Business Opportunities for using AI in Healthcare | Saurabh Bhatia | Skillshare

Business Opportunities for using AI in Healthcare

Saurabh Bhatia, Author & Teacher

Play Speed
  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x
10 Lessons (39m)
    • 1. 001-Introduction

      0:54
    • 2. 002-Benefits part 1

      2:53
    • 3. 003-Benefits 2

      3:56
    • 4. 004-Sample analysis

      6:55
    • 5. 005-AI intervention part 1

      6:04
    • 6. 006-AI intervention part 2

      3:21
    • 7. 007-Data collection

      4:08
    • 8. 008-AI Models

      3:23
    • 9. 009-Stakeholders

      3:06
    • 10. 010-Challenges n conclusion

      3:59

About This Class

This course would be immensely useful for both healthcare professionals as well as software professionals.
Healthcare professionals will be able to understand real life use-cases where they can use AI and ask their vendors to build AI enabled modules for them.
Software professionals, on the other hand, will understand the needs by seeing real life examples and will be in a better position to collect data and build AI modules for their customers.

In this course, I have covered the basic benefits of using AI, Sample analysis using a real life example, the stages in clinical care life cycle where AI can be used for intervention, What will be the data requirements and how to gather that data, AI modelling and extending those models to healthcare, how to bring the stake holders of healthcare together using AI and finally, the challenges for implementing AI in healthcare and how to resolve them.

