The Illusion of Mental Illness and Creativity

From “There Was No Couch: On Mental Illness and Creativity” by Jalees Rehman.

Before I started my psychiatry rotation in San Diego, I had been convinced that mental illness fostered creativity. I had never really studied the question in much detail, but there were constant references in popular culture, movies, books and TV shows to the creative minds of patients with mental illness. The supposed link between mental illness and creativity was so engrained in my mind that the word “psychotic” automatically evoked images of van Gogh’s paintings and other geniuses whose creative minds were fueled by the bizarreness of their thoughts. Once I began seeing psychiatric patients who truly suffered from severe disabling mental illnesses, it became very difficult for me to maintain this romanticized view of mental illness. People who truly suffered from severe depression had difficulties even getting out of bed, getting dressed and meeting their basic needs. It was difficult to envision someone suffering from such a disabling condition to be able to write large volumes of poetry or to analyze the data from ground-breaking experiments. The brilliant book “Creativity and Madness: New Findings and Old Stereotypes” by Albert Rothenberg helped me understand that the supposed link between creativity and mental illness was primarily based on myths, anecdotes and a selection bias in which the creative accomplishments of patients with mental illness were glorified and attributed to the illness itself. Geniuses who suffered from schizophrenia or depression were not creative because of their mental illness but in spite of their mental illness.

 

The whole essay is fascinating and well written, covering hypermedicalization, media bias, sham science, and more.

End of Life Care

The Compassionate Physician

By Ambrose Bierce

A KIND-HEARTED Physician sitting at the bedside of a patient afflicted with an incurable and painful disease, heard a noise behind him, and turning saw a cat laughing at the feeble efforts of a wounded mouse to drag itself out of the room.”You cruel beast!” cried he. “Why don’t you kill it at once, like a lady?”

Rising, he kicked the cat out of the door, and picking up the mouse compassionately put it out of its misery by pulling off its head. Recalled to the bedside by the moans of his patient, the Kind- hearted Physician administered a stimulant, a tonic, and a nutrient, and went away.

How Watson Works

100% lifted from a recent Fastcompany article. I have tremendous faith that this platform will work. The only question in my mind is whether the pipedream of making this technology ubiquitous (“a Watson in every pocket”) will happen. Big companies have big proprietary interests, and big armies of lawyers. But my fingers are crossed.

How Watson Works

IBM’s Watson computer begins trials in the health care industry this fall. The initial goal is to help oncologists make better decisions for cancer treatment; eventually, the computer will also aid in the diagnosis and treatment of other chronic diseases.

1. For well over a year, the Watson computers have been “trained” in science and medicine. Technicians feed Watson medical textbooks and journals, patient histories, and treatment guidelines.

2. At Memorial Sloan-Kettering Cancer Center in New York, doctors have begun using a Watson app
on a tablet to access the computer through the cloud. The doctor logs in to Watson and begins to input data and ask questions.

3. When the oncologist queries Watson about a course of treatment for a lung or breast cancer patient, the computer–with its ability to understand natural language–notes keywords in the query, such as the particular type of cancer and the genomic variant of the tumor.

4. Watson then springs into action, using its massively parallel processors to review millions of pages of text in seconds. It explores the patient’s medical history, medications, and other existing conditions. It then combines this information with recent data from the patient’s medical tests and may comb through studies of patient groups at Sloan-Kettering who have had similar types of cancer. It also reviews doctors’ and nurses’ notes, recent medical research, journal articles, and treatment guidelines.

5. Watson then generates hypotheses for treatment. On the tablet app, these appear as separate options with varying levels of confidence. For instance, Watson might score one treatment option–a combination of chemotherapy drugs–with a 95% confidence level, suggesting it would be the most sensible path. It might also highlight options with lower scores as alternative treatment courses. The doctor then weighs the options and makes the call.

 

What do you think? Hype or Reality?

Newsflash (not really): Data Management is crucial for Accountable Care

A recent HealthAffairs articleoutlines the results of a study of Large Multispecialty groups and their readiness for new payment reform. As provider groups prepare themselves to take on new risk models, the success factors for building successful models of accountable care are clear: results-based payment models for physicians, strategic goal setting and effective management (e.g. targeting ED high flyers or reducing readmissions) and of course, effective information management systems.

