CEDH · CASELAW;JUDGMENTS;CHAMBER;ENG — 15 janvier 2026

.s800EAC49 { font-size:12pt } .s5E1364CA { margin-top:0pt; margin-bottom:12pt; text-align:center; page-break-inside:avoid; page-break-after:avoid; font-size:14pt } .sBB9EE52A { font-family:Arial } .sC02E897A { margin-top:42pt; margin-bottom:14pt; text-align:center; page-break-inside:avoid; page-break-after:avoid } .s29100277 { font-family:Arial; font-weight:bold } .s34DFC730 { margin-top:0pt; margin-bottom:0pt; text-align:center; page-break-inside:avoid; page-break-after:avoid } .sA36B60A1 { font-family:Arial; font-style:italic } .s13C79B1A { margin-top:0pt; margin-bottom:18pt; text-align:center; page-break-inside:avoid; page-break-after:avoid } .s339D85E6 { margin-top:0pt; margin-bottom:14pt; text-align:center; page-break-inside:avoid; page-break-after:avoid } .s780F5245 { border:0.75pt solid #000000; clear:both } .sE77B86B8 { margin-top:0pt; margin-bottom:0pt; text-align:justify; padding-top:1pt; padding-right:4pt; padding-left:4pt } .sEC28DD31 { margin-top:0pt; margin-bottom:0pt; text-align:justify; padding-right:4pt; padding-left:4pt } .s54452357 { margin-top:0pt; margin-bottom:0pt; text-indent:14.2pt; text-align:justify; padding-right:4pt; padding-left:4pt; font-size:10pt } .s55E5497F { margin-top:0pt; margin-bottom:0pt; text-indent:14.2pt; text-align:center; padding-right:4pt; padding-left:4pt; padding-bottom:1pt; font-size:10pt } .s10950C61 { margin-top:0pt; margin-bottom:0pt; text-indent:14.2pt; text-align:justify } .s598389FB { margin-top:0pt; margin-bottom:0pt; text-align:center; font-size:14pt } .sF5E1C6CF { font-family:Arial; font-weight:bold; text-decoration:underline; color:#ff0000 } .sFE10DC93 { margin-top:0pt; margin-bottom:0pt; text-align:center } .sE208486F { font-family:Arial; color:#ff0000 } .s598389F8 { margin-top:0pt; margin-bottom:0pt; text-align:center; font-size:11pt } .s4ACA9207 { page-break-before:always; clear:both; mso-break-type:section-break } .s32563E28 { margin-top:0pt; margin-bottom:0pt } .sA43C3626 { width:28.35pt; font-family:Arial; display:inline-block } .s3AAE10DF { margin-top:14pt; margin-bottom:12pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:14pt } .s3CA22BA { font-family:Arial; text-transform:uppercase } .sD0A217A5 { margin-top:14pt; margin-left:18.45pt; margin-bottom:12pt; text-indent:-18.45pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s3936C9DD { width:11.78pt; font:7pt 'Times New Roman'; display:inline-block } .s9D48DD53 { margin-top:6pt; margin-left:21.25pt; margin-bottom:6pt; text-indent:7.1pt; text-align:justify; font-size:10pt } .sBB64854C { width:8.45pt; font:7pt 'Times New Roman'; display:inline-block } .sDBF374D0 { margin-top:0pt; margin-left:25.5pt; margin-bottom:12pt; text-indent:-17pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s5BDECA8 { width:5pt; font:7pt 'Times New Roman'; display:inline-block } .s3BF0B6C7 { margin-top:14pt; margin-left:25.5pt; margin-bottom:12pt; text-indent:-17pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s7C9EDFAD { margin-top:14pt; margin-left:34pt; margin-bottom:6pt; text-indent:-17pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s54B12A03 { width:6.99pt; font:7pt 'Times New Roman'; display:inline-block } .sE8F2C496 { width:5.11pt; font:7pt 'Times New Roman'; display:inline-block } .s71220C2D { margin-top:0pt; margin-left:18.45pt; margin-bottom:12pt; text-indent:-18.45pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s715E7C6D { margin-top:14pt; margin-left:25.5pt; margin-bottom:12pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .sDD998142 { margin-top:14pt; margin-left:17pt; margin-bottom:12pt; text-indent:-17pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:14pt } .s6B505E72 { margin:0pt; padding-left:0pt } .s6C2746C6 { margin-bottom:12pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-family:Arial; text-transform:uppercase; list-style-position:inside } .s6047437D { width:11.03pt; font:7pt 'Times New Roman'; display:inline-block } .s6B38CF35 { margin-top:0pt; margin-left:34pt; margin-bottom:6pt; text-indent:-17pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .sF54F3725 { margin-top:0pt; margin-left:42.55pt; margin-bottom:6pt; text-indent:-17.05pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:10pt } .sDBC81028 { width:4.83pt; font:7pt 'Times New Roman'; display:inline-block } .s65DDED6B { margin-top:14pt; margin-left:42.55pt; margin-bottom:6pt; text-indent:-17.05pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:10pt } .s7AE800C3 { width:4.28pt; font:7pt 'Times New Roman'; display:inline-block } .s7ED160F0 { text-decoration:none } .s33165EBA { font-family:Arial; font-size:8pt; vertical-align:super; color:#0069d6 } .s55F67FD3 { margin-top:0pt; margin-left:51.05pt; margin-bottom:6pt; text-indent:-17.05pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; line-height:113%; font-size:10pt } .s3970C00F { width:8.17pt; font:7pt 'Times New Roman'; display:inline-block } .sCD82236A { margin-top:14pt; margin-left:51.05pt; margin-bottom:6pt; text-indent:-17.05pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; line-height:113%; font-size:10pt } .s320E5A8E { width:5.95pt; font:7pt 'Times New Roman'; display:inline-block } .s4DDA3AA3 { font-family:Arial; font-weight:bold; font-style:italic } .sFABD3260 { margin-top:14pt; margin-left:62.35pt; margin-bottom:6pt; text-indent:-19.8pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:10pt } .s16F6432D { width:7.9pt; font:7pt 'Times New Roman'; display:inline-block } .sA6765C01 { margin-top:14pt; margin-left:62.35pt; margin-bottom:6pt; text-indent:-11.3pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:10pt } .sFA550501 { width:6.3pt; font:7pt 'Times New Roman'; display:inline-block } .sFF8BF293 { width:8.05pt; font:7pt 'Times New Roman'; display:inline-block } .s7A7B69A2 { margin-top:0pt; margin-left:62.35pt; margin-bottom:6pt; text-indent:-11.3pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:10pt } .s3CAF9CA4 { width:8.72pt; font:7pt 'Times New