About this benchmarkRheumatoid arthritis · USRheumatologists · USAI answers generated with a US rheumatologist persona4 AI systemsPre-ACR baseline, 30 October 2026
03 · Patient Profiles
Does AI favour the same therapies for every patient?
The same therapy. Different patient contexts. Different AI positioning.
Upadacitinib
Not a controlled comparison
The two profiles contain different questions and different comparators. This is not a controlled comparison, and it says nothing about clinical value or superiority.
After TNF-inhibitor failure
120 comparative answers · 30 questions in this profile
Favoured in 16 of 32
- Favoured16 · 50%
- Equal11
- Conditional4
- Other1
- Comparator favoured0
Cardiovascular & thromboembolic risk
68 comparative answers · 17 questions in this profile
Favoured in 0 of 24
- Favoured0 · 0%
- Equal5
- Conditional8
- Other4
- Comparator favoured7
Profile explorer
Choose a patient profile.
MTX intolerance / monotherapy
Clinical context
Comparative questions framed around patients who cannot take methotrexate, mostly about monotherapy.
MTX intolerance / monotherapy: tocilizumab is favoured in 18 of the 24 answers in which it was compared.
Therapies assessed
TocilizumabFavoured in 18 of 24
- Favoured18 · 75%
- Equal1
- Conditional2
- Other3
- Comparator favoured0
AbataceptFavoured in 1 of 12
- Favoured1 · 8%
- Equal2
- Conditional1
- Other2
- Comparator favoured6
AdalimumabFavoured in 1 of 12
- Favoured1 · 8%
- Equal2
- Conditional2
- Other0
- Comparator favoured7
EtanerceptFavoured in 2 of 8Small base
- Favoured2 · 25%
- Equal0
- Conditional1
- Other1
- Comparator favoured4
SarilumabFavoured in 3 of 8Small base
- Favoured3 · 38%
- Equal1
- Conditional1
- Other2
- Comparator favoured1
After TNF-inhibitor failure
Clinical context
Comparative questions framed around patients whose disease responded inadequately to a TNF inhibitor.
After TNF-inhibitor failure: upadacitinib is favoured in 16 of 32 comparisons and rituximab in 14 of 48.
Therapies assessed
RituximabFavoured in 14 of 48
- Favoured14 · 29%
- Equal11
- Conditional14
- Other9
- Comparator favoured0
UpadacitinibFavoured in 16 of 32
- Favoured16 · 50%
- Equal11
- Conditional4
- Other1
- Comparator favoured0
BaricitinibFavoured in 7 of 32
- Favoured7 · 22%
- Equal11
- Conditional3
- Other4
- Comparator favoured7
TofacitinibFavoured in 0 of 32
- Favoured0 · 0%
- Equal12
- Conditional2
- Other6
- Comparator favoured12
AbataceptFavoured in 3 of 28
- Favoured3 · 11%
- Equal12
- Conditional0
- Other3
- Comparator favoured10
SarilumabFavoured in 0 of 24
- Favoured0 · 0%
- Equal10
- Conditional3
- Other4
- Comparator favoured7
TocilizumabFavoured in 6 of 24
- Favoured6 · 25%
- Equal9
- Conditional1
- Other7
- Comparator favoured1
Pregnancy planning
Clinical context
Comparative questions framed around patients planning pregnancy.
Pregnancy planning: certolizumab pegol is favoured in 5 of 8 comparisons.
Therapies assessed
Certolizumab pegolFavoured in 5 of 8Small base
- Favoured5 · 63%
- Equal0
- Conditional2
- Other1
- Comparator favoured0
AdalimumabFavoured in 0 of 8Small base
- Favoured0 · 0%
- Equal1
- Conditional0
- Other3
- Comparator favoured4
EtanerceptFavoured in 1 of 8Small base
- Favoured1 · 13%
- Equal1
- Conditional2
- Other2
- Comparator favoured2
HydroxychloroquineFavoured in 2 of 8Small base
- Favoured2 · 25%
- Equal2
- Conditional3
- Other1
- Comparator favoured0
Cardiovascular & thromboembolic risk
Clinical context
Comparative questions framed around patients with cardiovascular or thromboembolic risk.
Cardiovascular and thromboembolic risk: no JAK inhibitor is favoured in any comparison; tocilizumab is favoured in 8 of 16, abatacept in 6 of 12 and sarilumab in 6 of 12.
