The clinical reasoning of general practitioners is the key to providing quality medical services to patients, but it is not suitable for traditional methods to evaluate this ability because it is an abstract ability that cannot be directly observed.

This paper aims to improve the efficiency of general practitioners' clinical reasoning training and enrich the evaluation of medical education by optimizing the evaluation method of curriculum teaching.

From September to December 2022, 38 master's degree students of the Medical School of Tongji University were divided into the in-service group (16) and residential training group (22). The course consists of three modules, namely core, case and result, and adopts blended teaching method to carry out teaching. The content of quantitative evaluation is different kinds of rating scale, while the content of subjective evaluation comes from multi-source feedback.

The final course scores for the core and case modules were statistically significant for both groups (P<0.05), but the first course scores were not (P>0.05). The quantitative evaluation scores of the whole class and the two groups of students in the first and last courses of the above two modules were compared, and there was statistical significance (P<0.001). Subjective evaluation can be divided into positive evaluation and improvement evaluation. The positive evaluation of the courses on consultation skills, physical examination, auxiliary examination, health management, and chronic disease management is relatively low, while the improvement evaluation is relatively high; the positive evaluation of the courses on doctor-patient communication, SOAP, multiple diseases coexirelatively high, while the improvement evaluation is relatively low.

Programmatic assessment can enrich the evaluation system of general clinical reasoning, promote students' construction of general clinical reasoning in order, and also found that "flipped classroom" is a form of procedural evaluation.