Interest and interpretation of hair analysis in mistreatment of hospitalized older people

Objective To identify the cause of the disappearance of some psychotropic drugs in the pharmacy of a geriatric long-term care unit. Methods From July 2017, many cases of missing drugs were reported by the day nurse of one geriatric long-term care unit, including midazolam vials while no patient had such a prescription. A misuse by an unidentified nurse was suspected. In order to exclude mistreatment, 31 blood samples were taken without the knowledge of the staff during the first week of November 2017 at the end of the night care in the 7 patients with the most frequent agitation crisis. Given the results, it was proposed to perform a hair testing in 5 of those patients. After strengthened safety measures to the dispensation of benzodiazepines in the unit, a second blood testing was conducted in February 2018 in 18 patients from two adjacent geriatric units. Hair testing was also performed in March and October 2018 to evaluate the evolution. Segmental hair analysis was performed when the length of the lock allowed it (≥ 3 cm). After decontamination and liquid/liquid extraction, screening and quantification of psychotropic drugs (anxiolytics, antidepressants, hypnotics, antiepileptics, opioids, other sedatives) were performed by triple quadrupole TSQ Endura (Thermofisher®). Results In the first blood analysis, one of the 7 patients was found positive for midazolam (0.9 ng/mL) and OH-midazolam (0.6 ng/mL). This patient was found initially negative with a TSQ Access MAX triple quadripole (Thermofisher®), significantly less sensitive than the TSQ Endura, showing the interest of sensitive devices in these cases of mistreatment. According to this result showing the administration of a non-prescribed molecule in this patient, hair testing performed in 5 of those patients showed the presence of non-prescribed midazolam in 3 of them. The concentrations were very low for 2 (between 1 and 5 pg/mg) probably in accordance with single administration and was found higher for the 3rd in accordance with more regular administration (40 to 120 pg/mg). Were also found zolpidem (4 times), carbamazepine (4), clobazam (3), lamotrigine (2), tramadol (2), nefopam (2), zopiclone (1), alprazolam (1), levetiracetam (1), citalopram (1), venlafaxine (1), codeine (1), each patient having between 2 to 9 non-prescribed molecules in his hair. From the 18 blood samples analyzed 3 months later in February 2018, 3 were still positive to a non-prescribed medicine, 1 zopiclone (while this molecule was no longer available in the unit), 1 clobazam (not available) and 1 paroxetine (still available for the nurse staff). The second hair analysis performed in March 2018 on the 5 same patients analyzed in November showed the disappearance of all the benzodiazepines and “Z-drugs” (midazolam, alprazolam, zolpidem, zopiclone) except clobazam present in 2 patients (already present in November, both with decreased concentrations). Carbamazepine (4), lamotrigine (1) and levetiracetam (1 but not the same patient than in November) were still present and two antidepressants appeared [mirtazapine (2), venlafaxine (1), both at low concentrations], with a possibility of administration mistake since patients in the same room received these treatments. After reporting to the prosecutor, press conference carried out by the general manager of the AP–HP to secure the delivery of all psychotropic drugs in the geriatric departments of our Institution, and suspension of a suspected nurse, the 19 patients collected in October 2018 for the third hair analysis showed the absence of non-prescribed psychoactive molecules in the 1 or 2-cm proximal segment, analyzed to avoid any radial diffusion of molecules from distal segment in these very elderly subjects. Conclusion This study shows the interest of a very sensitive toxicological analysis and the reliability of hair testing in these cases of hospital mistreatment, although its interpretation can be delicate considering possibility of drug administration mistakes.