Randomised controlled trialindependent cross-checkfull text, structured risk-of-bias assessment★ key study I-ORG-05 in-depth review
UK trial: tonsil removal in adults with repeated tonsillitis
Conservative management versus tonsillectomy in adults with recurrent acute tonsillitis in the UK (NATTINA): a multicentre, open-label, randomised controlled trial.
Overview
Wilson JA et al.. Lancet 2023 · PMID 37209706 · DOI 10.1016/S0140-6736(23)00519-6
Surgery roughly halved sore-throat days over 2 years, but 19% had bleeding complications.
Does the study support the specific claim? No · Methodological quality: high
Plain-language summary
Study explained simply
- What did the researchers want to find out?
Adult tonsillitis
- How was the study done?
RCT
- How many people, animals or samples were studied?
453
- What was actually found?
Fewer sore-throat days
- How large was the observed effect?
IRR 0.53
- What do the results mean in everyday life?
For frequent tonsillitis, surgery helps despite risks
- What are the study's limitations?
Open label; self-reported outcome; many switched groups; missing weekly data
- What does it say about Elawady's specific claim?
Contradicts 'should not be removed'
- Do the results apply to Germany and Europe?
European
- What can explicitly not be concluded from this study?
Very long-term risks
Scientific in-depth review
1. Full reference
- Authors
- Wilson JA et al.
- Journal
- Lancet
- Year
- 2023
- Identifier
- PMID 37209706 DOI 10.1016/S0140-6736(23)00519-6
- Verification
- verified – PubMed efetch 2026-10-25
- Retracted
- no
- Corrections
- none
- Extent checked
- full text
- Conflicts of interest and funding
NIHR Health Technology Assessment programme (12/146/06); funder had no role. Author disclosures: research grants (NIHR, Cancer Research UK etc.), travel support (ENT Scotland), leadership role in ENT UK; no industry conflicts reported.
2. Research question
Tonsillectomy vs conservative management in adults with recurrent tonsillitis
3. Study design and population
- Study design
- Randomised controlled trial – Pragmatic multicentre open-label randomised controlled trial (NATTINA; ISRCTN55284102), 27 UK hospitals, 1:1 permuted blocks stratified by centre and baseline TOI-14 severity, with economic evaluation
- Population
Adults >=16 years newly referred to secondary-care ENT with recurrent acute tonsillitis meeting UK (SIGN) criteria (>=7 episodes in 1 year, >=5/year for 2 years, or >=3/year for 3 years); median age 23, 78% female, 90% White
- Country
UK
- Period
Recruitment May 11, 2015 - April 30, 2018; 24-month follow-up
- Sample size
4165 assessed, 453 randomised (233 tonsillectomy, 220 conservative); 429 (95%) in primary ITT analysis (224 vs 205)
- Inclusion and exclusion criteria
UK tonsillectomy guideline criteria; full criteria in appendix (not in file)
4. Methods and endpoints
- Exposure / intervention
Immediate elective tonsillectomy within 8 weeks (cold dissection 53%, bipolar 39%)
- Comparator
Conservative management for 24 months (self-administered analgesia, ad hoc primary-care antibiotics or ED attendance)
- Endpoints
Primary: number of sore throat days over 24 months (weekly text message, post-operative days included). Secondary: TOI-14, SF-12, adverse events, costs, QALYs
- Statistical methods
Mixed-effect negative binomial regression (centre random effect, baseline severity fixed) with return rate as exposure; ITT; planned per-protocol and per-treatment and two unplanned sensitivity analyses; cost-utility with multiple imputation and bootstrapping
5. Actual results
- Results
Sore throat days: median 23 (IQR 11-46) vs 30 (14-65); IRR 0.53 (95% CI 0.43-0.65; p<0.0001). Per protocol IRR 0.42 (0.31-0.55); per treatment 0.73 (0.59-0.90); missing weeks = no sore throat 0.62 (0.51-0.75). TOI-14 at 12 months 4.3 vs 21.7 and at 24 months 4.7 vs 15.4 (p<0.0001); SF-12 better with tonsillectomy. Cost difference GBP 488 (349-626), QALY gain 0.12 (0.09-0.14), ICER GBP 4136/QALY. Adverse events related to surgery: 191 in 90/231 (39%); bleeding 44 (19%), serious bleeding 34 (15%), readmission 47 (20%; 37 for bleeding), infection 6 (3%); no return to theatre; no deaths. 157 (35%) did not receive allocated intervention (74 conservative participants requested a switch to surgery).