Transcripts

1. 001-Introduction: hello and welcome to the schools on business opportunities for using air in healthcare. I'm Dr Sort of idea and I'm going to be your instructor for schools before we move ahead a little bit about me. I'm a doctor. I'm a specialized in medical informatics. I'm also certified in artificial intelligence, big data and fire technologies. I'm certified in our high tech HIPAA and the child processes. I'm an entrepreneur and I build software courses. You can find out more about me from Lincoln. The link is available in the Slave has a domain export in good health care in a. I am going to lay out the falling topics in the schools who can be better fitted in health care, how to go about applying here in health care, business opportunities and challenges and solutions. So without further ado, let's go and start. Of course, 2. 002-Benefits part 1: So let's talk about the benefits off Using the benefits of using AI are available to both patients and providers off health care services. In this course, If I use the word provider, it will include all medical staff like doctors and nurses, as well as healthcare institutions like hospitals. It will also include non medical stuff, a hospital. It's so let's see what other benefits to patients. The first and the most important benefit of the patients is in the form of better care. If the artificial intelligence cannot really improve the quality off care that the patient is going to get that stuff really worth it, the next possible benefit is less a number off complications and the recurrence off any particular disease. Commonly, what we face in the medical world is complications after a surgery or after treatment of a particular disease or the recurrence of disease. This causes misery to the patients apart from initial cost of treatment. So if artificial intelligence can help there, it will be like nothing else. The third thing is cost effective coping models. So currently there is no really good way off assessing what is the efficiency off a doctor and the provider to treat a particular patient and therefore the patient's Jews go pay models as per their financial condition, or whichever way they want to pay the premium. If they pay a higher premium, they can have a lesser out of pocket expense. If they can pay a lower premium right now, then they'll have higher out of pocket expense at the time off. Admission are medical need, so this is the basis off the copay bottles right now. But artificial intelligence can help change. This can help change in deciding the copy model for a patient based on is held, his risk factors and the other things which are not taking care off right now. But they can be understood by the neural networks off artificial intelligence. So what other benefits took care providers and stakeholders? And when I say stakeholders, I mean insurance companies, pharmacies and others. So first of all, there's lesser readmissions and lesser workload because, okay, I will be able, tow, intervene and improve the recurrence rate or the complication rate higher performance of providers. This means a provided can give better service more service in lesser cost, so this can increase the performance and the remuneration will also improve. Cost saving interventions will increase profits. So if there are interventions which suggests there cost saving, then there is more money left as a profit for the hospital. So these are the perceived benefits off a I and in the score. Subsequently, we're going to see how these benefits can be brought about in real life. 3. 003-Benefits 2: so continuing from the previous section, Let's talk about the costs related benefits. It has been proven without doubt, that dude heard off clinical outcomes dependent patients in situations outside the hospital . I'll give you a bullet to digest that. So did you read it again? 2/3 off clinical outcomes dependent patients is situations outside the hospital, not inside. So it's not really the doctors work or the nurses work or the Impatient Time, which controls 2/3 of clinical outcomes. What controls them are things like living conditions off patients are living in healthy in Veron's or not. Are they? Are they over the poverty line? Are they able toe afford good carrick home or not? Need off logistic support, like transported there in the fall sick if they want to come to the hospital, is there easy availability of transport or not? Is the patient really a winding to come to the hospital just because it's difficult for him to get a transport? Or is the transport too expensive? Cooked food is another issue, so a patient usually is expected toe take some kind off food in certain diseases, which is different from the routine food that a patient that a person consumes. So is the special food available to him or her or not, financial condition of the patient? Is the patient able to afford the kind of care or not? Is he or she avoiding to come to the hospital merely because it is going to be expensive? Or you have to do, ah, a lot of out of pocket expense to pay for the visit so he or she is not willing to to visit the hospital, moving ahead, using artificial intelligence and machine learning, the hospital can gain access to the above factors and then the cycle Rickman interventions in high risk cases. So what kind of into inches we'll talk about them in a subsequent section detail about what can I do to intervene in such high risk cases? But the point here being that when the above interventions are done, the readmission read complications and costly in with investigations, especially the ones which are repeated again and again. They can be brought down significantly, and this feel improved revenue of the hospital, the bottom line of the hospital, the clinical outcome for the provider and for the patient. This will save the costs and have a better quality off life. So what is the business opportunity? Okay, let's see. The first thing is, do you really want to approach artificial intelligence to predict in hospital or out of hospital solutions? Remember the tutor to wanted issue to third off. The outcomes depend upon situation off the patient outside the hospital, while 1/3 off them depend upon the situation inside the hospital. What's tend to the patient inside the hospital? So which is the lower hanging fruit? I think it's pretty obvious that the lower hanging fruit is what is happening outside the hospital, and artificial intelligence will get much more data and ability to intervene at that stage . Finally, some mines two pointer for you before end section. Can you think of any other benefits which are available to the patient or the hospital? Apart from the ones we have listed out here, can you make a list of these additional benefits that can accrue due to the use of AI outside the hospital, specifically outside the hospital? Now we'll move to the next section often example, and we'll do a simple analysis 4. 