PATIENT OUTCOMES WITH ANTIBIOTIC THERAPY OF LYME DISEASE

莱姆病（LB），第二个最重要的新兴疾病和最 在美国常见的媒介传播疾病，是由感染 螺旋体，伯氏疏螺旋体。 在马里兰州，LB正在迅速蔓延 发病率逐年上升，报告病例每年翻一番。 感染会引起典型的皮疹和“流感样”症状， 慢性和复发性关节炎、心脏炎和神经系统异常。 LB的诊断和治疗都很困难，尤其是对 慢性阶段，仍然不确定和有争议。 这个项目 建议通过以下措施提高LB患者的护理质量： 研究变异导致的患者结局差异 抗生素治疗。 不适当的实践模式将是 确定和教育干预措施，以改善治疗做法 将被评估。 通过电话采访报告医生和所有LB病例 从1993年到1997年向马里兰州卫生部报告，我们将 获取有关抗生素处方、患者依从性和 药物的不良反应和不同药物的临床反应 用于治疗游走性红斑（EM）、莱姆关节炎和其它LB的方案 综合征 要衡量的结果是：疾病的改善或治愈 预防并发症。 变量是病人，疾病- (e.g.,确诊或未确诊病例、疾病阶段、疾病 综合征）和治疗相关的（例如，药物、药物组、剂量、长度 和治疗途径）。 在头三年，估计有1,200至 1，600例LB病例将被纳入研究并随访3年。 除了简单的列联表分析外， 将采用逻辑回归分析来保持混杂因素的影响 在计算风险和相对风险时保持不变。 在第4年和第5年，将进行两次医师培训干预， 评价：全地区干预措施将比较 在该州部分地区执业的医生经常收到邮件， 讲座和提供与那些在另一个国家执业的LB“热线” 地区没有这些。培训曝光 此外，随机 这项研究将比较单一干预措施对一半人的影响。 800-1,000名医生在头3年报告了LB病例， 剩下的400到500人 将评估两种干预措施的效果 通过对临床实践模式的观察， 随机化研究中的终止试验。 最后，算法将 构建以协助马里兰州医生管理患者， 怀疑LB

Lyme borreliosis (LB), the 2nd most important emerging disease and most common vector-borne disease in the USA, is caused by infection with the spirochete, Borrelia burgdorferi. In Maryland, LB is spreading rapidly and increasing in incidence with reported cases doubling each year. Infection causes a characteristic rash and "flu-like" symptoms and chronic and recurrent arthritis, carditis and neurological abnormalities. LB is difficult to diagnose and it's treatment, particularly therapy of the chronic stages, remains uncertain and controversial. This project proposes to improve the quality of care in patients with LB by investigating differences in patient outcomes resulting from variations in antibiotic therapy. Inappropriate practice patterns will be identified and educational interventions to improve treatment practices will be evaluated. Using telephone interviews with reporting physicians and all LB cases reported to the Maryland Department of Health from 1993-1997, we will obtain information on antibiotics prescribed, patient compliance and adverse reactions to the drugs and clinical responses to different regimens for treating erythema migrans (EM), Lyme arthritis and other LB syndromes. Outcomes to be measured are: improvement or cure of illness and prevention of complications. The variables are patient-, disease- (e.g., confirmed or unconfirmed cases, stage of disease, disease syndrome) and treatment- related (e.g., drug, drug group, dose, length and route of therapy). During the first 3 years an estimated 1,200 to 1,600 cases of LB will be admitted to the study and followed for 3 years. In addition to simple contingency table analysis, multiple linear and logistic regression will be employed to hold the effects of confounders constant while computing risks and relative risks. During the 4th and 5th years two physician training interventions will be evaluated: the area-wide intervention will compare outputs from physicians practicing in part of the state who receive frequent mailings, lectures and provision of a LB "hotline" with those practicing in another area not having these. training exposures. In addition, a randomized study will compare the effects of a single intervention on half of the 800-to-1,000 physicians reporting LB cases during the first 3 years with the remaining 400-to-500. Effects of both interventions will be assessed by observations on clinical practice patterns and by pre- and termination-testing in the randomized study. Finally, algorithms will be constructed to assist Maryland physicians managing patients with suspected LB.