# EULAR Study: Shorter Documentation Time in 82% of Consultations

Real-world study in EULAR Rheumatology Open (2025) evaluates Nixi AI in a German rheumatology clinic: 82% of consultations ran faster and patient acceptance was high.

- Author: PD Dr. Johannes Knitza, Consultant Rheumatologist · Philipps-University Marburg (MD, PhD)
- Published: 2025-12-01
- Updated: 2025-12-01
- Read online: https://www.nixiai.ai/en/insights/eular-study-nixi-ai-rheumatology-real-world-evidence

A new study published in **EULAR Rheumatology Open (2025)** provides the first
real-world evaluation of **Nixi AI** in a clinical rheumatology setting,
offering compelling evidence that ambient AI scribes can significantly reduce
administrative burdens without compromising patient trust.

> **Time reduction:** The rheumatologist reported that the AI scribe reduced consultation time in 
>   **82% of cases (72 of 88)**.

> **High acceptance:** Most patients indicated they would welcome the use of the scribe in future
>   visits.

## Real-World Results: Time Savings and Accuracy

Conducted at a large German rheumatology outpatient service in August 2025,
the study evaluated the use of Nixi AI during routine consultations. The
results were reported by the treating rheumatologists and verified through
patient questionnaires.

The impact on clinical workflow was immediate and positive:

- **Time reduction:** The rheumatologist reported that the AI scribe reduced consultation time in **82% of cases (72 of 88)**.
- **Neutral impact:** In 15% of cases, there was no change in time usage.
- **Minimal disruption:** Increased time was reported in only 3% of cases.

Regarding accuracy, the study noted that while errors occurred in 19% of
encounters, they were exclusively classified as "minor". These errors were
primarily misspelled medication names or incorrect sex detection, the latter
being a specific automated classification feature rather than a clinical
documentation failure.

**Rheumatologist feedback — perceived time impact**

| Category | Percentage |
| --- | --- |
| Saved >5 minutes | 11.4% |
| Saved 2–5 minutes | 70.5% |
| Neutral | 14.8% |
| Added 2–5 minutes | 2.3% |
| Added >5 minutes | 1.1% |

*Source: Knitza & Aries, EULAR Rheumatology Open (2025). n = 88 consultations.*

## Patients Prefer the "Heads-Up" Approach

One of the most significant findings of the study was patient acceptance.
There is often a fear that AI might make medicine feel robotic, but this study
suggests the opposite.

Questionnaires completed by 108 patients revealed:

- **More engagement:** Patients perceived that their physician spent **less time looking at the screen** and **more time engaging with them** directly.
- **High acceptance:** Most patients indicated they would welcome the use of the scribe in future visits.
- **Positive attitudes:** 56% of patients held a positive attitude toward AI in healthcare, with another 14% being "very positive".

## The Future of Rheumatology Documentation

The authors, Johannes Knitza and Peer Aries, concluded that these preliminary
findings align with broader data showing meaningful time savings and high
acceptance rates for ambient AI. As the first study of its kind in this
specialty, it highlights how tools like Nixi AI can help clinicians return to
the core of their profession: interacting with patients rather than keyboards.