Roman'; display:inline-block } .s1F6AC3E7 { font-family:Arial; font-size:11pt; font-style:italic } .s1B1D10B4 { margin-top:14pt; margin-left:19.85pt; margin-bottom:12pt; text-indent:-19.85pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s7B868B6A { width:9.85pt; font:7pt 'Times New Roman'; display:inline-block } .s6F863F49 { width:6.51pt; font:7pt 'Times New Roman'; display:inline-block } .s5B1AF0FB { margin-left:36pt; text-indent:-18pt; text-align:justify; font-family:Arial } .s2B9FE6BC { width:7.33pt; font:7pt 'Times New Roman'; display:inline-block } .s2D4532D { width:8pt; font:7pt 'Times New Roman'; display:inline-block } .s8990723E { width:10.67pt; font:7pt 'Times New Roman'; display:inline-block } .s706500A1 { margin-top:0pt; margin-left:36pt; margin-bottom:0pt; text-indent:-18pt; text-align:justify } .s60C44206 { font-family:Symbol } .sCFEF29CB { width:11.41pt; font:7pt 'Times New Roman'; display:inline-block } .s448F0C15 { margin-top:14pt; margin-left:18pt; margin-bottom:12pt; text-indent:-18pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid; font-size:14pt } .sF0C78780 { margin-top:14pt; margin-left:17pt; margin-bottom:3pt; text-indent:-17pt; text-align:justify; font-family:Arial } .sFBC99493 { font-style:italic } .s743F3A55 { margin-right:0pt; margin-left:0pt; padding-left:0pt } .s5165BC52 { margin-left:17pt } .sE5BF05B1 { width:2.33pt; font:7pt 'Times New Roman'; display:inline-block } .s31FAC864 { margin-left:17.05pt; text-indent:-17.05pt; list-style-position:inside } .sE5C1F6E3 { width:3.33pt; font:7pt 'Times New Roman'; display:inline-block } .s6E90AEA1 { margin-top:0pt; margin-left:51.85pt; margin-bottom:0pt; text-indent:-17.85pt; text-align:justify } .sEB913C4D { width:11.26pt; font:7pt 'Times New Roman'; display:inline-block } .sD27F934 { margin-left:51.05pt; text-indent:-17.05pt; text-align:justify; font-family:Arial } .s4B65A808 { margin-left:34pt; text-indent:-17pt; text-align:justify; font-family:Arial } .s2D9C6089 { margin-top:12pt; margin-bottom:12pt; text-indent:14.2pt; text-align:justify; page-break-inside:avoid; page-break-after:avoid } .s69DCC830 { margin-top:36pt; margin-bottom:0pt } .sC986E16F { font-family:Arial; color:#ffffff } .sA7186A9D { width:36.9pt; font-family:Arial; display:inline-block } .s7B051ABA { width:164.45pt; font-family:Arial; display:inline-block } .s1E019DFF { width:46.56pt; font-family:Arial; display:inline-block } .sA49AD2E4 { width:177.11pt; font-family:Arial; display:inline-block } .s39E5096F { margin-top:0pt; margin-bottom:14pt; text-align:center } .s6DB91820 { text-align:center } .sE3D764B7 { width:128.18%; margin-right:auto; margin-left:auto; border-collapse:collapse } .s57395D7A { height:28.35pt } .s8377CE32 { width:5.52%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .s2EF62ED2 { margin-top:0pt; margin-bottom:0pt; font-size:12pt } .sE1A7A04C { font-family:Arial; font-weight:bold; color:#424242 } .sAFA1961F { width:18.06%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .s7B21DC38 { width:13.28%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .sF84D9FA7 { width:16.48%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .s4157D923 { width:23.34%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .s43A25E32 { width:23.32%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top; background-color:#dfdfdf } .s30D4263B { height:27.8pt } .sF9B21C5C { width:5.52%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .s3E421A79 { width:18.06%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .s2E0FEBCE { width:13.28%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .s5DCEAEB { width:16.48%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .sE23565A4 { width:23.34%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .s67B697B1 { width:23.32%; border:0.75pt solid #838383; padding:1.4pt 5.03pt; vertical-align:top } .s580FC71E { height:0.95pt } .s1479B0A7 { height:13.85pt } .s9793A85B { margin-top:0pt; margin-bottom:0pt; text-indent:14.2pt } .sF6A12959 { width:33%; height:1px; text-align:left } .s85226119 { margin-top:0pt; margin-bottom:0pt; text-align:justify; font-size:10pt } .s653E6C45 { font-family:Arial; font-size:6.67pt; vertical-align:super; color:#0069d6 } .fixListIndent { list-style-position: inside } FIRST SECTION CASE OF MAGHERINI AND OTHERS v. ITALY (Application no. 32707/19)   JUDGMENT   Art 2 (substantive) • Use of force • Death of applicants’ relative after being immobilised and maintained in a prone position for approximately twenty minutes during police intervention • Absolute necessity of prolongation of restraint in a prone position not substantiated • Positive obligations • Guidelines in force at the time lacked clear and adequate instructions on placing individuals in a prone position to reduce to a minimum the risks to the health and life of the person concerned • Failure to train law-enforcement officers to ensure they possessed requisite level of competence when employing immobilisation techniques, such as prone position, that could pose a threat to life Art 2 (procedural) • Ineffective investigation • Non-compliance with independence requirement   Prepared by the Registry. Does not bind the Court.   STRASBOURG 15 January 2026   FINAL   11/05/2026   This judgment has become final under Article 44 § 2 of the Convention. It may be subject to editorial revision. In the case of Magherini and Others v. Italy, The European Court of Human Rights (First Section), sitting as a Chamber composed of:   Ivana Jelić , President ,   Erik Wennerström,   Raffaele Sabato,   Frédéric Krenc,   Alain Chablais,   Artūrs Kučs,   Anna Adamska-Gallant , judges , and Ilse Freiwirth, Section Registrar, Having regard to: the application (no.   32707/19) against the Italian Republic lodged with the Court under Article   34 of the Convention for the Protection of Human Rights and Fundamental Freedoms (“the Convention”) by ten Italian nationals (“the applicants”), whose details are indicated in the appended table, on 27   May 2019; the decision to give notice to the Italian Government (“the Government”) of the complaints concerning Articles   2 and 3 of the Convention; the parties’ observations; Having deliberated in private on 2   December 2025, Delivers the following judgment, which was adopted on that date: INTRODUCTION 1.     