| Verified against primary record | |
| Statutory basis of the count | Law of 4 June 2007, Article 4, in force 1 July 2010[4] |
|---|---|
| Stays by patients not insured in Belgium (2014) | 94,740[3] |
| Of those, patients not domiciled in Belgium | 37.8 per cent, about 35,860 stays[3] |
| Share of all hospital stays | About 1.3 per cent, or about 1.53 per cent including S2 cases[3] |
| Largest countries of residence | Netherlands 61.2, France 9.6, Germany 3.8 per cent[3] |
| Average length of stay, non-residents | 5.5 days[3] |
| Median cost of a classic stay | EUR 3,682 for foreign residents, EUR 3,147 for Belgian residents[3] |
| Prior authorisation requests (2023) | 103 received, 17 authorised, 86 refused[1] |
| Independently reported | |
| Incoming patients under the Directive (2016) | 27,457, the third highest in the European Union[2] |
| Bands apply only to the rows beneath them. The Observatory figures count hospital stays, not individual patients, and describe discharge records for 2014 as published in the 2015-2016 annual report, the most recent one located. The cost figures are drawn from hospital billing files and include supplements and non-reimbursable items, so they are not a price list and not comparable with the separately sourced S2 cost figures. Figures from reports up to 2012 rest on domicile alone and are not comparable with later ones. | |
Medical tourism in Belgium is measured more directly than in most European countries, because Belgian law established a standing body whose first statutory duty is to count patients treated in Belgian hospitals who are not insured in Belgium. The resulting figures show a flow dominated by neighbouring countries, concentrated in routine rather than exotic procedures, and amounting to a little over one per cent of hospital activity.
The Patient Mobility Observatory
The Observatoire de la mobilite des patients, or Observatorium voor patientenmobiliteit, was created by the Law of 4 June 2007 amending legislation to promote patient mobility. Article 4, paragraph 1 establishes it jointly at the National Institute for Health and Disability Insurance and the Federal Public Service for Public Health. The provision took effect on 1 July 2010 and the Observatory began work in 2011. It has thirty members drawn from employers, trade unions, insurers, care providers, government bodies and regional and community ministers, and is required to submit an annual report before 1 July.[4]
Article 4, paragraph 2 gives it five tasks: to collect data on patients not insured by a Belgian insurer who are treated in Belgian hospitals, including their treatment and country of origin; to collect data continuously on waiting times for Belgian-insured patients; to facilitate and support agreements with foreign insurers; to develop expertise on pricing treatments offered to foreign markets; and to advise government on infrastructure, planning and staffing in the light of the number of patients treated who are not insured in Belgium.[4]
The pairing of a counting duty with a pricing and planning duty is what distinguishes the Belgian arrangement. The same law allows prices for patients who are not insured in Belgium to be set by royal decree, with the Observatory supplying the expertise.[3]
Despite the statutory duty to report every year, the most recent annual report located is the one covering 2015 and 2016. Any figure presented as current Belgian data should be checked against that publication date.
How many patients and from where
Working from the Resume hospitalier minimal, the compulsory minimum hospital discharge dataset, the Observatory recorded 94,740 stays in 2014 by patients not covered by a Belgian insurer. Of those, 37.8 per cent, or roughly 35,860 stays, involved patients not domiciled in Belgium. Together these amounted to about 1.3 per cent of all hospital stays, rising to about 1.53 per cent once stays reimbursed under the S2 route are added.[3]
Among the non-resident stays, the Netherlands accounted for 61.2 per cent, France for 9.6 per cent and Germany for 3.8 per cent. The average length of stay was 5.5 days. Day hospitalisation made up 40.8 per cent of the stays, ambulatory emergencies 33.3 per cent and classic admissions 25.9 per cent. Most care was planned rather than urgent: 88.4 per cent of day-hospital stays and 65.2 per cent of classic stays were planned, though close to a third of classic admissions arrived as emergencies.[3]
The earlier report covering discharge years 2008 to 2011 found a similar shape at a different scale, with non-residents accounting for about 1.5 per cent of inpatient stays and 1.3 per cent of day cases, some 113,388 inpatient and 87,135 day-case stays in total. The top ten countries of residence covered 93.4 per cent of the flow, with the Netherlands at about 57.4 per cent and France at about 20.5 per cent, followed by Luxembourg at 4.6 per cent and the United Kingdom, Italy and Germany at about 2.6 per cent each. The diagnoses and procedures were common ones: musculoskeletal disorders, spinal disc lesions, cataract surgery and colonoscopy.[4]
The two sets of figures are not directly comparable. Reports up to discharge year 2012 identified the relevant population by domicile alone, while the later work uses insurability codes, and the Observatory says so explicitly.[3]
Cost and contracts
The median cost of a classic hospital stay was 3,682 euro for patients resident abroad against 3,147 euro for Belgian residents, about 17 per cent higher. The Observatory draws these figures from hospital billing files, which include supplements and items that are not reimbursable, and notes that the S2 cost figures come from insurers and are not directly comparable.[3]
On its contracts mandate, the Observatory received 111 agreements, of which 89 were analysed as genuine contracts for the care of foreign patients and ten were cross-border border-zone agreements. It reported that a guide issued in 2013 produced clearly better contracts from 2014 onwards, while four provisions remained problematic: the authorisation and registration status of the parties, the clause safeguarding capacity for domestic patients, the number of patients covered, and arrangements for care after discharge.[3]
