Reporting

The main strengths of the identified RCTs and non-randomized controlled clinical trial were noted to be due to clear reporting. All 13 trials described the study objectives, provided clear patient inclusion and exclusion criteria, and all but one¹² summarized baseline characteristics and distribution of potential confounders between study groups. Eleven studies listed the main outcomes in the Methods section, with descriptions or references to how the outcomes would be measured;^{12–15,17–23} however, two studies did not describe the methods for measuring the outcomes.^16,24 Most studies provided some detail about how SRP was performed, including the tools used during the procedure and number of sessions to complete treatment. However, one study stated that SRP was completed within 30 days of the baseline visit but did not provide further detail,²³ and three studies did not describe any methods for SRP.^12,14,24 The study by Sant’Ana et al.²⁴ also did not describe the “professional prophylactic” intervention provided to both treatment and control groups, so it is unclear how this may have contributed to response to therapy. The results were generally reported well; all studies summarized the results for the main outcomes by presenting mean values for each group. Estimates of the random variability in the data (standard deviation or standard error) were presented in all but one of the studies²³ and actual probability values were provided in all but four studies.^13,16,18,24 Likewise, patient loss to follow-up and the number in each study group when it occurred were reported in the majority of studies.^{12–17,19–22,24} Some aspects were infrequently reported in the included clinical trials; none provided the simple outcome data that contributed to those mean values, and adverse events potentially associated with the study interventions were not addressed by 10 studies.^{12–15,17,18,20,22–24}

External Validity

The external validity of the studies was influenced by the choice of patients to include in the studies, the methods for patient selection, and the environments in which the studies were conducted. Six studies described the source population or methods for selecting patients.^{13,14,17–19,24} However, seven studies did not clearly describe methods regarding patient recruitment or selection,^{12,15,16,20–23} and six studies did not provide reasons for refusal in patients who declined to participate in the study.^{13,16–18,21,24} The study by Sant’Ana et al.²⁴ identified eligible pregnant women with periodontitis from an antenatal care program, and the study by Kapellas et al.¹⁹ included a convenience sample of Indigenous Australians; however, it is possible that individuals who are already participating in a health care program, or who are easily accessible to or cooperative with health care providers, would exhibit different health-related behaviours than people who do not do these things. In general, people who agree to participate in clinical trials may be more likely to be health conscious, for example brushing and flossing regularly, so this consideration applies to most studies, particularly those that did not specify how patients were recruited or selected. One study that recruited patients with rheumatoid arthritis was conducted in an orthopedics department,¹² and another that included patients with hyperlipidemia was conducted in a cardiac and renal transplant centre;¹⁴ it is unclear whether the majority of patients with these specific health conditions would normally attend or receive the level of care provided at these types of facilities. All of these factors contribute to uncertainty around whether the patients who were approached or those who agreed to participate in these studies would be representative of the larger patient population.

Internal Validity

The internal validity of the studies was influenced by the study designs and methods for analyzing the results. Blinding patients to the intervention they were receiving was not done in any study, though this was likely impossible due to the nature of the interventions. Potential performance bias can still be minimized by blinding outcome assessors to the patient’s study group; this practice was described in six studies^{13,17,18,21,23,24} and not mentioned in seven.^{12,14–16,19,20,22} A common strength for all studies was that they recruited all patients from the same source over the same period of time, and 12 of the 13 studies measured the outcomes at consistent time points that were the same for both the treatment and control groups.^12–21,23 One study referred to follow-up at the “2^nd visit”, the timing of which may have varied within and between groups.²⁴ Twelve of the included clinical trials were RCTs, but two did not describe methods for randomization.^16,20 Likewise, allocation concealment was not described in nine studies.^{12–14,16,20–24} The study by Sant’Ana et al.²⁴ did not randomize patients to study groups; rather, allocation was based on patient choice. Despite a lack of randomization in this study, there were no significant differences between the treatment and control groups at baseline in any of the measured periodontal parameters and other baseline characteristics. Likewise, most of the RCTs demonstrated that baseline characteristics were well balanced between study groups. However, In the RCT by Khare et al.¹² the baseline OHI-S and BOP scores of the SRP group were significantly higher than those of the control group, indicating that the SRP group started the study with a poorer periodontal status. In the study by Sexton et al.²³ the SRP group was significantly younger than the control group. In both cases, these intergroup differences could have impacted response to therapy. Regarding the data analyses, all studies used appropriate statistical tests to assess the main outcomes, none appeared to perform any unplanned, retrospective analyses, and six clearly reported either analyzing all patients (none lost to follow-up)^{12,14,16,21,22} or analyzing an intention-to-treat population using the last observation carried forward.¹³ Compliance with treatment was not a concern given that most studies performed SRP once at baseline; however, four studies only analyzed post-treatment data from patients who were available at a follow up visit,^15,19,20,24 and the population analyzed was unclear in three studies.^17,18,23 This may not accurately account for confounding variables (such as the age difference between groups in one RCT²³) or reflect the true difference between treatment and control groups. Losing patients throughout the study can be especially impactful for studies with small sample sizes, but five studies reported an a priori power calculation to determine the necessary sample size required to detect a difference between groups in the non-periodontal primary outcomes.^{13,15,17–19} Four of these studies maintained the necessary sample size was after attrition^13,15,17,19 while one study reported the power of the study at randomization but did not discuss accounting for attrition in this calculation, and the number of patients lost to follow-up was not reported.¹⁸

