INSERM, Centre de Recherche en Epidémiologie et Santé des Populations, UMR1018, hôpital Paul Brousse, Université Paris Saclay, Villejuif Cedex, France.
Département de médecine générale, Faculté des sciences de la santé Simone Veil, Université Versailles-Saint-Quentin-en-Yvelines, Montigny le Bretonneux, France.
Pharmacoepidemiol Drug Saf. 2020 Apr;29(4):444-452. doi: 10.1002/pds.4921. Epub 2020 Feb 17.
Antidementia drugs (cholinesterase inhibitors and memantine) are still widely prescribed despite their controversial effects and 2011 guidelines that no longer encourage their prescription. The objective was to assess which factors remained determinants of antidementia drug prescriptions.
A cross-sectional study was performed in 2013. Patients suffering from dementia, aged 65 and over, identified in the French national health insurance database were included. Because we anticipated a high correlation between age, comorbidities, and health care use, we first identified the patients' health status by a latent class analysis. Second, we performed adjusted logistic regression models. The explanatory variables were patients' health status, gender, prescription of nonpharmacological treatments (physical and speech therapies), prescription of psychotropic drugs, and access to health care.
Among the 3873 patients included, 38% received antidementia drugs. Three latent classes of patients with different health status were identified. Patients with poor health status received significantly fewer antidementia drugs (P < .001). Patients with speech therapy or antidepressant drugs received significantly more antidementia drugs (P < .001), whereas patients with physical therapy received significantly fewer antidementia drugs (P = .006).
Antidementia drugs were less likely to be prescribed for patients with poor health status. This result is encouraging for these frail patients who are more vulnerable to the adverse effects of treatments. At the same time, this result encourage targeting specifically patients in good health status for the use of a decision aid, in an attempt to limit prescriptions by involving patients and families.
尽管抗痴呆药物（胆碱酯酶抑制剂和美金刚）的疗效存在争议，且 2011 年的指南不再鼓励使用这些药物，但它们仍被广泛开具。本研究旨在评估哪些因素仍然是抗痴呆药物处方的决定因素。
本研究为 2013 年进行的一项横断面研究。在法国国家健康保险数据库中识别出年龄在 65 岁及以上、患有痴呆症的患者。由于我们预计年龄、合并症和医疗保健使用之间存在高度相关性，因此我们首先通过潜在类别分析确定患者的健康状况。其次，我们进行了调整后的逻辑回归模型。解释变量为患者的健康状况、性别、非药物治疗（物理治疗和言语治疗）的处方、精神药物的处方和医疗保健的可及性。
在纳入的 3873 名患者中，38%接受了抗痴呆药物治疗。确定了具有不同健康状况的 3 个潜在类别患者。健康状况较差的患者接受抗痴呆药物治疗的可能性显著降低（P <.001）。接受言语治疗或抗抑郁药物治疗的患者接受抗痴呆药物治疗的可能性显著更高（P <.001），而接受物理治疗的患者接受抗痴呆药物治疗的可能性显著更低（P =.006）。
健康状况较差的患者更不可能开具抗痴呆药物。对于这些身体脆弱、更易受到治疗不良反应影响的患者来说，这一结果令人鼓舞。同时，这一结果鼓励针对健康状况良好的患者，专门使用决策辅助工具，试图通过让患者及其家属参与来限制处方。
