The difficulty lies, not in the new ideas,

but in escaping from the old ones...

-John Maynard Keynes

CHAPTER TWO (See Summary)

[edit] Content, Connectivity, and Communityware

One of the most interesting questions our e-Patient Scholars Working Group came up with in our several years of deliberations was this: "What is it that our professional colleagues don't understand about the importance of e-patients?" Part of the answer comes from the work of John Seely Brown. ^[1] He notes that when established experts first consider the effects of new information technologies and the cultural transformations they produce, they will typically attempt to do so from within the cultural constraints of their established professional paradigm: their customary point of view from the professional perspective. We were no exception to this rule.

When we first began our explorations of the emerging world of the e-patient, we were all operating from a deeply held set of assumptions about healthcare, many of which were unconscious. But our findings were so unexpected that, time after time, we were forced to consider alternative points of view, including some that we had initially considered unacceptable. This in-depth process of reassessing the medical model that all our professional members had trained in, and had long taken for granted, was among the most difficult, confusing, and contentious part of our deliberations. But as Brown discovered, and as we came to agree: ^[2]

'Really substantive innovations-the telephone, the copier, the automobile, the personal computer, or the Internet-are quite disruptive, drastically altering social practices. [So] our established experts may be those least capable of helping us find our way safely through the disruptive social innovations these new technologies will require'.

We finally came to this consensus conclusion with respect to e-patients: The e-patient revolution cannot be adequately understood in terms of our older medical constructs, including many still taught in medical school and reaffirmed in continuing medical education. Those who attempt to view the experience of today's e-patients through the cultural lenses of yesterday's professionally-centered clinicians are doomed to failure. They will be blinkered if not blinded by what Brown calls their professionally-centered tunnel vision.

Our most helpful insights came from a growing awareness that e-patients use the Internet in three fundamentally different ways: to access content, connect with others, and collaborate with others in ways never possible before.

[edit] Patient-Driven Healthcare: Seven Preliminary Conclusions

[edit] 1. e-Patients have become valuable healthcare resources and providers should recognize them as such.

The e-patient revolution has already changed the experiences of millions of patients and family caregivers across the globe. But it has had little impact on the workflow at most clinical facilities. Many clinicians have become somewhat more Internet-friendly, though some still stiffen when a patient presents a print-out from a health site or journal article. At most clinical facilities, one would struggle to find a single formal change reflecting e-patients' new abilities.

Few clinicians ask patients and caregivers what they already know about their illness, what other sources they have consulted, and which have been most valuable. Few clinicians ask their patients, "What do you think we should do?" Few clinicians introduce patients with the same disorder to one another. And few clinicians advise their newly diagnosed patients with diabetes, Bell's palsy, or breast cancer that they might find useful support and guidance in a face-to-face or online support community.

Many provider organizations still do not ask patients for their e-mail addresses. Few ask patients to provide in-depth in-their-own-word critiques of the services they receive. Fewer still invite patients to join their clinical work teams, their advisory boards, or their governing councils-even though many would be well qualified.

Thus for our first generation of e-patients, going to the doctor's office is all too frequently a Kafkaesque experience of going back in time to when the Internet did not exist. We must learn to invite our capable fellow-citizens to join us in our common healthcare efforts. And such collaborations are especially important for patients with a chronic disease.

As Bodenheimer and his colleagues have noted, patients with chronic conditions make so many of the key day-to-day decisions about their illnesses that they are already effectively managing most of their own medical care. ^[4] When clinicians acknowledge and support their patients' essential role in self-management, providing them with the skills, tools, and support they need to manage their care even more effectively and supporting their efforts to provide helpful care for themselves and others, they exhibit fewer symptoms, demonstrate better outcomes, and require less professional care.

[edit] 2. The art of "empowering" patients is trickier than we had thought.

Some 20th century clinicians and health educators once believed that it might be possible to "empower" patients by providing them with targeted medical information, "prescribing" selected medical content in much the same way that clinicians currently order prescription drugs. We now know that empowering patients requires a change in the patient's level of engagement, both with their illness and with their clinicians, and in the absence of such changes, clinician-provided content has few, if any, positive effects.

Some studies have even suggested that such top-down "targeting" of patients by professionals may be harmful. When Elizabeth Murrayand her colleagues at the Cochrane Collaboration reviewed the medical literature for studies in which chronic disease patients had been targeted by professionally-developed interventions via the Internet, they found that even though the targeted patients showed an increased knowledge of their condition and an increased perception of social support, the interventions had no effect on the patients' sense of self-efficacy or on their health-related behaviors. ^[5] And these patients had worse outcomes than those who were left alone.