The strengths and limitations of individual clinical trials are provided in Table A5.

Scope and Purpose

All four included evidence-based guidelines^25–28 had clearly described objectives, scope, and intended users and target populations.

Stakeholder Involvement

The American Dental Association guideline development group had broad representation from relevant groups, including research and methodology experts.²⁵ The guideline by Tonetti et al.²⁶ described member affiliations but not specific job titles. The group responsible for the Ministry of Health Malaysia’s guideline²⁷ included representatives from clinical practice and the government; however, it was unclear whether a methodologist was included to provide guidance on best practices for evidence searches and synthesis. One guideline did not provide any details about those involved in development.²⁸ All publications described the target users of the guideline. None of the guidelines considered patient input; one group acknowledged that this process was ideal but not feasible for the development of their guideline, for unspecified reasons.²⁷

Rigour of Development

Two guidelines^25,26 were based on separate SR publications^8,9 (see Table A4 for details of their strengths and limitations). The other two guidelines^27,28 used systematic methods to identify evidence from multiple databases and listed search terms and dates. The Ministry of Health Malaysia guideline also reviewed reference lists of key articles to search for publications not identified from the electronic database search;²⁷ however, neither guideline specified whether grey literature was included in the search.^27,28 These two guidelines also lacked clear descriptions of how evidence was selected, as inclusion and exclusion criteria were not provided.^27,28 In three guidelines, the body of evidence was evaluated, evidence statements were assigned a quality or certainty level, and these evidence statements were clearly linked to recommendations.^25–27 However, the guidelines referred to relying on expert consensus to develop recommendations, but none described methods used during this process. One of the four guidelines explicitly considered evidence related to adverse events;²⁵ another guideline reported that adverse events were addressed in another guideline from the same European Workshop on Periodontology.²⁶ An external review process was described in three guidelines,^25,27,28 and a plan for updating the guideline was presented in two cases.^25,27

Of note, the full text of the original HealthPartners Dental Group guideline could not be obtained for review, and a guideline summary²⁸ from the National Guidelines Clearinghouse (NGC) was used for this report. Several details regarding the methodology of guideline development were not provided in the summary. Guidelines summarized by NGC are considered evidence-based, and it is possible that more detailed methodology would be presented in another source that would change a critical appraisal of this guideline.

Clarity of Presentation

The recommendations from three of the guidelines were somewhat ambiguous, as they did not describe the elements that would influence treatment choices when considering SRP as initial treatment,²⁵ what combination of treatments should be provided given that SRP should not be the sole modality for patients with periodontitis,²⁶ or how to tailor periodontal treatment to patients’ risk factors for disease progression.²⁷ However, these three guidelines addressed a range of periodontitis management options across the guidelines as a whole and clearly presented key recommendations in a visually identifiable way.^25–27 The majority of the content in the HealthPartners Dental Group guideline summary was presented in the Major Recommendations section, making it unclear which portions of the text reflected evidence summaries, expert opinion or commentary, or recommendations.²⁸

Applicability

Applicability considerations were infrequently described; potential barriers and facilitators to implementation of the recommendations were not described in two guidelines,^25,28 one of the four guidelines specifically addressed cost considerations,²⁷ and none provided additional resources to assist guideline implementation. Two of the four guidelines presented or referred to auditing criteria.^27,28

Editorial Independence

Each guideline addressed potential conflicts of interest of guideline development group members, but none provided an explicit statement that the recommendations were developed without undue influence from the funder.

The strengths and limitations of individual guidelines are provided in Table A6.