Look Ma, I’m building an ACO

What was interesting is that even among the ‘advanced groups’ in the study – meaning those who operated on a majority of risk-based contracts (71 percent on average), these success factors are hardly fully implemented. For example, the risk-based groups still doled out about half of their compensation (to both PCP’s and specialists) in volume-based contracts (compared to 85 percent of the FFS group). The management strategy was a little bit better, with 82 percent of those practices having implemented some kind of targeted policies or programs for readmission, compared to 70 percent of the FFS group.

Where did the Risk-based advanced group really blow their FFS counterparts out of the water? Data Management. While both groups fared well in having a shared EMR or a results management system (radiology/lab tests integration), 100 percent of the risk-based groups had a data warehouse and analytic software, 60 percent had disease registry systems, and 70 percent had practice variation analysis software. Only 10, 20, and 10 percent of the FFS practices, respectively, had implemented those systems.

 So what does this mean?

Quite simply, advanced data management capabilities are at the heart of an effective accountable care model. Nothing we didn’t know before – but still a green light for the analytics industry and vendors who prove themselves in the space in the coming months.

EDIT – This article just came out. ACO’s are expected to spend $1-4 million depending on their size.

http://www.ihealthbeat.org/articles/2012/9/26/acos-to-spend-up-to-4m-for-technology-data-requirements.aspx

Room for Improvement: Personal Health Records

(Please note – this also appears on Health2.0 Austin)

Sometimes, everything seems to happen at once. The week before last, I had a string of conversations with a big software vendor, a national health plan, a small developer firm, an academic research department, and a fellow industry analyst that all touched one one common topic: Personal Health Records (PHR). Then, last week I attended National Health IT Week, which the ONC kicked off with the big “Blue Button” initiative announcement to put data back into the hands of the patient or caregiver. Finally, after reading about AHIMA’s new guide to PHR, I decided it was time to revisit my notes and cobble together some observations from the last two weeks.

Unfortunately, the story remains the same as it’s been for the last couple of years: Despite the growth in consumer-driven health through a glut of apps, sensors, open data initiatives and a slowly changing culture encouraged by health reform, the dial on personal health records uptake has not really moved a tick. The percentage of Americans using PHR still hovers around 10 percent. Here are just some of the reasons that have come up in some of my conversations in the last two weeks.

Lack of Awareness

As ONC “Consumerista” Lygeia Ricciardi pointed out at the National Consumer Engagement Summit on September 10^th, most people are still unaware that they can simply ask for their records. Moreover, as several panel members discussed, patients and clinicians alike misunderstand HIPAA laws. People persist in believing that HIPAA merely restricts access to information when in fact, it was designed to ensure secure access. As long as this is the case, we’ll see limited voluntary uptake.

Patients, not Patience

Sadly, many untethered PHR platforms still require manual entry of medications, treatments, appointments, and other data points. Not only is it silly to expect correct spellings, doses, and values from this process – it’s silly to assume people have the time and patience to do this in the first place. Best case scenario, they are forced to enter their own data when something goes wrong and they’re backed into a corner, and manual entry is the only way to manage their own information.

#ConsumerFriendly

Unless you are lucky enough to be a Kaiser Permanente member, your PHR is probably something that has only been out for a couple of years. While health plans like Aetna and United have made good progress in tying together their networks’ information pipelines, as long as patients access a piecemeal health care system that sees payers, providers, and vendors competing to provide their own PHR platform, it will be confusing and difficult for people to use. I spoke earlier this week with someone who tried to set up a HealthVault account, and she told me that after requiring her to set up a Microsoft Live account (she has a Mac) and a Microsoft Healthvault account, she got a series of registration error notices, had to restart her computer, and then had to wait for a confirmation e-mail. While Microsoft’s advanced market position is well-deserved in most industry observers’ books, so too is their notoriety among consumers.

 Lack of Trust

For the time being, people still trust their doctors the most. In an integrated delivery system, doctors have an interest in promoting self-management behaviors such as tracking, assessing, or proactively scheduling appointments. While payers have this same interest for financial reasons, they don’t carry the same cachet of trust in most peoples’ minds. Now, you might think there is a weak link between provider trust and PHR uptake, and you be partially right – but the bottom line is that people trust brands, whether it’s in banking or retail. Sure, healthcare is always a little bit different, but a site that is branded with a reputable health system’s name, or even referred to a patient by a trusted doctor might make more progress than we’ve seen to date.

