Does the Vacuum Bell Actually Work? What the Evidence Shows

Yes, and the odds depend almost entirely on two things you can identify before you start: age and consistency. Younger cartilage is more pliable, so adolescents typically respond faster than adults, and sustained daily use over months is the dominant factor in the result. In a Swiss cohort at University Hospital Basel, about 80% of 140 patients showed significant improvement, and Scaife and colleagues (2025, n=240) reported a successful outcome in 66%.

Key points

• Pooled across every patient regardless of age or consistency, Scaife and colleagues (2025, n=240) reported a successful outcome in 66%, and Haecker and colleagues (n=133) reported comparable results.
• Two things separate responders from non-responders: starting while the cartilage is still pliable, and using the device consistently over months rather than weeks.
• Cincinnati Children's Hospital reports vacuum bell therapy may remove the need for surgery in up to 35% of patients who use it, with the strongest results in children aged 6 to 12 with a mild deformity.
• The immediate lift is real and not disputed: sternal elevation under the device has been confirmed thoracoscopically during surgery (Haecker 2016).
• The main failure mode is not the device. Zhou and colleagues (2024, n=65) found 60% of children paused treatment for more than two weeks at least once, a lapse in the consistency the outcome data rewards.
• After 20 years of clinical use, Haecker and colleagues (2024) report that no validated international guidelines for the therapy exist.

The short answer, expanded

Vacuum bell therapy uses a silicone suction cup and a hand pump to create negative pressure over the sunken part of the chest. The sternum and the attached rib cartilage lift while the device is applied. That part is not controversial and has been directly observed: during minimally invasive surgical repair, surgeons have watched the sternum rise under the device through a thoracoscope (Haecker 2016).

The real question is whether repeated lifting produces a change that persists once the device comes off. Drawn from about 20 years of clinical use, the honest answer is that it does for some people and does not for others, and the published data now identifies fairly clearly which group a given patient is likely to fall into.

This is not a treatment with a single success rate. It is a treatment with a strongly age-dependent and adherence-dependent success rate. Reporting one number without those two variables attached would mislead, which is part of why the figures quoted around this therapy vary so widely.

What the evidence actually shows

The largest recent outcome cohort

The larger published cohorts are consistent with one another. Graded by improvement in the chest wall, the reported rates were as follows:

These figures pool every patient in each cohort together. They include late starters and patients who stopped inside two years, both of which work against the result. The pooled number is therefore a floor, not an estimate of what a well-selected, consistent patient should expect, and these cohorts do not report a separate response rate for that subgroup.

The more useful question is what separates the two groups. Two things dominate. Younger cartilage is more pliable, so adolescents typically respond faster and more completely than adults. And consistency of daily use over months, rather than weeks, is the single largest factor a patient actually controls.

This is a single-centre retrospective study, which limits how far the numbers generalise. The direction and size of both effects are consistent with the underlying biology: costal cartilage is more pliable before it begins to ossify, and a treatment that works by sustained gentle deformation depends on both pliability and time.

The youngest cohort

Luo and colleagues (2022, n=139) studied preschool children (mean age 4.6 years) in Frontiers in Pediatrics, tracking chest wall depth with a three-dimensional scanner. Of that group, 30.9% (43 children) reached the termination criterion of a residual depth under 3 mm with a cosmetic result. Two factors independently predicted complete correction: initial depth, where shallower did better (odds ratio 0.69), and the length of the treatment period (odds ratio 1.58). This was short-term follow-up, so it says nothing about whether those corrections held over years.

What families report

Zhou and colleagues (2024, n=65) surveyed caregivers. Pooling responses regardless of age, the caregiver-rated effect was excellent in 12.3%, good in 46.2% and moderate in 41.5%. The survey drew an 84% response rate (Zhou et al. 2024). These are subjective parental assessments, not imaging, and parents who have invested two years are not neutral observers. Read the figure as a satisfaction measure rather than an efficacy measure.

The most useful figure in that paper is not an outcome at all. Zhou and colleagues also report that thirty-nine of the 65 children, or 60%, paused treatment for more than two weeks at least once. Set against the finding that more than 24 consecutive months of use quadrupled the odds of response, the shape of the problem becomes clear. The same survey linked petechiae, the small red pinpoint spots that appear when capillaries rupture under suction, with reduced treatment effectiveness (p=0.046).

The institutional position

Cincinnati Children's Hospital, which offers the therapy and has no product to sell, states that it may eliminate the need for surgery in up to 35% of patients who use it, with results reported as very positive in children aged 6 to 12 with a mild defect. The device is typically worn for a set period each day over one to two years.

Two further cohorts

A Swiss cohort in Basel reported significant improvement in approximately 80% of treated patients across a range of ages. Scaife and colleagues (2025, n=240) report a headline figure of 66%; the endpoint that figure measures is not reproduced here.

