February 20, 2026
by Ergün Cakal
What is the adequate level of healthcare in prison? How are judges to make such assessments (and how well placed are they to do so anyway)? How is a prisoner’s own refusal of treatment to be weighed in that assessment? When does a lack of healthcare require that a prisoner be released? These are some critical questions raised and grappled with by the European Court of Human Rights (the Court) in Fernandez Iradi v. France– concerning the ailing ETA leader, Mr. Fernandez Iradi.[1]
Read more: Assessing the adequacy of healthcare in prison: The deference to medical evidence in Fernandez Iradi v. FranceThe applicant, Juan Ibon Fernandez Iradi, was detained in 2003 and convicted by French authorities in 2008 and 2009. He was sentenced to a total of 30 years in prison for terror-related offences, including the shooting of a French police officer. While in prison, in 2012, he was diagnosed with multiple sclerosis, which has since required constant medical attention. Since his diagnosis, state-commissioned medical expert opinions have rendered the applicant’s detention conditional on the provision of a range of specialist neurological, psychological and physiotherapeutic treatments. The applicant argued that treatment had not been prompt enough to begin and had been partial and periodic thereafter.
Accordingly, Iradi complained to the French courts to obtain a suspension of sentence (meaning a release from prison). His attempts were initially successful with a first national instance ordering his release, based on the two medical expert opinions. This ruling was reversed on appeal by the prosecution at a second national instance, which found instead that the prescribed treatment was being adequately administered. His appeal to a third national instance, namely the Court of Cassation, failed for technical reasons. The applicant’s attempt to be released failed.
Iradi applied to the European Court of Human Rights. He argued that, in violation of his Article 3 right to not be subjected to torture, inhuman and degrading treatment, he was not provided with an adequate level of health care (adequate according to the expert opinions), and that this should compel his release in order for him to receive such treatment. He also complained about a specific instance in 2012 when he was taken to see a hospital neurologist in restraints. The Court deemed that complaint inadmissible as it had been entered too late (beyond the four-month period). France argued against finding an Article 3 violation, pointing to the adequacy of the treatment he received, although it characterised the expert opinions as mere ‘recommendations’ and underscored the lack of deterioration in the applicant’s health. Iradi argued that the fact that his condition had not notably deteriorated (as his motor skills and autonomy had not been affected) should not exonerate the authorities from their responsibility. France also argued that the applicant had himself accepted some of the treatment whilst having rejected the psychological treatment on offer in prison. No further details are provided of the applicant’s choice in this respect. An annual neurological follow-up, as recommended by the medical experts, had also notably not been arranged by the French authorities.
At his own request, Iradi was transferred to a prison in San Sebastián in 2022, where he continues to serve the remainder of his sentence with the possibility of periodic releases.
The majority ruled that France had failed to provide adequate healthcare according to the standards set by its national courts, particularly with respect to the authorities’ failure to provide neurological treatment in certain periods. Instead of assessing adequacy from its own standpoint, the majority based its assessment of adequacy on the opinions of the national medical experts, asking what they prescribed and was that provided. These expert opinions played a crucial part in the national courts’ rulings – as they did for the majority of the European Court of Human Rights.
On the whole, in its ruling, the European Court reiterated its central principles on the question of prison healthcare under Article 3:
Importantly, the majority rejected the applicant’s argument that he should have been released due to the inadequacy of the care he received in prison, noting that the applicant’s own rejection of the psychological treatment and that his health had not sufficiently deteriorated. The majority maintained that the necessary circumstances which would have justified his release were not established by the applicant. It awarded the applicant 10.000 EUR compensation plus costs.
A majority of six judges found a violation of Article 3, with concurring opinions from Judges Felici, Zünd, Sârcu (jointly) and Judge Gnatovskyy (individually), and a dissenting opinion from Judge Elósegui. In their opinion, Judges Felici, Zünd and Sârcu emphasised that the fact that the applicant’s health had not deteriorated was not relevant to the outcome. The judges therefore chose to emphasise the provision of the treatment not the health outcomes. Judge Gnatovskyy noted the complexity of assessing the quality of health care and the applicant’s selective refusal of certain treatments in order to be released. He also mentioned that France had not raised the argument of non-exhaustion of domestic remedies – as it meant that the Court could not consider this ground for inadmissibility. Judge Elósegui could not find that Article 3 had been violated, highlighting the adequacy of the healthcare, the applicant’s own refusals, and that his health had not materially deteriorated. She also considered that the security regime was justified based on the applicant’s history of political violence.