The case concerns the death of the applicants’ relative, R.M., in the course of a police intervention and the alleged ineffectiveness of the ensuing criminal investigation.   It raises issues under Article   2 of the Convention. THE FACTS 2 .     The applicants’ details are set out in the appended table. They were represented by Ms   A. Mascia and Mr   F. Anselmo, lawyers practising in Verona. 3.     The Government were represented by their Agent, Mr   L. D’Ascia, and by Mr   M. Di   Benedetto,   State Attorney. 4.     The   facts of the case, as established   by the domestic courts and submitted by the parties, may be summarised as follows. I.         The events of 3   march 2014 5 .     Between 1.00 and 1.20   a.m. on 3   March 2014, several calls were made to the emergency services in Florence reporting an individual, later identified as R.M., exhibiting signs of agitation and serious mental distress. In particular, they reported that someone was running down the street crying desperately for help. He was reported to have taken a mobile phone from someone in a restaurant and to have been shouting that he was being followed and that “something bad was going to happen” as he ran down the street. 6.     When the first carabinieri patrol, comprising Officers S.C. and D.A., arrived on the scene, they found R.M. kneeling on the pavement with his arms outstretched. 7 .     According to a timeline established by the first-instance court and confirmed at subsequent levels of jurisdiction, the subsequent sequence of events was as follows: “- At 1.21 a.m. [S.C.] called 112 by radio and said that [R.M.] had been stopped. [S.C.] was transferred to 118 and the operator [told him] that ... the ambulance [was on its way]. - at 1:21:36 a.m. Carabiniere [L.M.] called the 118 service from the operations centre asking for an intervention and reported: ‘there’s a man completely out of his mind, bare chested, screaming, there are already cars on the spot, we are trying to calm him down ... if you send them too ... he’s really out of his mind’, and the 118 operator classified the intervention as relating to a psychiatric [episode]. In the first phase, according to witnesses, the carabinieri tried to calm R.M. down verbally and then [attempted to] restrain him[.] All of them [were] standing up for [the first] few minutes;” 8 .     The evidence in the case file shows that a second carabinieri patrol, comprising Officers A.D. and V.C, arrived on the scene and intervened in the operation. The four carabinieri surrounded R.M., who tried to escape and, in resisting the officers, hit D.A. in the head. Then, according to the first ‑ instance court’s timeline: “- at 1.23 or 1.24 a.m. R.M. was placed on the ground and handcuff[ed]; at this point he was kicked two or three times by V.C.; - from 1.24 to 1.28 a.m. R.M., in a prone position and handcuffed, continued to struggle for a few minutes (according to [one] witness [for] three, four or five minutes; according to [another] witness, five minutes); - at 1.27 a.m. a third patrol was called, while Carabiniere D.A. proceeded to identify those present; - [from] 1:27:08 [to] 1:28:28 a.m. [one witness recorded] video footage, in which R.M.’s movements and desperate cries could be perceived; - then, at about 1.29 a.m., R.M. stopped speaking[;] another four minutes passed before the ambulance [staffed by Red Cross volunteers, including first responder C.M.] arrived at 1.33 a.m.; - at 1:33:38 a.m. [a member of the] 118 team called the control centre to request the intervention of an ambulance with a doctor; - at 1.44 a.m. an ambulance with a doctor arrived.” 9 .     Before the doctor reached the scene, the third patrol car with two additional officers arrived. 10 .     When the ambulance with a doctor and a nurse on board arrived the doctor recognised that R.M. was in cardiac arrest, turned him over, began performing cardiopulmonary resuscitation (CPR) and administered adrenaline. 11.     At 2.12   a.m. the doctor called the emergency services, to report that the person being rescued was in a state of cardiac arrest. R.M. was being taken to hospital by ambulance. 12.     R.M. was officially declared dead at 3   a.m. at the hospital. II.       The INVESTIGATION into r.m.’s death A.    The first steps in the investigation 13 .     At 3.05   a.m., Officers V.C. and A.D. took a witness statement (assunzione di sommarie informazioni ) from the Red Cross volunteer C.M. at the hospital, in the emergency room where R.M. had been taken. 14 .     Between 3 and 3.10   a.m. Captain C., the officers’ superior, informed the public prosecutor of R.M.’s death. They agreed to entrust the investigation to the investigations unit ( nucleo investigativo ) of the carabinieri . 15 .     At 3:10   a.m. Captain C. and Officer S.C. registered a criminal complaint and took a statement from a witness whose mobile phone had allegedly been stolen by R.M. and who had been present at the scene of the incident. 16.     At 3:30   a.m. the head of the investigations unit ordered officers on duty at the carabinieri scientific department to photograph the scene and carry out first investigative acts. 17 .     Between 4 and 7   a.m. officers from the carabinieri investigations unit took witness statements. B.    The opening of proceedings 18 .     On 3   March 2014, the public prosecutor’s office at the Florence District Court opened proceedings against persons unknown. The public prosecutor instructed the carabinieri and the State police to carry out investigative activities aimed at reconstructing in detail the events surrounding R.M.’s death. 19.     