Page 107 - GTM-4-3
P. 107

Global Translational Medicine                                  Comparative analysis of MIF and CF techniques



            tooth loss due to periodontal reasons (TLP).  At the   3. Results
                                                   34
            tooth level, multilevel binary logistic regression with
            generalized estimation equations was used to relate the   3.1. Characteristics of the study population
            independent factors and covariates to the binary outcomes   There were 40 males (45.5%) and 48 females (54.5%), with
            (e.g., TLP, BOP at T1). Raw odds ratio (OR) and 95%   an average age of 63.1 ± 13.8 years, ranging from 30 to
            confidence intervals (CI) were obtained from the Wald’s   87 years at baseline. Each patient contributed, on average,
            χ  statistic. Then, multiple models were estimated to   1.13 teeth to the database, resulting in a total sample of 99
             2
            adjust by potential confounding factors (at patient’s level,   teeth (40 in the MIF group and 59 in the CF group) treated
            such as age, sex, smoking, diabetes mellitus, periodontitis   by 84 post-graduate residents.  Table 1 summarizes the
            diagnosis, number of maintenance visits, and duration of   demographic characteristics of both groups, the type and
            follow-up, as well as clinical data, including the number   location of tooth, and the operator level.
            of walls, antibiotic use, the type of membrane applied,   The sample included 88 patients who underwent either
            furcation involvement, number of walls of the defect, defect   MIF (n = 36, consisting of 30% SPPF incision technique,
            dimensions, and the type of grafting material). Quantitative   47.5% modified papillary preservation incision technique,
            outcomes (e.g., CAL gain, PD reduction, KT change, and   and 22.5% papillary preservation incision technique) or
            GR at T1) were analyzed using linear regression models   CF (n = 52) procedures. The mean follow-up period after
            estimated with generalized estimation equations to control   treatment was 42.0 ± 30.1  months, ranging from 2 to
            the within-subject dependence of teeth. Beta coefficients   163 months (median: 34; interquartile range: 19–57).
            and 95% CIs were reported. As previously mentioned,
            multiple models were estimated. The significance level   In terms of group homogeneity, no significant differences
            used in the analysis was 5% (α = 0.05).            were found between the groups regarding patient-level
                                                               covariates, such as sex, age, diabetes mellitus, smoking, and
              To ensure sufficient statistical power to detect a clinically   periodontitis diagnosis (staging and grading), as well as the
            valuable difference between the MIF and CF groups, a   number of follow-up visits (p>0.05). Group homogeneity
            power analysis using a post hoc estimation was performed.   information is presented in Table S2.
            A sample size of 99 independent teeth provides 72.3% power
            at a 95% confidence level to detect an OR of 3 as significant   No significant difference was found in the treating
                                                                                   nd
            using a logistic regression model (an OR of 3 is equivalent to   residents’ level, with 2 -year residents predominately
            comparing rates of 50% and 25%, for example, for BOP rates).   contributing to both groups (MIF: 47.6%, CF: 52.3%).
            However, since teeth were not independent observations, the   3.2. Analysis of changes in clinical outcomes
            power was adjusted to account for the two-level structure
            of the data. Each patient provided an average of 1.13 teeth.   3.2.1. Effect of procedure complexity on clinical
            Assuming a moderate within-subject correlation of 0.5,   outcomes
            a correcting coefficient of 1.06 was obtained. Therefore,   The mean CAL gain was found to be 2.17 ± 2.18  mm
            a sample of 99 dependent teeth was equivalent to  93   for the CF group, in contrast to only 0.59 ± 3.43 mm for
            independent observations, yielding an estimated power of   the MIF group (Figure 3). Significant differences in CAL
            70.0% under the same previous conditions.          gain remained between the groups even after adjusting

            Table 1. Demographic characteristics of both groups, the type and location of the tooth, and the operator level
            Demographic characteristics            Total                     CF                      MIF
            Number of patients                      88                       52                       36
            Age, mean (years)                     63.1±13.8                 63.16                    63.45
            Sex (M/F)                              40/48                    25/28                    15/20
            Follow-up time, mean (months)         42.0±30.1                 44.85                    38.5
            Number of teeth                         99                       59                       40
            Type of the tooth (%) (M/P/C/I)    65.7/19.2/4.0/11.1        86.4/6.8/5.1/1.7        35.0/37.5/2.50/25.0
            Arch (%) (Max/Man)                    44.3/55.6                62.5/37.5                32.2/67.8
            Operator level (R1, R2, R3)           13, 42, 29               9, 22, 19                4, 20, 10
            Note: R1, R2, and R3 refer to 1 -year residents, 2 -year residents, and 3 -year residents, respectively.
                                                          rd
                                           nd
                                st
            Abbreviations: CF: Conventional flap; M/F: Male/female; M/P/C/I: Molar/premolar/canine/incisor; Max/Man: Maxilla/mandible; MIF: Minimally
            invasive flap.
            Volume 4 Issue 3 (2025)                         99                          doi: 10.36922/GTM025080015
   102   103   104   105   106   107   108   109   110   111   112