Population A total of 144 patients were enrolled from 6 university clinical centers (Sapienza University of Rome, Department of Periodontology, Rome, Italy ; University of Rijeka, School of Dental Medicine, Rijeka, Croatia; University of Zagreb, Department of Periodontology, Zagreb, Croatia; Jagiellonian University, Dental Institute, Krakow, Poland; Medical University of Sofia, Faculty of Dentistry, Sofia, Bulgaria; Medical University of Plovdiv, Faculty of Dentistry, Plovdiv, Bulgaria). The extension in time from patient recruitment to end of the protocol was January 2016 to November 2018.
The sample size was determined considering the patient level as statistical unit and probing depth (PD) as the primary outcome measure. In order to detect a mean difference of 1 mm of PD between groups with a standard deviation of 1 mm, maintaining a power of 85%, a significance level (alpha) of 0.05 and a drop-out of 25%, 12 participants per group in each center were required.
The ethical committees of each center approved the protocol of this study. The protocol followed the ethical principles of Helsinki Declaration on the good conduct of clinical trials. All patients were thoroughly informed on the treatment details and on the possibility of being excluded from the study if missing any of the 3 months follow-up appointments or showing poor compliance in general. After verbally accepting the conditions of the protocol, patients signed the informed consent prior to study initiation. Figure 1 shows step by step the study protocol that each patient followed.
Study design, randomization, allocation concealment and blinding This study was projected as a controlled randomized multicenter triple blinded 12-month clinical trial to compare the treatment of residual periodontal pockets using standard non-surgical mechanical debridement with and without hyaluronic acid gel. The detailed description of the treatment and evaluation is described in the following paragraphs.
The operators were trained and calibrated on treatment application and on PD measurement prior to study initiation. A preliminary session of PD measurements on 3 periodontal patients revealed an intra-class correlation coefficient ≥ 0.75 for each operator and among different operators.
An electronic web-based online randomization system was used for this study including a simulation version of the software for operator training prior to trial initiation. This system had been approved by the Austrian Agency for Health and Food Safety (AGES). The system parameters were set to randomization block size of 4, treatment code selection to random, using 24 treatment codes per center, and using minimization method. Centers were included as a factor for randomization process with assignment probability of 0,125 for each center (8 centers x 0,125 = 1.0) as the randomization ratio was 1:1 corresponding to treatment/ control allocation ratio.
Each operator had a username and password to perform the randomization process for each patient. An accurate audit trail showed each user log entry consisting of login-ID, center number, investigator name, event code, date, and time of which the event occurred. The system had trial specific features (role-based access control) where trial coordinators could manage each user assigned specific role determining the user's access rights. For the present study, the investigators role was to randomize without the ability to view the trial design or randomization design. The access to full control, view randomizations, un-blind randomization, was only granted to the independent trial coordinator.
Furthermore, the system provided high flexibility and security including verification for the personal email for each investigator to guarantee that each investigator will perform the randomization personally, setting a pre-defined randomization limit, the possibility to include multiple users under each center, and an effective email notification and confirmation upon each randomization performed. All transactions on the system were logged and an audit trail could be analyzed at any time by the trial monitor.
The allocation concealment was done by means of independent computerized randomization sequence process which can only be accessed by the independent trial coordinator. The investigators were provided with online login and were instructed to perform the randomization only after finishing the baseline measurements and mechanical debridement of selected tooth. To ensure that, investigators were requested to write down the exact time on provided case report forms upon finishing debridement. The computerized time stamp, electronic logs of sent allocation served as a method for monitoring allocation concealment. Furthermore, having the randomization as parallel with stratified (stratification by center) permuted blocks, served as a second layer of preventing knowledge of upcoming patient's assignment.
Clinical operators were kept blind until the moment of adjunctive treatment application allocated by the randomizer. Patients remained blinded during the entire follow-up. Microbiological evaluation and statistical interpretation of the results were made blinded on the treatment performed.
Treatment procedure Each center was provided with the same amount of HA , Eppendorf tubes, curettes , tooth paste , tooth brushes ** and interdental brushesǁ. Patients were instructed to use the modified Bass technique for brushing twice daily with the provided toothpaste free of chlorhexidine and to use the interdental brush once daily. They were instructed not to use any type of mouthwash.
All patients underwent a full mouth debridement completed within 48h in two consecutive visits. The debridement was performed using ultrasonic devices followed by curettes for all pockets of more than 4 mm. The ultrasonic treatment was repeated to ensure all surfaces were clean, followed by polishing.
At reevaluation, a tooth entering the inclusion criteria was selected. After application of local anesthetic, removal of plaque and polishing of the tooth surfaces was performed followed by the local administration of hyaluronic acid gel for the test group and local administration of anesthetic for the control group. The tip of blunt application needle mounted on the syringe containing HA was delicately inserted in the pocket until resistance was felt. The injection was performed until the remnants started getting out of the pocket. Patients were advised not to drink or eat for 2 hours. This adjunctive treatment was performed at baseline and at the 3 months control. For both groups cleaning of supragingival tooth deposits and polishing was repeated at 3, 6, 9, and 12 months.
Statistical analysis Numerical variables were reported as median (IQR) while categorical variables were reported as frequency (percentage). McNemar test was used to test for differences in proportions between time points within groups and Chi-squared test was used for differences in proportions between groups. For numerical variables, time-points differences were assessed with the Kruskal-Wallis test and pairwise comparisons were performed with Mann-Whitney U test and Wilcoxon signed rank test (for unrelated and related measurements respectively). Holms correction was used to adjust for multiple comparisons. A significance level of 0.05 was used for all statistical tests. All the analytical tests were performed using the statistical software R (version 3.6.0).