004-Sample analysis: So in this section we're going to do a sample analysis. Awful example patient. The example is that there are two identical patients with similar medical history and treatment going on in the hospital. The difference exists in financial condition and living conditions. So that's assume that both of these patients met with a road traffic accident for for some reason, the fracture, their legs. And now they're in the hospital for the treatment of that no one off them. After receiving the treatment, he has been sent home because a fracture cases not really kept in the hospital for months on end. So one is being looked after by family and has adequate money for caretakers, while the other one on the right side of your screen is struggling along with war finances . So which one off these two is more likely to end up with complications? What do you think about it? You can pause the video and think about that. What will be the effect off those complications on the outcome in case of good patients, if any, So the possible outcomes for this kind of case and the 1st 1 in the biggest will is readmission to the hospital. Patients who are not being looked after properly enough or who have poor finances are more likely to have a complication and get to be admitted in the hospital. The result off this can be a poor final outcome. When I said the first outcome, it waas when the patient was first treated and discharged, but the patient was not fully okay. He still had a plaster cast. He was still moving around with the help off crutches or any other moving device, and one off them landed up with greater complications and he had to be re admitted. For this particular patient may be the final outcome of treatment is border than the other one. Maybe he had a situation very is born would not join. Maybe he had an infection in his in this injury site, so the final outcome can be poor. This would also deserve in greater cost of frequent. A second admission with the complication like an infection, for example, will cost greater money in antibiotics and admission costs. And lastly, there will be a fall in perceived quality off life. It's quite possible that the complication may lead to a permanent disablement of patient in one way or the other. It may not be that the patient is maimed for life, but at the same time the complication will result in persistent pain, disfigurement, and the patient will feel that his quality off life for the remaining part has Bean compromised. So once again just to do a recap, the possible outcomes for the patient with lesser ability to look after himself. I don't you to finances are living conditions. Outside the hospitals are readmission complications, poor final outcome, greater cost of treatment and fall in perceived quality of life. Now what does this mean for the provider for the hospital, first of all, it means low patient satisfaction. A patient who was having situations which are bad outside the hospital never really blames his own situations for the complications he's got. It's always assumed that the hospital and the doctors have given a lower quality of treatment, so the satisfaction rate of the patient goes down even though the doctors have given exactly similar treatment to a different patient who is having a bigger up. This leads to a decreased performance off provider. The provider hair will be seen as having lesser efficiency because a case got complicated. Two cases came. One case became all right. The 2nd 1 got complicated, so it's a 50% complication rate. That's not good for the hospital. And the biggest irony is that hospital is not the one who is causing the complication. It is not the one that can control the situation. The long term problem for this is increased in premiere for the patients from insurance overall been throughout the year when it is seen that the number of patients having similar problems, what is the number of them getting complications? The insurance realizes that patients who are prone toe a particular disease. I mean, anybody can be thrown for a road traffic accident here, but let's say chronic diseases like diabetes and complications of diabetes or heart disease . So what's the insurance? Realizes that a significant number of patients have complications in the poorer final outcome. The premium for those kind of insurances can be taste in some countries. This is against the law, while in others it is not so, depending on that the patient was already poor. Who's already not having good financial and logistics situation will also have to give ah higher premium for his insurance the next year. There will also be a decrease in the reimbursement rate for providers. This is because they have been perceived as having a higher complication. This is because the complication was happening outside the hospital and the provider could not do anything about it. But even then, the reimbursement given to them by insurers will be lesser, because the insurance will feel that it was possible to do a better job for this patient. But you did not do it. So both the provider, as well as the patient, suffers in this case. So there's the business of what you did. The business opportunity here is nobody is really gathering social data financial later and family data, which can have a lot off impact on the outcome off a particular patients pregnant. Currently, most of the electronic medical records they capture the date of it in the hospital or whatever his street taking the doctor does. If the doctor misses out on certain things, there is no other way to capture the data, so you will find that most of the medical records are deficient in this kind of data. And even if this data is provided to an AI system, it will not yield a satisfactory prediction model because the data sets are incomplete. So here is the business opportunity for on souvenirs to provide a system where social, financial and family data which can possibly impact the outcome of cases can be collected, related and used for ah high This prediction using a a point of wonder for you before we sign off from this section. Think of possible examples where situations not ordinarily handle where doctors can affect the outcomes. Think about these possible examples where situations outside the hospitals get affect the outcomes. 5. 