The same effect has been observed in other professional efforts supposedly intended to "empower" patients, (e.g., telemedicine, Information Therapy, patient handouts, etc.). Unless the underlying relationships between clinicians and patients are fundamentally transformed, professional interventions do not empower patients. ^[6]

For the earliest evidence of this insight, consider the work of White Paper Advisor Kate Lorig and her colleagues at Stanford's Patient Education Research Center (http://patienteducation.stanford.edu). Beginning in the early 1980s, as part of their efforts to help patients learn to manage their health more effectively, Lorig and her colleagues discovered that when it came to empowering arthritis patients, "Content did not matter." The success of Lorig's early programs came not from the specific medical content they provided (e.g., "rheumatoid arthritis is a disease of the connective tissues in which inflammation affects the joints and other organs of the body"), but from successfully communicating the idea that patients really can effectively manage their own illnesses. This concept was imparted through "patients-really-can-run-the-show" messages, encouragement, support, training and rehearsing, as well as proactive modeling in the art of devising and carrying out an action plan based on one's own agenda, be it losing weight or confronting an inadequate physician. ^[7]

Just as Lorig's model would predict, many professionally developed IT systems designed to "empower" patients have just the opposite effect. In one classic analysis of such a system, the late Diana Forsythe, a medical anthropologist at the University of California-San Francisco School of Medicine, described what happened when a distinguished team of medical informatics professionals set out to build a computerized patient-education system for migraine patients. Her paper, "New Bottles, Old Wine: Hidden Cultural Assumptions in a Computerized Explanation System for Migraine Sufferers," ^[8] should be inscribed on the cubicle walls of all those who aspire to develop online systems for patients.

The senior developers in Forsythe's study made four assumptions: (1) that the patients knew nothing about migraines, (2) that what they needed to know was what their doctors wished to tell them, (3) that the only reason the clinicians hadn't provided the information patients needed was a lack of time, and (4) that while the patients had much to learn from their neurologists, the clinicians had nothing at all to learn from their patients. Forsythe concluded that all of these hypotheses were mistaken.

Many of the migraine patients had read extensively in the medical literature and most knew a great deal about their condition, including many things that their physicians did not know, e.g., exactly what happened during a migraine attack and the best practical ways to manage it. The things the patients wanted to know were radically different from the things the neurologists wanted to tell them.

Other patients were perfectly capable of supplying much of the information their fellow-patients needed. A long list of "nonmedical" questions concerned patients greatly but was not taken seriously by the physicians. The neurologists systematically ignored these questions.

When patients attempted to bring up concerns that their neurologists did not see as strictly medical, they often appeared not to 'hear' them, or "attempted to pass the matter off as a joke." Thus the clinicians effectively ignored or disregarded all but a few "approved" types of questions. The neurologists' failure to answer their patients' questions had little to do with time. They offered little explanatory material and dominated the patient interviews so completely that patients had few chances to ask questions.

The system's final version did not allow patients to raise the questions that the clinicians habitually ignored. It did not permit patients to describe their experiences in their own words, but forced them to choose from a limited list of doctor-defined options. It did not allow patients to present the many things that they knew that the clinicians did not. It did not permit patients to inform other patients on topics that they knew more about than the doctors. It did not permit patients to provide the doctors with useful feedback on their services. It also failed to provide a way for patients to give their doctors important feedback and help keep them abreast of the literature.

Like many clinicians today, the developers were unwilling to address the important issues of knowledge, power, and control among doctors, non-physician health professionals, and patients, that are vital to building any effective IT system. So in the end, they built a technically impressive system that reflected all the same profound cultural defects of care the patients had experienced repeatedly in the clinic.

Forsythe's untimely death-she died while mountaineering in Alaska-deprived us of one of our most insightful observers in the emerging field of e-patient studies. But we can honor her memory by remembering what she taught us. "In practice," she once told me, late at night after a long AMIA (American Medical Informatics Association) meeting, "most medicine is ten parts culture to one part real science. From a medical anthropologist's point of view, a good deal of what happens in healthcare is invisible to most clinicians. Imagine that." ^[9]

[edit] 3. We have underestimated patients' ability to provide useful online resources.

Several years ago, a visitor to my Web site posted the following question: ^[10]

Medical Web sites created by persons who are not physicians are unreliable and should, de facto, be considered medically unsound. Don't you agree?

