AI in CDI:

Fact vs Fiction

In healthcare, AI has become a major buzzword, and for good reason: It holds real potential to transform the industry.

AI healthcare market growth between 2019 and 2020

$3.9 billion to $6.7 billion1

Projected AI market growth between 2024 & 2030

$32.3 Billion to

$208.2 Billion1

524%

Growth1

Amid this growing hype, it’s essential to evaluate AI critically. We’re here to help.

AI is a broad term that can refer to many different types of algorithms and training models…

and not all are as powerful as the next. So, how do you separate fact from fiction in the CDI universe? Let’s explore what AI really means, how it’s being used in CDI, and debunk the top falsehoods with facts.

A simple definition

Algorithms that can perform tasks that require intelligence, like understanding languages, recognizing images, or making decisions using data.

Is it fiction or fact? Let's explore the truth about AI for CDI teams.

Fiction

More people can help us capture all of the missing or incorrect diagnoses.

Fact

Even high performing CDI teams can’t review every data point of every chart — no human can.

Even if hypothetically, a CDI team achieved 99.5% chart accuracy, that remaining .5% could potentially add up to millions in missed revenue for a large health system — as well as improved quality of care scores.

Many CDI teams have technology in place that prioritizes charts, but still doesn’t address all charts. No matter your team’s benchmark, AI can meaningfully help improve revenue and quality capture.

Fiction

More people can help us capture all of the missing or incorrect diagnoses.

Fact

There are 30,000 data points per patient chart and 156,000+ ICD-10 codes.

It’s simply not feasible for the human eye to review that many data points for every discharge in a given shift, let alone catch every single missing or incorrect diagnosis.

Advanced clinical AI can process and analyze all of that information, then present its recommendations to CDI professionals for validation.

Fiction

AI is going to replace my CDI team.

Fact

AI is here to make the lives of CDI teams easier, not replace them.

44% of CDI professionals did not agree that they were adequately staffed for the workload they had.2

With AI doing the analysis, CDI pros can practice at the top of their license and evaluate the most relevant information AI has identified.

Fiction

AI and new tech investments cost too much. Our hospital simply can’t afford it.

Fact

Not all technology investments have to come with a big price tag.

CDI teams can explore a couple of paths here: First, look for solutions that can offer an attributable ROI.

For example, explore clinical AI that sits in the prebill stage, but after your team’s CDI workflows and processes. That makes any findings directly attributable to the AI, and gives the CDI team an opportunity to approve and validate.

Second, consider new models like contingency-based models that share in risk and reward by only paying for approved and billed findings from the new revenue that is captured.

FINDING THE RIGHT AI

SmarterPrebill is clinical AI that helps CDI teams review 100% of patient charts to fully capture the value of the care delivered — both quality and revenue. Even with other technology and/or services in place, SmarterDx has been able to find an average of $2 million in annual net new revenue per 10,000 patient discharges for its hospital and health system clients.

1. AIPRM 50+ AI in Healthcare Statistics 2024, 2024
2. ACDIS “Industry Overview Survey” Results, 2023

Contact Us

Let’s get differentially better

Ready to learn more? Get in touch with us today and we’ll set you on the path to revenue integrity.

Thank you! We will be in touch soon.
Oops! Something went wrong while submitting the form.