005-AI intervention part 1: in this section, we will see how the artificial intelligence interventions can help patients in the Kreuter's. The place is to intervene. Can be too, either within the hospital or post discharge. We're going to see in the section how the artificial intelligence could intervene and reduce the readmission rate for the visions. What other business opportunities inside the hospital. Let's talk about these one by one CDSs, which stands for clinical decision support system. Currently, there are a lot off. CDSs is available in the market in doctors used also do not. All the they're not very happily, but at the same time they're forced to use it sometimes. But these cdss is our old and they are based on the knowledge that already exists. The beauty off AI lies in the fact that new knowledge, which waas existing but not available in a concrete form to the medical staff, can be related using neural networks and brought it into the CD S is so the clinical decision support systems, which are using either but Beijing learning or they're using the existing knowledge, can be augmented with neural networks and artificial intelligence. The next thing is the rules engine currently a lot off clinical decision support is utilized by simple rooms building into this office. There's a basic rule if the patients B M Eyes about 25 sent him into the obesity clinic. If the blood sugar is about this levels and new to the diabetes clinic, so on and so forth. So these rules get triggered when a certain value is noted into the electronic medical record system. But there situations where these may or may not be appropriate or these may differ from country to country, geography to geography and also the deferment based on region, religion, ethnicity and culture. So these dual engines can be upgraded using a where all these factors can be taken care of by itself and still treatment. The best clinical intervention automated documentation is another area where he has got a big rule. It is now well known that artificial intelligence is already providing automated contract management systems in the legal departments off various companies, so this very ability can now be brought it into the hospitals where the clinical case sheets and the documents which have been created through the time when the patient was in the hospital, can be correlated and created into a discharge somebody for the patient. Currently, even in the advance hospitals, the patient is expected to collect a discharge, some weed 36 to 48 hours after the discharge, or sometimes a descent on lane later on. The next biggest thing is claim rejection prediction, and this is actually a very big business opportunity. A lot of hospitals have multiple Bayer engagements. They have several different stakeholders were going to pay for the patient's treatment and different stakeholders and testicles here. I mean, the insurance companies, they have different requirements or documentation. A lot off claims get rejected not because they were not properly done in a clinical way, but because the documentation was in problem. So artificial intelligence on one hand can help in the clinical workflow to serious is while on the other hand, it can predict the claim rejection in various documents off a patient. When it is realised that the document can be falling short off standards with respect to the particular player, which is going to find this particular treatment, so claim rejection prediction and taking an intervention, re step their toe, complete the record so that the claim is not rejected is going to be a big boost for hospitals and did what only hospital stay prediction is another big area. Hospitals have limited resources. There are certain number of bags. They have a certain number off equipment. They. So if a particular patient is going to use a particular equipment predicting the length off , stay off the patient in the hospital is going to help the hospital plan their resources in a much battery cost. Off treatment Prediction. Hospitals these days have different models for treatment, especially off court cases. So typically, there will be a package available for Let's a cardiac bypass surgery, and this particular package will be costing a certain amount. And it will be including off certain days off. Hospital states that in a moment of investigations, certain treatments, if anything, falls out off that the patient is expected to pay for it. But at the time off admission, there is no way to predict whether the spaceship is more likely or less likely to have expenses outside the package, so artificial intelligence can be very helpful in finding out if a particular patient is high risk case for having additional expenses and the patient can be prepared in advance to keep the cash reserves in case of additional requirement of money, the treatment should not be compromised. Lastly, use off device prediction. Every hospital contains certain number of devices, which can be used only by one patient at a time, for example, of until it. If a patient is on ventilator, we know any other patient cannot use it intermittently. Hey, I can be very helpful in understanding that how long is a Is a particular patient likely to use that precious resource like a ventilator? And when can the ventilator be really will do subsequent patients? 6. 006-AI intervention part 2: No, let us see the out of hospital business opportunities. Most of this lays in the fin services, and, as I said in the earlier section, and I made you think about it, think in terms often ambulance. If there's a cardiac case who requires to have certain kind off medicines given to him within one hour, and that's called the Golden Hour, then if the hospital is at a distance off 30 minutes or 45 minutes from the patient's home , or if he knows that there's a certain time of the day when they're the rush, it's rush hour, and there will be traffic jams encountered. Will it be more sensible to send the ambulance along with a doctor who can administer the treatment there itself, or will it make more sense to bring the patient back to the hospital and then it minister the treatment? Both cases are welded. Both situations are valid, but they require differentiation. So there is no blanket solution available that with every ambulance, we cannot send a doctor who's goingto start the treatment. It missions home itself and at the same time, all the time, said Justin Ambulance, which cannot start the critical treatment righted the home itself is also not a good idea. Think about pharmacy pharmacy opportunities Like today, we have online medications, which are reaching your home, but at the same time, sometimes they're costly and the cost off sending them to your home is also included in that. So can be a be utilized to see. How can the pharmacy supplies be made better? This is. There's also situation where people who use veterans, hospitals or ex servicemen contributed health schemes that there are medicines which are falling short and they're not available at the time. So patients who have to take chronic medicines regularly, like diabetics they happen, sometimes bite from overexcited. Higher cost can be used to predict how many patients will require how much America medicines and dosages at what time so that the procurement of medicines can be smoothing. Note caretakers for the patient. Usually there are countries like India where usually it's the family members, where the caretakers well in the west, where a lot of people live alone, especially in at one stage, they have to hire a caretaker can be used to find out if the spaceship is a highly Skase because of the absence of caretakers are because of the expense off hiding a caretaker. Similarly, we have talked earlier about food and home equipment availability. Complication prediction is another thing. As we're talked with the schools, 2/3 off complications are happening because of reasons outside the hospital. And if all these reasons have been captured in a system than the way will do more reliably , predict that if this particular patient is more likely to have complications and what can be done to prevent that? So let's have a point to ponder. Can you write a 500 word proposal about any of these friends services that have been mentioned stating, How would you like to utilize AI for creating a benefit for the patient? 