Call9 (YC S15) Delivers On-Demand Doctors in Emergency Situations
techcrunch.com
techcrunch.com
1. Nursing home staff should be able to recognize typical and atypical signs of cardiac arrest ('should' being a key part of this).
2. My county's 911 service (serving 300,000 people) has the following response times: 6.0min for BLS (Basic Life Support) services, 6.1min for ALS (Advanced Life Support). 15 minute response times would be... a whole new world.
3. Impending cardiac arrest, I'd be curious as to what the 'difference made' would be. Administering ASA (aspirin)? Not emergently significant. Nitro? Typically self-administered, and again, I'd question the quality of a nursing facility that didn't feel comfortable administering such drugs based on vitals and symptoms alone, without needing the assistance of an ED physician (granted, several in our area as a policy/liability concern will immediately phone 911 as their 'treatment plan' for these situations - but similarly, these same facilities would be on the low end of the bell curve for utilizing a service like Call9).
4. Similarly, a nursing facility that wasn't vigorous in training and utilization of CPR/AED (because that's the number one hope for an arrest patient, high quality chest compressions - in the absence of tele-EKG for administering cardiac drugs for certain arrhythmias and dysrhythmia - again, if your 911 arrival time is 15+ minutes, perhaps...)
I can definitely see a use case for anything that moves from "proactive" (nor necessarily in a positive, often actually a negative sense) ED / 911 visit called by a skilled nursing facility. Indeed any increase in the use of a community outreach / service for high risk patients is a good thing.
I did read that you are indeed going to supply EKG and US to clients, and this helps - but I'd be curious about the value of some interventions. A presentation of impending cardiac arrest should be a first call to 911 (I realize your example presented with constipation and stomach pain). And in the context of having to talk a nurse through an EKG, what interventions do you really expect them to be able to reliably perform (start an IO/IV)? I am intrigued though, to see how this could grow, but it seems to me that your target might want to be 'urgent care' as much as emergency medicine.
Of course, where 911 response truly is that slow (and I know that it can be in several parts of the country), then anything that helps the patient's prognosis cannot be a bad thing.
Protocol with 911:
* Person has suspected heart attack symptoms.
* 911 is called, short triage on the phone.
* Ambulance is dispatched.
Protocol with this new startup:
* Person has suspected heart attack symptoms.
* Skype with doctor, short triage on Skype.
* Doctor quickly trains nurse (conveniently available on site) to take EKG (presumably while patient is grasping his chest).
* Nurse takes first EKG in her life.
* Doctor reads EKG, and calls 911.
Who in their right mind would call a gatekeeper instead of 911 in an emergency?
1) When you call 911, you talk to a gatekeeper. That gatekeeper is a phone dispatcher, not a doctor. Talking to a doctor is not a delay in care--it's the reason people call 911 in the first place.
2) Since Call9 works with facilities such as nursing homes and has access to patient information, using Call9 actually saves time in the dispatch process.
3) As described in the article, it's not always that clear to the person calling 911 has "suspected heart attack symptoms". In the given example, he presented with abdominal pain. That is why having the doctor be the first point of contact is so important.
Wait. The facility shares PHI (protected health information) with Call9? Have the facility residents consented to this?
Too much time is taken while care providers ask each other vague questions about "do you know what he/she has?" when in my head I'm thinking "You should KNOW this. Or have it readily available!"
Your service might be useful in non-emergencies. It's basically like a family doctor visit on Skype. There are a number of companies offering this, but it's not clear how this helps with outcomes. (For example, http://nyti.ms/1fyWXeK)
For healthcare facilities, like nursing homes, you are basically providing a doctor on call via telemedicine. That is already heavily used. Many hospitals employ radiologists, neurologists, etc who can view diagnostic images and data remotely and provide consultation to the medical staff on site.
What I think is dishonest (and could cost lives) is advertising your service to the general public as an alternative to 911, and especially claiming that it is faster than 911.
It is true that putting a doctor first in the process adds another step before calling 911. But it is a meaningful interaction that improves the quality and efficiency of the medical care (both by having an ER doctor assess the patient, and by having information ready so that the 911 call can go faster). It’s not about cutting out steps, but reorganizing the system to optimize care and save lives.
As it exists now, calling 911 for an ambulance sets off a chain of 9 human-to-human information transfers and averages over an hour for you to get to the hospital and be treated by a doctor. Many articles and research papers are out there showing the inefficiencies of the current system.
