The switch to electronic records vastly increased the amount of time to complete notes.
Is there something inherent in medical notes that typing them should take longer than hand writing them? Or, is the software just awful? (I assume the second, which makes me wonder why nobody is disrupting this area)
Both. It's extraordinarily complex. The data entry is highly structured. All the electronic record software I've seen has probably ~100 pages worth of forms, a big subset of which usually need to be filled out for each patient visit.
It's not being "disrupted" because it's extremely capital intensive to enter this market. The requirements seem intractable:
- Structured data with very high level of granularity.
- Patients often don't fit the mold exactly. It's hard to classify something that is going on with a patient; there are 68,000 ICD-10 codes.
- Data needs to be interoperable with existing and competing systems.
- Every hospital has their processes and documentation practices. As a result, most electronic record software is highly customizable, and a company representative is usually on call or on site to continuously customize it.
- Software needs extreme backwards compatibility. You need to be able to read records that are tens of years old.
- Need to be able to print everything.
- Need to meet regulatory requirements and security audits. Regulatory requirements vary by state and country.
- Healthcare institutions are inherently conservative with their technology, and for likely good reason. Plus they don't upgrade equipment very often. If they just spent $1 billion upgrading their entire IT infrastructure, they're not going to consider your startup's solutions for awhile.
- Patients don't like it when their provider is focused on a computer or taking notes instead of on them.
- Different UI for each provider (e.g., primary care physician vs. pharmacist vs. radiologist)
- Need to communicate with a wide variety of medical hardware that doesn't have good standard protocols, and much of which is very old.
The main problem is that hospital administrators and doctors have very different incentives. Administrators want to bill as much as possible, whereas doctors want to take care of patients. However, administrators are the ones who make purchasing decisions so EMRs are built for administrators, NOT doctors.
People say that EMRs are awful (And I used to think that). But the truth is that they're just not built for doctors, they're built for the hospital. And they're doing a great job at increasing billing, to the detriment of doctors productivity.
As an analogy, imagine a tenant / landlord relationship with a sink problem. The tenant asks the plumber to change the whole sink whereas the landlord asks the plumber to find the cheapest hack to fix it. The plumber has no choice but to listen to the landlord. The tenant and their friends blame the plumber for doing a shit job. But hey, soon after, a startup knocks at the door of the tenant and offer to change their sink for the best-of-the-market.. but the tenant would have to pay himself so he politely declines and stick to his old shitty sink.
My father is in his 60's. He literally cannot type. EMR laws have increased his day by 2 hours. In regards to scribes... It's PHI which is dealing with HIPAA, having medical training to understand terminology, etc. Hiring someone to understand "Arthroplasty, acetabular and proximal femoral prosthetic replacement" is expensive. That was just a hip-replacement, imagine dealing with something like cancer. I realize he's somewhat of an exception, but people that spend their day talking to patients rarely get the amount of typing practice that people who have the time to browse the internet do.
In regards to the second question, that's what I do. If you've ever worked in any corporate setting that has industry-specific software, convincing the company to change it is near impossible. "Time saving" is incredibly difficult to demonstrate and "cost saving" can take 2 years to show anything meaningful. Software for a complicated field ends up being complicated no matter how hard you try to not be the java-megastructure catch-all application, which means it will always take some amount of training. This means you have to convince someone who just lost 2 hours of their day to spend more time learning a new application. Welcome to the shit-show of corporate software.
As my father loves to counsel: "It's hard to get the nine-to-fivers to listen to the five-to-niners".
Is there something inherent in medical notes that typing them should take longer than hand writing them? Or, is the software just awful? (I assume the second, which makes me wonder why nobody is disrupting this area)