7. 007-Data collection: So in this section we talk about the data requirements for having an effective artificial independence. We had talked earlier that there is a requirement for debt asserts to be complete without a good, incomplete data set. It's difficult for any kind of a engine, or I'll go to them, too. Make any 11 predictions. So what kind of data do we need? The clinical data is already available to some extent in the electronic medical records. This can be diagnostic data, a prognostic, better descriptions, etcetera. But the remaining three letter that you see are very sketchily available in the current systems. These are typically ease off excess off medical facility. Is the hospital too far? What is the distance between the patients work and hospital and home in hospital? Is there Is there availability off caregivers? Is the patient himself capable to drive or not? What are the possible scenarios with the patient will be stuck at his place away from the hospital in an emergency financial letter like affordability of treatment. Can the patient afford a certain type of treatment? Is the solvent or not access to better performing providers? Or so the reverse situation where the patient goes to an inefficient or more costly provider just because that Reuters close to his home is the patient being to insurance or self being. And usually this data is captured. But what needs to be captured is what kind of plan the patient his own. Is he on a plan where he's doing adverse reaction for himself by having ah, high GOP high out of pocket himself in exchange for the low premium? In this scenario, the patient tends to a wide coming to the hospital. And lastly, family did like non modified with genetic data. Is it available? Hasn't been collected properly. What if I had been living situations? The patient's own attitudes, psychology and support? All this data is not really gathered today, and it will do a lot of good for the future of the patient care that if this data is collected, the animal system will be able to help in the following cases. If the debt actually showed in the previous slide is collected, it can identify high risk. Patients can also identify early intervention points. It can recommend actions generate for intervention. I mean, identifying an intervention point sometimes is not enough because even then the doctors order medical stuff, doesn't know what to do, so they also needs to recommend in action collision off results. Will intervention was done? Were since when it was not that this is important. It's important to show that an intervention helps and this is a kind off feedback for the system itself, and it will be a positive feedback to the system can do more off those interventions and recommendations, and this weather system will learn for future optimization, off interventions and recommendations based on previous visits. So the ability to improve clinical outcomes using both clinical as well as non clearing a letter improves the providers. Performance improves the outcome for the patient and Botham stakeholders will be happy. So what is the business opportunity here? A software that conformed the providers about their lagging performance due to poor outcomes compared with the providers. And this comparison can be learned within the same hospitals between various doctors. Within the same community would be in various hospitals and within the same country between various hospitals, communities, a CEO's or other type off hospital chills. Now that you know what kind of debt as important can you lie down, Fired at the points that you have learned are important for a intervention. But they have never been collected from you or anyone you know, while visiting a doctor or a hospital in the past. 8. 008-AI Models: in this section. We're going to talk about artificial intelligence model, so let us see what other types off AI models available. There are three main types of model which we'll talk about Miss Scripted Model's predictive models impressed Victor models. So descriptor morals. They used that aggregation and data mining to provide insight into the past. An answer. What has happened so essentially this? This means learning from the past. Predictive models use statistical models and forecasts techniques to enlist in the future and answer what could happen. So essentially, they're predicting the future possibilities and descriptive models. They use optimization and simulation algorithms to advise on possible outcomes. An answer. What should we do? So essentially, they're recommending an intervention. No, let us extend these models in health care. So in health care, if the model is descriptive, that means it is learning from the past. Then it has to utilize the existing data to correlate with the outcomes. And this way it also has to find out the deficiencies in data models. So, as an example, the media patient whose data record is complete and its outcome is also recorded. The A. I can see that these were the data points, and this was the outcome. But to compare it with the similar patient who has had a similar disease, the AI needs to find at least one more record, which has as much data completion as this one. If it doesn't, then it can point out all those records for the data was incomplete. To be able to make it comparable toe the pilot data in predictive models where the prediction is done for the future possibility. The prognostic data and statistical models need to be used for prognostic debt essentially means what we thought is going to happen to the patient and what actually happened to the patient. So the prognosis will be utilized for future predictions In prescriptive models, the system is expected to recommend an intervention so optimizing the statistical models to include high risk factor reduction essentially, whatever has been seen in descriptive directive models that will be utilized for prescriptive models because the AI will be able to see certain kind of behavior or certain internal situation is high risk and at what point and intervention should happen so as to prevent the happening off a particular idea situation. So where is the business opportunity. The first thing that needs to be done is harvesting the current data states and to determine deficiencies in that data. Then advising the hospitals demand additional data collection. Trump's off prep Reuters. Unless the hospital knows that what data is important for directing a better outcome, they will not be able to collect it, and they will not be ableto ask this off. Providers toe get therefore that data collection in this software systems further, it will be important to advise this software companies regarding what additional later to include to make this off a ready. This will also improve the cell ability off a soft. 9. 