The collective burden of rare diseases, especially rare tumors, is substantial. The rarity itself poses major challenges in studying these diseases, hampering efforts to understand the pathophysiology, define clinical characteristics, natural history, and in the conduct of meaningful clinical trials that eventually lead to the development of effective therapies.
Desmoid tumors (also known as aggressive fibromatosis), are locally invasive, slow-growing soft-tissue tumors with an annual incidence 2-4 cases/1,000,000 per year; annual incidence in the US is estimated at ~900 new cases. These tumors can be sporadic or be associated with familial genetic syndromes, such as familial adenomatous polyposis. Desmoids are phenotypically highly heterogeneous, disease can be either asymptomatic, thus requiring no intervention, or it can be associated with severe loss of organ function and significant morbidity. At present there is no FDA approved systemic treatment for desmoid tumors. One of the criteria used to determine whether a novel agent is approved for standard clinical use is the ability to demonstrate clinical benefit. Definition of clinical benefit relies on characterizing the natural history of the disease. Relatively little in known in the case of desmoids.
In order to address this, we have initiated a collaboration with the Fu Jen Catholic University in Taiwan to use a record linkage approach with the Taiwan National Health Insurance Research Data Base (NHIRD) and other databases to identify clinical characteristics of desmoid and better define the natural history of this disease. Taiwan National Health Insurance Research Data Base (NHIRD) was established in 1995 and is one of the largest population databases in the world covering 99.5% of the residents in Taiwan. It contains data on 23 million people and includes details pertaining to clinical diagnoses, outpatient and inpatient procedures, and prescription medications. Data generated from this descriptive epidemiology project will be used to inform clinical trials of promising novel therapeutics for the treatment of desmoids. The pilot project could become the prototype of studying rare diseases using Big Data. This project has been supported by a SPECTRUM pilot grant and a grant from the Desmoid Tumor Research Foundation (DTRF). DTRF is working with the National Organization for Rare Diseases (NORD) to participate in a natural history study (NHS) database project through a cooperative agreement with the FDA. I am a member of the Natural History Registry committee and results of our project will also inform this effort.
—Shivaani Kummar, MD
JUNE 16, 2016
Team Science: The Foundation of Population Health Science
When you scan the tables of contents of many of the leading biomedical journals, one of the things you might note first is how many authors there are on each of the papers—10 is typical and often the number is more like 50. Wow, this must be team science!
Alas, this is not team science. This is the response of some in the research community to the need for very large numbers of subjects to assess such results as gene-wide associations—evidence some gene contributes 1% or more to the risk for a common disease. In other words, these are rather banal clinical examinations on steroids, in which geneticists and clinical trialists have pooled their (common) data sets together to achieve a statistically meaningful result in a highly varied population.
So what is this team science everyone’s raving about. Well picture yours truly arriving in Chicago two decades ago, occupational physician and epidemiologist, invited to join a research group studying health disparities. They needed someone who knew about work I imagine. At the head of the table, Nancy Adler, world-renowned health psychologist from UCSF, and chair of this MacArthur funded network. And around the table are a dozen others: a sociologist from Columbia, a neurobiologist from Rockefeller, a statistician from Princeton, and an economist from there too, two other physicians (thank god! I was getting lonely), one from Yale, another from London, though neither had seen a patient in years. But for the fact these (weird) people were so smart and nice, and the promised future meeting venues so alluring, I might have tried to slide out unnoticed, for in truth I had no idea what I could contribute—they were speaking foreign languages! It was like the tower of Babel.
In actuality it took literally about two years before I did finally get what was going on and opened my mouth for the first time. And it was a good thing I stuck around. Collectively we pooled all the extant theory and data on SES and health and arguably resurrected (with the help of some brilliant trainees like David Rehkopf!) a field that was languishing from neglect. We produced a lot of articles and several books people still cite from time to time. But way more importantly we achieved something even Nancy didn’t appreciate at first: this was team science!
“Team” does not refer to numbers but to scope. “Science,” perhaps ironically, strains each participant’s prior notion about what science is. I confess now that at first I thought the psychologists were obsessed with big boring correlation tables; the economist seemed to consider the epidemiologists and the qualitative approaches of Kathy Newman, our sociologist, nothing short of voodoo. In truth, it was an extraordinary lesson in what a group of scholars from diverse disciplines could do that none alone could even begin to imagine. Population Health Sciences as we envision it today had not been born yet.
