Stay Home

This post was originally featured on HIStalk.

AliveCor just announced AliveInsights, an integrated telemedicine solution that lives on top of AliveCor’s smartphone ECG. AliveCor is advertising that a cardiologist will read one’s ECG within 30 minutes.

Everyone is fighting to reduce costs through telemedicine. Beyond regulatory issues, the greatest challenge has been delivering medical devices and equipment to patients’ homes so that doctors can receive that data. AliveCor is making interesting moves in this space by providing an integrated telemedicine service with their ECG hardware.

How and when will all basic medical services be delivered remotely? I see two distinct models emerging.

HealthSpot has been making waves recently for setting up a private "health station" or "health kiosk" that costs less than $15,000 to install. Right now, they’re piloting the kiosks in ERs and a few retail clinics. I suspect their vision is to install the kiosk in every retail clinic and school throughout the country. The kiosk only consumes about 44 square feet and can be operated by a medical attendant for just $30,000 – $40,000 / year. The kiosk includes a stethoscope, otoscope, dermascope, thermometer, BP cuff, scale, and pulse ox reader. It can support most simple PCP visits.

HealthSpot isn’t employing physicians, but selling these kiosks to provider organizations so that the providers can extend their reach virtually through HealthSpot. This is interesting because there could be HealthSpot kiosks in Maine with Mayo Clinic branding. Although this makes sense for prestigious provider organizations in the short term, I can’t help but think that this practice will ultimately lead to a dilution of brand and prestige in the medium to long term. It will become increasingly more difficult for patients to recognize brand value when most provider interaction is virtual.

American Well, Teladoc, Ringadoc, and others are taking the more extreme view that patients shouldn’t leave their homes at all. They are also employing or contracting with physicians directly as opposed to selling into existing provider organizations. This model presents the challenge of getting equipment to patients in their homes. I suspect this problem will solve itself in the next 2-3 years. Amazon, Google, and others are investing an enormous amount of capital and resources in bringing same-day delivery to the masses.

I can foresee a future in which the tele-mid-level providers will see patients virtually, determine what additional equipment is needed, and rent the necessary equipment to patients for a follow-up visit later that day, or at worst, the next day. Amazon and Google would gladly support this model to drive enormous volumes of shipping traffic. If the at-home diagnostics suggest more serious complications, the mid level would loop in a higher level provider or specialist.

I’ve spoke to HealthSpot’s CEO Steve Cashman about this threat. For now, HealthSpot is taking baby steps, but recognizes what may become an existential threat. I’m sure they’re planning for this future. It will be interesting to see how things play out in the "stay out of the hospital" telemedicine space.

Who's Eating Healthcare?

This post was originally featured on HIStalk.

As Robert Cringley recently noted, computers empower unparalleled discrimination. Before insurance companies could calculate rates on an individualized basis, they calculated rates based on population pools. They simply didn’t have the computing power or prowess to discriminate at the individual level. As a result, the healthy financially supported the unhealthy by average premiums across population pools.

In the 1990s, the cost of computing fell to a point that payers could discriminate. So they did. Payers could easily identify patients that would incur high costs based on a relatively simple set of questions about one’s health. For many patients, payers were so concerned that healthcare costs would be so high that they’d prefer not to take on any risk at all and simply refuse to insure the patient. This has been the controversial norm for the better part of the last 15 years.

One of the most important clauses of the Affordable Care Act is that which mandates that payers cannot deny coverage for any reason. Payers must price that risk. In many cases, they expect that the costs of care will be so large that they are distributing those costs across their entire insured populations. Paul Levy recently noted that this is happening to such a degree that many healthy individuals are seeing their premiums increase under the Affordable Care Act.

This is direly ironic. Computers, the ultimate discriminatory tool, can no longer discriminate. Over the 15-20 years of the discriminatory cycle, healthcare costs have systematically outgrown the GDP as obesity has risen to become the #1 killer in the US. Coupled with the fact that no one is allowed to be uninsured (most of whom weren’t healthy to begin with), premiums are, on average, increasing for a substantial percentage of society.

Is this a classic case in which computers are eating healthcare, or are we as a society eating ourselves?

Anesthesiologists are Moving up the Value Chain

Dr Leng is right. The future of anesthesia isn't about competing with CRNAs and computers. Cost pressures dictate that eventually, mid levels and computers will win. Instead, anesthesiologists should look for what else that can be done - and that only they can do - given that more junior staff can do what anesthesiologists once had to do.