009-Stakeholders: in this section will talk about bringing the stakeholders together. Improved health care outcome Benefits All stakeholders in the stakeholders here are patient provider PCP or the primary care provider Insurance and the Cure. Using AI, we can identify points where an intervention would alleviate the current problem as well as a potential problem, which has been identified as a higher institution. Eat stakeholder must be made aware of the actions that he or she needs to take and its impact in the overall picture of patients. Clerical up. So let's let's assume a doctor who's being told that he needs to find out how many people are living in the family of the patient and what are their timings. And will they be able to bring the patient to the hospital on a regular basis for a checkup or north over the daily somebody or not in the home toe, figure out whether the patient is able to take his medications on time or not. These things that the wanted doctors do not usually together this kind of data points and usually they're time is more expensive. But if the hospital is made aware that what will be the impact off gathering this kind of data and how will the patient's clinical outcome improves? Then the hospitals will be able to take some action and collectors data, whether through doctors or to other person, a very defined benefit in terms off our way that is, return on investment as well as the clinical endpoints helped the stakeholders work towards a predetermined definitely cool. So what it means is stick orders. Lake, hospital and insurance will have to work towards over notifying clinical. And there will be a clinical situation in which we can say this is a desirable outcome and they will also be a clinical situation very well said. This is an undesirable outcome it could have been through into so defining. Such outcomes is going toe help. The providers and other stakeholders like the insurance, work towards a definitely gold instead of just saying this patient was treated better than the other patient. But not able to say how and not they will do. Therefore little in the cost of treatment in the cost of reimbursement for a kid, This business opportunity in training sessions to be conducted for hospital staff, insurance companies, software builders as well as patients and their caretakers. There's a huge business opportunity for people who understand this, and once they're found out what data needs to together, it can be utilized toe provide training to all these stakeholders so that they can actively collect, monitor and utilize the doctor that has been collected. So can you think off some challenges that will prove to be a hurdle in fulfilling the vengeance? 10. 010-Challenges n conclusion: in this section, we're going to talk about the challenges toe a implementations. The telling regarding patients are disengaged. Patient is the most harmful thing for a patient. Yes, for the for himself, the patient. The patients are not usually aware of the potential pitfalls off their disengagement with the primary care provider in the hospital. A lot of times, patients do not tell the true situation to the hospitals. I difficult of embarrassment or shame, or simply because they don't think it is important to be told. So the hospitals need to ask him pointed questions regarding the shortcomings in and out of the hospital care people are sometimes not forthcoming about de financial situation, especially if they are in a border situation and they don't have any pretensions. A lot of times patients will not discuss co morbidity is co morbidity means and a similar on existing in listen the patient. So, for example, a patient who comes having who are vision and is not telling the hospital or the doctor about his existing deputies is doing a huge disservice to himself because the patient needs to be examined for devotees also, and in person, contact may be necessary to engage or disengage patients. So patients who think that doctors are going to ask more questions they're going to prescribe more medications going toe us for more visits. It's simply because they want to make more money. It's important that they be convinced that the money part will not change. In fact, this is going to actually aggravate the situation of the hide information and do not give the complete set off information to do so in person contact with the patient, maybe of critical importance. Incomplete data improvement off Emma's or AM ours or the software operating companies is necessary because social in financial later which is not really being collected right now needs to be collected so that the I can make predictive and interventional in the morning. More time needs to be spent by personal collective of committing the physics longer. This is not a desirable situation because the sweet this means more people will have to be hired by the hospital. More expense needs to be done in the EMR in the software's, so that that I can be collected so the hospitals are going to resist this particular action . It depends on the service provider to devise a solution where lesser expense and lesser need off additional man power can be shown, and yet the data can be collected for the identification off. 11 data As per country culture in geographical regions is important because every data point is not really went in every geography, religion, culture or country technology of great barrier. There's a cost of upgrading to newer hardware or a software, and this should be brought down by easing out the rules. Off Engagement with software provide little cloud providers. Skill personal will also need to be hired or existing workforce may need to be re skilled to optimize the cost off a origin's to conclude the schools There are four important points that I would like to talk about. Two. Remember 2/3 of outcomes depend upon out of hospital situations. Appropriate data collection can help a identify patients with high risk situations. He Aikins had just timely interventions to prevent unwanted situations. Technical, financial, personal and data related challenges will have been resort for effective air conditioning. I hope you like the course and you have been able to learn what you came to learn here for . Thank you