Edem Anaglo

Automating Clinical Billing in Healthcare

As a cornerstone of Elephant Healthcare’s financial health, clinical billing profoundly influences the sustainability of our primary care network. This case study will show how we resolved two chronic pain points in the existing system: substantial unbilled care resulting in revenue loss, and the excessive workload placed upon our billing officers.

TOOLS

Figma, Slab, Miro

ROLE

Design Lead

Discovery

We researched billing challenges at our facilities in Ghana and Kenya. Our goal was to confirm reported problems and find the exact needs of billing officers. We focused on three key areas:

The officer's billing workflow.
The patient's billing experience.
How price lists are set and managed.

Research Methodology

A contextual inquiry helped us to combine direct observation with interviews to intimately understand clinician workflow needs, behaviours and pain points.

We researched billing challenges at our facilities in Ghana and Kenya. Our goal was to confirm reported problems and find the exact needs of billing officers. We documented three key areas: the officer's billing workflow, the patient's billing experience, and how price lists are set and managed.

Research Methodology

We used contextual inquiry for deep learning, combining direct observation with interviews to understand clinician workflow needs, and pain points.

What we found

5 – 7% of patients cannot afford full services, requiring officers to seek approvals for discounts.

Manual Concessions

Manual Concessions

5 – 7% of patients cannot afford full services, requiring officers to seek approvals for discounts.

Officer Risk

Officers often pay small patient deficits out of pocket because the system offers no alternative.

Officer Risk

Heavy Workload

Billing officers manually track all care events and managing collections on every invoice.

“Yesterday, I saw three patients who were unable to fully pay for their medications. It would be nice if we had a way of making some exemptions for our patients”

Billing officer

Methodist Hospital, Aburaso

Mapping out the current user journey

We followed this by mapping the patient and billing officer journeys. This gave the team a shared, high-level understanding of the interactions between a patient, the clinicians and the billing officer, revealing key pain points and opportunities for improvement.

Working backwards with hard questions - PR/FAQ

To align the team and define final requirements, we used a PRFAQ framework. The PR framed the feature's hypothetical launch to secure stakeholder alignment, and the FAQ sessions that followed focused on customer questions and stakeholder concerns.

This process ensured the entire team and stakeholders were clearly aligned on the feature's scope, potential limitations and any other constraints.

Taking a step back and rescoping

Taking a step back to rescope

The outcome of our PR/FAQ suggested that automation by triggering real-time care events presented shortcomings we had not considered. Two main focal points needed to be addressed.

Complete automation was not feasible at that stage. For efficiency, the billing officer must still intervene manually in the following scenarios:

Complete automation was not feasible at that stage. For efficiency, the billing officer must still intervene manually in the following scenarios:

Complete automation was not feasible at that stage. For efficiency, the billing officer must still intervene manually in the following scenarios:

Updating the patient's prescription based on its availability.

Updating the patient's prescription based on its availability.

Updating the patient's prescription based on its availability.

Assigning insurers based on the patient's health insurance coverage.

Assigning insurers based on the patient's health insurance coverage.

Assigning insurers based on the patient's health insurance coverage.

Accounting for clerical errors made by clinicians.

Accounting for clerical errors made by clinicians.

Accounting for clerical errors made by clinicians.

Switching between price types based on the facility's price list.

Switching between price types based on the facility's price list.

Switching between price types based on the facility's price list.

Service and drug prices vary based on the patient's insurance (national, private, or none). The officer must manually select and apply the correct price list.

Service and drug prices vary based on the patient's insurance (national, private, or none). The officer must manually select and apply the correct price list.

Service and drug prices vary based on the patient's insurance (national, private, or none). The officer must manually select and apply the correct price list.

Taking a step back to rescope

The outcome of our PR/FAQ suggested that automation by triggering real-time care events presented shortcomings we had not considered. Two main focal points needed to be addressed.

Complete automation was not feasible at that stage. For efficiency, the billing officer must still intervene manually in the following scenarios:

Updating the patient's prescription based on its availability.

Assigning insurers based on the patient's health insurance coverage.

Switching between price types based on the facility's price list.

Some service and drug prices varied based on the patient's insurance (national, private, or none). The officer must manually select and apply the correct price list.

Prototyping

We defined the solution by focusing on core user stories and high-impact features. Using our in-house component library, I skipped low-fidelity work and moved straight to high-fidelity mockups. All designs were optimised for our core device, the Tigi Tablet.

Automation flow

This flow depicts how an automated invoice behaves when triggered by real-time care events during a patient's visit.

Manually creating an invoice if a care event is not triggered

In the event a real-time care event is not triggered, the billing officer has the option of manually creating an invoice and populate it based on the care received by the patient at the facility.

Updating an invoice

Based on insights from our PRFAQ, we accounted for each scenario where the billing office updates the invoice by changing its attributes and finalizing it before collection begins.

Taking payment

This flow depicts how the billing officer takes payment once the invoice is finalised.

Exempting Payment

Billing officers are allowed to make payment exemptions based on the type of visiting patient.

Waiving Payment

The billing officer is afforded the option to waive a given amount off the transaction amount when payment is being taken.

Refund

In the event a patient is over-billed, this feature severs as a contingency to rectify the issue.

Creating transparency via an audit log

A Payment History was essential to resolve the lack of transparency by tracking all payments, timestamps, cashiers, and amounts for a given invoice.

Validation

We stress-tested the solution under real-world scenarios (e.g., pricing, processing time, insurance coverage) by conducting a soft launch at several facilities.

Users praised the Payment History for easy audits and appreciated the integration of Medical History with patient invoices

The new dispensing process did not fully solve pilferage. Moreover, users also reported confusion over specific financial terms.

We stress-tested the solution under real-world scenarios (e.g., pricing, processing time, insurance coverage) by conducting pilots at selected facilities.

Users praised the Payment History for easy audits and appreciated the integration of Medical History with patient invoices

The new dispensing process did not fully solve pilferage. Moreover, users also reported confusion over specific financial terms.

Outcomes

We updated the final prototype based on research and soft launch feedback. While it was too early to measure overall success due to varied clinic processes, initial analytics (first two months post-launch) showed positive outcomes.

We discovered that:

Average waiting time

-12 mins

At our dispensaries, this was substantially reduced reduced.

Finalized invoices

+48%

Across our territories this was almost doubled.

Revenue increase

+24%

We recorded an increase in revenue from collections alone.

We discovered that:

Average waiting time