Current 911 system STEP 1: You call 911
STEP 2: Cellular service carrier connects you to an operator (4.5% busy signal rate)
STEP 3: Operator with no medical background takes your basic information and location
STEP 4: Operator types that info into a processing/relay system
STEP 5: EMS Dispatcher reads the relayed info
STEP 6: EMS Dispatcher radios paramedics to go to scene [Steps 1-6: average of 3.9 mins]
STEP 7: Paramedics type address into dashboard GPS system
STEP 8: Paramedics drive to scene [Step 7-8: average of 9 mins]
STEP 9: Paramedics treat, stabilize, and package you
STEP 10: If advanced procedure or medication is needed, paramedic must call a designated physician in a local emergency department to gain approval.
STEP 11: Nurse or other employee answers phone and then finds a physician who is concurrently taking care of other patients in the ER
STEP 12: ER doctor hears your case from the paramedics
STEP 13: ER doctor approves advanced procedures or asks for clarifying information [Step 9-13: average of 14.9 mins]
STEP 14: Paramedic transfers you to the hospital
STEP 15: Paramedics drop you off at an ER (may not be the same ER who approved the advanced procedure) [Step 14-15: average 12.2 mins]
STEP 16a (CRITICAL): In critical situation, paramedic with 2 years of training and no access to your medical records tells what they know about your case to an ER physician STEP 16b (URGENT/EMERGENT): In noncritical situation, paramedic tells the ER triage nurse about your case
STEP 17: ER triage nurse tells the treating nurse assigned to you about your case
STEP 18: Treating nurse tells an MD about your case
STEP 19: You wait for an MD [Step 15-19: average 24 mins]
STEP 20: An MD finally sees you
As physicians who worked in the emergency departments of the world’s leading medical institutions we were able to see first hand the inefficiencies of the current emergency care system leading to poor patient outcomes and avoidable costs.
TLDR - 1) 911 is an outdated system in many ways 2) there is room for improvement.
STEP 2: <1% busy signal is a goal of and attained by a large number of PSAPs.
3: Many PSAPs have their dispatchers trained to the EMT level.
4 through 6 are happening concurrently, not as a single step flow.
6: Many PSAPs have a goal that the initial unit dispatch is within 60 seconds of call reception.
7: I've not seen a system yet that relies on ALS units punching an address manually into a TomTom. AVLS (automatic vehicle location system), and mobile data terminals pull the exact information that is entered into 911 (which whilst still prone to error, is an issue with Call9, too) and use GPS which is linked into maps + county GIS overlays.
8: Highly dependent on location. My county has an average 6min response time.
10: Most advanced procedures and medications are governed by offline medical control ("here's our protocol for using this drug / treatment. Indications, contraindications, dosing regime") - the goal is that the need to contact a physician is the exception, not the rule.
If the time taken for 'online medical control' is 14.9 minutes I'll be at the hospital already, because that's where definitive care happens. I, as a medic, am not going to wait on hold and have a leisurely chat with a physician for nearly quarter of an hour on scene. This is a gross distortion of reality. If I do need to call the ER doc, there will be one available, or the charge nurse will know that one needs to be found, now.
16a: at an SNF if we need access to medical records, we can get them. If a patient uses the same healthcare system, they're available at the hospital. Pretending like we're stabbing in the dark and shrugging our shoulders is devaluing the care that paramedics and EMTs do down to the 'ambulance driver' stereotype. Both are trained in patient assessment and history taking, even if not medically qualified to the same level as a physician. In 16b: In a non-critical situation AND there are insufficient resources, then triage is the most appropriate care situation. If you're non-emergent and there's capacity in the ED you're going to a bed.
17/18: As opposed to... what? Unless Call9 claims to be the sole definitive (and remote) care provided, there's going to be a transfer of knowledge.
Your example has it taking a quarter of an hour to get to the patient, an undefined amount of time treating, stabilizing and packaging the patient, supposedly before requiring an online sign-off and discussion with a physician lasting another quarter of an hour before finally loading the patient for another nearly quarter hour trip to the hospital, while some paramedic with no situational awareness gives a nurse a hand off to relay to the physician.
What really happens on scene in a cardiac arrest: BLS providers show up first, AED, CPR, bag patient, potentially with advanced airway, LMA, etc. ALS shows up, determines quickly whether to stabilize on scene or load-and-go, CPR still in progress. A call is made to the hospital, with no discussion with a physician, only the charge, giving pertinent information, and ETA.
"Code Blue, Ambulance, Bed A3, 4 minutes out."