Anesthesiologists must move up the value chain. That will likely manifest pre, intra, and post op:

Pre-op: Devote more time and energy to the pre-op processes. That may even mean working with surgeons (if they can), planning and communication with the patient before they arrive for the procedure, work with OR staff and pre-op nurses in the prep processes, and ensure that all pre-op checklists are being adhered through (even endoscopic scope cleaning checklists).

Intra-op: Support the most complex intra-operative anesthesia cases.

Post-op: Monitor patients more closely in the PACU and help with discharge planning to ensure the fastest, safest recovery.

In a nutshell, anesthesiologists will move up the value chain into a more managerial role. Most providers are by definition doers, not managers. As mid levels continue to proliferate throughout every avenue of care, doctors, particularly anesthesiologists, are poised to become the grand overseer of the entire peri-operative flow.

Assuming a managerial role will introduce a new set of challenges, primarily around communication and collaboration between anesthesiologists and their mid levels. Pristine EyeSight helps solve that exact problem.

Pristine EyeSight is Pristine's HIPAA compliant, secure, 1st person audio / video streaming solution built for Google Glass. Pristine Glass wearers can share what they're seeing, hearing, and saying securely to any authorized device - iPhone, iPad, Android phone, Android tablet, Mac, or PC. EyeSight is the ultimate communications tool.

Anesthesiologists and their mid levels have felt tension as they compete over the same cases. The problem hasn't been the mid levels, but the fact that the anesthesiologists didn't have the tools to assume a more managerial role. Pristine EyeSight is the ultimate communications tool that will connect CRNAs and anesthesiologists so that anesthesiologists can move up the value chain. As anesthesiologists move between the PACU, the outpatient clinic, and pre-op staging areas, they'll use EyeSight to beam into ORs through their mid levels.

The future of anesthesia looks brighter than ever before.

Tackling the Top Safety Issues of 2014 Through Pristine Glass

This post was originally featured on The Pristine Blog.

The ECRI Institute just released a list outlining the top 10 safety issues facing hospitals in 2014. The Joint Commission will be looking at these issues with particular focus over the coming year as they audit hospitals.  This list helps us understand the problems that are currently plaguing hospitals and how we can leverage our solutions to address them.

Eyeware computers, including Glass, present incredible new opportunities to improve patient safety at lower costs, thus driving value. Our solutions on Glass solve several of these acute problems by enabling new abilities which can transform existing procedures. In short, we help deliver lower overall costs and safer medical care.

6) Inadequate Reprocessing of Endoscopes and Surgical Instruments - This has been an acute challenge because scope processing staff can't use their hands to interact with digital or paper checklists. This problem is particularly difficult to deal with because the cleaning process for many of these scopes involves 40-50 steps that can take over one hour. Each scope per manufacturer and model has to be cleaned according to a unique set of instructions. With just a few models of scopes, the system is highly prone to mistakes; it's very easy to miss a single step. Using Pristine CheckLists, scope processing staff are able to utilize checklists per manufacturer and model with their voices. Additionally, because Pristine CheckLists support images and videos, staff can take pictures and videos at each step along the way without using their hands.

10) Retained Devices and Unretrieved Fragments - Everyone has heard of horror stories in which a sponge, towel, or other item was left inside of the patient during the surgery. Many vendors are developing solutions that attempt to tackle this problem. Given that there are literally dozens of opportunities for error per surgery, the ultimate solution will probably be multi-pronged. Using Pristine EyeSight, we're helping provider organizations understand what causes mistakes. And using Pristine CheckLists, we can help nursing staff maintain an accurate count as they place items inside and remove items from the body cavity.

Glass and other wearable computers will leverage context to drive value. The practice of medicine is defined by specific actions of hospital personnel in particular contexts, often in conjunction with other providers and workers. Pristine's apps on Glass will support a future of medicine in which both computers will assist humans as contextually necessary.

How We Recruit Talent

This post was originally featured on the Pristine Blog.

At Pristine, we don't believe in paying for job listings on sites such as LinkedIn, Monster, GitHub, and StackOverFlow. All of these sites, and many others, use the same model: pay a fee, usually $350-$450, to post a job for 30 days, and pray that good candidates apply. After posting the job, we can try to optimize the ad, but at the end of the day, it's still a matter of praying that good candidates find the listing, read it, and apply.

There are a few problems with this model:

1. There are no incentives for performance. The job sites run off with our money regardless of the success of the job listing.

2. Exceptional candidates have to find the job listing among dozens of job listing sites. The probability that a given exceptional candidate will find our job listing is slim.

3. It's hard to stand out. Yes, there are ways to make the title of the ad stand out, but at the end of the day, our job ad is just another blue link on a page with 24 others. We can't even include a picture of Google Glass to appear on the index of our listing to try to draw attention. We are bound by a monotonous wall of blue text.

