A plea for controlled trials in medical informatics.
A plea for controlled trials in medical informatics.
复制标题
呼吁进行医学信息学对照试验。
DOI：
10.1136/jamia.1994.95236170
复制
发表时间：
1994
期刊：
影响因子：
--
通讯作者：
McDonald,CJ
中科院分区：
文献类型：
--
作者：
Tierney,WM;Overhage,JM;McDonald,CJ
Investigators from McMaster University recently reviewed all published controlled trials of practical applications of electronic medical record systems in physician decision making’and quality assurance. 2 Although these trials effectively used high-quality research methods to assess the effects of information systems on health care delivery and, in a few instances, on patient outcomes, the number of such controlled trials was shockingly small. The reasons for this paucity of high-quality controlled trials are many, with some being more obvious than others.A controlled trial is difficult to conduct, especially when the site is the clinical arena, the intervention is as unwieldy as an electronic information system, and the outcome is as difficult to define and measure as the quality of health care or its cost. Controlled trials and observational studies are extremely difficult to perform when the subjects are physicians and the intervention is information. How does one prevent bias and contamination? What kinds of physicians would allow themselves to be manipulated in a controlled trial? House officers and academicians would allow themselves to be manipulated which explains their prevalence among the subjects of controlled trials in medical informatics. The barriers to doing such controlled ‘trials in the “real world” are enormous and increasing. A research project may decrease the operational efficiency of a facility at a time when changes in health care financing require increases in efficiency. Those investigators, such as McPhee and his colleagues, 3 who have succeeded in overcoming these barriers are to be lauded.
