Kashner T Michael, Rush A John, Crismon M Lynn, Toprac Marcia, Carmody Thomas J, Miller Alexander L, Trivedi Madhukar H, Wicker Annie, Suppes Trisha
Department of Psychiatry at the University of Texas Southwestern Medical Center, Dallas, TX 75390-9086, USA.
Psychiatr Serv. 2006 May;57(5):648-59. doi: 10.1176/ps.2006.57.5.648.
Disease management systems that incorporate medication algorithms have been proposed as cost-effective means to offer optimal treatment for patients with severe and chronic mental illnesses. The Texas Medication Algorithm Project was designed to compare health care costs and clinical outcomes between patients who received algorithm-guided medication management or usual care in 19 public mental health clinics.
This longitudinal cohort study for patients with major depression (N=350), bipolar disorder (N=267), and schizophrenia (N=309) applied a multi-part declining-effects cost model. Outcomes were assessed by the Inventory of Depressive Symptomatology and the Brief Psychiatric Rating Scale.
Compared with patients in usual care, patients in algorithm-based care incurred higher medication costs and had more frequent physician visits, although these differences often became smaller with time. For major depression, algorithm-based care achieved better outcomes sustainable with time but at higher agency and non-agency costs (mixed cost-effective). For bipolar disorder, patients in algorithm-based management achieved better outcomes at lower agency costs (cost-effective). For schizophrenia, patients in algorithm-based care achieved better outcomes that diminished with time, with no detectable difference in health care costs (cost-effective).
Cost outcomes of algorithm-based care and usual care varied by disorder and over time. For bipolar disorder and schizophrenia, algorithm-based care improved outcomes without higher costs for health care services. For major depression, substantively better and sustained outcomes were obtained but at greater costs.
已提出采用药物治疗算法的疾病管理系统，作为为重度和慢性精神疾病患者提供最佳治疗的具有成本效益的手段。德克萨斯药物算法项目旨在比较在19家公共心理健康诊所接受算法指导的药物管理或常规护理的患者之间的医疗保健成本和临床结果。
这项针对重度抑郁症患者（N = 350）、双相情感障碍患者（N = 267）和精神分裂症患者（N = 309）的纵向队列研究应用了多部分递减效应成本模型。通过抑郁症状量表和简明精神病评定量表评估结果。
与接受常规护理的患者相比，接受基于算法护理的患者药物成本更高，看医生的次数更频繁，不过随着时间推移，这些差异往往会变小。对于重度抑郁症，基于算法的护理随着时间推移能实现更好的可持续结果，但机构和非机构成本更高（成本效益混合）。对于双相情感障碍，基于算法管理的患者以较低的机构成本实现了更好的结果（具有成本效益）。对于精神分裂症，接受基于算法护理的患者随着时间推移实现了更好的结果，但医疗保健成本没有可检测到的差异（具有成本效益）。
基于算法的护理和常规护理的成本结果因疾病和时间而异。对于双相情感障碍和精神分裂症，基于算法的护理改善了结果，而医疗保健服务成本没有增加。对于重度抑郁症，虽然获得了实质上更好且可持续的结果，但成本更高。
