Patient Satisfaction in Medical Tourism

Patient satisfaction in medical tourism is the most frequently published quality figure in the sector and the least informative, because it is almost always collected by the treating provider, at the destination, before the clinical result of the treatment is known.

Satisfaction is not experience, and neither is outcome

Three distinct things are routinely conflated. Patient satisfaction assesses patients’ expectations and whether those expectations were met, whereas patient experience considers whether, or how often, various aspects of care such as clear communication with providers occurred.^[1] A clinical outcome is a third thing again: whether the operation worked.

The difference matters because satisfaction is relative to what the patient expected, which varies by nationality, by price paid and by what the patient was told before travelling. Two patients receiving identical care can report different satisfaction, and a provider can raise reported satisfaction by managing expectations rather than by improving care. Experience measures, treated in patient-reported experience measures, were developed partly to avoid that.

Why the published figures cannot be pooled

Two recent assessments describe the same structural problem. A 2014 review commissioned in the United Kingdom concluded that little is known about the longer-term health outcomes of medical tourists beyond incidental reports of complications, and that little is known about which patients choose to travel and why.^[2] A 2026 systematic review records that there is a lack of systematic, national data on the true scope of medical tourism and its associated harms, that all quantitative data it included was cross-sectional and collected through surveys, that studies conducted within the destination country create sampling bias, that research focuses on single centres, and that prospective studies with generalisable samples are needed.^[3]

Two specimen studies show what that means in practice. One used an instrument the authors built themselves, a five-point scale across three phases of the journey with internal consistency reported between 0.736 and 0.945, administered in hospitals in one Korean city during April and May 2018 to patients who were in stable enough conditions for the survey. There was no comparison group and no long-term follow-up.^[4] The other used four borrowed seven-point items, one of which asked respondents to rate agreement with being satisfied with staying in this hospital, among 205 Chinese exchange students in Malaysia, with no stated response rate and no treatment of selection bias.^[5]

Both are legitimate within their stated scope. Neither supports a statement about satisfaction among medical travellers generally, and because the instruments, populations and timing differ, their results cannot be averaged into one.

The timing problem

Satisfaction in this sector is typically measured at the destination, at or near discharge, which is the point at which a surgical result is least knowable. Infection, graft failure, scarring, revision need and functional outcome declare themselves over weeks to months, usually after the patient has flown home and left the provider’s reach.

The consequence is a systematic asymmetry. Satisfaction is captured from everyone who completes treatment, while complications are captured, if at all, by a different health system in another country that has no connection to the provider’s dataset. A provider’s satisfaction figure and its complication rate are not drawn from the same population, and the first cannot be read as evidence about the second.

Satisfaction is not a proxy for clinical quality

The general literature on this question is more cautionary than the sector’s marketing use of satisfaction scores would suggest, and it must be reported with its conditions attached. A study using a United States household medical expenditure panel from 2000 to 2015, with 92,952 respondents, examined satisfaction with clinicians against subsequent mortality. After applying an adjustment recommended for the survey instrument used, hazard ratios for the second, third and fourth satisfaction quartiles compared with the first were 1.28 with a confidence interval of 1.01 to 1.62, 1.43 with an interval of 1.12 to 1.82, and 1.57 with an interval of 1.25 to 1.98.^[6]

The authors’ own statement of what they found is narrow and should be carried with the figures: the association of higher patient satisfaction with clinicians with higher short-term mortality was evident only after the recommended adjustment, and was evident in women but not men.^[6] This is a finding about satisfaction with individual clinicians in one national survey under one adjustment method, in one sex. It is not a finding about hospitals, about surgery or about medical travel, and it is widely over-claimed.

What it does establish is that the relationship between satisfaction and clinical outcome is not reliably positive, which is enough to make a high satisfaction score unusable as a substitute for outcome data.

Reading a provider satisfaction claim

Who collected it, and were they the treating provider. When was it collected relative to the procedure. What instrument was used, and has it been validated. How many patients were eligible, how many responded, and what happened to the rest. Was anyone who did not complete treatment, or who returned home with a complication, capable of appearing in the sample. Where those cannot be answered, the figure describes a provider’s own feedback collection and nothing wider.

See also

• Patient-reported experience measures, the standardised alternative to satisfaction
• Patient-reported outcome measures, what measuring the result looks like
• Medical tourism surveys, the sampling frames behind these studies
• Medical tourism indexes, where satisfaction scores are used as inputs

References

1. Agency for Healthcare Research and Quality. What is patient experience? Last reviewed March 2025. Verified against primary record: agency definition page opened and read. Retrieved 2 October 2026.
2. Lunt N, Smith RD, Mannion R, et al. Implications for the NHS of inward and outward medical tourism. NIHR Journals Library, January 2014. Independently reported: publicly commissioned peer-reviewed research. Retrieved 2 October 2026.
3. Nichol B, Ray D, Tanner L, et al. Decision-making about bariatric and cosmetic medical tourism from countries with universal healthcare: a rapid systematic review. Globalization and Health, 2026;22:45. Independently reported: peer-reviewed systematic review. Retrieved 2 October 2026.
4. Development and validation of a medical tourism satisfaction instrument, DOI 10.1186/s40887-018-0024-2, 2018. Independently reported: peer-reviewed article read for instrument design and recruitment; the fetched record did not return author names or journal title, so it is cited by digital object identifier. Retrieved 2 October 2026.
5. Li L, Ragavan NA, Patwary AK, Baijun W. Measuring patient satisfaction on medical tourism. Frontiers in Public Health, 2022;10:893439. Independently reported: peer-reviewed article read for instrument and sample. Retrieved 2 October 2026.
6. Jerant A, Fiscella K, Fenton JJ, Magnan EM, Agnoli A, Franks P. Association of clinician denial of patient requests with patient satisfaction, and companion analysis of satisfaction and mortality. Journal of General Internal Medicine, 2019;34(8):1459-1466. Independently reported: peer-reviewed article read for hazard ratios, adjustment method and stated conditions. Retrieved 2 October 2026.

Sourcing note: the agency definition, the two reviews, the two specimen studies and the satisfaction and mortality analysis were opened and read on 2 October 2026. The hazard ratios are reproduced only together with the adjustment condition and the sex restriction stated by the authors, because the finding is routinely quoted without either. One specimen study is cited by digital object identifier because the retrieved record did not carry its authors or journal title. No pooled satisfaction rate is published, because the instruments and populations in the available studies are not comparable.