Multiple Correspondence Analysis of the Influence of Heterogeneity in Sample Composition on the Investigation Results of Primary Chronic Disease Management Quality

Background The assessment of chronic disease management capabilities in healthcare institutions is an important element of chronic disease management, but the researchers did not pay attention to the impact of sample heterogeneity on the scientific validity of survey results when using relevant scales, which requires our attention. Objective To explore the impact of the heterogeneity of sample composition on the evaluation results of the scale and put forward countermeasures based on the investigation of chronic disease management quality in primary medical institutions in Chengdu. Methods In February 2022, a total of 889 medical workers was selected from 46 primary care institutions in 23 prefectures and districts (counties) of Chengdu City by using multi-stage stratified cluster sampling method. The Chinese version of Assessment of Chronic Illness Care (ACIC) was used to assess the quality scores of chronic disease management in the medical institutions. Multiple correspondence analysis was used to explore the impact of sample heterogeneity caused by different genders, years of experience, professional titles, educational levels and job composition on the scale scores. Results The total score of chronic disease management competence of medical workers was negatively correlated with their education level and positively correlated with their professional title (P<0.05). The decision support dimension score of chronic disease management competence was negatively correlated with the working years of subjects (P<0.05), the information system dimension score of chronic disease management competence was negatively correlated with working years and positively correlated with education level (P<0.05). Multiple correspondence analysis showed that the higher educational level was correlated with lower total score of chronic disease management competence, intermediate titles tended to give the highest scores. For the decision support scoring dimension, medical workers with more working years tended to give higher scores. For the information system scoring dimension, medical workers with more working years tended to give higher score, and those with college/bachelor degree tended to give the highest score. Conclusion The heterogeneity of sample composition will have a definite impact on the results of the primary chronic disease management quality evaluation scale, which means that the evaluation will biased, especially not conducive to fair inter-institutional comparisons. The above suggests that the development of the scale needs to strictly regulate the scope of application. The reliability and validity test of the scale and its promotion need to attention to the composition of the sample to ensure its homogeneity, so as to ensure the objectivity and validity of the quality evaluation of primary chronic disease management.