On 4   March 2014, the public prosecutor appointed an expert to carry out toxicological tests to determine whether there had been any narcotic and/or alcoholic substances in R.M.’s system when he died and, if there were, whether they could have caused or contributed to his death. The prosecutor further appointed two forensic medical experts to determine the cause of death. 20 .     On 30   April 2014, the public prosecutor’s office opened an investigation against four officers, S.C., D.A., A.D. and V.C, for offences referred to in Articles   110 and 584 of the Criminal Code relating to manslaughter committed with others ( omicidio preterintenzionale in concorso con altri ). The Red Cross volunteers and medical and nursing staff who intervened to assist R.M. were investigated for offences under Article   113 of the Criminal Code relating to cooperation in causing death by negligence ( cooperazione nell’omicidio colposo ) and Article   589 of the Criminal Code – causing death by negligence ( omicidio colposo ). C.    The forensic medical expert reports 21.     At a hearing held on 22   May 2014, an initial consensus was put forward by the medical experts appointed by the prosecutor in conjunction with the experts representing R.M.’s relatives, who identified a complex, combined mechanism of intoxication, cardiac dysfunction, and asphyxiation as the probable cause of death. 1.      The forensic expertise on the cause of death 22 .     On 11   September 2014 the forensic medical experts appointed to establish the cause of death submitted their report. According to the experts, the results of the toxicological-forensic examination indicated that R.M. had been a habitual cocaine user and that, at the time of his encounter with the carabinieri , he had been in a state of acute cocaine intoxication which had triggered a behavioural and pathological condition classified as “excited delirium syndrome”. Since the mid-1980s the relevant literature had been increasingly replete with papers and case reports of individuals exhibiting behavior attributable to this syndrome, although the term used to describe it was more recent. The joint occurrence of delirium, psychomotor agitation and excitement distinguished this syndrome from others, and among the triggering factors were the use of narcotics such as cocaine or alcohol, and the presence of a mental disorder. 23 .     Defining the clinical picture of such a syndrome was considered increasingly important, given the fatal outcomes that could result from it, to the point that in the years prior many associations of clinicians had developed guidelines to be followed in order to make accurate diagnoses, prevent deaths and improve the knowledge of both law-enforcement and health professionals on the issue. In that latter connection, the report noted that, despite the low reporting rate of deaths associated with the syndrome – partly explained by the fact that some deaths took place without witnesses – the majority of reported deaths occurred during or after intervention by law-enforcement or health professionals. It had also been noted in the literature that deaths often occurred after a period of apparent calm, during or after restraint manoeuvers by law-enforcement officers and attempts by the subject to free him or herself. This had led to growing attention being paid to the methods of arrest, immobilisation and restraint employed by law enforcement when confronted with individuals in such a state. The report mentioned guidance issued by some associations of physicians as of 2009, which included a recommendation that excited delirium syndrome be viewed as a medical emergency and that the police and healthcare services be alerted simultaneously. Another recommendation when dealing with such subjects was that the use of force should be kept to a minimum and medical intervention should be provided as quickly as possible to limit the release of catecholamines and the development of metabolic acidosis, both of which were important factors leading to the death of individuals experiencing excited delirium. 24.     The report then reviewed studies investigating the particular role that immobilisation in a prone position might play in leading to the death of individuals experiencing the above-mentioned syndrome. Some of those studies showed that the prone position had interfered with respiratory mechanics, while other studies stated that the evidence of the causal connection was less certain. In certain cases where the heart had already been weakened by drugs such as cocaine, this had led to fatal consequences. In other studies autopsies had revealed signs of asphyxia in the causes of deaths, referred to as “positional asphyxia”, arising from breathing difficulties when individuals were immobilised in certain positions. In that context, the report noted recommendations on restraint in United Kingdom prisons issued in 2005, which stated, inter alia , that for agitated or violent subjects the prone position should be avoided unless strictly necessary, that any form of chest compression should be avoided, and that in the event of a medical emergency the prisoner should be released immediately to allow medical staff to provide timely assistance. 25 .     Turning to the manner of R.M.’s death, which was recorded as cardio ‑ circulatory arrest, the experts noted that the circumstances were consistent with those reported in the literature: the death had occurred after the subject had been immobilised by law enforcement and after a period of apparent calm. An electrocardiogram (ECG) performed on R.M. showed asystole, in which cocaine intoxication had undoubtedly played an important causal role. Another element of undoubted importance was that the dynamics of the arrest and subsequent immobilisation and R.M.’s struggle to free himself had induced an additional release of catecholamines. There was also evidence of asphyxiation. The experts could not establish whether and to what extent direct compression of R.M.’s neck, reported by witness testimony and supported in the evidence by a faint but bilateral haemorrhage in the lateral cervical region, had contributed to the asphyxia. However, the experts were certain that the asphyxia was relatable to R.M.’s having been held in a prone position for a prolonged period, which reduced respiratory dynamics. In summary, R.M.’s acute cocaine intoxication, the stress caused by his immobilisation and attempts to free himself, and the prone position in which he had been held were all factors that had contributed, jointly, to his death. 