/// This is the 2nd article of the three-part series on ClinOps debt. Click here to read the 1st part, where we explain what clinical operators do. ///

In its early days, a digital health startup is defined by its search for product-market fit (PMF). It's a race against time to validate the business before the money runs out. So the team needs to move fast and can't get too attached to anything. 

On the ground, that might look like releasing a feature they know is buggy. Or doing a task in a way they know won't be possible once the number of patients increases.

The team knows they'll have to undo or fix these things later. But they have decided that's ok in exchange for something more important: time. 

For the lucky few who find PMF, the new challenge is keeping up with rapid growth (i.e., scaling).

Remember all those things they knew they'd have to undo or fix later? It's later. The bug they used to encounter once every few months they now face multiple times a day. Things are breaking. Patients are unhappy. Customers are churning. Morale is low.

They have to pay back the time they bought for themselves earlier. With interest.

In tech, this interest is called "technical debt." We think healthcare startups also pay interest on clinical processes they don’t embeded into their tech stack. We call this Clinical Operations (ClinOps) debt and argue that, for healthcare companies, it's as important as tech debt.

What is ClinOps debt?

In an earlier article, we described ClinOps as the role responsible for scaling the practice of medicine by translating clinical protocols into reliable, scalable, and repeatable processes.

ClinOps debt is the price you'll have to pay later for cutting corners when setting up those processes today. In other words, it's the result of prioritizing scrappy, manual processes over perfect, streamlined, automated processes.

“Healthcare uses humans as routers, forcing workers to toggle between disparate systems, Lane said. "They copy, they paste, they manipulate data, they become robots. They click and type and extract and import, all day long – and it’s one of the leading reasons that one out of every three dollars spent in the industry today is spent on administrative costs." - Sean Lane

The price of ClinOps debt:

• More than a third of the U.S. healthcare costs are administrative (about $800 billion or $2,500 per person; that's equivalent to a 24-year Netflix subscription for everyone in the country)¹
• 57% of clinicians are worried that they will burn out due to the number of repetitive tasks required in their role²
• An average of 58% of staff time is spent on repetitive tasks such as data entry and documentation²

These numbers are behind many healthcare workers' gripes with their current jobs. And ClinOps debt is a key driver.

When ClinOps debt starts to stack, it decreases job satisfaction, productivity, team morale, and quality of care. This eventually leads to the company losing money and patients.

ClinOps debt is a major reason for high employee churn and low job satisfaction, especially in this digital age. With the current way we treat healthcare professionals, chances are high that they will leave their jobs soon and start working in non-patient-facing jobs. This would make the war for talent in healthcare even worse. 

It's time care organizations start taking ClinOps debt seriously.

“The more steps and activities of the care delivery process are built into the software applications and automated, the higher the efficiencies gained and the lower the errors against the validated process. This benefits patients, care team and payers.” - CareOps Lifecycle

How does ClinOps debt emerge?

To answer this question, let's take a look at two examples with two different outcomes:

Company A is building a virtual-first behavioral health clinic. Since they built an MVP, their patient numbers steadily went up.

As with most MVPs, they designed the clinical processes to generate direct user feedback rather than fully automate them. So the onboarding flow looks like this:

• John is a 26 years old male struggling with anxiety. He fills out a Google Form on the website to let them know he's interested.
• Mary, the ClinOps person, wakes up in the morning and sees the form John submitted. She asks Alex, the care coordinator, to call John to learn more about what he's looking for.
• During the conversation, Alex asks John a couple of questions to make sure that the services they provide will be safe and effective for him.
• Alex determines that John is a good fit for their services, so she books him for an appointment with a provider, Fernanda, and adds him to the EMR.

Nine months later, Company A raised $15M and signed a contract with a Blue Plan. This means their total number of patients will drastically increase in a few weeks. 

There's no time to automate this onboarding process, so the product team decides to use a "bandaid approach." They promise Mary, Alex, and the rest of the ClinOps team that they will work on a scalable solution very soon. In the meantime, Alex will keep having to call every patient who fills out a Google Form. 

In the months after raising Series A, the product team continues to focus on shiny new objects that mention "AI" instead of streamlining the onboarding process or incorporating Standard Operating Procedures (SOPs) into the EMR. 