So what trends will it take to turn this around? The same industry rhetoric about growing consumer-centeredness and improved patient engagement have been discussed ad-nauseum for the last few years without a tangible outcome. Here are a few of the possible (and perhaps more concretely definable) drivers:

• Modern Design: Impeccable UI/UX is now table stakes. Clunky design, multiple logins and passwords simply won’t cut it. Only with the rise of mobile and tablet apps did the trend of health care design even emerge. It may take another iteration before it really takes hold and vendors design tools with people – not their competitors – in mind.
• Integration: Again, we don’t like to use things that take hard work just to set up. Data pull is just the beginning – information needs to be more than an ASCII file in order to be useful – it needs to be sortable and modular in order for developers to funnel it into usable, practical platforms. And once this is achieved, let’s please give patients data they can make sense of.
• The Role of the Caregiver: 30 percent of US adults are now caregivers, and as 10,000 seniors turn 65 each day this number will skyrocket. People won’t want to log in to a separate portal for themselves, their parents, and their children or spouse. As we all get used to these things, the grumbling demand for vendor-agnostic solutions will eventually rise to a clamor and some innovative companies will respond with software that doesn’t underachieve.

Do you think PHRs will eventually take off, in their current form or in a new iteration? What else might it take for this to happen? Let ‘s discuss below!

Cancer and Health IT: Promising Opportunities Ahead

National Health IT week is nearly upon us. And while the developments in Health IT over the last several years have been well documented, more specialized areas, such as cancer care, have been flying under the radar of many technology executives. This is understandable, as cancer happens to be one of our most complicated health challenges and one of our most specialized and evolving areas of medicine.

As a growing disease burden and the second-leading cause of deaths in the US, cancer is rarely mentioned on the common list of chronic conditions that has become part of our national dialogue: diabetes, CHF, hypertension, asthma, depression. The facts remain the same: 28 million people worldwide are impacted by cancer, with nearly half of those cases happening here in the US. The disease is among our most costly, running over $260B annually in medical costs and lost productivity according to the National Institutes of Health.

Fortunately in this case, the lack of a spotlight does not mean there has not been progress in the field. Quite the contrary. The next 12 months will see substantial advances in the information technology available to help patients, survivors, families, and others affected by cancer. Here are just a couple of general areas where we can expect to see some breakthroughs in the next 12 to 18 months.

Enterprise Cancer Research Platforms

 In the last year there has been a surge in the number of analytics platforms in health care. Claiming to help organizations make sense of the big data revolution, dozens of vendors have been riding the wave of momentum created by the nation’s acceptance of accountable care: clinical decision support, business intelligence, CRM to name a few. Many of these are still early in the hype cycle phase; those who were on last year’s HIMSS showroom floor will understand the explosion in the number of these supposed “solutions.” Yet above the market fray in cancer research, there has quietly emerged a revolutionary application of these concepts.

Enabling cancer scientists to join up multiple data sources has produced a learning system that blurs the traditional barriers between “research” and “treatment.” To name a few of these sources: Longitudinal clinical information from treatment records (surgical, therapeutic, diagnostic), molecular testing (tumor assays, genetic and other biomarkers), patient-reported outcomes (surveys, histories, demographics, behavioral). With the IT power to “mash up” these disparate data, we can now form powerful profiles for multiple uses, from scientific or public health research, to health plan case management design, to better treatment. This last category is most important: with an end-to-end profile, patients can be matched to customized treatments or informed of relevant clinical trials. The more the system is used, the more it learns and the better it gets. Importantly, the involvement of the private sector has been the instrumental force behind avoiding the kinds of failures we’ve seen in this area to date.

Here are just a couple of specific examples to watch for in the coming months. The first is IBM’s Watson engine. IBM made a splash last year when they announced they were unleashing Watson on cancer. In a pilot launched at Memorial Sloan Kettering, researchers are using Watson to crunch medical literature into the fold. Watson’s also been a part of a case-management driven pilot through Wellpoint which applies the same principles to systemic challenges of disparities in treatment, outcomes, and costs. These organizations have announced that these pilots will be expanding with broader applications in 2013.

