Clinical and theoretical context
In the CBT model used in this literature, threatening interpretations of bodily sensations, repeated checking, avoidance and reassurance-seeking can sustain distress. The purpose is to change the relationship with uncertainty and related behaviours, not to prove that illness is impossible. That distinction is especially important for a conversational product: repeated reassurance may be immediately satisfying while working against the treatment model. This is a design concern inferred from the model, not a directly measured harm of Ella. Appropriate medical assessment and treatment of physical illness remain distinct from psychological work on persistent health anxiety.
Sources: Axelsson & Hedman-Lagerlöf, 2019; Axelsson et al., 2020
How strong is the comparative evidence?
The 2019 meta-analysis found benefit for CBT against varied non-CBT controls; its estimate became smaller after adjustment for publication bias. Recruitment and comparison conditions also moderated results (Axelsson & Hedman-Lagerlöf, 2019). The important synthesis is therefore support for CBT within studied contexts, rather than a universal recovery percentage. A May 2026 network meta-analysis broadened the evidence and found CBT and several related approaches superior to waitlist. Reported component associations included exposure/response prevention and cognitive restructuring (Lai et al., 2026). Indirect rankings and waitlist comparisons cannot establish that one approach is best against all credible alternatives, or that an isolated feature reproduces a complete intervention.
What a digital trial actually tested
Axelsson et al. (2020) compared a therapist-guided internet programme with protocol-matched face-to-face CBT and met the prespecified noninferiority boundary at twelve weeks. Human support was part of the online condition. This supports a defined guided format, not equivalence between professional treatment and a chatbot. The sample was relatively highly educated, and important clinical risks were excluded. Later assessments were useful, but the study was not adequately powered to establish long-term noninferiority. “Comparable in this trial” is more accurate than “online therapy is always just as effective.”
Sources: Axelsson et al., 2020
Durability and remaining uncertainty
The CHAMP follow-up provides unusual evidence extending to eight years, with a small average health-anxiety advantage for adapted CBT over standard care. Long follow-up is valuable, but substantial missingness and deaths complicate interpretation. The cohort included some participants under eighteen, so it is not wholly adult evidence (Tyrer et al., 2021a). A published corrigendum replaces a subgroup table; the original clinic- and age-specific estimates should not be reused (Tyrer et al., 2021b). These publications represent one cohort, not two replications. Neither durability nor average benefit establishes that every participant improved or that further medical care became unnecessary.
Sources: Tyrer et al., 2021a; Tyrer et al., 2021b
Responsible implications and research needs
A proposed tool could help people recognize patterns of checking and prepare a balanced conversation with a healthcare professional. It should not label a new symptom as anxiety or promise that someone is medically safe. Those boundaries follow from the difference between health-anxiety treatment and diagnosis. Product studies should assess health anxiety, functioning, checking behaviour, appropriate medical-care use and deterioration; fewer consultations alone would not establish benefit. They should test the actual reassurance and escalation behaviours of the system over time. None of these sources directly evaluates Umbrella Journal, Ella or Nuggets.
Sources: Axelsson et al., 2020; Lai et al., 2026; Tyrer et al., 2021a
References
Links open the original papers or official sources. Access labels describe the material assessed for this review.
Axelsson, E., & Hedman-Lagerlöf, E. (2019). Cognitive behavior therapy for health anxiety: systematic review and meta-analysis of clinical efficacy and health economic outcomes. Expert Review of Pharmacoeconomics & Outcomes Research, 19(6), 663–676. 10.1080/14737167.2019.1703182
Access assessed: Indexed complete PubMed abstract and metadata; full article not appraised.
Axelsson, E., Andersson, E., Ljótsson, B., Björkander, D., Hedman-Lagerlöf, M., & Hedman-Lagerlöf, E. (2020). Effect of Internet vs Face-to-Face Cognitive Behavior Therapy for Health Anxiety: A Randomized Noninferiority Clinical Trial. JAMA Psychiatry, 77(9), 915–924. 10.1001/jamapsychiatry.2020.0940
Access assessed: Publisher full text.
Lai, L., Liu, Y., Axelsson, E., Tyrer, P., Li, Y., Lu, Y., Shi, C., & Ren, Z. (2026). The comparative effectiveness of psychological interventions for health anxiety: systematic review and a network meta-analysis of randomised controlled trials. The British Journal of Psychiatry, FirstView, 1–10. 10.1192/bjp.2026.10647
Access assessed: Publisher abstract, PubMed abstract and metadata; full article not assessed.
Tyrer, P., Wang, D., Crawford, M., Dupont, S., Cooper, S., Nourmand, S., Lazarevic, V., Philip, A., & Tyrer, H. (2021a). Sustained benefit of cognitive behaviour therapy for health anxiety in medical patients (CHAMP) over 8 years: a randomised-controlled trial. Psychological Medicine, 51(10), 1714–1722. 10.1017/S003329172000046X
Access assessed: Indexed complete PubMed abstract and publisher bibliographic record.
Tyrer, P., Wang, D., Crawford, M., Dupont, S., Cooper, S., Nourmand, S., Lazarevic, V., Philip, A., & Tyrer, H. (2021b). Sustained benefit of cognitive behaviour therapy for health anxiety in medical patients (CHAMP) over eight years: randomised controlled trial – CORRIGENDUM. Psychological Medicine, 51(10), 1723. 10.1017/S0033291720001932
Access assessed: Publisher correction text indexed in search; bibliographic record verified.
This evidence review is educational. It does not provide an individual diagnosis or treatment plan. Read the review methods, access limits, and clinical-review status.