Simplified Oral Hygiene Index (OHI-S)

Two RCTs evaluated OHI-S scores after treatment with SRP alone²⁰ or in combination with OHI¹² compared with no treatment. In the RCT by Khare et al.¹² the baseline OHI-S score of the SRP group was significantly higher than that of the control group, indicating that the SRP group started the study with a poorer periodontal status overall which may have impacted this group’s response to therapy. Despite this baseline discrepancy, this study also showed that OHI-S scores were significantly lower in the SRP group than in the control group at three months.¹² The RCT by Koppolu et al.²⁰ reported a statistically significant reduction in OHI-S scores from baseline in the SRP group, while scores increased in the control group over the same period. This study did not report a statistical comparison between groups.²⁰ Neither study that evaluated this outcome commented on what constitutes a minimal clinically important difference in OHI-S score, yet Koppolu et al.²⁰ described plaque reduction in the treatment group as “satisfactory”.

Probing Depth (PD)

PD was evaluated in 11 RCTs^{12–15,17–23} and one non-randomized controlled trial.²⁴ General trends for the PD results from the clinical trials are presented in Table 2.

Table 2

Clinical Trial Results for Probing Depth.

Eight RCTs found that PD improved after SRP treatment at follow-up time points ranging from four to 28 weeks after baseline.^{13–15,18,20–23} This was signified by either a statistically significant reduction in mean PD in mm,^{13–15,18,20,21} or a significant decrease in the proportion of sites with PD greater than 4 mm or 5 mm.^22,23 One study also showed a significant increase in the proportion of sites with PD ≤ 3 mm, demonstrating an overall decrease in PD severity as the PD distribution shifted to the less severe category at follow-up.²² In some cases, PD improvement was only observed when PD was more severe (≥ 4 mm) at baseline.¹⁵

In these same eight RCTs, PD did not change^14,18,22 significantly increased,^13,20,21 or significantly decreased over time in the control group.^15,23 These latter two studies offered SRP and OHI to the treatment group and OHI alone to the control group, suggesting that OHI provides some benefit for patients with periodontal disease. However, SRP and OHI were significantly more effective at improving PD than OHI alone in these two studies.^15,23 Likewise, five of these studies statistically analyzed intergroup differences at follow-up and found that PD was significantly smaller in the SRP group than the control group,^{13,15,18,21,23} and in one study this finding depended on the initial severity of periodontal disease.¹⁵ Intergroup differences were not analyzed statistically in three of these eight RCTs.^14,20,22

Two RCTs^12,19 did not statistically analyze changes from baseline in either study group but showed that the SRP group had significantly smaller PD¹² or significantly fewer sites with PD of at least 4 mm¹⁹ than the control group at three months.

One RCT showed that there was no change from baseline in PD at two months in either the SRP or control group, and no difference between these groups at two months.¹⁷ However, the mean PD in both groups at baseline was less than 4 mm, and the authors discussed an observed reduction in the prevalence of patients with PD greater than 4 mm, suggesting that perhaps the benefit of SRP was greater in a subset of patients with more severe periodontal disease. Finally, one non-randomized study showed that PD worsened in both the SRP and control groups; the authors attributed this to the fact that the study patients were pregnant women, suggesting that periodontal deterioration may be expected during pregnancy.²⁴ However, there was a significant difference between the SRP and control groups at follow-up, leading the authors to conclude that SRP may mitigate periodontal disease progression during pregnancy.²⁴

Clinical Attachment Level (CAL)

CAL was evaluated in one SR,⁹ nine RCTs,^{12–15,17–19,21,23} and one non-randomized controlled clinical trial.²⁴ General trends for the CAL results from the clinical trials are presented in Table 3.

Table 3

Clinical Trial Results for Clinical Attachment Level.

The SR by Smiley et al.⁹ meta-analyzed the results from 11 RCTs and found that SRP was associated with a statistically significant improvement in CAL of 0.49 mm as compared with no treatment when measured at least six months after baseline. No details were provided in this review regarding the duration or frequency of SRP treatment.

Of the 10 individual clinical trials that evaluated this outcome, six found that the SRP group demonstrated significant improvements from baseline in CAL.^{13–15,18,21,23} CAL improvement was reflected in statistically significant reductions in the mean CAL in mm^{13–15,18,21} or the proportion of sites with a CAL greater than 2 mm.²³ One of these studies showed that a significant change from baseline in the SRP group was limited to patients with an initial PD of at least 4 mm.¹⁵ This study also found that, in the SRP group, a greater proportion of measured sites had a less severe CAL (1 to 2 mm) and fewer sites had more a severe CAL (at least 5 mm) compared with baseline; no such CAL severity shift was observed in the control group.¹⁵

In these six RCTs, CAL did not change,^14,15,18 significantly increased,^13,21 or significantly decreased from baseline in the control group.²³ This last result was from the same study that noted significant decreases in PD from baseline in the control group, who received OHI alone.²³ Four of the six RCTs found a significant difference in CAL between the SRP and control groups,^13,15,18,21 and in one study this finding was limited to the subgroup of patients with more severe periodontal disease at baseline.¹⁵

Of the remaining four clinical trials, two RCTs^12,19 did not statistically analyze changes from baseline but showed that the SRP group had significantly smaller CAL¹² or significantly fewer sites with CAL of at least 3 mm and PD of at least 4 mm¹⁹ as compared with the control group at three months.