The second involves the Moffitt Cancer Center in Tampa, Florida. Last year Oracle partnered with Moffitt to develop and commercialize a health research informatics platform. Moffitt’s CEO stepped down to run this new entity, M2Gen, and it has been going strong ever since. A rapidly expanding base of consented patients (90k+) across a consortium of 17 sites, combined with strategic business partnerships with pharmaceutical companies (four expected partnership announcements in addition to an existing deal with Merck) could hold the potential to transform the clinical trials matching process, and to speed up the way we discover new cancer drugs.

Patient-Focused IT Resources

One step closer to the day-to-day lives of people and families managing a cancer diagnosis, there is an emerging set of more focused applications which aim to provide cancer patients, survivors and families with the help they need. While the app market will continue to churn out tablet and smartphone programs that provide information, tracking features and networking outlets for members of the cancer community, there are two areas where we’ll see some big developments in the next year on a more sophisticated level.

The first focuses on cancer survivors, for whom treatments and surgeries may come with side effects or complications, for whom fear of recurrence can lead to uncertainty and confusion about seeking help such as follow-up care. For these hundreds of thousands of individuals, the Institute of Medicine identified Survivorship Care Plans (SCPs) – customized, comprehensive guides to health care issues after cancer – as a critical component of improving long term outcomes. In an effort to spur adoption of SCPs the American College of Surgeons’ Commission on Cancer (CoC), which vets cancer treatment facilities in the US, has mandated that all accredited cancer facilities, which treat 70 percent of cancer patients in the US, must start providing patients SCP by 2015.  Yet in a tightly scheduled office visit, clinicians are often ill-equipped (and not reimbursed) to assemble the diagnostic, surgical, and treatment history required to complete a care plan.

This is where the application of IT will play a role.  Partnership efforts with the CoC and the American Cancer Society, led by the Lance Armstrong Foundation, are underway to automate the process of completing these resources. In one pilot study, the LIVESTRONG Care Plan will be auto-filled using electronic health records and data from cancer registries, in partnership with the National Cancer Database. The focus is to produce a set of guiding documents that promote replicability and easy adoption. With similar pilots already underway in states like Colorado and Minnesota, we will see the SCP emerge as a more common part of oncology providers’ workflow and a more widely adopted resources for patients and survivors nationwide.

The second area involves patient-facing clinical trials matching tools. In the last year alone, two portals won an ONC competition for innovative approaches to matching patients to clinical trials through websites. Those pilots have begun to commercialize and will develop further in the coming months. The ONC has also been active in the state of Texas, spurring stakeholders to become active in health IT projects that benefit cancer patients. While it is too early to divulge all of the details, suffice to say there is momentum building around a project to incorporate IHE profiles into a data extraction pilot in Texas that will make medical record data usable by researchers to provide a transparent, up-to-date clinical trials matching process for patients to use.

The Tip of the Iceberg…

 Without going into a substantial level of detail, here are just a few more things to consider:

• Federal Policy: The ONC has shown an increasing interest in cancer. Dr. Farzad Mostashari spoke a few weeks back at the NIH/National Cancer Institute (NCI) summit and stressed the need for innovation in research and treatment. In addition to ONC support for clinical trials and a budding pilot project in Texas, there have been several app challenges (here is the latest) focused on cancer prevention and treatment. And, there are also the Stage 2 meaningful use requirements that incentivize clinicians to collect and transmit data to state cancer registries.
• Cancer Registries: Innovation is happening left and right with state registries. This extends far beyond involvement with survivorship care plans and the MU Stage 2 reporting described above. At the aforementioned NCI summit in August, over a third of the posters involved some kind of pilot project to expand communication between themselves and doctors, patients and hospitals. At a national level, the continued development of the rapid quality reporting system holds a lot of potential for better integration of cancer centers into ACO’s, health insurance exchanges and health information exchanges.
• Personal Health Records: During a recent call with Sean Nolan, architect of Microsoft Healthvault, we discussed the emerging integration of cancer-related apps into the PHR ecosystem. Now that data from medication suppliers like CVS and Walgreens, data from the Blue Button initiative and payers, and patient-reported data from Bluetooth enabled devices to smartphone apps are swirling around in the mix, there is some serious potential to link up these data sources.
• Payers: In addition to Wellpoint’s involvement with Watson, Aetna just announced a similar partnership with Philadelphia-based Eviti to use a knowledge platform to provide case management guidance to doctors and nurses within the provider network. This is of note to an IT audience for two reasons: One, Aetna’s HIE platform, Medicity, is the basis for how clinicians will access these guidelines, paving the way for a potential learning HIE that others have been working on for a while. Two, as the undisputed payer leader in the App world, this sets up some cool integration to plug in provider-facing reference apps focused on oncology guidelines.