The pioneer cohort and the long view

The modern device was developed by an engineer who had pectus excavatum himself and who corrected his own chest over roughly two and a half years of daily use. His subsequent cohort of approximately 450 patients, aged 2 to 61, established the protocol still in general use: a minimum of 30 minutes twice daily during an initial period, extending to several hours daily thereafter (Haecker 2016). The device design received FDA clearance in May 2012 and carries CE certification. Reviewing 20 years of clinical application, Haecker and colleagues (2024) concluded that the therapy is safe and a potential alternative to surgical repair in carefully selected patients, and that in children under 10 it appears to represent a reasonable first step. They also noted plainly that no validated guidelines for the therapy exist.

Who this works best for

The published predictors are consistent enough to state directly.

Strong candidates

• Children and younger adolescents. Cartilage that has not begun to ossify responds to sustained deformation, which is why this group typically responds faster than adults.
• Shallow to moderate depth. Initial depth independently predicted complete correction in the preschool cohort (Luo et al. 2022).
• A flexible chest wall. Where the sternum visibly lifts under the device, the mechanical precondition is satisfied.
• Realistic about duration. The people who succeed are measured in years, not months.
• Consistent. Sustained daily use over months is the dominant factor in the outcome. Consistency here is a dosing requirement, not a virtue.

Reasonable candidates

• Adolescents aged 12 to 17. Response rates fall after 11 but the cohorts still include responders. Expect longer timelines.
• Adults with a flexible chest wall. The pioneer cohort included patients up to 61 (Haecker 2016). Adult response is slower and less complete, and published adult timelines commonly run 12 to 24 months or longer.
• Patients who have declined surgery. A formally accepted indication (Loufopoulos et al. 2021).
• Children too young for surgery with moderate to severe deformity, as a holding and improving measure (Loufopoulos et al. 2021).

Who this does not work for

This section matters more than the one above it. About half the largest published cohort did not respond, and the people in that half share recognisable features.

• Rigid chest walls. If the sternum does not visibly lift under suction, the mechanical precondition is absent, and time will not supply it.
• Severe deformity in an adult. A deep, long-standing, ossified chest is a poor target for a device that works by gradual remodelling. Severe cases in adults are usually a surgical conversation, and the honest advice is to have that conversation.
• People who will not use it consistently for years. This is the largest single failure group, and it is not a character judgement. Zhou and colleagues (2024) found 60% of children paused for more than two weeks at least once. If two hours a day for two years is not realistic in your life right now, the outcome data does not apply to you.
• Anyone expecting change in weeks. The immediate post-session lift is temporary and partially reverses within hours. Mistaking that for lasting progress ends more treatment courses than any biological factor.
• Pectus carinatum. A protruding sternum is the opposite deformity and needs compression bracing, not suction. A vacuum bell is not indicated.
• Specific contraindications. Cardiac conditions, coagulopathies or bleeding disorders, vasculopathies, skeletal disorders affecting the thoracic wall, and active skin disease over the treatment area. Anyone in these categories needs clinical clearance before starting.

Anyone with symptoms suggesting cardiac or pulmonary compromise, meaning breathlessness disproportionate to exertion, palpitations, chest pain or exercise intolerance, should be assessed by a physician before considering any treatment.

The one part of this that is time-sensitive

Most of this decision can wait. One part cannot.

Costal cartilage is more pliable in younger children and stiffens through adolescence. A therapy that works by sustained gentle deformation depends on that pliability. This is not a marketing observation, it is the pattern that dominates the outcome literature: the same daily suction acts on more responsive tissue in a child than in an adult, and children who begin later change less. Cincinnati Children's reports the strongest results in children aged 6 to 12 with a mild deformity.

Nothing bad happens if a family takes four weeks to think. But if a child is 9 or 10, the odds available now are meaningfully better than the odds available at 12, and no later decision recovers that. If a child is already past that window, this is not a door that has closed. The cohorts include adolescent and adult responders. It means timelines are longer and the case for a proper assessment now rather than next year is stronger, not weaker.

What this means in practice

Four things follow from the evidence above.

Age is the variable you cannot get back. Every month of delay in a child under 11 moves them toward the group with materially worse odds. If a young child has pectus excavatum and the chest is flexible, the case for starting a conservative trial early is stronger than the case for watchful waiting, and it is worth raising with a paediatrician directly.

Adherence is the variable you can control, and it is the one that fails. The pattern of paused treatment in the Zhou cohort, set against a therapy whose published cohorts run for years, identifies the actual failure mode of this therapy. It is not that the device does not work. It is that most people stop using it. Because consistency is the factor a patient can actually change, anything that sustains it acts directly on the strongest lever available: Haje and colleagues (2021, n=115) found that structured support raised adherence from 58% to 83%.

Measure properly or you will misread your own progress. Take measurements in the morning, before the day's first session, at least 20 hours after the previous one, using the same landmarks, lighting and posture every time. Post-session measurements capture temporary lift and tell a flattering story that collapses later.

This is a trial, not a commitment. Six months of consistent use is enough to know whether a chest is responding. If it is not, that is real information, and surgery remains available. Conservative treatment does not close any doors.

