International Patient Visits and Unique Patients

International patient visits and unique patients are two different quantities, and the difference between them is the most common source of inflation in medical travel figures. A hospital that reports treating a given number of international patients in a year may be reporting the number of people, the number of visits they made, or the number of episodes of care recorded. The three are not the same, and the gap between them is large.

The distinction, in the words of those who maintain the data

Statisticians call this a mismatch between the administrative unit and the statistical unit of interest. The international guidance on using administrative data for official statistics lists the alignment of administrative units with statistical units of interest as a specific limitation, illustrating it with the difference between a tax unit and an enterprise, and notes that the units and variables an administrative data source contains are defined by administrative rules and may therefore not be identical to those required by a statistical office.^[3]

The United States national inpatient dataset states the problem about itself in plain terms. It contains discharge-level records, not patient-level records, which means that individual patients who are hospitalised multiple times in one year may be present in the dataset multiple times; and there is no uniform patient identifier available that allows a patient-level analysis.^[2] This is not a flaw that careful analysis can work around. The identifier that would be needed does not exist in the data.

The European hospital activity collection shows the same thing from the reporting side. In the country notes accompanying its definitions, Romania records that a person could be included multiple times for every episode of discharge.^[4] Double counting also arises because transfers between hospitals are treated differently in different countries, so the same patient journey produces a different number of discharges depending on where it happens.

A worked example of the size of the gap

The clearest published illustration comes from waiting list statistics rather than from medical travel. At the end of May 2026 England recorded 7.3 million incomplete referral to treatment pathways, described in the release as approximately 6.2 million unique patients.^[1]

The gap is about 1.1 million, or roughly eighteen per cent of the patient figure. One person can hold several pathways at once, and the release distinguishes the two counts precisely because the difference matters. Where a published figure of that kind is quoted as a number of people waiting, the error is not marginal.

It is worth noting what makes that example useful: the publisher states both numbers. An international patient figure from a hospital or a promotion body typically states one, without saying which.

Four quantities, often conflated

In medical travel the confusion has more layers than in a domestic system, because arrival and treatment are separately countable. Four quantities are routinely reported as though interchangeable.

A unique patient is one person, however many times they attend. A visit or admission is one episode of attendance, of which one person may have several. A procedure is one intervention, of which one visit may contain several. And an arrival is one journey into the country, which may cover several visits or none at all, since a person may travel once and attend repeatedly, or attend without having travelled.

A figure drawn from immigration or tourism data counts arrivals. A figure drawn from hospital administrative data counts episodes. A figure drawn from a billing system counts invoices. None of them counts people unless a patient identifier links the records, and in the datasets examined above no such identifier is present.

No source located for this entry states, for any published international patient figure, which of these four quantities it reports. That absence is itself the practical finding: where the unit is not stated, the figure cannot be compared with another, and a year-on-year change in it may reflect a change in recording rather than in activity.

See also

• Hospital discharge data in medical tourism, where the episode count comes from
• Administrative data in medical tourism research, and the unit mismatch in general
• Medical travellers and resident foreign patients, the other great conflation
• Medical tourism statistics, the overview of what is measured

References

1. NHS England. Statistical Press Notice: Referral to Treatment waiting times, May 2026. Published 9 July 2026. Verified against primary record: official statistics release opened and read. Retrieved 2 October 2026.
2. Agency for Healthcare Research and Quality, Healthcare Cost and Utilization Project. Introduction to the HCUP National Inpatient Sample, 2021. Verified against primary record: dataset documentation opened and read. Retrieved 2 October 2026.
3. United Nations Expert Group on National Quality Assurance Frameworks. Module for Quality Assurance when using Administrative and Other Data Sources to produce Official Statistics. March 2025. Verified against primary record: international statistical guidance opened and read. Retrieved 2 October 2026.
4. Eurostat. Health care activities: definitions and data collection specifications annex. Verified against primary record: statistical metadata annex opened and read. Retrieved 2 October 2026.

Sourcing note: the two official statistics releases and the two statistical standards were opened and read on 2 October 2026. The worked example and the two self-descriptions come from domestic datasets, which are used here because they document the distinction explicitly in their own words; no international patient dataset examined for this site does so. The phrase mismatch between administrative and statistical units is the terminology the international guidance uses, and the transactions-against-people framing common in commentary is described rather than quoted, because no statistical authority read for this entry puts it that way.