How MediRAG Reduces Patient Support Costs by 50%

A medium-sized Spanish hospital with 300 beds handles between 200 and 400 non-urgent telephone enquiries every day. Half of those calls are repetitive — appointment confirmations, preparation instructions, medication questions, visiting hours. MediRAG, part of the IgeraSolutions platform, trains on the hospital's own protocols and appointment data to answer these questions automatically, freeing nursing staff for clinical work and reducing operational costs by up to 50%.

Why are Spanish hospitals overwhelmed with repetitive enquiries?

A medium-sized Spanish hospital operates with limited administrative staff relative to patient volume. At peak hours — Monday mornings, after diagnostic result deliveries, before and after surgical procedures — the telephone lines saturate. Patients wait on hold for 10 to 20 minutes to ask questions that have a single, standardised answer written in the hospital's own procedure documentation.

The five most common non-urgent enquiry types at a typical 300-bed hospital account for approximately 50% of total call volume:

• Appointment confirmation and rescheduling requests (22% of calls)
• Pre-procedure preparation instructions — fasting, medication, bowel preparation (16% of calls)
• Questions about medication — whether to take regular medication before a procedure, post-discharge instructions (7% of calls)
• Department visiting hours and access policies (4% of calls)
• How to collect test results or access the patient portal (3% of calls)

Every one of these questions has a definitive, documented answer in the hospital's internal protocols. A nurse answering these calls is applying clinical knowledge to answer administrative questions — an expensive and inefficient use of their training. The operational cost per handled enquiry runs between €3 and €6 when salary, infrastructure, and management overhead are included.

What types of questions does MediRAG handle?

MediRAG is trained on the hospital's own protocol library, not on generic medical information. This means every answer is specific to that hospital's procedures, not a general health information chatbot. Representative examples from deployed installations:

• "What do I need to do before my colonoscopy?" — Returns the pre-procedure protocol with dietary restrictions, bowel preparation instructions, and medication guidance, citing the protocol version and approving clinician.
• "What are the visiting hours for the oncology ward?" — Returns the current visiting schedule with any restrictions that apply to the specific ward.
• "How do I cancel and reschedule my appointment?" — Guides the patient through the rescheduling process with direct links to the booking system for real-time slot availability.
• "Can I take my metformin on the morning of my procedure?" — Returns the anaesthesia pre-operative medication protocol and flags that specific medication management questions require confirmation from the patient's physician, then routes an alert to the relevant department.
• "How do I access my test results online?" — Provides step-by-step instructions for the hospital's patient portal with the relevant login procedure.

How does MediRAG handle questions that require clinical judgment?

MediRAG operates with a clear distinction between informational queries — those with a documented answer in hospital protocols — and clinical queries that require professional judgment. The system is designed to answer the former and escalate the latter.

When a query falls outside the protocol library or touches on individualised clinical assessment, MediRAG identifies this and routes an alert to the appropriate nursing station or department with the full conversation context. The nurse or physician who receives the escalation already has the patient's question in front of them and can respond immediately — without the patient having to repeat themselves.

This intelligent escalation is what distinguishes MediRAG from a simple FAQ bot. It handles the routine volume automatically and makes the human interventions that do occur more efficient, not less.

What are the GDPR and regulatory compliance considerations?

Healthcare is one of the most heavily regulated sectors for data processing in Spain. MediRAG is designed to operate within these constraints:

• No patient record access: MediRAG is trained exclusively on generic hospital procedures and publicly available protocol information. It does not connect to electronic health records (EHR) or access any individually identifiable patient data. This removes the system from the GDPR special category data processing requirements for health data.
• LOPDGDD compliance: The system is configured and hosted in compliance with Spain's Organic Law on Data Protection and Digital Rights Guarantees, including data minimisation, purpose limitation, and retention restrictions.
• EU AI Act classification: Because MediRAG provides only informational responses based on hospital protocols and does not make or assist in making medical diagnoses or treatment decisions, it does not fall into the EU AI Act's high-risk AI system category for healthcare. Patients are always clearly informed they are interacting with an automated system.
• Conversation data: Patient queries submitted via the system are retained only for the minimum period required for quality assurance and system improvement, with full anonymisation applied before any aggregate analysis.

What is the patient experience like?

Hospital patient experience research consistently shows that the two primary sources of patient dissatisfaction in the administrative journey are waiting times on the telephone and receiving inconsistent information from different staff members. MediRAG addresses both directly.

Responses are delivered in under 10 seconds via the patient's preferred channel — web, mobile app, WhatsApp, or SMS. The answer is always drawn from the same source document, so it is consistent regardless of when or how the patient asks. An 87% patient preference for receiving instant answers via chatbot over waiting on hold has been reported in patient satisfaction surveys at hospitals using the IgeraSolutions platform.

For hospitals with diverse patient populations, MediRAG supports queries in multiple languages. Spanish and Catalan are standard; English, Arabic, and other languages can be enabled for hospitals with specific community needs.

What is the financial case for a 300-bed hospital?

The ROI calculation for a medium-sized Spanish hospital is conservative and based on documented operational parameters:

• Direct cost reduction: €3.20 per automated enquiry × 250 enquiries/day × 250 operating days/year = €200,000 saved annually
• Nursing capacity recovered: 2 hours/day per nurse × 5 nurses on administrative rotation = 10 hours/day, equivalent to the output of 1.25 full-time administrative staff without additional hiring
• Patient satisfaction effect: A +15-point improvement in NPS generates an estimated 8–12% increase in patient referrals through word-of-mouth, measurable in revenue over 12–18 months
• Staff burnout reduction: Repetitive administrative call handling is a documented contributor to nursing burnout. Removing this workload has qualitative retention benefits that are difficult to quantify but real in terms of reduced recruitment costs

Frequently Asked Questions

Does MediRAG access the hospital's electronic health records?

No. MediRAG is trained exclusively on generic hospital procedures, department protocols, and publicly available patient information leaflets. It does not connect to EHR systems, patient administration systems, or any database containing individually identifiable patient information. This is a deliberate design decision that keeps the system outside the GDPR special category data processing requirements for health data.

How does the appointment booking integration work?

MediRAG connects to the hospital's appointment scheduling system via API to retrieve real-time slot availability. Patients can check available appointment times and initiate rescheduling requests directly through the chatbot interface. The system does not modify appointment records autonomously — any changes are confirmed through the standard booking workflow and logged in the scheduling system.

What channels does MediRAG support?

The system supports web chat (embeddable on the hospital's patient portal), mobile app integration, WhatsApp Business API, and SMS. Most Spanish patients in the 35–65 age group already use WhatsApp for healthcare communication. The hospital chooses which channels to activate based on its existing patient communication infrastructure.

How long does implementation take?

A standard deployment for a 300-bed hospital with four to six specialties takes six to eight weeks. The most time-intensive phase is the initial protocol audit and digitisation — many hospitals have their procedures documented in formats that need to be converted before indexing. Hospitals with a well-maintained digital protocol library can go live in four weeks.

How are patients told they are communicating with an AI system?

At the start of every conversation, MediRAG identifies itself as an automated patient information assistant. The introduction message is customisable by the hospital and is displayed prominently before the patient submits their first query. Patients can request to speak with a human staff member at any point in the conversation, and the system escalates immediately with full context.

What happens when the hospital updates a protocol?

The hospital's protocol management team uploads updated documents via the IgeraSolutions management portal. New versions are indexed within minutes and immediately replace the previous version in the response system. The system logs the document version used for every response, creating an auditable record that confirms patients received information based on the current approved protocol at the time of their enquiry.