4. The most important reason that job listings aren't great at finding exceptional candidates is that the best employees are usually already employed. By definition, exceptional talent is always employable, and thus almost always employed.

So here's how we recruit talent:

We dig through hundreds - perhaps thousands - of profiles on Linkedin, AngelList, and GitHub by hand. We skim through the profiles of every candidate, looking at previous work experience and types of projects they've worked on. We ignore education entirely. We generally don't spend more than 10 seconds on the first pass at a given profile. We employ slightly different strategies for each site. These are our favorite sites because they have large numbers of candidates that we can effectively target:

On LinkedIn, most developers don't list their technical skills in detail (so that we can search for them). Thus,  searching by language or technology stack leads to poor results. Instead, we filter against companies that we know use the technologies that we use and that are located in Austin. If we find a candidate that we think might be qualified, we message them. We never use LinkedIn InMail (LinkedIn's built-in mail service) unless we cannot find another form of contact information. Talented developers smartly ignore all recruiting related InMails on LinkedIn. Most talented individuals provide links from their LinkedIn to their Twitter, personal website, or some other online profile. We dig through these other profiles to get a better sense of who they are and what they're capable of, and to find an alternate channel to reach out. If all else fails, we even resort to guessing their email address for their current job by using the Peep tool and some basic intuition. We only resort to LinkedIn InMail when all else fails.

AngelList - AngelList is an excellent recruiting tool because of the self-selection effect. Unlike LinkedIn, which is filled with garbage, AngelList has a high proportion of talented individuals who're hungry and eager to work at startups. On AngelList, we use the talent filters, "looking for" filters, and geography filters to find qualified candidates. Once finding a potentially qualified candidate, we reach out through AngelList (which sends an email). Talented individuals tend to respond to AngelList messages because the nature of the site lends itself to more friendly responses. That may not hold true as AngelList continues to scale and the overall quality of the applicant pool decreases, but it does for now.

GitHub is our favorite recruiting website because it provides direct links to projects that developers have worked on. We can very quickly assess developers without calling them. GitHub provides utilities to filter by location and programming language, which while not ideal, are sufficient. Perhaps the best aspect of GitHub is that most developers tend to list their email address publicly, making it very easy to reach out.

So, once we've found a potential candidate, how do we go about engaging them? We've A/B tested our opening messages. We've tried emails that detail why we're magical and revolutionary, and we've tried emails as short as "You look like a rockstar. We make awesome stuff. Call now." As in most facets of life, a middle ground is best. Our opening message is concise, sincere, and provides just enough of a tease to intrigue candidates:

"Hi, my name is Kyle Samani. I'm Founder and CEO of Pristine, a well-funded Capital Factory startup in Austin that's developing apps for surgery for Google Glass.

I was just perusing your LinkedIn / blog / AngelList / GitHub. [write comment something relating to their blog/profile/linkedin/angellist]. You rock. You're exactly what we're looking for.

Do you want to spend the rest of your life [something related to their current job], or do you want to invert healthcare delivery models through Glass?

PS, feel free to learn more about me, the company, and our recruiting practices."

Why is this message so effective?

1) I introduce myself as Founder and CEO. Most talent is sick of hearing from recruiters. They will take the request a lot more seriously coming from the CEO.

2) We let them know we're funded, and in a well-established institution in Austin (Capital Factory). This lets them know that we can pay them, and that we're legit. No one wants to work for a loser.

3) We tease them with "Google Glass for Surgery." If that doesn't intrigue them, nothing will.

4) After intriguing them, we make a sincere, informed comment about something they've published on the Internet (blog post, GitHub project, tweet, etc).  That shows that we took the time to read about them online and get to know them a bit before asking for their time.

5) Then we pull a classic Steve Jobs by asking them to dream about the future. It's a teasing ask. It lets them know we want them, but puts the ball in their court.

6) The "PS" clause establishes one of our company values up front: transparency. We make it easy for them to learn more about me, the company, and if they're so inclined, the very message they just finished reading.

Our hiring process is extremely time consuming. I don't enjoy it, but I do enjoy the results of it. It's perhaps the single most important tactic we employ. Our team is mind-blowingly awesome, and I wouldn't have it any other way. I can't do the scouting forever. When I no longer have time to do this, I will ensure that whoever does will scour the Internet just as fiercely as I do. As the company grows, we must scale our talent, and we'll do whatever it takes to achieve that goal.

PS, here's my inspiration for Pristine's recruiting practices. This is one of the best blog posts ever written on any subject.