more so since the term used goes beyond criticism, or admonition. Even if one overlooks the legal aspect of the term (“penalise”), the expression shows not so much the fear of harshness from superiors as, even worse, the consequences on career prospects. The administration has a long memory: in the mind of the person making the statement, an “incident” is less important with regard to its immediate repercussions than for its subsequent impact in terms of promotion, choice of post etc. because of a note in the officer’s file that they are “the individual who, on one occasion…” What is the nature of these “incidents” against which one needs to protect oneself in order to avoid lasting consequences? Of course, some incidents correspond to those which can occur in custody: assault, blows, violence etc. However, these do not constitute the kind of incident in question: in such scenarios, officers are victims; there will be no consequences whatsoever in terms of “penalisation” of their subsequent careers. The possibility in question is, of course, that of flight on the part of the person with whom one is entrusted, the possibility of escape. As we know 121, there are very few escapes from French prisons. However, it is true that chances of escape are obviously greater when prisoners are removed from prison in order to appear in court and, above all, in order to go to hospital; either in areas between the vehicle and the destination (getting out of the van), or indeed inside the latter premises (for example during hospitalisation without consent called for by the prefect). Of course, the risk is small, although figures are not available: fifteen to twenty escapes for a number of removals from prison that is not recorded122. Many of those which do occur happen for reasons for which the prison staff are not responsible (breaking of a window in a hospital room for example). At the most, there probably remain between five and ten cases annually for which insufficient vigilance on the part of the officers may have contributed to the escape. This small number does not justify the precautions taken, described above. Discomfort and humiliation for tens of thousands of people and refusal of medical treatment on the part of several hundreds, are a high price to pay for the “penalisation” mentioned by the officer quoted. However this is obviously not the way that the officers concerned see things. For their part, they escort a few hundred people a year. If a single one of them runs away, their career is compromised and there results a feeling of inferiority with respect to colleagues who, for their part, will not have to endure the opprobrium of failure in a professional act. From this point of view, the officer is right. But what seems unreasonable, and is at the origin of this excess of precautions, which the Contrôleur général cannot approve, is the way in which such thinking can become established. 121 There were seven escapes in 2010; four in 2011. The rate of escapes in relation to the number of prisoners is one of the lowest in Europe. 122 The number of instances of removal from prison in order to go to hospital varies according to the number of doctors operating in the prison medical consultation and outpatient treatment unit (UCSA) (the more specialists available there, the fewer the instances of removal from prison for hospital visits) and the equipment available in the latter (a high-performance radiology apparatus decreases the number of removals from prison for the purpose of investigations by this means): in one relatively well-equipped remand prison, with around five hundred prisoners, the number of instances of removal from prison in order to go to hospital is in the region of half a thousand per year, that is to say as many removals as there are prisoners. Removals from prison in order to appear in court obviously vary according to the status of the institution: they are common in remand prisons, above all for un-convicted prisoners, and rare in prisons for definitively convicted prisoners. If by way of hypothesis, one were to estimate the number of transfers to hospital in the course of a year as being roughly equivalent to the total number of prisoners (according to the same ratio concerning un-convicted prisoners given above), the annual number of removals to hospital would be in the region of about a hundred thousand. The rate of escape of prisoners thus taken to hospital would therefore be around two in 10,000. 56