The study protocol was approved by the institutional review board of the Medical Association North Rhine, Düsseldorf, Germany (No. 2013384).
Participants and date of data assessment
Data assessment for the present study took place from April to June 2018. Participants were former participants of the DMS V (10) which took place in 2014 and comprised representative samples of 12-year olds (N = 1468, now 16-year olds), younger adults (aged 35–44; N = 966, now aged 39–48) and seniors (aged 65–74; N = 1042, now aged 69–78). 91 % of the participants of that study agreed being contacted for further studies. The present study aimed to assess random selections of N = 333 within each of the age groups, thus to a total of 999 participants.
Information regarding age, sex, and highest degree of education were available by the data entries of the DMS V. After their participation in the DMS V study, the participants were included in a panel and contacted every six months for address and contact maintenance.
The current contact data of the participants were always requested and updated if necessary. For those persons who did not have a telephone number, an internal search was carried out first.
Subsequently, the participants were contacted in two concrete steps with regard to their telephone number.
The search for the telephone numbers was then completed. Of the 3057 target persons from the DMS V, telephone numbers were already available for 1897 persons. The search described above was carried out on 1020 persons. Further 140 telephone numbers could be identified, so that from altogether 2037 persons telephone numbers for this study were present.
Structure of the survey
For the current survey the authors of the present study revised, updated and complemented the survey conducted in 2007 (9) which based on a patient information brochure of the German society for Periodontology.
Participants were asked 5 types of questions to be considered in this analysis:
-
Open-ended questions with no selection of answers provided by the interviewer
-
Single choice queries where participants had to chose one out of several alternatives
-
Multiple choice queries where participants were asked to chose all alternatives which they thought to be correct; no limits were given regarding the number of selections
-
Statements for which participants should decide whether they considered them to be right or wrong
-
Filter questions leading to prespecified further questions depending on the answers given by the participants
These questions related to
-
the definition of periodontitis (block 1; open-ended vs. single choice),
-
consequences of periodontitis (block 2; open-ended vs. multiple choice),
-
risk factors of periodontitis (block 3; open-ended vs. multiple choice),
-
prevention, early recognition and prevalence of periodontitis (block 4; statements which were right or wrong),
-
devices indispensable for proper oral hygiene (block 5; open-ended vs. multiple choice),
-
aspects to look out for regarding proper oral hygiene (block 6; open-ended),
-
locations particularly important to clean (block 7; open-ended).
Other questions whose analysis would go beyond the scope of the present paper and which will therefore be evaluated in a subsequent analysis related to specific attitudes, health experiences, and health behaviors of the participants. Appendix 1 shows the whole structure of the interview including interviewer instructions and highlights those questions considered in the present analysis.
Instruction of the interviewers
Long-term permanent employed professional interviewers specialized in computer assisted telephone interviews (CATI) conducted the interviews. In order to minimize interviewer effects on answering behavior they received no information regarding the correct or expected answers. None of the interviewers was related to dentistry.
Furthermore, interviewers had no other information about the individual participants as their name and that they were former participants of the DMS V. Interviewers received a specific and detailed face to face training for the present study regarding techniques of questioning and categorization of answers given to open-ended questions.
All interviewers accomplished several test-interviews prior to interviewing real participants.
Interviewer instructions were refined according to the results of the test-interviews.
Interviewers categorized answers to open-ended questions during the interview by allocating them to pre-specified categories. When open-ended questions were followed by single- or multiple-choice queries regarding the same content the pre-specified categories corresponded to the choices presented there. Answers not corresponding to any of the pre-specified categories were saved in text form. RJ and RD jointly discussed whether these answers could be re-categorized into one of the pre-specified categories (e.g. in question 1 answers like “bleeding of the gum” were re-categorized into “gingival inflammation”), or summarized within a new category, or should be kept as “miscellaneous” answers. Decisions for pre-categorization were only made if both authors agreed in doing so. Most answers to most questions, however, were already covered by the pre-categorization and thus could be easily categorized by the interviewers. For quality assurance, the supervisor in the CATI studio monitored 10 % of the interviews.
Statistical data analysis
All analyses were run by means of SPSS 26 (IBM, Amrok, USA). Frequencies of positive answers to any item were computed for each age group (16 yrs, 39–48 yrs, 69–78 yrs). When frequencies were below 10% or above 90% no further analyses were run regarding differences between age groups, males and females or highly vs. lower educated participants. In all other cases the effect size Cramer’s V regarding differences between age groups and regarding differences between males and females and highly vs. lower educated within age groups were computed by the cross tables procedure of SPSS. Effect sizes of Cramer’s V = 0.1, Cramer’s V = 0.3, Cramer’s V = 0.5 are considered small, medium or large, respectively. All data are presented along with the 95% confidence intervals (20); recommended method for single samples).
Informed consent and data availability
The written consent obtained from the study participants to participate in the study did not include making their data publicly available. Therefore, the datasets generated and analysed during the current study are not publicly available.