26 .     With regard to the time of death, it was noted that, according to the chronology reconstructed by the investigating authorities, R.M. appeared to have been alive at 1.33   a.m., when the first ambulance arrived. One of the first responders, who had approached R.M. to check his vital signs, put her hand under his mouth and apparently felt him breathing. The experts noted the unscientific nature of the manoeuvre, which they found to have been undoubtedly incorrect, and concluded that it was likely that the R.M. had already been in an unconscious state at that stage, as no words or movements were reported. Between 1.35 and 1.44   a.m. R.M. went into cardiorespiratory arrest. The doctor who arrived at 1.44   a.m. had attempted to resuscitate him, but to no avail. R.M. had already been dead when he arrived at the emergency room and the death, or at least the irreversible phase that preceded it, had occurred between 1.35 and 1.44   a.m. The experts stated that the asphyxiation would in all likelihood not have occurred if R.M.’s respiratory activity had been facilitated by placing him in a supine, seated or standing position. They also noted that the first responders, based on the available evidence, had limited themselves to establishing the presence of vital signs by unscientific means, without taking any measures to facilitate R.M.’s respiratory dynamics. 27 .     They concluded that the prolonged maintenance of R.M. in a prone position, with the resulting asphyxia, in view of his concurrent state of intoxication and adrenergic dynamics, had reduced the chances of recovery and thus of avoiding death. They emphasised that the mortality rate of cardiac arrest is high and that the chances of success of CPR with a defibrillable rhythm decreased by 7-10% with every minute that passed without intervention. It could be concluded with near certainty that the lethal outcome of the situation had already been determined by the time of the doctor’s intervention. 2.      The toxicology report 28.     On 5   January 2015   the toxicology report prepared by the expert appointed by the prosecutor was submitted. It found that R.M. had consumed a significant quantity of cocaine, resulting in acute intoxication at the time of the impugned events. This finding could have been indicative of toxic psychosis and a state of “excited delirium”. III.     The criminal proceedings A.    Committal for trial 29 .     On 11   November 2014 the public prosecutor’s office requested the committal for trial of the four carabinieri officers and the three Red Cross volunteers for causing death by negligence under Article   589 of the Criminal Code ( omicidio colposo ) and Article   113 of the Criminal Code relating to cooperation in causing death by negligence ( cooperazione nell’omicidio colposo ). In the charge relating to the officers, the negligent conduct was identified as having maintained R.M. in a prone position following his immobilisation and handcuffing, which reduced his respiratory activity. 30 .     On 21   November 2014 the preliminary investigations judge of the Florence District Court set the preliminary hearing for 8   January 2015. On that date the applicants joined the proceedings as civil parties. 31.     On 3   February 2015 the preliminary investigation judge, upon the confirmation of the indictment issued by the public prosecutor, ordered that the seven accused be committed for trial. V.C. was also committed for trial for causing bodily harm aggravated by abuse of power. B.    Proceedings in the Florence District Court 32.     By judgment of 13   July 2016 the Florence District Court found Officers V.C., S.C. and A.D. guilty of the offences under Articles   113 and 589 of the Criminal Code for having, in complicity, caused the death of R.M. on 3   March 2014 by cardiorespiratory arrest brought about by acute cocaine intoxication associated with asphyxia. In particular, after having immobilised and handcuffed R.M. they had held him to the ground from 1.30 to 1.45   a.m. in a pose likely to reduce his respiratory dynamics. As a result, the Florence court sentenced V.C. to eight months’ imprisonment and S.C. and A.D. to seven months’ imprisonment. The court acquitted D.A. and the Red Cross volunteers. 33 .     The court began by examining the facts of the case, tracing the sequence of events that had occurred on the evening at issue and the circumstances that had culminated in the intervention of law enforcement. Those included R.M.’s having grabbed a taxi driver by the neck saying things such as “you’re one of them too, you can’t do this to me”, R.M.’s having thrown around road signs, harassing a passer-by, attempting to stop and get into passing cars, shouting at the top of his voice with his chest exposed, entering a restaurant and taking someone’s mobile phone, and then breaking the glass door of a restaurant that he had entered. The court then established a timeline of the events that followed (see paragraph   8 above). 34 .     As to the cause of death, following a review of the medical experts’ reports submitted during the trial, the court concluded that R.M.’s death by cardio-circulatory arrest could be attributed to a combination of factors, each of which had contributed to a state of catecholaminergic stress, and that those were: acute cocaine intoxication; immobilisation by the police in an attempt to restrain him and, on the other hand, his attempts to free himself both before and after immobilisation; and the prone position in which he had been held, albeit without any thoracic compression, from the time he was handcuffed. 