CASE STUDY
Strmeamling Claims Reimbursement
We enabled our client to maximize insurance reimbursements by deploying chatbot-assist billing code selection.
Frontline caregivers often have only a few minutes between patient visits – yet within that time, they must complete critical administrative tasks that enable hospitals to receive payment.
While the specifics of this process vary by institution, the process generally involves submitting detailed claims to insurers, outlining the services rendered and associated costs. Although much of the process has been digitized, caregivers are still required to manually identify and input treatment codes from extensive databases or digital catalogs. This manual step is a significant source of error.
According to industry sources such as Change Healthcare and the American Medical Association (AMA), coding errors are responsible for approximately 15% to 25% of all claim denials. One major hospital system, our client, was experiencing a $900 million impact to operating capital as a result of these denials, prompting them to engage Fei Business Advisory for strategic guidance and technical support.
In under six months, our team designed and deployed an AI-powered chatbot that reduced coding errors by 65%, delivering an annualized increase of $600 million to operating capital. The solution enabled frontline caregivers to quickly and accurately identify treatment codes using plain English terms, and was optimized for mobile use across caregiver devices—allowing them to complete tasks seamlessly between patient visits. In addition to dramatically reducing errors, the solution also returned valuable time to caregivers, allowing them to focus more on patient care.
Fewer coding errors
By enabling caregivers to use plain English to identify treatment codes, the chatbot significantly improved coding accuracy and reduced the risk of claim denials.
Reduced claim denials
Fewer errors translated directly into recovered working capital, saving the hospital system $600 million annually by reducing denied claims and rework.
Greater caregiver efficiency
The mobile-friendly solution allowed caregivers to complete administrative tasks between patient visits, freeing up more time for direct patient care.
CASE STUDY
Programmatic Contract Assessments
We enabled our client to reduce legal costs and timelines during supplier contracting.
Effective third-party management programs encompass an end-to-end, integrated approach to assess risk during the contract process. However, skillset limitations or workforce shortages frequently require the use of legal services to review new or amended agreements, driving up costs and increasing timelines.
FeiCorp engaged with several client Procurement Teams to help reduce costly bottlenecks associated with legal review. By collaborating with in-house legal departments, category managers, suppliers, and product managers, FeiCorp built a custom solution for one client that produced a risk score for each new agreement, allowing the organization to introduce process for specific outcomes.
The solution used existing contracts that served as precedents for comparison against new agreements. By transforming existing contracts to vector embeddings, we leveraged machine learning models to measure the cosine similarity between new and existing contracts. FeiCorp created a framework to assign risk scores when new contracts deviated by certain thresholds, allowing software engineers to further fine-tuned the model for greatest accuracy.
We observed this solution resulted in 20% reduction in legal costs and +30% reduction in time to contract year-over-year for this client. Our client Procurement Department was able to move faster, increase product and service assortment, and boost revenue by 10% over the same period, resulting in greater customer satisfaction and overall company growth.
Prevent damages
Our solution helped in-house legal teams hone in on adverse contract clauses that prevented $7 Million in damages.
Reduce legal bottlenecks
The impact from our risk framework decreased legal fees by 20% and time to contract by 3 weeks on average.
Focus on top-line growth
Semi-automating contract review allows Procurement Teams to focus on high value add product and service assortment.
CASE STUDy
Clinical Workforce Optimization
We supported a large west coast hospital system reimagine workforce capacity to improve patient experience and allow caregivers to focus on higher-level responsibilities.
Staffing Coordinators are responsible for caregiver and practitioner work schedules at hospitals. The attention to detail required in their work is imperative to daily hospital operations. Our client had multiple legacy systems for creating schedules that added complexity and friction to the process. Lack of integration required Staffing Coordinators to update individual systems, fragmenting communication channels and resulting in miscommunication due to overlooked data input. Inadequate staffing, missed deadlines, preventable delays, and inconsistent levels of service began impacting client reputation. Patient satisfaction scores declined from 88% to 82% over a trailing 5yr period and the frequency of extended work shifts for Nurse Managers, Charge Nurses, and other unit caregivers (RNs, CNAs, etc.) increase from 12 occurrences/year to 16 occurrences/year over the same timeframe, resulting in higher caregiver turnover and several cases of burnout.
Fei Business Advisory supported this client by developing a scheduling platform to optimize workforce capacity. We developed our product strategy in partnership with Staffing Coordinators and front-line caregivers to align the organization on a shared vision of the solution. After rigorous market research and user testing to inform the overall experience and design, Fei Business Advisory launched a single-screen scheduling prototype in a little over 9mos. The solution integrated key data from each legacy system, while adding a predictive scheduling feature that allowed front-line caregivers to anticipate staffing shortages and mitigate with temp workers.
We built the predictive model by polling front-line caregivers to help us identify the variables that impact staffing. Drawing from our talent pool, we leveraged best in class data scientists to rank the these variables using a distributed gradient-boosted decision tree model (XGBoost) that served as the key data for our model. We optimized these data using an LLM (ChatGPT-4o) and programmed constraints around caregiver hand-offs, a common source of clinical errors, and labor costs.
After a 3mo testing phase, Fei Business Advisory had some preliminary results. The prototype predicted staffing shortages with 91% accuracy and we extrapolated that it would decrease extended work shifts from 16 occurrences/year to 14 occurrences/year at current labor costs. Most importantly, we observed patient satisfaction scores increase from 82% to 85% during the 3mo testing period driven by fewer patient handovers per hospital stay. We left our client with extensive documentation on the product for further integration and UI improvements that ultimately led to an extension in our contract and special recognition as a favored technical consultant.
Predict staff shortages
Our solution predicted staffing shortages with 91% accuracy in a forward-looking 3mo window, enabling our client to meet additional demand at lower rates.
Decrease long work shifts
The impact from better prediction helped out client reduce extended works shifts (over 12hrs) by 14% from 16 occurrences/year to 15 occurrences/year.
Increase patient satisfaction
Patient satisfaction scores increase from 82% to 85% during the project timeline driven by fewer patient handovers per hospital stay.