RT, IV, pharmacy, imaging, are all ready. A team is ready to continue CPR. Physician is there. All staff with the exception of the physician move patient from gurney to bed, begin work, while physician gets information from ALS, and then begins directing the code.
We've done A LOT of user research including days of time on the back of ambulances, days of sitting in nursing homes and watching how a typical 911 call (not a Call9 call) works, countless conversations and late nights with world-leading experts in EMS care, and years of being an EM doc on the receiving end. . Paramedics are NOT just ambulance drivers. They are some of the most valuable members of our society (and surely undervalued/underpaid). But THE reason I left my job to pursue Call9 with my team, is because I have felt if only a trained EM doc could have been in the field, than many of my patients with unfortunate outcomes could have decreased morbidity/mortality. With Call9, we can do that, and save patients who are not otherwise being saved.
So, you and your team seem to be very competent with no tolerance for BS that endangers those you care for. But be careful not to project what you do onto the whole system. The system varies in efficiency, how much people care, steps in between, and so on. A whole spectrum. An outfit like yours might have less need for this service but I've experienced many that could use it.
That doesn't even count the whole "help less competent people in a nursing home assess a potentially, deadly situation" part. I'm totally with them on that.
I agree, the ideal world and the real world are two very different things. Though still, given the length of your average RN course (or even LPN, since as we're talking real world, many nursing facilities are supervised by one RN over many LPNs) versus EMT... certainly it "should" be better. You do make a very good point about the nature of chronic versus acute patients.
2. Thurston (http://www.co.thurston.wa.us/medic1/stats.htm) - I'm not entirely sure about the measurement of response time, would have to clarify with a dispatchers.
3. Thank you for calling out this. Agreed, ASA is very significant, and should not be discounted. I meant that emergently it's not going to improve patient condition, but I should have acknowledged that's only one part of the treatment plan. NTG, too, at least here, is typically administered or assisted by BLS with numerous contra-indications (brady/tachycardia, hypotension, etc), but most patients I have seen are instructed by their physician to self-administer with symptoms "similar to previous cardiac event". I'm presuming right ventricular infarction on your example patient?
As said elsewhere, I would like to see a service like this succeed, and apologize if I come off 'skeptical' in some of my comments!
Their 911 claims (and the people who attack them on HN) - all this could easily be just viral marketing. :)
It was common to arrive for a call of "general illness" only to find the patient in respiratory failure or cardiac arrest with no CPR in progress.
I see this as providing a fast way for clueless staff to up-triage patients.
My pet hate is always "Do you have a history on the patient or what's been going on?" was almost invariably answered with:
a) "No, this is my first day", or
b) "No, I just got on shift"
Oh. Good.
For regular people, this service is like a family doctor visit on Skype. Sometimes it can be handy, but it remains to be seen how this impacts outcomes statistically. (I wouldn't use this myself.)
For healthcare facilities, like nursing homes, this is like providing a doctor on call via telemedicine. Many hospitals already do this with radiologists or neurologists for example.
What is misleading is their marketing jump from these potentially useful scenarios to advertising as an alternative to 911 for the general public, and even claiming that they are faster than 911 when in fact they are a gatekeeper to 911.
"Since the 1970s, arriving within eight minutes 90% of the time has been the gold standard for determining the quality of an EMS system."
"NFPA 1710 states that first responders and BLS units must arrive on scene within a four-minute timeframe 90% of the time for all incidents. The ALS crew that must respond within eight minutes."
Source: http://www.jems.com/articles/2012/02/great-ambulance-respons...
The survival rate of sudden cardiac arrest is 8%, and it kills about 350,000-400,000 people in just the US. It's our #1 killer.
The difference between a life saved and a life lost is whether non-medical bystanders can intervene and perform CPR or use an AED within the first couple of minutes.
That definitely helps. But is not unique to Call9. Call 911 and you'll get dispatcher-assisted CPR. Counting compressions with you, talking you through the process.
i) do you have plans to gather data? Correct research would be tricky, but have you considered collaborations with researchers / statisticians? It'd be fascinating to see some robust numbers after a year
ii) in England a considerable amount of 999 calls are either alcohol or mental health related. You've limited the "custer base" to schools and care homes and etc, but how will you handle such calls? For example: you get a call for someone who has engaged in severe (requiring surgery and hospital admission) self harm but who has no suicidal intent. Do you have procedures for that kind of call?
Thanks for patiently answering questions! HN is sometimes somewhat hostile. :-/
I doubt it. I would never put my life in the hands of an insurance like app which may or may not work; say where data is not available. You can call 911 over all cell network even if yours is not available/no sim/unregistered phone.