35 .     With regard to the reasons for the use of force by the officers, the court considered that R.M.’s conduct had to be taken into account in order to define the context in which the officers had acted. It was clear from numerous eyewitness accounts that the police intervention had stemmed from the need to stop him in order to ensure his safety and that of the public, and also to identify him, as he was suspected of having committed offences for which arrest could have been envisaged. The court thus found that the handcuffing and immobilisation on the ground of R.M. carried out by the carabinieri had to be considered justified. Once handcuffed, it had also been logical and correct that they had maintained him in that position until he had calmed down, for a period of time that common sense would suggest was appropriate to his condition.   However, the Court considered that – although the carabinieri had been justified in having handcuffed R.M., in placing him in a prone position and in keeping him in that position for a period of about two minutes (from 1.29 to 1.31   a.m.) necessary for observation – their keeping him in that prone position for longer than that and even after he had fallen silent had been unlawful. 36 .     The court considered that the basis for its assessment should be the behaviour expected of a “model police officer” who had undergone training based on the knowledge disseminated and in use at the time of the event. The handcuffing and prone positioning procedures used on R.M. were in accordance with a circular issued in 2008 (see paragraphs 59-61 below) relating to immobilisation and handcuffing. The court then cited another circular that had been issued on 30   January 2014, which included references to the prone position (see paragraphs 62-65 below), but found that that it had not yet entered into force at the material time and that no training had been delivered to the officers with respect to that circular. It noted that, during his testimony, Captain C. had acknowledged that the latter guidelines dealt with the risk of positional asphyxia associated with the prone position, which, prior to the date of issue of the 2014 circular, had been the subject of general knowledge but had not been mentioned in previous regulatory texts or been the subject of specific training. Nor could it be argued that the defendants should have relied on their own personal knowledge, because that would have meant allowing officers to disregard orders, with the consequent risk of arbitrary conduct. The court therefore ruled out the possibility of liability for the breach of written precautionary rules. At the same time, the court emphasised that the officers had not employed any thoracic compression or exerted pressure on any other parts of R.M.’s body in excess of the limits established in a scientific study submitted by the civil parties. The court noted in that connection that the asphyxia identified by the medical experts had not been caused by compression but had rather been positional. 37 .     Nevertheless, the court determined that a negligent failure to act ( colpa di tipo omissivo ) was at issue. Such a conclusion could be drawn from the holding of R.M. in a prone position by the officers even after R.M. had entered a state of calm – that is after 1.31   a.m. – in particular in view of the absence of vital signs, which required the officers to have intervened or to have called the emergency services instead of passively waiting for the ambulance with the doctor to arrive at 1.44   a.m. Their actions, in the court’s view, indicated an “obstinate conviction” that although R.M. had become motionless and silent he still needed to be restrained, regardless of his actual behaviour/condition. The offence was constituted by their failure to act promptly to change the position in which R.M. was being held, since the medical experts found that the asphyxiation had been linked to the officers holding him in a prone position. The officers had had a duty of care towards R.M. once he had been handcuffed and restrained, particularly because they themselves had created a source of danger. The lawful alternative behaviour which would have been warranted in the court’s view would have been to ask questions or move R.M., thus making it possible to detect his loss of consciousness and to intervene by placing him in a supine or sitting position, unfastening his handcuffs, allowing the ambulance personnel to monitor him and, if necessary, to administer CPR. Had they acted thus there was a high logical and statistical probability that the incident would not have occurred. As indicated by the medical experts appointed by the public prosecutor, the chances of success of CPR decrease by 7-10% with every minute that passes without intervention. 38 .     The court also examined the allegations, raised by the civil parties to the proceedings, that attempts had been made to mislead the investigation and manipulate evidence. It found that V.C. and A.D.’s conduct had not been unlawful. They had legitimately obtained information from the Red Cross volunteer C.M. as the events had just taken place, they had not yet been under investigation, their intervention had been ongoing and they had had to be proactive to assist with the reconstruction of what had happened. During the trial, the officers’ hierarchical superior, Captain C., reported that he had learned that V.C. and A.D. had taken statements from C.M. in the emergency room on their own initiative. That was also considered legitimate, given that at the time they had still been unaware that the investigation had been assigned to the investigative unit of the carabinieri . The court did not find evidence that the volunteer C.M. had been coerced into providing a favourable version of events. She had subsequently been heard on two further occasions, in normal and