After some consideration, I discussed the matter with White Paper Advisor John Grohol, psychologist, author, and Webmaster of the highly respected comprehensive mental health site for patients and professionals Psych Central (http://www.psychcentral.com). Grohol is also the author of the definitive book on the field. As it happened, John had recently selected a short list of the "best of the best" online resources in mental health from his extensive directory of mental health Web sites. His rating scale took into account four areas in each Web site: content, presentation, ease of use, and overall "experience." The quality of the content was weighted more heavily than either presentation or ease of use. We decided to address the question above by determining whether the developers of Grohol's 16 "Top Sites" had been professionals or patients. Our conclusions: ten (62.5 percent) of the 16 top sites were produced by patients, five (31.25 percent) by professionals, and one (6.25 percent) by a group of artists and researchers at Xerox PARC. ^[11]

We observed that the patients who produce these sites certainly don't know everything a physician might know, but they don't need to. Good clinicians must have an in-depth working knowledge of the ills they see frequently and must know at least a little about hundreds of conditions they rarely or never see. Online self-helpers, on the other hand, will typically know only about their own disease, but some will have an impressive and up-to-date knowledge of the best sources, centers, treatments, research, and specialists for this condition. Smart, motivated, and experienced self-helpers with hemophilia, narcolepsy, hemochromatosis or any number of rare genetic conditions may well know more about current research and treatments for their disease than their own primary care practitioner. And when it comes to aspects of illness that some clinicians may consider secondary-e.g., practical coping tips and the psychological and social aspects of living with the condition―some experienced self-helpers can provide other patients with particularly helpful advice. The things clinicians know and the things self-helpers know can complement each other in some interesting and useful ways.

We concluded by quoting online patient-helper Samantha Scolamiero, founder of the BRAINTMR mailing list: ^[12] "Group members-lay folk and professionals alike-have moved beyond the old, obsolete mindset that holds that only certain 'qualified' medical professionals may create and disseminate medical information. We lay folk are learning that we are qualified through our experience, our knowledge, and our concern. We now see that we are capable of contributions no professional can make and that by linking our efforts [with those of professionals] in a coordinated team, we can advance the well being of all."

To better understand the power of online help provided by patients, consider the case of Karen Parles. Parles was a research librarian at a major New York art museum until January 1998, when she learned she had advanced lung cancer. "My doctors told me my cancer was incurable, that even with chemotherapy I had only a year or so to live," she recalls. "I'd never smoked, I have two great kids, and I was only 38. So the whole thing came as quite a shock. I was pretty overwhelmed at first. But as soon as I could, I went onto the Internet, looking for information. ^[13]

"I'm pretty good at finding things online. But even so, I had a hard time locating all the information and people I needed. There was great stuff out there, but it was scattered across dozens of different sites. There was no comprehensive site that provided links to all the best online information for this disease.

"I found a great support group, the ACOR LUNG-ONC mailing list. ^[14] The patients on the list answered my questions, suggested useful sites, and gave me a lot of invaluable support. I can't overemphasize the importance of their help in those early weeks and months. It gave me instant access to the wisdom and experience of hundreds of other lung cancer patients."

Then a friend told Karen about a surgical team developing a new treatment for her type of cancer at Boston's Massachusetts General Hospital. "I went to Boston to see them and I was pretty impressed," she says. "But having a lung removed by an unproven procedure still seemed frightening, so I shared my fears with my LUNG-ONC friends. I heard right back from eight or ten others who'd had a pneumonectomy. They assured me that I could do it and encouraged me to give it a shot."

"After I recovered from the surgery, I got to thinking," Karen recalls. "I'd found my life-saving treatment by a combination of Internet-smarts, luck, and personal contacts. Others might not be so lucky. So I decided to create an online resource just for lung cancer, a single, centralized site where patients could find links to everything they needed to help them get the best possible medical care, a place where they could learn to manage their disease in the best possible way."

Parles launched her site Lung Cancer Online (http://www.lungcanceronline.org) in 1999. It lists and describes the best sites containing information about lung cancer. She encourages newly-diagnosed users to craft their own crash course in their disease. Visitors find guidelines and databases to help them find the top-rated lung cancer specialists and the best medical centers for each type of lung cancer. There is also practical advice with survivors' stories, up-to-date information on the latest clinical trials, and guides to mining the medical literature. The site also refers visitors to a variety of online support groups.