Well now it has. Of course our culture doesn’t change in the middle as fast as it does at the cutting edge. We still have appointment and promotion rules that favor adherence to singular traditions of scholarship. We still have research funding strategies in which every proposal gets evaluated by a monolithic “team” representing the singular mission of an Institute or Agency. Indeed we need only glance at the success-stories in the private sector to realize we have a big task ahead of us if we are to solve the actual problems of our society like the one Dr. Adler brought us together to consider twenty years ago. Her legacy is what PHS intends to emulate and extend.
—Mark Cullen, MD, PHS Director
MARCH 1, 2016
"Open data".... What does it hold for academia?
It seems that “data” and “big data” are everywhere in the air. I just returned from Sacramento where leaders of the major State agencies and the governor’s office were convening to discuss the use of their data at an “Open Data Fest,” follow on to the meeting of the same name we hosted here last month. The original focus of this movement, now sweeping across the country, was to encourage states and counties to “get their data out” in ways that would satisfy the increasing needs of providers and consumers of services—including health of course—to have easy and open access to relevant information. And with access naturally came the idea of sharing these data across silos—health and education, justice and housing, social services and immigration, just to name a few folks here might be interested in. Hard to disagree with any of that!
But “open data” can only be as interesting as the analyses that can be performed with them, which led to the obvious question: How can we in academics help, while at the same time gaining access to critical sources of information previously available only to researchers who invested years of relationship building. Our own growing interest in such data sparked the beginning of a dialogue that now reaches all the way to the Governor’s mansion. How might academics in California, indeed right here at Stanford join this effort. How might we someday get access to these “open” data in a protected form that is granular, linkable and useable for population health analyses.
I would love to report that I came home with a trove of datasets to pop on our servers, but alas nothing is ever quite that simple. That said, I can declare with great satisfaction that the dialogue has begun in earnest, and on mutually respectful terms. Initially aghast at the idea we might want identified or linkable data, there is now a greater appreciation for what such access could provide the State and its people, and how it might be achieved without compromising privacy and security. The conversation morphed from one of “wonderment” to one of “how can we get this achieved.” One particularly hopeful direction was talk of developing an office to oversee academic data needs and requests, to coordinate such requests at a statewide level rather than investigator by investigator, agency by agency. While we will have to await this next step before the process will radically change, we at Stanford have already developed both the good will and connections that augur well for the future.
In the meantime, our data team at PHS will be angling to identify and connect all of the state-level data that are already available somewhere on campus, and we will happily field any requests. If we cannot find it for you, we will use the opportunity to further press our contacts to oblige or at least lay out the as-is situation—good, bad or ugly. Stay tuned!
—Mark Cullen, MD, PHS Director
JANUARY 1, 2016
Plans for the new year
As 2016 begins, I wanted to share a few highlights with friends and members, and alert those who may want to be notified of upcoming opportunities.
While many of you participated in our annual symposium in October, those who missed the event can view the slide presentations and keynote speaker videos here.
In the spotlight were the presentations by the leaders of our 10 working groups, the core research engines of our Center, through which resources for research and training will flow. We encourage you to review the working agendas of each of these groups, and perhaps give some thought to which ones jibe with your own scholarly focus.
We will also soon announce the 17 recipients of our 2016 pilot grants, who will collectively receive a remarkable $520,000, owing to the largesse of Spectrum and Stanford Health Care. (7 grants were awarded for Population Health Sciences projects, 7 to the Learning Health Care Innovation Challenge, and 3 for Community Engagement projects. All were aligned with the missions of our working groups.) We also invite you to see the presentations and posters of last year’s grantees at Spectrum’s Innovation Symposium on Monday, Jan. 25, 2016, 9-11:30am, LKSC Bldg, Berg Hall.
And the opportunities are just beginning. Starting in January, each of the working groups will meet quarterly. Invitations to these meetings will be sent out to members this week. A monthly speaker program, with speakers chosen by the groups, will commence in the spring, with rotating venues to optimize access for our members from all seven schools. And keep an eye out for some newer innovations still—an intranet to facilitate the exchange of resources and information across the spectrum of our work, and—soon to be unveiled—the PHS “Shark Tank”—where working group members will pitch their strongest proposals for funding and support.
In the meantime, we wish everyone a fabulous holiday season and break, and welcome those of you who have not yet thought about where in the PHS firmament you might find academic opportunity. Please use a few moments of the respite to visit our website, look around, and sign up to become a member, here.
We look forward to working with you in 2016!