- Absolute risks
Median sore throat days over 2 years 23 vs 30; post-operative bleeding 19% (serious 15%), readmission 20%; average 14 days of post-operative pain expected.
- Authors' conclusion
Immediate tonsillectomy clinically and cost-effective in adults with recurrent acute tonsillitis.
6. Strengths
- Largest adult tonsillectomy RCT; pragmatic, multicentre
- Central randomisation with concealed allocation
- Weekly prospective symptom capture over 24 months, including post-operative sore throat
- Pre-planned and unplanned sensitivity analyses consistent
- Economic evaluation
7. Weaknesses
- Open label; self-reported primary outcome without swab confirmation
- High crossover/non-adherence (35% did not receive allocated intervention)
- Missing weekly data, more in conservative group (80% returns: 62% vs 53%) (authors)
- Secondary outcome completion only 44-53% at 12-24 months
- Sample size reduced from 510 to 444 (power 85%)
- Patients who declined had worse symptoms; may limit generalisability (authors)
- TOI-14 English version not formally validated (authors)
8. Risk of bias
- Orienting tool
- RoB 2
- Type of assessment
- structured assessment based on the full text by a single reviewer (no formal double review)
- Full formal assessment
- no
- Overall risk of bias
- some concerns
- Rationale
Structured single-reviewer judgement from full text; a formal assessment would require two independent reviewers and the protocol.
| Domain | Judgement | Rationale |
|---|---|---|
| D1 randomisation process | low | Central web-based randomisation with independent statistician schedule; balanced baseline. |
| D2 deviations from intended interventions | some concerns | Open label; 35% non-adherence including 74 conservative participants switching to surgery; ITT appropriate for effect of assignment and likely biased towards null. |
| D3 missing outcome data | some concerns | 5% without any primary data; incomplete weekly returns, differential between arms; sensitivity analyses consistent. |
| D4 measurement of the outcome | some concerns | Self-reported sore throat days by unblinded participants; expectation effects possible. |
| D5 selection of the reported result | low | Registered protocol published; primary outcome as prespecified. |
Full-text review (25 October 2026): Added eligibility criteria, patient flow and 35% non-adherence, per-protocol/per-treatment IRRs (0.42/0.73), TOI-14 differences, cost-effectiveness (GBP 4136/QALY), detailed adverse events (serious bleeding 15%, readmission 20%), funding/COI; RoB 2 domains added, overall unchanged (some_concerns).
9. Assessment of the statistical interpretation
About half the sore throat days with tonsillectomy over 2 years (absolute difference in medians 7 days), robust across sensitivity analyses; benefit must be weighed against frequent post-operative bleeding (19%) and readmission (20%). Applies to adults meeting guideline frequency criteria, not to occasional sore throats.
10. Relevance to the specific claim
Contradicts blanket rejection of tonsillectomy (C-ORG-01); also quantifies harms (bleeding).
11. Independent contrary findings
12. Transferability to Germany and Europe
Directly European.
13. Overall methodological judgement
Strong randomised evidence that tonsillectomy benefits adults with frequent recurrent tonsillitis, with substantial but non-fatal harms; conservative management remains an option chosen by informed patients.
| Multidimensional assessment | |
|---|---|
| Authenticity of the source | verified |
| Methodological quality | high |
| Risk of bias | some concerns |
| Precision of results | high |
| Directness | direct |
| Reproducibility | partly replicated |
| Consistency with the overall evidence | consistent |
| Regional transferability | high |
| Adequacy of the interpretation (by the archive) | not applicable |
14. Open questions
Very long-term outcomes (beyond 2 years), including immune or respiratory effects; outcomes in patients who declined (more severe).
15. Linked fact checks
- Adenoids, tonsils and the gallbladder are organs that should not be removed; correcting the diet should be the foundation of treatment. – Studies contradicting or limiting the claim
- No working organ should be removed (e.g., a thyroid with nodules that still functions); one should stay away from the scalpel, and Elawady is 'against any operation, including caesarean section'. – Studies contradicting or limiting the claim