In addition, visitors have had access to Karen herself. "Because I'm an experienced lung cancer survivor, I've found that many patients and family members want to interact with me personally, by phone or e-mail or in person," she explains. "So in addition to keeping my site up-to-date, I spend many hours each week helping other patients, sharing my story, and providing hope and encouragement." Many of the patients who have come to rely on her counsel and Web site refer to Karen as their mentor, their role model, or their hero.

With characteristic modesty, Parles urged us to emphasize the important roles her support group, her extensive network of other lung cancer patients, and her network of expert medical advisors all play in maintaining and updating her site. "I work collaboratively with the Webmasters and members of all the sites and groups I link to. I depend on others for the questions I receive and the resources I hear about. So lungcanceronline.org serves as a conduit for many voices."

[edit] 4. We have overestimated the hazards of imperfect online health information.

In the early days of the World Wide Web, we all worried about what would happen to patients who found incorrect, outdated, or self-serving medical advice online. We now know that they rarely come to harm. Two comprehensive efforts have researched whether e-patients are being injured as a result of bad online information. In the first, a distinguished international team that included White Paper Advisor Alejandro Jadad of the University of Toronto conducted an extensive review of the worldwide medical literature, looking for reported cases of patients who had died because of bad online information or advice. They found that for the first decade of the Internet's existence, only a single case had been reported and the evidence was by no means conclusive. ^[15]. In the second, White Paper Advisor Gunther Eysenbach and his colleagues established the Database of Adverse Events Related to Internet Use (DAERI), appealing to clinicians worldwide to report case studies of harm to patients resulting from bad online information and offering a fifty-euro "bounty" for each report. As of September 2004, Dr. Eysenbach reported that in the first four years of the project, only a single case of a possible fatality was reported. ^[16] The project has since been discontinued.

By contrast, in the "off line" world of medical care, a survey by the National Patient Safety Foundation found that 42 percent of U.S. adults reported that they had experienced a medical error-either personally or to a relative or friend. ^[17] The Institute of Medicine has estimated the number of hospital deaths due to medical errors at 44,000 to 98,000 annually. ^[18] A 2004 HealthGrades, Inc., study puts this number at 195,000. ^[19] A collaborative national surveillance project conducted by the Centers for Disease Control and Prevention, the Food and Drug Administration and the U.S. Consumer Product Safety Commission reported in 2006 that adverse reactions to common medicines sent at least 700,000 Americans to emergency rooms annually. ^[20] The experts who compiled and analyzed the data concluded that this was probably a conservative number because so many severe drug reactions go undiagnosed.

Professor Barbara Starfield of Johns Hopkins University concluded that 225,000 U.S. deaths annually can be attributed to iatrogenic causes-medical mistakes, adverse reactions, hospital-based infections, etc. ^[21] This number would make healthcare hazards our third leading cause of death, surpassed only by heart disease and cancer. The HealthGrades, Inc., researchers concluded that: "Medical errors seem to be the elephant in the room that no one wants to acknowledge or talk about. The lack of recognition and acknowledgment of the seriousness and urgency of the problem fosters a culture of denial and complacency." ^[22]

We can only conclude, tentatively, that adopting the traditional passive patient role and putting oneself in the hands of a medical professional may be considerably more dangerous than attempting to learn about one's medical condition on the Internet. We suspect that the over-generalized warnings some clinicians have given their patients about the "bad quality of health information on the Internet" will in retrospect be considered counterproductive and ill advised.

Why did many clinicians assume that the Internet would be so much more dangerous for their patients than it actually turned out to be? We don't believe that this can be entirely explained by self-serving motives. Clinicians are most familiar with closed information systems-such as electronic medical records and the systems that report laboratory values. Closed information systems rely on supply-side quality control measures: Only the duly authorized can add or edit information; the accuracy and currency of data is closely monitored; and there is little tolerance for error. Thus the data in closed systems is assumed to be universally accurate.

By contrast, the Internet is an open information system, albeit one that provides access to many closed information systems. Anyone can publish anything they wish. The Internet, therefore, will always contain a good deal of inaccurate, outdated, and self-serving information. But while this would be a major defect in a closed information system, in an open information system like the Internet, it is merely a system characteristic.

Open systems rely on demand-side quality control measures. Thus sophisticated Internet users understand that one does not accept any random bit of online information (e.g., spam) with the same degree of unquestioning acceptance that might be appropriate when checking one's bank balance or looking up a patient lab value. Unlike some patients at their doctor's offices, patients who use the Internet don't leave their critical faculties at the door. And even though the Internet contains imperfect information, it also offers its users many valuable tools that can help them tell the good from the bad.