I hope Call9 understands the emergency side of the business esp around PSAP? Otherwise it won't scale and will be liable should their technology fail to save a life.
A problem with mobile phones is location. Unlike a home phone which is tied to your physical home address, how will the emergency personnel know where you are if you pass out before telling them the address? Cell tower location find can be up to 100m away from you. Satellite requires 3+ direct satellite triangulation.
Then there is the routing issue. So the app connects to a Physician in a hospital? But hospitals don't own Ambulances. Ambulance companies are private in the USA. So a hospital will in turn contact the Ambulance company. Additional piping that can be avoided with a better model. Say; if Call9 connected directly with an Ambulance company that is running a Home Visit Program (HVP) ie. Paramedicine, then use the EMT to triage the situation using industry Protocols/ICD. Then, Call9 would be starting to solve Hospital Readmission Penalties/Rates. That's a big fish worth $$$ per patient and thus a great disruptive business model.
Now of course this is all easier said than done since the red tape to get this going is the major hurdle. The technology part is easy. I've been down this rabbit hole ;-)
A dedicated line gets you "Oh, this is Call9".
They still need to get pt demographics. ALS/BLS. The dispatcher still needs to find you an available ambulance. Dispatch it. Get the crews out and on the road.
Yes, the infographic breaks up a lot of that information, but the call receiving time is in many locations mandated to be within 60 seconds of reception to unit toned. And your service doesn't cut that to zero by virtue of "we have a dedicated line to an ambulance service".
And if they don't have that? Because that WILL happen. "Ambulance Co B, respond in place of Ambulance Co A, no units available".
"Call 911." And you've burned that time because you need to go back to the start.
Depending on the location, when you dispatch ALS/BLS ambulances this way, whose medical license are the EMTs and paramedics working under? Does the county medical protocol director like/want this?
In cases of cardiac arrest, you're going to typically want more than an EMT and a medic on an ambulance. CPR is, as you know, a man-power intensive activity.
The three things you mention are not really problems in a 'modern' PSAP. "911, what are you reporting?" "Confirmed cardiac arrest". Even without that, in most areas around here criteria-based dispatch is used to determine ALS/BLS with an erring towards ALS if there's any question.
In my county (Washington state), the protocol states (and is met with 94% reliability that from call reception for EMS, dispatchers should have toned the appropriate unit(s) with 60 seconds of call reception, and those units should be en route within 2 minutes of that (even for in quarters response).
SCA survival rate of 16% in Seattle, vs. 3% in Alabama or 5.5% in Toronto.
Still a long ways to go, even for the best regions.
But then again, I'm sure Call9 would rather be compared to 'how we fare versus the best EMS systems in the country, not the worst'. :)
Sensitivity: What's the OR, relative delta, or NNT in calls for changing prehospital acuity or intervention to a higher than initially suspected level?
Specificity: How often does the early contact result in a less intense, on-site, or diverted response?
So many questions! Would love to hear what you're finding.
(edit: copy)
However, I have met Tim before and can attest to his interpersonal candor on problems with EMS, and the reputation for rigor of his medical training. In short, I think he's a smart, hard working guy who is legitimately committed to solving a real problem.
Given my own experience (as an EMT, working in the ED as an admin and medical student, and as a federal emergency response officer in a previous life), I do think solutions along the lines of higher fidelity information on scene will be useful for patients and providers.
There will need to be piles and piles of evidence before we can conclude that products like this will provide the outcomes we want in our health system. In the meantime, I'll try to appreciate the obstacles I know this team has overcome to get as far as they have.
We've done our best to research the stats and present a thoughtful representation of the current state of EMS. We have many official and unofficial EMS advisors who have spent tireless hours to make sure this is true. I assure you, nothing on our site or in our advertising is meant to be misleading - the 911 system is antiquated and could benefit from an overhaul. Like many large systems, there are a lot of good, smart people involved, but change is difficult. We hope to work from the outside in, and save lives/improve outcomes in the process.
Research and academic integrity is important to me and my Call9 colleagues.
I think that this is not a startup, but a psychology experiment that needs to find out just how gullible we are.
Interesting service.
My first question is what differs Call9 from any medical first responder?
My second question: Thinking long term. Where do you forsee the barriers of our reactive emergency services in the U.s. from making traits of Call9 more standard.
Right now, when EMTs get to the site, they don't know anything about the patient. With Call9, EMTs can begin work right away upon arrival.