Over time, the ClinOps team gets used to the "coming soon" promise. As the number of patients grows, they have to evolve their workflows and SOPs to keep up. Some SOPs are documented in Lucidchart, while others live in Miro or Google Docs. Because the SOPs live in different places, it's unclear which SOP each care team member follows.

The patient numbers continue to grow by the day, and Mary has no time to translate clinical protocols into SOPs. There is also no time to train care team members. If care team members have questions, they need to reach out to her directly so she can tell them what they should do.

“Low performers are not successful at building their (digital) paper processes into their software applications. Operations happen by reading protocols off of a PDF or flowchart.” - CareOps Lifecycle

In the beginning, the ClinOps team could onboard patients in just two days. It now takes five to eight days, much longer than the company advertised. So naturally, patients start posting complaints on social media and leaving angry reviews.

Eventually, the lack of clinical and operational investment takes its toll. Most on the ClinOps team decide to leave the company and start working for another virtual-first care provider that listens to its clinical staff. Patients are churning and requesting refunds. Finding new talent to replace the people that left is hard. The problem is getting bigger every day.

Company B is also building a virtual-first clinical for behavioral health therapy (it's not a niche market). A rockstar team of ex-One Medical employees founded the company and raised serious amounts of money with just a Powerpoint presentation. No MVP, no proof, no customers. Just a rockstar care team that wanted to do everything “right” from the begining.  

To prepare for launch, the product team is hard at work on buying, building, and integrating a suite of fancy tools. They're using Athena Health for their EMR, Notable for automated intake, DeepScribe for note-taking, and hard-coding different workflows to automate and standardize all aspects of the care delivery process. 

Meanwhile, the ClinOps team's biggest fear is that they'll be unable to meet demand. So they're hiring a ton of therapists, coaches, and physicians. And because they want providers to be able to focus on patient care, they're staffing up on administrative staff even faster.

The CEO kept postponing the first big marketing campaign to ensure they could handle the patient influx. Once the campaign was launched, a handful of patients found their way to the landing page. The company burned too much money too fast, couldn't find product-market fit, and shut down.  

In this example, the problem is that the company was building things "for scale" too early and too extreme. 

Most engineering resources were spent integrating with 3rd party solutions instead of focusing on things that will make them stand out. It was difficult to change and iterate on the clinical processes because they were hard-coded. They had way too much clinical staffing available for the number of patients they actually had.

Closing thoughts 

Like tech debt, ClinOps debt is not necessarily bad. Sometimes (e.g., when building an MVP), it's even necessary. But what is bad is letting it pile up or not acknowledging that it exists. 

Healthcare startups must find the right balance between investing in clinical operations and product features. Adding a new feature that allows patients to connect their Smartwatch to your app is cool. But what's even cooler is helping your clinical operators automate and standardize (part) of their care processes so they can work more efficiently. 

Digital health companies searching for a fast track to profitability and higher margins should consider that 58% of their clinical staff's work is manual, repetitive, and boring. Given that the care team and care delivery process is the product, it is alarming to know how little investment they effectively receive.

Over the next few years, the clinical labor shortage will become increasingly problematic. Attracting and retaining the best clinical operators will be crucial for a company's success. Nurses and doctors went to school to work with patients and provide care, not to be data entry clerks. Companies that take that seriously will have a significant advantage. 

Now that you know what ClinOps debt is, the next question arises:

• What are the different types of ClinOps debt?
• What symptoms do I need to look for?
• What are ways to avoid ClinOps debt? 

The good news is that we will tackle these questions in our next article. The bad news is that we’re still working on it so you’ll have to wait :-).

If, in the meantime, ClinOps debt resonates with what you’ve seen in practice please reach out to rik@awellhealth.com. We would love to hear more about your challanges and ways you overcame them.

--

Quick intro: we’re Thomas and Rik, building Awell - a low-code platform allowing care teams to design, implement and optimize care flows in days, not months. CareOps grew out of our years spent improving CareOps at innovative providers..

—

¹ https://www.reuters.com/article/us-health-costs-administration-idUSKBN1Z5261

² https://digitalhealthwire.com/notables-state-of-automation-2022-report/