This added value of connectivity and communities is enormous. As David Reed has put it, "What we see, then, is that there are really at least three categories of value that networks can provide: the linear value of services that are aimed at individual users, the 'square' value from facilitating transactions, and exponential value from facilitating group affiliations. What's important is that the dominant value in a typical network tends to shift from one category to another as the scale of the network increases." ^[23]

Pattie Maes, one of our most highly regarded IT visionaries, has defined a whole new category of online quality-control systems she calls communityware. ^[24] In addition to helping users find the high-quality online information they seek, communityware systems help loose online networks of people with common interests accomplish tasks that formerly required centralized control and coordination: choosing the best doctor or medical center, for example, deciding on the most appropriate treatment for their disease, or providing advice and referrals for other patients. Communityware systems, Maes notes, are driving a major power shift from established authorities to online communities, not just in medicine, but in every sector of society. ^[25]

The software that enables online support communities-listservs™, Web forums, chat rooms, etc.-provide one example of communityware. The proprietary PageRank algorithm used by Google to identify Karen Parles' site as the number one resource for lung cancer is another. ^[26] Online databases that provide user reviews and ratings (for examples, see the seller reputation databases at ebay.com, the product reviews at www.amazon.com, and the patient drug reviews at www.askapatient.com) are another form of communityware. Different forms of electronic communication can also serve as platforms for communityware; think of e-mail networks, blogs, instant messaging, text messaging, moblogging, wikis, P2P file sharing networks, and other collaborative technologies.

[edit] 5. Whenever possible, healthcare should take place on the patient's "turf."

White Paper Advisors John Lester, Dan Hoch, and their colleagues did not begin their groundbreaking e-patient project Braintalk by requiring patients to come to their clinical offices at Massachusetts General Hospital. They went to where the e-patients were and observed what they were doing. And once they had learned what online support communities were and how they operated, they offered to help them do what they were already doing.

Similarly, White Paper Advisors Alan and Cheryl Greene, developers of drgreene.com, did not launch their groundbreaking pediatric site by designing a technically sophisticated "consumer health informatics" system they imagined their patients might like. Instead, they asked their patients what they wanted-and gave them what they asked for. It was because they understood that they could accomplish their purposes more effectively by going to the e-patients' familiar "turf," rather than by forcing patients to come to their familiar professional terrain, that these pioneering developers were able to make such important contributions to the online healthcare commons.

In their book The Support Economy, ^[27] Shoshana Zuboff and James Maxmin draw an important distinction between the traditional realm of professional/organization space (e.g., the clinic and the hospital) and the rapidly evolving realm of individual/consumer space (e.g., the e-mail inbox, the search engine, the Web site, and the virtual community) in the emerging information age economy.

Professional/organization space is the traditional turf of clinicians and researchers. It is completely within the medical professional's realm of control. When operating in professional/organization space, patients must conform to their clinicians' "rules, procedures, standards, requirements, and schedules, not the other way around." ^[28] And while this may feel perfectly natural to most clinicians, Zuboff and Maxmin warn that the experiences, capabilities, and needs of our first generation of e-patients have changed so dramatically that professionals who still insist on operating exclusively within organizational space simply "do not have the capacity to perceive [these] individuals' [new reality], let alone to respond to their [real] needs… even as a perpetual drumbeat of consumer-oriented rhetoric claims the opposite." ^[29]

Zuboff and Maxmin define individual/consumer space: "Individual space is always the individual's space, and the choices that exist there are always the individual's choices. Each individual sets the rules of relationship… [and] these relationships cannot be constructed mechanistically. They are necessarily intimate and authentic. They build over time, based on mutual respect and interpersonal trust." ^[30]

Individual/consumer space is another name for what we've been calling self-help cyberspace, the domain within which autonomous e-patients find and use the online health resources of their choosing-friends and family, patient-helpers, online communities, medical Web sites, and unstructured e-mail with clinicians. Individual/consumer space-or self-help cyberspace-always lies within the e-patient's realm of control.

Most of the most successful online health resources developed to date-that is to say, those that e-patients like best (and we would argue that this is the only appropriate standard)-operate in individual/consumer space. e-Patients, not professionals, are in charge. Most have been developed by patients. And those developed by professionals have almost always relied heavily on patient direction and guidance.

