DigitalReasoning.com Artificial Intelligence-Powered Oncology Software Value Creation & ROI Highlights CASE STUDY In 2016, HCA partnered with Digital Reasoning to conduct a wide variety of data science experiments. The most promising experiment was tested and piloted. In 2017, a staged roll-out scaled our software across the entire enterprise to provide 162 hospitals with AI-powered cancer care. In one year, results were impressive. The overall system-wide volume growth was very strong. The oncology volume growth rate tripled from the prior year (2016), adding more than 10,000 cancer patients nationwide from Jan. 2017 to March 2018 (Figure 1). Figure 1. Oncology Volume Growth Validated net new oncology patients directly attributed to software (system-wide; Jan 2017 to March 2018) Across all HCA’s markets, the results were positive. For ease of comparison, Figure 2 shows an example of a single market/MSA that compares to a health system with one major hospital, cancer center and a mix employed and affiliated physician groups. This site saw 788 net new patients directly attributed to the software from Jan. 2017 to March 2018 (Figure 2). Figure 2. Oncology Volume Growth - Single MSA Validated net new oncology patients directly attributed to software (Single market / MSA; 1 hospital & cancer center; Jan 2017 to March 2018) Breast # Retained 10,000 8,000 6,000 4,000 2,000 0 Lung Complex GI Colon Total 10,227 1,274 1,613 2,223 5,117 Breast # Retained 1,000 800 600 400 200 0 Lung Complex GI Colon Total 788 145 219 282 142
2
Embed
Artificial Intelligence-Powered Oncology Software · Direct Patient Interaction Physician Relations Patient ID & Triage Data ntry 70% of time report reading, prioritizing, classifying
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DigitalReasoning.com
Artificial Intelligence-Powered Oncology Software Value Creation & ROI Highlights
C A S E S T U DY
In 2016, HCA partnered with Digital Reasoning to
conduct a wide variety of data science experiments.
The most promising experiment was tested and
piloted. In 2017, a staged roll-out scaled our software
across the entire enterprise to provide 162 hospitals
with AI-powered cancer care. In one year, results
were impressive.
The overall system-wide volume growth was very
strong. The oncology volume growth rate tripled
from the prior year (2016), adding more than 10,000
cancer patients nationwide from Jan. 2017 to March
2018 (Figure 1).
Figure 1. Oncology Volume Growth
Validated net new oncology patients directly attributed to software (system-wide; Jan 2017 to March 2018)
Across all HCA’s markets, the results were positive.
For ease of comparison, Figure 2 shows an example
of a single market/MSA that compares to a health
system with one major hospital, cancer center and a
mix employed and affiliated physician groups. This
site saw 788 net new patients directly attributed to
the software from Jan. 2017 to March 2018 (Figure 2).
Figure 2. Oncology Volume Growth - Single MSA
Validated net new oncology patients directly attributed to software (Single market / MSA; 1 hospital & cancer center; Jan
2017 to March 2018)
Breast
# Retained
10,000
8,000
6,000
4,000
2,000
0Lung Complex
GIColon Total
10,227
1,2741,613
2,223
5,117
Breast
# Retained
1,000
800
600
400
200
0Lung Complex
GIColon Total
788
145
219282
142
DigitalReasoning.com
Figure 3. 12 Month Volume Growth – Single MSA
Validated net new oncology patients directly attributed to software (One major hospital with mix of owned and affiliated referring physician groups; Jan 2017 to Dec 2017)
Figure 3 provides an even easier comparison for
a single market oncology program by limiting
the same aforementioned market analysis to one
calendar year. 2017 saw our software discover, help
navigate and retain 540 net new oncology patients
for that hospital and cancer center (Figure 3).
Improved productivity and speed-to-treatment
show how AI helps Nurse Navigators via dynamic
triage, prioritization, and care complexity models.
Prior to implementation, navigators and care
coordinators would comb through pathology reports
to find positive results, triage, match to pathways
and document follow-up – all now automated. After
implementing our solution, the amount of time
spent directly interacting with patients doubled
(Figure 4). Navigator caseload increased by 50 to
250%, equivalent to hiring 114 new navigators. By
spending more time navigating and coordinating
care, more patients were able to receive the benefits
of nurse navigation and average speed to treatment
decreased by 5 days.
Figure 4. Navigator Time Analysis
% Time Spent (Before) % Time Spent (After)
Navigators’ Time Spent:Pre/Post Study
PRE - DIGITAL REASONING POST - DIGITAL REASONING
Direct Patient Interaction Physician Relations
Patient ID & Triage Data Entry
70% of time report reading, prioritizing, classifying into
pathways and documenting- non-value add activities.
65% of time now spent on outreach, patient navigation