[Item selection for perinatal maternal health literacy scale]

To establish a more reliable and accurate scale of perinatal maternal health literacy during perinatal period by selecting items form existing scale.

Two rounds of inquiry were performed by 14 experts on mother and children to evaluate the necessary and importance of 56 items by Delphi method, in which 50 items were retained. Th en we proceeded a cross-sectional survey in 350 woman who just gave birth 1-3 days before. 10% of them were selected to retest aft er 1 week. Based on these data, we used 6 different methods to select items and kept those that could pass by at least 3 different methods. The methods used in data analysis were Chi-square test, correlation coefficient method (2 kinds), factor analysis, Cronbach α coefficient method and the retest reliability method.

The Person correlation coefficient was 0.507 (P=0.004). By using the 6 statistical methods, we deleted 9 items through Chi-square test, 25 items through correlation coefficient method 1, 1 item through correlation coefficient method 2, 19 items through Cronbach α coefficient method, 8 items through factor analysis and 37 items through retest reliability method. In the end, 33 items were retained for the novel scale of maternal health literacy during perinatal period.

Simplified novel scale is acquired, which need to do large efforts before extensive use, such as large sample survey, reliability and validity test.