Data analysis

Partial questionnaire completion was not possible. In some questions, the responses were restricted to a specific population, e.g., only dentists assisting patients at the time the survey was responded. Weighting of items or propensity scores were not used. The responses ‘I’d rather not say’, ‘I don’t know how to answer’, and ‘Does not apply’ were considered missing data, resulting in variable numbers of respondents for different questions. Words used in the open question about the feeling associated with the pandemic were translated to English. Singular and plural words, lowercase and capital letters were grouped together. Nouns and adjectives also were grouped for consistency, e.g., anxious/anxiety. The decision was to keep either the noun or adjective considering the most frequent form used. The words were categorized as negative, neutral, or positive feelings. Descriptive statistics were used to identify frequencies and distributions of variables. Proportions were compared using chi-square tests. The numbers of weekly patients assisted before and during the pandemic were analyzed using a paired t-test with unequal variances (α = 0.05). All analyses were performed in Stata 14.2 (StataCorp, College Station, TX, USA).