**4. Discussion**

The validation of the CBB questionnaire for health care personnel in nursing shows adequate discrimination of items. Cronbach's alpha for this scale was 0.089, which shows its high internal consistency.

With respect to the factor analysis, four factors were extracted from the revised model, which differed from the original structure of the Brief Burnout Questionnaire [41]. This model was proven to generate better fit of the data than the original. The percentage explained by this model was 51.86%, emphasizing the first factor, where all the items loaded over 0.55 and explained 31.77% of the variance. This factor, called job dissatisfaction, clusters indicators in two dimensions, burnout factors and burnout syndrome. This factor compiles items that refer to the balance between job expectations and reality, and how much enjoyment the individual finds in the job [21]. This coincides with the proposal made by Moreno et al. [35], in their questionnaire for evaluating professional burnout in doctors, where a factor referring to the loss of job expectations was included. Similarly, the second factor, made up of four items, groups indicators corresponding to the relationship the worker establishes with fellow workers and superiors at work. This factor, which is called social climate, responds to a cluster that may be due to the importance in developing burnout of chronic stressful interpersonal situations in the workplace [6]. The third factor grouped seven items, which in the original questionnaire were in the burnout syndrome scale, except for one, which was on the consequences scale. The cluster of these items is called the personal impact factor and refers to the direct consequences exhaustion has on different areas of an employee's life.

Finally, the fourth factor, called motivational exhaustion, combines five items that in the original model were part of the burnout syndrome scale. The questions grouped under this factor of motivational exhaustion refer to the absence of job growth and stimulation for development in the job position. These are aspects that promote work demotivation, generate distress [17], and are among the challenges health care personnel face most frequently [18].

Although this four-factor model showed adequate fit, after making the corresponding respecifications according to theoretical and statistical criteria, items 2, 16, 3, 13, 17, and 11 were eliminated, so all 21 items in the original questionnaire were not retained. The items that were finally kept in each of the factors in the Brief Burnout Questionnaire Revised were items 6, 14, 19, and 20 for job dissatisfaction; 4, 8, and 9 in the social climate factor; 1, 7, 15, and 21 in personal impact; and 5, 10, 12, and 18 in the motivational exhaustion factor.

The model fit improved considerably this way, and also showed consistency in the validation sample. Configural and metric invariance of the model across the type of job (permanent/temporary) is also assumed, and invariance in all cases (configural, metric, scalar, and strict) across gender. Given the divergence found when clustering items, inquiry into the adequacy of the structure reported by the CBB-R for nursing personnel will have to continue. The multifactorial construct of burnout shown here, which differs from the one reported by the authors of the original study, shows the need for further study of the internal structure of the evaluation instruments in this construct in the population studied [50]. In the process of adapting and validating instruments for certain populations, it must be known whether the factor structure coincides or not with the terms of the original version, as the job characteristics of each sample partly moderate the conditions where burnout appears. The proposed model also includes an analysis of all the burnout risk and protection factors now known as found in the theoretical review.