—Mark Cullen, MD, PHS Director
NOVEMBER 1, 2015
Our initiatives take off
Welcome to the Population Health Sciences website. For those of you new to the site, come explore for a few minutes to check out our active programs and ambitious plans. For members and friends who have been here before, please take a look at the new videos of our keynote speakers from the annual symposium last month, and the links to all of the slides and presentations. But please don’t blink—big changes are coming soon to the site.
Even faster and more impressive than the changes here are the efforts of our nascent working groups (see, Initiatives). Spearheaded by the prodigious efforts of the group co-chairs and fueled by a level of member enthusiasm I would not have dreamed possible, each if making headway to define the key challenges in their arena and setting goals on shorter and longer horizons; a few have already homed in on a real moonshot. For example, our working group on Sex and Gender in Health Research has thrown down the gauntlet to fill the huge void to define a measure of gender suitable for population level study of impact on health. Harder than it sounds, and probably more important to our understanding of health over the life course than any of us suspect.
There were so many ideas generated by the first few meetings of the mHealth group that plans are afoot to set up a “collaboratory,” where experts with ideas for health interventions exploiting the penetration of smartphones across the planet can draw on the insights of behavioral scientists and the technical genius of the engineers to develop new solutions for population health. Imagine what’s possible here!
The working group on Gene-Environment Interaction, pairing geneticists, social epidemiologists and social scientists have laid out ambitious plans to explore mechanisms by which such diverse factors as the microbiome, social stressors and pollutants impact the expression of the genome, probably from the moment of conception. That moment is also the starting point for the working group focused on the First 1000 Days of Life. The Group, launching its efforts with an already robust trove of data and research in progress proposes to avail the opportunities for new data, new collaborators and new technologies to address the pathways by which health—for better or worse—are patterned during this critical period in the life-course.
I could go on, but take a look at the presentations themselves. As you contemplate where you might find value in all of this—research and training opportunities, colleagues and support—plan to come back soon. Over the coming months, we will be showcasing new datasets and access to populations of interest and will have additional staff to support efforts aligned with our mission - using data from large populations to change care at the bedside, guide population level decision making and discover new pathways by which health and disease are determined. And if you haven’t yet joined us a member, what are you waiting for?
—Mark Cullen, MD, PHS Director
OCTOBER 1, 2015
What is Population Health Sciences?
Everywhere I go — around campus and beyond — three questions inevitably surface when I describe my new job: What is population health sciences anyway? Why has it all of a sudden become so salient? And why of all places at Stanford, a bastion of lab science and fundamental methods?
The first question is easy. Population Health Sciences refers to the analysis of data derived from any large population — defined by geography or membership in some large group — to three ends: better care of individuals; enhanced decision-making for public interventions for health; and discovery of underlying causes of health and disease in a population. In this way, PHS helps to bring lab discoveries to the bedside and public, while at the same time aiming to explainpatterns of health and disease observed in the population. This is an endeavor demanding knowledge of how our genome “plays out” under the influence of our social, physical and behavioral environments across the life-course.
Why has the idea of population health gotten so hot all of a sudden? The most proximate cause is probably rooted in the dramatic changes in our health care system, stimulated by the Affordable Care Act. This new law requires our health care system to serve not only individuals but entire covered populations, requiring new revenue streams for a previously neglected domain of health care.
But perhaps more deeply, PHS has emerged because of spectacular advances in knowledge and methods. Electronic records and claims data create troves of previously untapped information about patients, linkable to the enormously rich administrative data held by government and business. At the same time, the explosion in biology enables not only full sequencing of the genome but extensive biologic measures of phenotype—the epigenome, “omics” and RNA — allowing direct observation of fundamental processes over time. None of this would matter, though, but for the spectacular gains in biomedical data science and high-performance computing, facilitating the analysis of these fabulous new sources of fundamental information.
Given the above, it should come as little surprise that Stanford would be poised to lead. All of the basic component disciplines — from the social sciences to the basic biology to the data sciences and computing — are “best in show” here. Our campus enjoys perhaps the lowest “walls” separating schools and departments. And because we have not previously enjoyed a cohesive effort in population health sciences, the level of pent-up energy to forge the new trans-disciplinary teams required to take on these challenges is almost explosive. More than 350 faculty, scientific staff and trainees representing all seven schools have joined this initiative in little over a month.
In the meantime, browse through our website and be on the lookout for upcoming resources and study populations that you can leverage to help fuel our effort!
—Mark Cullen, MD, PHS Director
PHS sends out a bulletin by email highlighting our most important news quarterly. Join our mailing list or browse our previous bulletins below.
Visit our YouTube channel to see the latest videos from Population Health Sciences