Therapies assessed
TocilizumabFavoured in 8 of 16
- Favoured8 · 50%
- Equal4
- Conditional3
- Other1
- Comparator favoured0
AbataceptFavoured in 6 of 12
- Favoured6 · 50%
- Equal0
- Conditional5
- Other1
- Comparator favoured0
SarilumabFavoured in 6 of 12
- Favoured6 · 50%
- Equal1
- Conditional5
- Other0
- Comparator favoured0
UpadacitinibFavoured in 0 of 24
- Favoured0 · 0%
- Equal5
- Conditional8
- Other4
- Comparator favoured7
BaricitinibFavoured in 0 of 20
- Favoured0 · 0%
- Equal0
- Conditional8
- Other0
- Comparator favoured12
TofacitinibFavoured in 0 of 20
- Favoured0 · 0%
- Equal0
- Conditional6
- Other1
- Comparator favoured13
EtanerceptFavoured in 7 of 16
- Favoured7 · 44%
- Equal4
- Conditional4
- Other1
- Comparator favoured0
AdalimumabFavoured in 4 of 12
- Favoured4 · 33%
- Equal4
- Conditional4
- Other0
- Comparator favoured0
Across profiles
Follow a therapy across the four profiles.
Not a controlled comparison
Each profile has its own questions and comparators. Read a therapy across profiles as a description of AI answers in different contexts, never as a controlled comparison or a ranking.
| Therapy | MTX intolerance / monotherapy | After TNF-inhibitor failure | Pregnancy planning | Cardiovascular & thromboembolic risk |
|---|---|---|---|---|
| Abatacept | Favoured in 1 of 12 | Favoured in 3 of 28 | Not shown | Favoured in 6 of 12 |
| Tocilizumab | Favoured in 18 of 24 | Favoured in 6 of 24 | Not shown | Favoured in 8 of 16 |
| Tofacitinib | Not shown | Favoured in 0 of 32 | Not shown | Favoured in 0 of 20 |
| Adalimumab | Favoured in 1 of 12 | Not shown | Favoured in 0 of 8Small base | Favoured in 4 of 12 |
| Upadacitinib | Not shown | Favoured in 16 of 32 | Not shown | Favoured in 0 of 24 |
| Baricitinib | Not shown | Favoured in 7 of 32 | Not shown | Favoured in 0 of 20 |
| Etanercept | Favoured in 2 of 8Small base | Not shown | Favoured in 1 of 8Small base | Favoured in 7 of 16 |
| Sarilumab | Favoured in 3 of 8Small base | Favoured in 0 of 24 | Not shown | Favoured in 6 of 12 |
“Not shown” means the therapy was not compared in that profile, or was compared fewer than 8 times there.
Explore how AI represents your therapy across your priority patient profiles.
Schedule a meetingWhat this means
Why profile intelligence matters to brand, medical and market-access teams
Positioning, evidence communication and access arguments are built for specific patients. In this benchmark, the therapies AI answers favour differ between the four profiles shown, so a single market-wide reading can hide where a therapy is favoured and where it is not.
What limits this reading
Four profiles are shown. Each has its own questions, comparators and bases, and profiles overlap, so no figure is added across them and none is a controlled comparison.
The Theralexis capability
A Theralexis assessment measures the patient profiles that matter to your portfolio, with the comparators and questions you choose.
Your portfolio
Discuss your competitive AI representation
See how AI positions your therapy for the patient profiles your teams prioritise.
About this benchmark
What was measured
AI representation analysis of 1,816 answers to US rheumatologist-style open questions and 960 answers to comparative questions from four web-enabled AI systems (ChatGPT, Claude, Gemini, Perplexity), collected 30 October 2026 with a US rheumatologist persona. Visibility = share of answers naming a therapy. Neither visibility, order of mention nor head-to-head verdicts indicate clinical efficacy, safety, preference or recommendation.
Pre-ACR baseline (v1.1 corrected analytical view, 30 Oct 2026). Re-measured after ACR Convergence 2026.
Sources and scope
- Publisher
- Theralexis
- Market
- Rheumatoid arthritis · United States
- Persona
- AI answers generated with a US rheumatologist persona
- Open questions
- 454 questions · 1,816 answers
- Comparative questions
- 240 questions · 960 answers
- Measured
- Published
- AI systems
- ChatGPT (GPT-5.4-mini), Claude (Claude Haiku 4.5), Gemini (Gemini 3.1 Flash-Lite), Perplexity (Sonar)
- Head-to-head judge
- gpt-4.1-mini