Summaries of the initial scale

The research group categorized the initial index pool into three dimensions: knowledge, practical skills, and leadership. The preliminary index pool comprised three primary items, 14 secondary items, and 81 tertiary items (Supplementary Table 1).

Scale modification via the Delphi methodDelphi round one

Twenty experts participated in the first round from 24 experts we initially contacted, yielding a response rate of 83.3%. The majority were male (13/20, 65%), aged over 50 years (14/20, 70%), working in CDCs or research institutions (19/20, 95%), and having a background in public health and preventive medicine (18/20, 90%). 95%(19/20) of the experts held a master's degree or above, 90% (18/20) possessed senior professional titles, and 85% (17/20) had at least 20 years of work experience (Table 1). The average Cr values for the three primary dimensions were 0.935, 0.933, and 0.908, respectively (Supplementary Table 2), representing a high authority.

Table 1 Characteristics of Delphi experts

Across all indicators, experts’ mean importance scores ranged from 3.32 to 4.95, with CV values ranged from 0.04 to 0.24, and full mark proportions ranging from 0 to 95% (Table 2). According to the critical value method and experts’ suggestions, one secondary item and three tertiary items were deleted, three secondary items and 30 tertiary items were revised or merged, and one primary item, two secondary items, and 11 tertiary items were newly added (Supplementary Table 3). A new primary item named "Personal Quality" was introduced, encompassing two secondary items (professional qualifications and professional quality). Kendall’s W was 0.285 (P < 0.001), indicating an acceptable coordination among experts (Supplementary Table 4). The revised scale after the first Delphi round comprised four primary items, 15 secondary items, and 73 tertiary items (Supplementary Table 5).

Table 2 Experts’ scores in the first round of DelphiDelphi round two

Twenty experts participated in the second round from 27 experts contacted, resulting in a response rate of 74.1%, with similar basic characteristics to those in the first round (Table 1). The average Cr values of four primary dimensions were 0.938, 0.925, 0.933, and 0.945, respectively (Supplementary Table 6), representing a high authority.

Across all indicators, experts’ mean importance scores ranged from 3.85 to 4.85, with CV values ranged from 0.07 to 0.18, and full mark proportions ranging from 10 to 85% (Table 3). Based on the critical value method and experts’ suggestions, three tertiary items were deleted, two tertiary items were revised, four tertiary items were merged into two, and one tertiary item was newly added (Supplementary Table 7). Kendall’s W was 0.192 (P < 0.001), indicating an acceptable coordination among experts (Supplementary Table 8). The revised scale after the second Delphi round comprised four primary items, 15 secondary items, and 69 tertiary items (Supplementary Table 9).

Table 3 Experts’ scores in the second round of DelphiEvaluation and validation of the scaleParticipants’ characteristics

A total of 155 subjects were enrolled in the pilot survey for item analysis and EFA, and 476 subjects were enrolled in the formal survey for CFA (Table 4). Age, gender, education level, years of work related to infectious diseases, position, and postgraduate supervisor qualification did not differ statistically between subjects in the two surveys. In contrast, major of bachelor/college, professional title, job type, monthly income, experience in prevention and control of infectious diseases, and number of participations in outbreak response varied between the groups.

Table 4 Basic characteristics of participants in the pilot survey and formal surveyItem analysis and EFA results

The group with the lower 27% of total scores had significantly lower mean scores than the group with the upper 27% for all indicators (Supplementary Table 10), indicating a high degree of discriminant validity. Cronbach’s Alpha values of the four primary dimensions exceeded 0.893 in the pilot survey (Supplementary Table 11). The corrected item-total correlation coefficients between individual items and the overall scale score ranged from 0.33 to 0.78, with all P < 0.001 (Supplementary Table 12). Cronbach’s Alpha of the total scale remained stable and exceeded 0.98 when any individual item was deleted.

The criteria for conducting EFA were met because the KMO values of four primary dimensions were 0.940, 0.929, 0.943, and 0.868, respectively (all P < 0.05). The cumulative variance explanations for the four dimensions were 77.463%, 73.976%, 81.174%, and 68.654%, respectively. Three rounds of EFA and item adjustment were conducted to ensure all items fitted the preset theoretical frame (Supplementary Table 12−20). In summary, four tertiary items were deleted based on item analysis or EFA, two secondary items and two tertiary items were merged based on item analysis or EFA, and one tertiary item was moved to another secondary item based on EFA (Supplementary Table 21). The revised scale after item analysis and EFA comprised four primary items, 14 secondary items, and 64 tertiary items (Supplementary Table 22).

CFA results

Cronbach’s Alpha values of the four primary dimensions exceeded 0.898 in the formal survey (Supplementary Table 11). All CR values and AVE values exceeded 0.8 and 0.5, respectively, representing a satisfactory convergent validity (Table 5). The standardized factor loading of the items ranged from 0.630 to 0.977. The path diagrams with standardized factor loadings for the four dimensions are shown in Supplementary Figure 1 to 4. Good discriminant validity was observed in each dimension, with the square root of AVE exceeding the correlation coefficients between factors (Supplementary Table 23−26). Among the seven model fit indices, all four dimensions satisfied at least five criteria, indicating a good fit effectiveness in CFA (Supplementary Table 27).

Table 5 Confirmatory factor analysis and convergent validity of the scaleWeight determination of each indicator through AHP

The expert judgment matrixes for the three hierarchies are shown in Supplementary Table 28−46. The consistency ratios were less than 0.1 for all the matrixes, indicating good consistency in each. The combined weight for the four primary items were 0.4064 for Knowledge, 0.2878 for Practical skills, 0.2082 for Leadership, and 0.0981 for Personal quality (Table 6). The secondary items with the top three combined weights were Knowledge of infectious diseases (0.1993), Knowledge of public health emergency management (0.1267), and Public health response to infectious diseases (0.1199). The tertiary items with the top five combined weights were Basic epidemiological knowledge of infectious diseases (0.0823), Criteria for judging aggregates epidemics and outbreaks of infectious diseases (0.0486), Mastering knowledge and skills of personal protection (0.0486), Command, coordination and control of health emergency (0.0413), and Analyzing the situation of public health incidents and proposing targeted prevention and control measures (0.0410). The final versions of the IDCCS in English and Chinese are presented in Supplementary Table 47.

Table 6 The weight values of final three-level scale