Development of datasets for basic medical information system / 浙江大学学报·医学版

<p><b>OBJECTIVE</b>To develop dataset for basic medical information system.</p><p><b>METHODS</b>All items of basic medical information system dataset were drafted by literature review, brain storm and group discussion. A Delphi study was conducted to evaluate and adjust the candidate items; then items were finalized based on comprehensive scores.</p><p><b>RESULTS</b>The final dataset developed through two-round Delphi study contained 20 first-level items, 102 second-level items, 40 third-level items and 88 fourth-level items. The response rates were 71.79% and 92.86% in round 1 and 2, respectively. The authority coefficient was both 0.77 and the coordination coefficient of the specialists' opinion was 0.239 and 0.516, respectively.</p><p><b>CONCLUSION</b>The dataset for basic medical information system determined from Delphi study is credible and can be used for development of related software.</p>

Comparison of mail and meeting forms in evaluation of Delphi study / 浙江大学学报·医学版

<p><b>OBJECTIVE</b>To evaluate and compare mail and meeting forms in evaluation of Delphi study.</p><p><b>METHODS</b>Delphi study by mail and meeting approaches was used to determine the health information dataset. Experts were required to grade the listed items through three indexes: importance, necessity and availability. Study duration, coefficient of variation of items, authority coefficient and coordination coefficient of the experts' opinion of two forms of study were calculated and compared.</p><p><b>RESULT</b>The study duration was four months through mail form and 2 days through meeting. Compared with the first round, the coefficient of variation decreased (P<0.001, all of the three indexes by two forms), and the cooperation index increased (P<0.005) in the second round. The experts' opinions were easier to be consistent through meeting than through mail(P<0.033). And the authority coefficient by meeting consultation (0.83 ± 0.05) was higher than that by mail (0.77 ± 0.03) (P=0.001).</p><p><b>CONCLUSION</b>Both mail and meeting forms of Delphi study can determine the health information dataset,but meeting consultation is better and requires shorter study duration.</p>