Maglione Margaret A, Stone Erin G, Shekelle Paul G
Health Division, RAND, Santa Monica, California, USA.
Am J Prev Med. 2002 Jul;23(1):43-6. doi: 10.1016/s0749-3797(02)00443-9.
We reviewed published and unpublished studies of mass mailings designed to increase utilization of influenza vaccine among Medicare beneficiaries.
Search of computerized indexes for published studies and the Medicare Peer Review Organization Health Care Quality Improvement Project database for unpublished studies. Study selection criteria were: use of a controlled trial design and use of mass mailings direct to Medicare beneficiaries to increase receipt of influenza vaccinations. Study selection and data extraction were performed in duplicate by physician reviewers with consensus resolution.
Six controlled trials of mass mailings were identified. One study, published in the Morbidity and Mortality Weekly Report (MMWR), reported a modest but statistically significant improvement in influenza vaccination rates among patients who received a letter relative to those who did not. This study was cited as justification for mail interventions performed in five subsequent studies, none of which reported clinically meaningful results. None of these five studies was published.
Mass mailings have at best had clinically trivial effects on increasing influenza vaccination among Medicare beneficiaries. Publication bias has contributed to continued use of a relatively ineffective intervention.
我们回顾了已发表和未发表的关于旨在提高医疗保险受益人群流感疫苗接种率的大规模邮件研究。
检索计算机化索引以查找已发表研究，并检索医疗保险同行评审组织医疗质量改进项目数据库以查找未发表研究。研究选择标准为：采用对照试验设计，以及直接向医疗保险受益人进行大规模邮件发送以增加流感疫苗接种率。研究选择和数据提取由医师评审员重复进行，并通过协商一致解决分歧。
确定了六项大规模邮件发送的对照试验。一项发表在《发病率与死亡率周报》（MMWR）上的研究报告称，收到信件的患者的流感疫苗接种率相对于未收到信件的患者有适度但具有统计学意义的提高。这项研究被作为随后五项研究中进行邮件干预的理由引用，然而这五项研究均未报告具有临床意义的结果。这五项研究均未发表。
大规模邮件发送在提高医疗保险受益人流感疫苗接种率方面充其量只有微不足道的临床效果。发表偏倚导致了一种相对无效的干预措施仍在继续使用。
