200 patients took part in the questionnaire and were examined using the PSI. Thereafter the participants were divided into two groups, non-periodontitis persons (group 1; PSI 0-2) and periodontitis persons (group 2; PSI 3-4). The answers were evaluated using a point system ranging from 0 to 8, based on known periodontal risk factors and their assumed degree of influence. Receiver-operating characteristic (ROC) curve analyses were applied to examine the overall discriminatory power, sensitivity and specificity, and corresponding cut-off points of the self-reported periodontal disease scale.

Materials and Methods

This clinical investigation was approved by the Ethics Commission at the Medical Faculty in Leipzig (337-13-18112013). Before commencing the double-blind, controlled clinical trial (), all study participants were informed of its content and the use of personal data and confirmed their voluntary willingness to take part in writing.

Patients

A new questionnaire was used as part of a preliminary dental examination to interview a total of 200 patients concerning the clinical indications and periodontal risk factors. The clinical follow-up examination was conducted by the Periodontal Screening Index (PSI). To include representative periodontitis patients a consecutive sampling was applied. The participants were divided into two groups of non-periodontitis patients (group 1; PSI Code 0,1,2) and periodontitis persons (group 2; PSI Code 3 and 4) for the first PSI classification (Perio 1). The sampling was finished until each group reached a number of hundred patients. To reduce the so-called center effect [25] the patients were all unknown to the examiners. Patients were also required to be at least 18 years to take part in the study. Patients undergoing a periodontal treatment, pregnant and disabled persons were also excluded from the study.

Periodontal Situation (PSI, PSR®)

The Periodontal Screening Index (PSI in Germany) or Periodontal Screening and Recording (PSR®) [12,13] was registered based on a WHO probe (Morita, Kyoto, Japan). The set of teeth was divided into sextants for the purpose of the investigation. The PSI scores (0 to 4) were recorded: score 0=healthy, score 1=bleeding, score 2=supra-/subgingival calculus, score 3=probing depths from 3.5mm to max. 5.5mm, score 4=probing depths greater than 5.5mm. We noted only the highest findings for each sextant. Persons with findings of code 0 to code 2 were classified as non-periodontitis persons (group 1), whereby codes 3 and 4 were considered a probable periodontitis persons (group 2). Two additional PSI classifications were defined in order to review the robustness of the screening test with respect to the prevalent severity of periodontitis. Classification Perio 2 comprised persons with a PSI code 0,1,2 and once code 3, indicating non-periodontitis, while persons recording two codes 3 or 4 on one occasion were periodontitis persons. In the third classification (Perio 3), only such patients as exhibited at least one code 4 were considered as periodontitis persons, while all others were evaluated as non-periodontitis persons.

All clinical recordings were performed by the same calibrated examiners. Examiners calibration was performed as follows: five adults, not enrolled in the study, were evaluated by the examiners on two separate occasions, 48 hours apart. Calibration was accepted if the millimetre measurements at baseline and 48 hours later did not differ more than 10 percent.

Questionnaire

The questionnaire was initially prepared by a retrospective selection of suitable items reviewed in current literature on the subject of periodontal risk factors and indicators. Research was conducted using PubMed, whereby only articles written in English were included. Systematic reviews and randomised, controlled studies were preferred. Based on this 16 questions were developed. The individual response options were assigned point values extending from zero to eight, based on their assumed degree of influence on the periodontal disease. To determine the correct wording and phrasing the questionnaire was given as a pretest to 20 patients not included in the study.

Statistical analysis

Statistical evaluation was conducted using SPSS for Windows, Version 22.0 (SPSS Inc., NY, U.S.A.) and BiAS. for Windows, Version 10.12. (epsilon-Verlag GbR, Hochheim Darmstadt, Germany). The sample size (200 participants) was calculated with an expected standard deviation of four score points, minimal different score point delta between 1.6 and 2, p-value of 0.05 and a power of 0.8 [24]. The categorised data was evaluated based on the Chi-squared test (question 2-6, 8-12, 15, 16), the precise test according to Fisher (question 1, 7, 13, 14). A two-sided review of significance was applied to each of the tests, whereby a p-value of <0.05 was assumed to be statistically significant for all statistical tests.

The distribution of total score was reviewed according to the Kolmogorov-Smirnov test in terms of normal distribution. The score did not exhibit a normal distribution (Kolmogorov-Smirnov test: p<0.05). Furthermore the Mann-Whitney U-test was applied in the comparison of scores based on the variety of periodontitis classifications (differentiated consideration of periodontally diseased persons) and box plots were created to elucidate the data.

ROC curves (receiver operating characteristic) were produced to illustrate sensitivity and specificity. The area under the ROC curve (AUC), which in a test without forecast reliability will be 0.5 and not more than 1, is a benchmark to measure forecast reliability. The cut-off point was defined as beyond a sensitivity of at least 80% with the greatest possible specificity in order to meet the requirements of a screening test.

Inclusion Criteria:

• Patients were all unknown to the examiners
• Patients were also required to be at least 18 years
• All study participants were informed of its content and the use of personal data and confirmed their voluntary willingness to take part in writing.

Exclusion Criteria:

Patients undergoing a periodontal treatment, antibiotic therapy, pregnant and disabled persons were also excluded from the study.