Developing and Validating an Item Bank for Alcohol Use Disorder Screening in the Chinese Population by Using the...

No prior work had resolved how to efficiently screen for alcohol use disorder severity within the Chinese population using modern digital assessment tools. Traditional paper-based questionnaires often impose a significant burden on respondents while failing to provide precise measurements across the entire spectrum of severity. That uncertainty drove the need for a more dynamic approach to clinical evaluation. Prior research has shown that item response theory provides a robust framework for developing adaptive assessments in various psychological domains. However, existing screening instruments frequently lack the necessary psychometric rigor to distinguish between varying levels of risk effectively. This gap motivated the development of a specialized item bank aligned with current diagnostic criteria. Researchers sought to leverage computerized adaptive testing to improve both the speed and accuracy of diagnostic screening. Establishing a validated tool is a prerequisite for enhancing clinical outcomes in diverse populations.

The aim of this study was to build an item bank for alcohol use disorder screening and derive a computerized adaptive testing version. Researchers sought to address the need for an effective and accurate method to detect individual severity levels. This project was motivated by the limitations of traditional screening tools that often lack precision. The team focused on creating a system that aligns with current diagnostic criteria for the Chinese population. By developing this adaptive instrument, they intended to reduce the burden on participants during the assessment process. The study addresses the challenge of maintaining high diagnostic accuracy while minimizing the number of questions asked. Establishing a validated item bank is essential for improving clinical screening practices. This work provides a foundation for more efficient behavioral health evaluations in diverse settings.

Review approach involved building an item bank from existing Chinese questionnaires based on established diagnostic criteria. The team recruited 915 participants representing both healthy and high-risk groups to complete the initial survey. Analysts evaluated the data using unidimensionality tests to ensure the items measured a single construct. They assessed item fit and discrimination parameters to refine the bank according to psychometric standards. The researchers applied differential item functioning analysis to identify and remove biased questions. They utilized the graded response model to structure the final item bank for adaptive administration. A simulation study followed, using the collected data to determine the reliability and validity of the new instrument. This approach allowed for a direct comparison between the adaptive version and the full-length questionnaire.

Key findings from the literature show that the adaptive test achieved a Pearson correlation of 0.95 with the full-length item bank. The criterion-related validity of the instrument reached 0.63. Researchers found that the system provides reliable results for subjects with a theta score exceeding -0.8. The adaptive version requires an average of only 16 items to complete the screening process. Predictive utility analysis demonstrated that the new tool outperformed traditional screening questionnaires. The final item bank successfully met all required psychometric measurement standards for item response theory. These results indicate that the adaptive approach effectively identifies high-risk individuals. The study confirms that the tool accurately measures the severity of the disorder in the target population.

The authors propose that the computerized adaptive testing version of alcohol use disorder screening effectively identifies high-risk individuals. This tool provides accurate measurements of disorder severity while maintaining high reliability across the tested range. Synthesis and implications suggest that the instrument performs better than standard existing questionnaires in predictive utility. The researchers state that the final item bank adheres to the requirements of item response theory. Their findings indicate that the system functions well within the graded response model framework. The study demonstrates that the adaptive approach reduces respondent burden by requiring fewer items for assessment. Clinical application of this tool may improve the efficiency of screening processes in healthcare settings. These results support the use of adaptive testing for precise behavioral health evaluations.