The Court’s practice in applying the above-listed principles, in this case as elsewhere, however, reveals three deep faultlines:
The primary importance of the judgment concerns how the Court positioned itself in its assessment of the adequacy of the medical care received. The Court readily recognises that assessing this is a difficult question and requires extensive medical knowledge (para 55) and that ‘it is not its task to rule on matters lying exclusively within the field of expertise of medical specialists and to establish whether an applicant in fact required a particular treatment or whether the choice of treatment methods appropriately reflected the applicant’s needs’ (citing Wenner v. Germany, No. 62303/13, § 58, 1 September 2016). The Court relied on the medical expertise produced for national courts to avoid assessing these questions anew. Whilst the Court’s deference to national courts or state-produced expertise are not novel in cases concerning the adequacy of healthcare in prison, there remains a critical question of when the Court makes such a deferential call. The deference in this case largely vindicated the interests of the applicant (although it did not see his release) but the deference in related cases concerning medical expertise, where the Court has followed medical opinions, has worked against applicants. The reasoning in Ramirez Sanchez v. France and Herczegfalvy v. Austria spring to mind – where states had an easier time referring to the assessments of state doctors in justifying coercive necessity. The question could readily be framed in terms of ‘epistemic authority’ as previously discussed in the context of home births. What would the Court have done if the French medical opinion had not made the applicant’s detention conditional on extensive treatment? It seems to me that the Court lacks in-house expertise to confidently assess questions of healthcare adequacy (read: there are no (medical) doctors in the building). There is a deep dependency here on state-produced evidence that remains problematic.
The Court’s Article 3 practice has another issue: namely, how it weighs the responsibility of an applicant in their own suffering. Applicants’ behaviour in the face of ostensible choice, their inaction or actions, often contributes to the harm they ultimately suffer. A contextual assessment is crucial to understand when applicants are being faced with a forced choice. How the Court recognises this (or not) signals the degree to which it views the applicant as deserving of protection. In Semenya v. Switzerland, the Chamber (poorly) reasoned that the applicant there had a choice to not suffer, in accepting or rejecting her treatment (para 215; see also dissenting opinion of Judge Serghides, paras 35-36). In N.D. and N.T. v. Spain, the Grand Chamber again problematically focused on the applicant asylum seekers’ ‘own culpable conduct’, emphasising that the applicants had crossed the border ‘in an unauthorized manner’ (para 206), in rejecting their claim. The applicant in O’Rourke v. the UK, who, having been evicted from temporary accommodation and rejected subsequent offers, found himself living on the streets for 14 months, saw his Article 3 arguments dismissed as he himself ‘largely responsible for his own deterioration’ (page 6). Assessments concerning the extent to which the applicant is responsible for their own situation that fail to contextualise such ‘choices’ (here, it was a prisoner refusing treatment in prison, which is generally of a low standard even in the ‘best’ of regimes) end up rendering the individual responsible for the suffering they endured at the hands of the state. This is a problem because protection should not require individuals to foresee and avoid victimization in the first place.
Relatedly, the Court was significantly (though summarily) drawn into the argument as to the deterioration of the applicant’s suffering – noting in rejecting the applicant’s arguments for release, that his health had not been manifestly deteriorating. That the applicant’s body had proven somehow resilient underplays the nature of the harmful acts, given his chronic condition and over-emphasises the need for visible individual impact. Such a logic distances as to discharge state responsibility for looming harm – which the prison, it ought to not be uncontroversial to point out, is not optimally placed to prevent. The harm will be realised at another time and another place – long after the Court has ceased to deal with the victim’s plight and assess the state’s responsibility.
The question of the extent to which the Court should accept security justifications is a related species of ‘deference’ which didn’t quite feature in this case. Until the final page of the judgment (of Judge Elósegui’s dissent), we do not read of the applicant’s history of political violence.[3] The Court, that said, has occasionally been explicitly reluctant to second-guess necessity-based arguments from states – whether medicalor operational – effectively characterising them as non-justiciable. If nothing else, the ‘practical demands of imprisonment’ clause would conceivably accommodate such justifications in the future as they have in the past – conceivably emphasising the security conditions necessitated by high-profile prisoners. This is a recurring phrase in the article 3 caselaw that finds no elaboration in the caselaw or commentary, often hanging on the end of the phrase ‘the unavoidable level of suffering inherent in detention’, similarly operating with thin reflection.
These concerns leave one with but a single (loaded) question: how do such justifications and rationales, as woven through the Iradi judgement, square with the absolute nature of the prohibition of torture, inhuman and degrading treatment which allows no justification whatsoever in principle, neither based on an individual’s conduct or policy considerations?
Iradi v. France is a reminder that the prohibition of torture, inhuman and degrading treatment remains a politically-charged and evidentially-fraught terrain for the Court – even where there is deference to expertise. Whilst the majority opted for a (largely) principled and restrained approach (and nothing less should be expected with Judge Gnatovskyy, a former president of the CPT, in the composition), faulty logics and slopes remain in the nature of such assessments. The limits of the Court’s evidentiary principles and practice (and resources) are on full display in such cases – which the Court does well to conceal here as elsewhere. Scrutiny of state-packaged evidence, however, a task at the fore of the Court’s protective function, requires much more.
[1] Euskadi Ta Askatasuna (ETA, translated as ‘Basque Homeland and Freedom’) was a Basque separatist group known for its use of political violence. It disbanded in 2018.
[2] I’ll keep a fourth, namely the uneasy appearance of the four-month rule as a statute of limitations for an Article 3 claim, in my pocket for a rainy day.
[3] The view that the majority’s denial of the applicant’s release implicitly is based on the applicant’s profile would be speculative.
Strasbourg Observers is an academic blog that discusses recent developments at the European Court of Human Rights. The editorial team of Strasbourg Observers is based at the Human Rights Centre of Ghent University and the Centre for Government and Law of Hasselt University.
Strasbourg Observers would not be possible without the valued work of countless guest bloggers. Visit the “Contact” page on our website to find out how to become one!