Systems that operate in individual/consumer space liberate patients from the confining, limited, and subservient roles they are typically forced to play when interacting with clinicians within organizational/professional space. For even though such limits are often invisible to health professionals, most patients feel intensely uncomfortable when forced to operate within the constricted roles allowed them in most contemporary healthcare institutions.

Zuboff and Maxmin warn that developers who continue to believe that it may be possible to develop systems that genuinely empower patients while remaining within the comfortable and familiar borders of professional/organization space are only fooling themselves. By way of explanation, they tell the story of the man who looks for his keys under the streetlight, instead of in the dark alley where he lost them, because that's where the light is. Like the man in the parable, "we professionals like to look for solutions within our accustomed organizational/professional space, even though the real area of concern is not to be found there, but in what seems to us the darkness, but in reality is simply a different world, the world of individual/community space…. [But] consumer solutions developed in organization space merely perpetuate the fundamental problems whose symptoms they seek to address... When we face discontinuity, the answers we seek cannot be found under the light from the lamppost. They are not here, but over there-in the darkness, through the looking glass, on the other side of what [we think we know]." ^[31]

[edit] 6. Clinicians can no longer go it alone.

A century ago, a physician's brain could hold all the medical information required to treat most known medical conditions; there was not that much to remember. Over the past century, medical information has increased exponentially. But as White Paper Advisor Richard Rockefeller has observed, the capacity of the human brain to store and process it has not. ^[32] As Donald Lindberg, director of the National Library of Medicine, explains, "If I read and memorized two medical journal articles every night, by the end of a year I'd be 400 years behind." ^[33]

It is no longer possible for medical professionals to stay completely up-to-date on a long list of medical conditions. It is difficult enough to do so for a single disorder. So unless e-patients are lucky enough to find a clinician who is a top specialist in their disease (and as we will see, a surprising number have done exactly that), they will be able to learn more from the Internet than they can in their doctor's office. And some will end up knowing more than their clinicians about many, if not most, aspects of their conditions. Understanding this development on an intellectual level is one thing. But as many of us can assure you, it is quite another to sit face-to-face with a patient newly diagnosed with a rare condition who has already spent countless hours reading up on it-while you can only dimly remember reading about it in medical school many years before.

We must admit that in the beginning, dealing with patients who knew more than we did was something of a shock. So we can fully understand and sympathize with colleagues who may be tempted to discount their patients' new abilities, to disparage their information sources, and to attempt to reassert their medical authority. We can also assure them that, in the long run, this will be neither possible nor necessary. For in the end, we're confident that our colleagues will conclude, as we have, that informed, activist e-patients are a blessing, not a curse.

Well-informed patients can help lift the burden of care from the shoulders of overworked clinicians. They can take their own vital signs during a clinical visit. They can do their own lab tests. They can examine their kids' sore ears with a home otoscope, listen to their own hearts to monitor an arrhythmia, administer their own peritoneal dialysis, give themselves injections, and can safely choose and self-administer many drugs currently available only by prescription. And, as we will see in the next chapter, they can help us upgrade our customer services and improve the quality of our clinical care, helping us avoid, detect, and manage medical mistakes. They can, in short, become our colleagues and our partners, not irresponsible children or mindless robots, capable of nothing more than obeying their doctors' orders without question or comprehension. To continue to prohibit patients from making the healthcare contributions they now can make would be an exercise in futility.

[edit] 7. The most effective way to improve healthcare is to make it more collaborative.

As we have seen, the emerging world of the e-patient cannot be fully understood and appreciated within the context of pre-Internet medical constructs. The 20th century medical worldview did not recognize the legitimacy of lay medical competence and autonomy. That is why its metrics, research methods, and cultural vocabulary are poorly suited to studying this emerging field. Deeply embedded cultural assumptions about the roles of physician and patient will often lead to a premature closure of reality.

The e-patient revolution we have been describing, encouraging and hopeful as it may be, is not a complete solution to our healthcare ills. We cannot simply replace the old physician-centered model with a new patient-centered model. Like software engineers faced with the challenge of creating a new computer operating system, we now find ourselves faced with the task of identifying and upgrading the areas in which our previous "operating system" has proved itself inadequate. We must develop a new collaborative model, akin to open-source software, that draws on the strengths of both the traditional and the emerging healthcare systems. We must find new ways to help 21st century clinicians and 21st century patients work together in 21st century ways. In the chapters that follow, we offer more suggestions on how we might accomplish this. n