Beyond these areas, there is ample potential for innovation and cost-savings related to imaging, secondary consultations, doctor communication, and patient engagement. Clearly, cancer-related health information technologies are thriving. Whether you work in the enterprise IT sector, in health insurance, public health research, natural sciences, or with apps, there is strong momentum in cancer.

It’s important to remember why this is – in addition to the numbers at the beginning of this post, cancer remains one of the most complicated diseases we know. It is expensive, and not always covered (despite the ACA.) It is uncoordinated, requiring patients to navigate between insurers, specialists, clinical trials, and other parts of the system. It is a brutal, awful disease. With so much room for improvement, oncology and cancer care is a perfect area for health IT innovators to unleash their hard work and drive some meaningful change.

Dr. Berwick’s Address to Harvard Med School Graduates

“You will soon learn a lovely lesson about doctoring; I guarantee it. You will learn that in a professional life that will fly by fast and hard, a hectic life in which thousands of people will honor you by bringing to you their pain and confusion, a few of them will stand out. For reasons you will not control and may never understand, a few will hug your heart, and they will become for you touch points—signposts—like that big boulder on that favorite hike that, when you spot it, tells you exactly where you are. If you allow it—and youshould allow it—these patients will enter your soul, and you will, in a way entirely right and proper, love them. These people will be your teachers.

Isaiah taught me. He was 15 when I met him. It was 1984, and I was the officer of the day—the duty doctor in my pediatric practice at the old Harvard Community Health Plan. My nurse practitioner partner pointed to an exam room. “You better get in there,” she said. “That kid is in pain.”

He was in pain. Isaiah was a tough-looking, inner-city kid. I would have crossed the street to avoid meeting him alone on a Roxbury corner at night. I’m not proud of that fact, but I admit it. But here on my examining table he was writhing, sweating in pain. He was yelling obscenities at the air, and, when I tried to examine him, he yelled them at me. “Don’t you f—–g touch me! Do something!”

I didn’t figure out what was going on that afternoon. Nothing made sense. I diagnosed, illogically, a back sprain, and I sent him home on analgesics. Then, that evening, the report came: an urgent call from the lab. Isaiah didn’t have a back sprain; he had acute lymphoblastic leukemia. And we didn’t have his phone number.”

more here.

Two Hundred Years of Surgery

Dr. Atul Gawande at the office

Dr. Atul Gawande’s latest article is available for free on NEJM.

In his usual analytical-turned-conversational manner, he reviews the evolution of surgery over the last two hundred years. He describes the roles that milestone advances such as anesthesia, antiseptics and technology played in making surgeries more successful and more common. In a telltale sign that he is at heart a huge measurement-geek, he even consistently measures these advances against the number of articles in the NEJM devoted to them over the years. Squeamish beware: there are some juicy/graphic descriptions of operations performed. Here’s an excerpt about the public health horizon and the opportunity to discover more surgical innovation by transforming the delivery system (emphasis added):

The increased safety and ease of surgery have produced an explosion in the volume of operations being performed — to at least 50 million annually in the United States alone. At the present rate, the average American can expect to undergo seven operations during his or her lifetime. This profound evolution has brought new societal concerns, including how to ensure the quality and appropriateness of the procedures performed, how to make certain that patients have access to needed surgical care nationally and internationally, and how to manage the immense costs. As early as the 1970s, researchers began documenting substantial rates of fatal errors in surgical care, wide differences in outcomes among institutions, and large disparities in access to care both within the United States and between countries. The science of effectively routinizing surgery for mass populations is still in its infancy, as it is for all areas of medicine. The Journal is entering its third century of publication, yet we are still unsure how to measure surgical care and its results. Experiments in the delivery of care will probably provide the next major advancement in the field of surgery.

 Note – also check out his TEDMED talk in April, where he covers surgery from the professional angle, discussing individual improvement, the role that teamwork can play, and the phenomenon of cowboy-ing in surgery.