The cost asymmetry is part of the decision

Surgical repair in the United States commonly runs 40,000 to 70,000 dollars, with uninsured cases reported up to 100,000 dollars, and involves a bar in the chest for two to three years plus a second operation to remove it. A vacuum bell system costs a few hundred dollars and can be stopped at any time. Insurers have in some cases covered the device as durable medical equipment. That asymmetry does not make conservative treatment the right answer. It does mean a six-month trial carries a very different downside to a surgical decision, and that is a legitimate input rather than a marketing point.

Not sure whether this applies to you?

The two questions this article cannot answer for you are whether your own chest is a reasonable candidate, and which device size fits it. Both depend on measurements only you can take.

The FormaChest assessment takes about four minutes. It screens your situation against the variables the research identifies as predictive (age, depth of the indentation, chest flexibility and symptoms) and returns an honest read on whether conservative treatment is a reasonable first step. That includes telling you when the answer is to see a surgeon first, which it does for roughly the situations described in the section above.

It also returns a vacuum bell size recommendation based on your own measurements, at no cost.

Sizing is normally not free or simple in this category. The usual process is to email photographs of your bare chest to a company and wait for a reply, or to print paper templates and tape them to yourself, with the manufacturer warning that self-measurement may produce the wrong model. The assessment asks for numbers you can take alone with a ruler. No photographs, no email exchange, no deposit.

Take the free assessment →

What the research does not tell us

Presenting this literature as settled would misrepresent it. The specific gaps:

No randomised controlled trials. Every study cited here is retrospective, observational or survey-based. There is no randomised comparison of vacuum bell therapy against surgery, against a sham device, or against no treatment. That is the single largest evidentiary gap in the field.

No validated guidelines. Haecker and colleagues (2024) state this explicitly after reviewing 20 years of use. Daily duration, session length, suction pressure and total treatment length vary between centres and between manufacturers, and no consensus protocol has been established.

Selection bias throughout. Patients who choose a conservative treatment and stay with it for two years differ systematically from those who do not. Published cohorts are enriched for motivated, supported patients, which likely inflates apparent efficacy relative to the general population.

Weak long-term durability data. Whether corrections achieved in childhood hold through adult growth and into later life is not well established.

Inconsistent outcome measures. Some studies use Haller index change, some use correction index, some use external depth, and some use parental satisfaction. These are not interchangeable, and headline percentages drawn from different endpoints should not be compared directly.

Tomas Titus is the founder of FormaChest (AUGUMENTUM SRL, Craiova, Romania). He had pectus excavatum and corrected it without surgery, documenting his own vacuum bell treatment on camera, taking measurements before each session rather than after. His chest depth went from 25mm to 4mm. That is one person's result at 19; published adult timelines run 12 to 24 months and results vary. He is not a clinician, and nothing in this article is a substitute for assessment by one.

Related questions

• What is pectus excavatum? The complete guide
• Is 25, 30 or 40 too old for a vacuum bell?
• How to measure your pectus excavatum at home
• Bruising and red marks from a vacuum bell: what is normal
• How much does the Nuss procedure cost?
• The Haller index explained

References

1. Scaife ER, et al. Vacuum bell therapy for pectus excavatum. 2025. n=240
2. Luo D, Cheng K, Yuan M, Xu C, He T. Efficacy and determinants of vacuum bell treatment in preschool children with pectus excavatum. Frontiers in Pediatrics. 2022;10:1008437. PMID 36313864
3. Zhou L, Deng F, Tian Y, et al. Questionnaire-based subjective evaluation of factors influencing vacuum bell treatment in children with pectus excavatum. Frontiers in Pediatrics. 2024;12:1467215. DOI 10.3389/fped.2024.1467215
4. Haecker FM. Vacuum bell therapy. Annals of Cardiothoracic Surgery. 2016;5(5):440-449. PMID 27747177
5. Haecker FM, et al. 20 years clinical application of the vacuum bell for conservative treatment of pectus excavatum. Journal of Thoracic Disease. 2024;16(8):5285-5298. PMID 39268139
6. Loufopoulos I, Karagiannidis IG, Lampridis S, Mitsos S, Panagiotopoulos N. Vacuum Bell: Is It a Useful Innovative Device for Pectus Excavatum Correction? Turkish Thoracic Journal. 2021;22(3):251-256. PMID 35110237
7. Haje SA, et al. Adherence in conservative treatment of pectus deformities. 2021. n=115
8. Scaife ER, et al. Vacuum bell therapy outcomes. 2025. n=240
9. Basel cohort. Vacuum bell treatment outcomes, Switzerland.
10. Cincinnati Children's Hospital Medical Center. Should I worry if my child's chest is sunken? Patient education resource.

Important

This article is for information only and is not medical advice. Pectus excavatum should be diagnosed and assessed by a qualified physician. If you have chest pain, breathlessness, palpitations or reduced exercise tolerance, seek medical assessment before beginning any treatment. Last reviewed: July 2026.