As with the findings for PD, one RCT showed that there was no change from baseline in CAL at two months in either the SRP or control group, and no difference between these groups at two months.¹⁷ Likewise, the non-randomized controlled trial that recruited pregnant women with periodontitis showed that CAL did not change in the SRP group and worsened in the control groups, and this difference between groups was statistically significant.²⁴

Plaque

One SR⁸ and seven clinical trials^{13,14,18,19,21,22,24} evaluated plaque-related outcomes (plaque index (PI),^13,14,22,24 number of teeth with plaque,¹⁹ or proportion of sites with plaque^18,21). The SR⁸ found that PMPR (scaling but not root planing) was associated with reduction in plaque levels, but that this improvement was not always significantly different from results for the control groups.

In five RCTs,^{13,14,18,21,22} SRP was associated with a decrease in plaque from baseline at one month,¹⁸ three months,^{13,14,18,21,22} or six months.¹³ There was a significant decrease¹⁴ or no change^13,18,21,22 from baseline in the plaque levels of the control groups. Furthermore, in three of these five RCTs that analyzed intergroup differences, plaque levels were significantly lower in the SRP group than in the control group.^13,18,21

The study of pregnant women with periodontitis found that professional prophylaxis and OHI, with or without SRP, did not affect PI scores at the second visit.²⁴ One RCT that found significant differences between SRP and control groups in other periodontal outcomes (PD, CAL, GI) did not observe the same results for the number of teeth with plaque at three months.¹⁹

Gingival Index (GI)

Six RCTs evaluated changes in GI after SRP treatment.^{12–14,19,21,22} Four studies analyzed changes from baseline and found a significant improvement from baseline in the SRP group at three months^13,14,21,22 and six months.¹³ In the control groups, GI worsened,^13,21 did not change,²² or improved from baseline (when the control group received OHI).¹⁴ All four studies that analyzed intergroup differences at follow-up found that GI scores were significantly different between the SRP and control groups at three months^12,13,19,21 and six months.¹³

Bleeding on Probing (BOP)

One SR⁸ and seven clinical trials^{12,13,15,17,18,23,24} evaluated the impact of periodontal treatment on gingival bleeding.

The SR⁸ identified some evidence that showed a greater reduction in gingival bleeding or inflammation after PMPR as compared with no treatment, but this finding was not consistent across all studies and the authors suggested that the magnitude of effect did not appear to be as great as for plaque-related outcomes; however, no statistical comparisons were presented to support this conclusion.

Five RCTs with follow-up time points ranging from four weeks to 28 weeks found that the percentage of sites with BOP significantly decreased from baseline after SRP treatment.^{13,15,17,18,23} There was no change^17,18 or an increase in BOP¹³ from baseline for the control group in studies where no treatment was provided,^17,18 but there was also a significant decrease in BOP in control groups that received OHI alone.^15,23

As with PD, BOP significantly increased from baseline in both study groups in the trial that recruited pregnant women with periodontitis.²⁴

All seven clinical trials analyzed intergroup differences at follow-up; BOP was significantly lower in the SRP group than the control group in six studies^{12,13,17,18,23,24} and there was no significant difference between groups in the RCT that found improvements in BOP in both groups.¹⁵

Gingival Recession

One RCT¹⁵ evaluated gingival recession, which significantly decreased from baseline with SRP when the initial PD was at least 4 mm. There was no significant change at four weeks in the SRP group when initial PD was 1 to 3 mm, or in any patient from the control group. In addition, there was no significant difference between treatment groups overall at four weeks.¹⁵

Periodontal epithelia surface area (PESA) and Periodontal inflammatory surface area (PISA)

One RCT¹³ evaluated PESA and PISA, which significantly decreased from baseline in the SRP group at the three month and six month follow-up visits.¹³ The control group, which did not receive any treatment, demonstrated significantly higher PISA scores at 3 months and higher PISA and PESA values at six months. Scores for both outcomes were significantly lower in the SRP group than the control group at both time points.¹³