more reassuring circumstances, having had time and the opportunity to clarify the events in her mind, without changing her version of events, at least in terms of certain essential aspects. Consequently, there was insufficient evidence to substantiate the assertion that the statements made by her in the immediate aftermath of the event had been influenced by the carabinieri . The court considered that, in any event, her subsequent status as a defendant had influenced her overall reliability, as it could not be discounted that C.M. had progressively reinforced elements intended to justify her actions in subsequent interviews. Lastly, the court noted that C.M. was an experienced first responder, trained in CPR and the use of a defibrillator, which she had previously used in other rescue operations, and was therefore trained to operate in emotionally stressful situations, even in the presence of deceased persons. There was no reason why the carabinieri should have questioned her psychological stability at the moment of the interview at the hospital. As regards the witness whose statement had been taken by Officer S.C. and Captain C., his statement was found to be free from all suspicion of manipulation as he had never changed his account of the events. The court noted that the officers had also searched certain locations frequented by R.M. for evidence related to the event, which it considered to have been an entirely legitimate action on their part, as otherwise they could have been accused of malicious omission if they had not made any effort at all to gather such evidence. 39.     D.A. was acquitted because at the time when his three colleagues were involved in the impugned acts he had left the scene to have his wound treated and to identify witnesses. 40 .     The Red Cross volunteers were also acquitted. Eyewitness statements and the defendants’ own accounts enabled the circumstances to be reconstructed unequivocally. This included the volunteers’ attempts to apply a saturation meter on R.M.’s finger, which once applied produced no results, and their requests to the officers to place R.M. in a supine position for more thorough monitoring. The officers refused, citing R.M.’s dangerousness. In the face of opposition from the officers, vital signs were checked by placing a hand under R.M.’s mouth. Amongst other things, the court noted that under Red Cross protocols rescuers can only intervene in a safe scenario. In the event that the situation is deemed to be unsafe, the rescuers have to request support from the operations centre. In this particular instance, it was clear that law-enforcement officers had classified the scenario as unsafe. The court decided to acquit them on the basis that the assessment of the risk within which they were forced to operate was not their responsibility. 41.     With regard the kicks given to R.M. by one of the officers, the judge ruled that that part of the case, which involved the offence of causing bodily injury, could not be examined owing to the absence of a criminal complaint. C.    Proceedings before the Florence Court of Appeal 42.     By a judgment of 19   October 2017 the Florence Court of Appeal upheld the conviction of V.C., S.C. and A.D. and the acquittal of D.A. and the Red Cross volunteers. 43 .     The court agreed with the assessment of the cause of death formulated by the medical experts, according to which R.M.’s death could be attributed to three causes that had brought on catecholaminergic stress, namely acute cocaine intoxication compounded by his attempts to free himself from immobilisation and the prone position in which he had been kept. That position resulted in a limitation of R.M.’s respiratory activity and, consequently, his oxygen supply, which was particularly critical given his pre-existing intoxication. 44 .     As to the justification for resorting to the use of force, the officers had intervened following numerous phone calls to the police. In the court’s view it must have appeared very likely, because of R.M.’s disturbed state and the number of violent acts against people and property reported up to that point, that he could continue to commit similar acts. Accordingly, in apprehending R.M., the carabinieri had been carrying out their duty of maintaining public safety, and thus ensuring that he was stopped and did not escape their control, to prevent him from harming not only others but himself as well. 45 .     The court first noted that the evidence showed that R.M had been in a severe state of physical and mental agitation when confronted by the carabinieri . Attempts to calm him down verbally had been useless: the carabinieri had initially approached him with their hands raised above their heads and tried to establish dialogue, but he had kept screaming, appeared delirious, and would not keep still. Four carabinieri had stood around him in a semicircle and followed him as he walked away. There was then what the court described as inevitable physical contact to try to stop him and attempt to handcuff him, which he had resisted, hitting one of the carabinieri in the process. The carabinieri , therefore, had had no other alternative but to apply force to immobilise him once the attempt to calm him down by intervening with raised hands and trying to talk to him had failed.   Even so, the court held that, considering R.M.’s evident pathological condition, the use of force had to have been kept to a minimum and that constant monitoring of R.M.’s condition had been imperative. 46 .     