As for long-term, we want this kind of service to reach everyone - that's certainly our end goal. But traditional emergency services have to completely alter their business and operational models to offer what Call9 offers.
This is completely not true. You really need to stop lying when advertising an emergency medical service that affects lives. Nothing personal.
When Call9 has a private transport arranged by the ambulance service, whose medical license do the EMTs and paramedics operate under? The county/hospital associated therewith? The ambulance service? Your physicians? And depending on the answer to that, what does the County MD have to say about that? Or the DOH EMS section for that matter?
If your ambulance service's EMS personnel are operating under county auspices, and that county has a medical protocol, how does that mesh in with instructions from your physician? What happens if your physician's instructions differ from the county protocol?
Because that could and will be an issue. An instruction from your physician could lead to a paramedic losing their job (and has happened, when physicians have instructed the on-scene medic to do procedure X when the paramedic is not authorized to do so). Do you have a clear policy for this?
"Right now, when EMTs get to the site, they don't know anything about the patient. With Call9, EMTs can begin work right away upon arrival."
This is, as someone else noted, a very hand wavy, dismissive and inaccurate description. You have access to patient history from facilities, and that is a fantastic resource.
But to say that when I roll up on scene that all I know is "Go to address X for sick person" is not remotely true.
"You'll be seeing a 65 year old male with sudden onset left arm numbness, shortness of breath, shoulder pain with no known cause, and no relief. Pt has a previous cardiac history and has taken prescribed nitro with no relief. Pt caregiver states that they have been in good health with no recent illness." is a sample 'short report' I might get en route. There's a good amount of detail there.
What's happening on the back end?
Two 911 operators listen to the call, the Call Receiver works through a flowchart to determine base complaint and appropriate response. In tandem, the Dispatcher begins assigning unit(s) and dispatching them over the air. After the initial tones to get the units rolling, whilst the call receiver continues to gather more information, the dispatcher is preparing a short report to give more detailed information to the responding units.
The system does no such thing. Criteria-based dispatch is a powerful tool. Many counties also require Fire/EMS dispatchers to be trained as EMTs.
Expert systems help ask the appropriate questions to establish the base complaint, and the level of service needed, the response level, lights/sirens, what-have-you.
See: http://www.kingcounty.gov/healthservices/health/ems/communit...
Additionally, a very large portion of 911 calls come from patients who should not be going to the emergency room. Going to the emergency room is always exposing yourself to risk (such as hospital infections), not to mention unnecessary stress and pain. For the elderly, or very young patients, or very ill patients, this undue (and often unnecessary) stress can have very detrimental effects on the patient. Call9 can prevent that portion of calls from going to the hospital.
So with call9, we either A: provide safer, better service on top of traditional 911, or B: we skip the emergency room altogether.
If the patient needs an ER visit, you are getting one, however it is initiated.
Community-outreach / in-call? Absolutely. I love that aspect of your service.
Your "A" and "B" benefits seem to be the same, not an either-or. The safer route in A comes from skipping the emergency room in B where applicable and appropriate.
I do like that you cut the waiting time to see an MD, regardless of the situation. Diagnosis is an important first step. But I'm curious about the 'medical kit' provided. Because diagnosis is limited to cases that don't require (for example) lab services, and for non-emergent cases, great. Prescribe the appropriate medication or treatment regime. For anything else, the outcome is likely to be the same - not much is changing in that "golden hour" (with all caveats thereto) in terms of definitive care.
I realize I may be coming across as cynical or skeptical in many of my comments, but I really do wish you success and would be very happy to talk with you further. I'm just raising some of the thoughts and questions that come to my mind as a first responder and, as I'm sure anyone will agree, anything that alleviates pressure on the 911 system is "A Good Thing(TM)".
If you were curious, like me, why Call9 didn't use call9.com, don't visit that domain to find out. It will redirect you to a malware website.
IMHO, Call9 should attempt to acquire the call9.com domain (or sue for it), since it clearly serves no decent purpose.
The "Request Access" button doesn't function properly on Firefox, due to the use of... an <input type="text" ...> there (not sure what the purpose of this is, but surely you could get by with just a <strong> or something.
'Murica, your shit is beyond repair.
Now, you want people to question that in a time of emergency? "Crap! Mom is having a heart attack! Should I call 911 or use Call9? It's supposed to be better! Shit, where is my smartphone, all I have is this landline. Crap, my smartphone is upstairs, I'll just use my tablet. Wait, I don't have Skype installed on my tablet. Dammit, I'll just call 911".