The Highlight of my Week

The man with the plan: Texas State Senator Kirk Watson, with Erin and I.

…was meeting State Senator Kirk Watson. He swung by our inaugural health2.0 kickoff event last night to show his support for our idea – to build a kickass health IT community in Austin.

Senator Watson has spearheaded the health reform movement here in Austin through taking on a list of initiatives. The movement and organization is called HealthyATX, and the goals include setting up a healthcare incubator in Austin, developing more innovative ways for health care delivery, and bringing a medical school to Austin. He’s already made some progress on the last one.

If I haven’t said it yet, this city is awesome.

Patient Consent, PHI, and Data Integration

I pinterest searched for 'PHI' and got the phi phi islands in Thailand...So here you go.

I’d like to draw your attention to two reports that were released this week. They highlight some of the wrinkles in the emerging area of the privacy of digitized patient information, data portability and patient centeredness in all of this.

NYCLU: Consent is King with PHI and HIEs

The New York Civil Liberties Union made headlines for releasing a report criticizing New York State’s information exchange policies. Specifically, the report says that patients are not given enough notification or authorization to share their personal health information (PHI), that information should not be shared across the board, but only for relevant information. There are a few different grievances aired:

Consent: Opt-in vs. Opt-out: NYCLU would prefer a complete opt-in model but recognizes the limitations (low uptake no info in case of emergencies). New York uses a hybrid, where patient information is included by default, but they can veto when a provider actually wants to access information. In the case of an emergency, there is a ‘break the glass’ provision where providers can act on the patient’s best interest in the event of an emergency. NYCLU offers this third model, “opt out with exception” as one to consider for HIEs and RHIOs.

Geographic Reach: Patients who want to include their record on a local RHIO, with doctors they may have known for years, may decide against sharing it with a more extended network of doctors and hospitals.

Information Completeness: Also known as ‘granularity,’ this is a tricky one. Patients may not want to share all of their information with all of their doctors. NYCLU provides a smattering of potential offensive cases: sharing history of a hemorrhoid with a podiatrist, psychiatry with a surgeon, etc. Recognizing that there are lots of opportunities for potential misuse or transmission of sensitive PHI, NYCLU recommends that HIEs mandate increased granularity of what information is available to who.

There are a whole slew of recommendations in the actual report. But the general takeaway is that the way current HIE’s are planned, patient consent and concerns about the privacy of their information is woefully overlooked by state governments, HIE governance and commercial vendor communities.

HealthAffairs: Include Psychosocial Information in EHRs

The other article is a HealthAffairs piece (login required) advocating for the integration of psychosocial health information into mainstream clinical EMRs. The premise is that data on behavior, mental health, self-assessments etc. is invaluable to the treatment of individual patients, preventative population health management and public health/biomedical research efforts. In the article, a panel of bigwigs from various academic institutions lay out a series of cogent points: this information is increasingly readily available with technology, it needs to be put into standardized formats for easy incorporation into broader medical software, and more broadly psychosocial considerations are at the core of effective treatment and research.

Fine. But a quick search of “privacy” only produced one hit:

            “It will be possible to collect data from a majority of patients on sensitive psychosocial issues only if patients are informed about what will be done with the data they provide and assured that their privacy will be protected. They need to be told what their information will be used for, who will see it and for what purposes, and how their confidentiality will be safeguarded.”

That’s it. The piece even goes so far as to pre-emptively counter some objections to the inclusion of psychosocial data, and the authors left out privacy concerns altogether.

What does this mean?

It’s just another example of how our healthcare system has people working against each other. Supposed patient-centered advocacy groups such as the NYCLU are reduced down to nagging whistleblowers who don’t seem to acknowledge the challenges of technical integration or the difficulty of developing intricate, piecemeal consent models on granular levels. Supposed health care experts from Harvard and the NIH blow off privacy concerns as a red tape issue without even paying lip service as to how patients – people – might feel about sharing their more intimate life details (sex, drugs, mental health) with doctors, nurses, administrators and researchers.

Where’s the middle ground? Are there any examples of health care systems who are at the cutting edge of technology, science and patient-centeredness? What about systems who have figured out the longstanding problem of behavioral health integration into the clinical and technological workflow without compromising the sensitivity of those issues? Would love to hear your thoughts.