However, the officers’ keeping of R.M., who had been in a seriously disturbed state, immobilised in a prone position had prevented him from receiving the oxygen he needed, which had proved to be lethal. The officers had failed to consider that R.M.’s behaviour after immobilisation – vocally expressing discomfort and asking for help, then gradually losing strength and speaking with a progressively weaker voice, and finally becoming unresponsive and falling silent – could have been a sign of breathing difficulties. The court found it significant that after R.M. had stopped moving one of the Red Cross volunteers who had arrived in the meantime had raised his eyelid without provoking any reaction on R.M.’s part, when such a reaction should have been immediate and instinctive. R.M.’s state of calm could therefore have been interpreted as a loss of consciousness, and that consideration should have obliged the officers to change his position. Furthermore, the court found no justification for keeping R.M. immobilised in a prone position when there were four officers present who could have restrained his hands and feet in a different position while still ensuring his safety and that of others. Despite that, R.M. had been kept immobilised in a prone position even after he had fallen still, as confirmed by several witnesses, including one of the police officers involved, albeit without any proven compression to the thoracic, scapular or back area. 47.     The court of appeal found that it had not been necessary for the officers to have had any medical knowledge to know that the prone position made breathing difficult even without direct compression to the chest owing to the compression of the front part of the pulmonary apparatus, especially when the body was in contact with a hard surface such as a road with the arms handcuffed behind the back and the legs pinned down. 48.     Nevertheless, the officers had kept him in the prone position, carelessly neglecting to consider that the obvious breathing difficulties R.M. was experiencing entailed a risk of asphyxiation, which had been, in the court’s view, foreseeable in the circumstances. The duty of care towards the individual under their authority and whose safety they were bound to protect had obliged them to consider that risk. Furthermore, the court reiterated that use of physical coercion had to meet the requirements of necessity and proportionality, and could be considered justified only to the extent that it was strictly necessary to achieve immobilisation. 49 .     In relation to the actions of the carabinieri in the immediate aftermath of the incident and the alleged manipulation of evidence, the court conceded that the fatal outcome of the police operation could, in all likelihood, have provoked significant worry on the part of the officers. Nevertheless, it found that the contested actions had not resulted in any detectable alteration of evidence. 50 .     In upholding the acquittal of the Red Cross volunteers, the court relied on the fact that their testimony had been coherent and consistent and that the officers had told them that R.M. was very agitated, aggressive and dangerous. The court noted that when one of the volunteers, C.M., had asked if they could change R.M.’s position in order for her to monitor him more closely, the request had been denied. That had been confirmed by Officer   V.C.’s testimony – he had told her that R.M. was dangerous despite his lack of movement. An officer who had arrived in the third patrol car (see paragraph   9 above) also reported that, when one of the volunteers approached R.M. to attach a saturation meter to his finger, another officer had warned her of the extreme agitation R.M. had exhibited up to that point, adding that he could not rule out the possibility of R.M.’s posing an immediate threat. The court observed that, in the meantime, Officer A.D. had been holding R.M.’s ankles and Officer V.C. had been holding his handcuffed wrists. D.    Proceedings before the Court of Cassation 51 .     By a judgment of 15   November 2018, deposited in its registry on 29   November 2018, the Court of Cassation reversed the judgment of the Florence Court of Appeal on the ground that the acts or omissions at issue did not constitute an offence. In particular, the court considered that the officers’ conviction was deficient in terms of the subjective element of the offence and the concrete foreseeability of the harmful event. This resulted in the officers’ acquittal. 52 .     The Court of Cassation first acknowledged that the lower courts had established and analysed at length all the factual elements on which their conclusions were based. It considered sound the reasoning on the basis of which the lower courts had found the intervention of the officers to have been legitimate. It reiterated the factual findings as to the dangerous conduct displayed by R.M. and his severe psychophysical agitation. According to the well-reasoned conclusions reached by the lower courts, the officers had had no alternative but to stop R.M. because he had appeared to pose a danger to himself and others, especially given that their attempts to calm him down had failed, that R.M. had attempted to flee and that he had hit an officer. 53 .     It noted that the determination of the three-fold cause of R.M.’s death was correct. His death had been caused by catecholaminergic stress resulting from the combination of three factors: acute cocaine intoxication; R.M.’s immobilisation and attempts to free himself both before and after immobilisation; and the prone position in which he was held, albeit without any chest compression. The judgment being appealed against did not contain any defect of reasoning regarding the existence of a causal link between the protraction of the prone position in which the R.M. had been held and the fatal outcome. 54 .     The court then examined the methods used by the officers to carry out the intervention and considered whether they could be considered excessive. The court considered the decision to introduce new guidelines for officers in January 2014 (see paragraphs 62 - 65 below) clearly indicated that there had been a need to impart new knowledge to law-enforcement personnel on safety risks when confronted with persons in a state of psychophysical agitation caused by illness or consumption of alcohol and/or narcotic substances, both for the individual involved and the officers. Such i