Implementing Technology to Relieve Provider Burnout: Tips for Health Systems

Research has shown administrative burden is one of the main causes of provider burnout. Punit Singh Soni, CEO of Suki, emphasizes that data entry is interruptive for clinicians, and many technology solutions are not scalable across entire health systems, leaving this issue unresolved throughout the industry. 

In this interview with Integrated Healthcare Executive, Mr Soni offers guidance for health system administrators to choose the best solution for their providers' needs. 

What are the main causes of burnout? What was the impact of COVID-19?

Clinician burnout is probably the biggest public health crisis in the country, but few people talk about it.

We have taken a very sophisticated group of people and overloaded them with so much administrative work that they have become glorified data entry clerks. These incredible professionals are spending time clicking boxes and inputting data, and then patients do not get the full attention of their doctors.

A lot of money spent in health care is not spent on clinical care but on the tools surrounding it. We spend a lot of money providing clinicians tools for documentation. And then, because these data are collected in a somewhat unstructured way, a lot of data is not useful for advancing clinical care.

I see a lot of issues in this space, but the singular issue that stands out is the huge impact on doctors themselves. In a recent survey, 88% of doctors said they would not recommend their profession to their own kids. Burnout can manifest itself in many ways, but the leading causes are significant documentation burden, complicated software that requires lots of clicks, and regulatory burden that adds more metrics to track.

COVID-19 has exacerbated burnout. Now, clinicians are dealing with a huge influx of patients in addition to changed work settings, virtual care, and all the issues they already had before. So the pandemic has only made burnout much worse.

Thank you, Punit. How else does burnout affect health care providers, patients, or the health system more generally?

If you are spending 2 to 3 hours per day documenting, clicking boxes, and checking things, then it is hard for you to love your job. Imagine spending 10 hours or 10 meetings per day talking to people, and then I asked you to write it down, attach numbers, and measure it. You probably would not like your job either.

We have added such an onerous amount of burden on these doctors that A) the job becomes much more frustrating; B) doctors find themselves hindered in clinical care; and C) the data generated is not used to help them. It becomes interruptive. We do not need to interrupt the doctor.

And then, at the end of the day, what patients see is the back of the doctor’s head. This is not because doctors want it to be this way; it is because doctors are trying to get through their day so they can see the next patient.

There is an impact on economics, clinical care, and data quality, as well as the actual retention and viability of the doctors themselves.

How can health systems and other health care stakeholders alleviate burnout?

The first question to ask is, how can health care technology companies help?

It is a classic answer to say that health systems are bureaucratic, and they do not really care about the doctor, which is why the system is the way it is. I do not think that is entirely true. There are a lot of well-meaning administrators and people who are trying to alleviate burnout.

The problem is that even if health systems were operating as the best versions of themselves, the technology they are using is from the 1990s. When I was building mobile games and platforms at Google, my developers had better technology than doctors have today—and doctors are saving lives.

I think health care technology has done a disservice to medicine. It is just a mishmash of all the regulatory requirements that must be pulled together and laid on top of a database. It is like taking a vice president of engineering and telling him there must be a comment written next to every line of code he wrote, and if the product fails, he will be sued. Poor technology has created a lot of burden on doctors.

Health systems can take a much more clinician-centric approach to their processes and programs. First, I would ask, what regulatory burden can be automated or offloaded, so the doctors can focus on clinical care? Second, I would ask health systems to consider adopting and using software that is user-friendly, similar to what we use in our personal lives.

Third, we should put the doctor at the center of the product. When an initiative is thought out, sometimes it feels like clinicians are an afterthought. The truth is, in health systems, patients come and go, but doctors are the ones who always are there. It would be useful to start with clinicians, and then work toward other goals.

Those are some of the ways I think you can at least take a stab at solving this problem.

In that same vein, what should stakeholders keep in mind when they are selecting technology or a vendor to help mitigate burnout?

There are a few things to keep in mind, especially considering the mistakes of the past.

When somebody is pitching a product to you in this space, it is useful to know if it is a technical product, service, or human-backed hybrid setup. This is important because a lot of products are just glorified humans in the back, filling out documents. That sounds fine, but it is not a scalable setup, so you end up with all sorts of quality issues. It costs a lot without resolving the problem.

These days, everything is about artificial intelligence. The question to ask is, which area of your product uses what artificial intelligence? Does it use a language processing algorithm? Do you have actual deep learning models? Do you have your own stack that you are building? Where is the data coming from? A lot of work is done to build a true machine learning model, so it is good to know which part of the product is driven by that vs what is just labels that mean nothing under the hood.

Third is to deep dive on data, because if it is truly an AI, machine learning product, then it must use a lot of data. Where is the data? Is the data being pulled out of the system? If it is, then are we following the right protocols for data security and integrity? How do you clean up that data so it can be fed into the model?

What health systems do is ask superficial questions around data security, which are important but may not get to the heart of the issues that machine-learning based products can bring up. If you are evaluating a health care technology product, focus a lot more on what the real AI is, making sure you know where the data comes from, and finally, building something super user-friendly that is scalable because it is pure software rather than a human-based system.

I think the big mistake of the past is people have pitched to health systems that they will solve their problem, and then they build all these custom solutions which are never scalable or repeatable. If this is a proper product—not a software service or custom product—do we have to make additional changes to deploy it several times?

As somebody who has built and run large-scale products in the consumer world, I think there is a gap between the kind of questions that health systems ask people they work with, and the reality of what they need to ask to build useful, scalable, thoughtful products.

Thank you for sharing that insight, Punit. Is there anything else that you would like to add?

In my opinion, the most interesting technology company ever built is going to be in health care. This is in part unfortunate because health care technology has missed the mark in helping and serving medicine earlier. But now, we have an interesting opportunity in front of us. There is a systemic change happening, accelerated a bit by the pandemic.

We should be able to rethink what health care technology looks like and make it invisible in a system. When you are building or evaluating a product, ask, who is your user? Whom are your product and company in the service of?

In the case of Suki, we are building a digital assistant for doctors. Every pixel in this company is in the service of the user. We can reduce clinical documentation burnout by 50% to 70%. We offer doctors a consumer product they are comfortable using in other areas of their lives, in a way that can take all that data and put it in the right places.

About Mr Soni

Punit Singh Soni is the CEO of Suki. Suki's mission is to make health care technology assistive and invisible so clinicians can focus on what they do best: taking care of patients. Before Suki, Mr Soni spent 8 years at Google in a variety of specialties, including research, product management for mobile applications, gaming, and software development for Motorola.