Separately from commercial contracting, organised border-zone cooperation with France moves a substantial number of patients. In 2012 about 15,670 patients crossed within these zones, 14,451 of them travelling from France to Belgium and 1,219 in the opposite direction.[3]
The European route
Under Directive 2011/24/EU the member state of affiliation must reimburse cross-border care where the treatment is among the benefits it covers at home, capped at the level of costs it would itself have assumed and not exceeding the actual cost of care received. Prior authorisation may be required only for care subject to planning requirements involving an overnight stay or highly specialised and cost-intensive infrastructure, for treatments carrying a particular risk, and for providers raising serious quality or safety concerns.[6]
Belgium’s own return for reference year 2023 records only 103 prior authorisation requests, of which 17 were authorised and 86 refused, with 25,155 euro reimbursed. It supplied no figures for reimbursement outside prior authorisation, because not all health insurance funds provided data. The refusal rate of 83.5 per cent is far above the European pattern, and on this evidence alone cannot be explained.[1]
In the other direction, the European Court of Auditors recorded Belgium as the third largest recipient of incoming patients under the Directive in 2016, with 27,457, behind Spain at 46,138 and Portugal at 32,895. Across the Union about 200,000 patients a year use the Directive, fewer than 0.05 per cent of EU citizens.[2]
Eurostat separately maintains a collection of hospital discharges of non-residents, with tables for in-patients, day cases, bed days and average length of stay, and reports country of residence for non-resident patients where member states can supply it. Individual cell values for Belgium were not retrievable for this article.[5]
Limits of the evidence
The Observatory is candid about the quality of its own inputs. Only 78 per cent of stays had the three complete billing files it relies on, about 18 per cent had none of them and 4 per cent had only one or two, while roughly 0.9 per cent of records carried unknown codes, which it judged acceptable. For S2 stays, about half could not be matched to discharge records at all, so the Observatory treats the country breakdown for that group as not meaningful. The discharge dataset also covers hospital admissions and emergency visits but not planned outpatient consultations, so care delivered entirely in clinics and consulting rooms is outside the count. Within the 2015-2016 report there is an internal inconsistency: the summary describes classic admissions as one fifth of non-resident stays where the detailed tables give 25.9 per cent.[3]
See also
- Cross-border healthcare, the wider framework Belgian figures sit inside
- Cross-border healthcare for border communities, the border-zone agreements with France
- Hospital discharge data in medical tourism, the record type the Observatory counts from
- Medical travellers and resident foreign patients, the distinction the 37.8 per cent figure turns on
- Average length of stay in medical tourism, the measure reported here as 5.5 days
- Medical tourism in the Netherlands, the origin of most patients counted in Belgium
References
- European Commission, Directorate-General for Health and Food Safety. Data on cross-border patient healthcare following Directive 2011/24/EU, reference year 2023. 15 May 2025. Verified against primary record: the report was opened and the Belgian rows for prior authorisation requests, authorisations, refusals and amount reimbursed were read, together with the note that not all Belgian health insurance funds supplied reimbursement data. Retrieved 23 October 2026.
- European Court of Auditors. Special Report 07/2019: EU actions for cross-border healthcare. 2019. Verified against primary record: the report was opened and the 2016 incoming patient figures for Belgium, Spain and Portugal and the Union-wide annual patient estimate were read. Retrieved 23 October 2026.
- Observatoire de la mobilite des patients. Rapport annuel 2015-2016. Institut national d’assurance maladie-invalidite. Verified against primary record: the report was opened and the 2014 stay counts, the domicile split, the shares of hospital activity, the countries of residence, the length of stay, the stay types, the median cost comparison, the contract analysis, the border-zone figures and the stated data-quality limits were read directly. Retrieved 23 October 2026.
- Observatoire de la mobilite des patients. Rapport annuel 2013. Institut national d’assurance maladie-invalidite. Verified against primary record: the report was opened and the statutory basis in the Law of 4 June 2007, Article 4, the five mandated missions, the composition and reporting duty, and the 2008 to 2011 discharge figures and countries of residence were read directly. Retrieved 23 October 2026.
- Eurostat. Hospital discharges and length of stay for inpatient and curative care (hlth_hosd), reference metadata. Last updated 15 December 2025. Verified against primary record: the metadata page was opened and the non-resident definition, the statement that country of residence is reported, and the list of non-resident tables were read. Retrieved 23 October 2026.
- European Parliament and Council. Directive 2011/24/EU, Chapter III, Articles 7 and 8. 9 March 2011. Verified against primary record: the chapter text was opened and the reimbursement rule in Article 7 and the three permitted grounds for prior authorisation in Article 8(2) were read. Retrieved 23 October 2026.
Sourcing note: this entry uses only official sources. The Observatory counts hospital stays rather than individual patients, and the figures quoted are stays throughout. The cost comparison is reproduced as the Observatory states it, including its own warning that the billing-file basis makes it unsuitable for use as a price. No attempt has been made to update the 2014 figures to the present, and no figure has been estimated for privately funded treatment arranged outside the hospital discharge system. The Eurostat non-resident collection is cited for its definitions and structure only; its individual cell values could not be retrieved.