Clinical outcomes of Single-Visit oral Prophylaxis: A practice-based randomised controlled trial

General dental practitioners (GDPs) routinely recommend and provide "scale and polish" or "oral prophylaxis" to patients attending their practices [1, 2]. This single-visit treatment consists of supra- and sub-gingival scaling to remove hard deposits, and polishing the teeth with a powered cup, or brush, and paste.

Different definitions of the term 'routine scale and polish' exist and its role in the management of periodontal disease is not specifically defined [3]. Scale and polish is intended to complement patients' self-care plaque-control methods and historically has become inextricably linked to the routine (six-monthly) dental check-up, even if a patient has no, or low, risk of developing periodontal disease [2‐5]. Oral hygiene instruction may be provided in conjunction with the scale and polish to encourage positive oral health behaviour change and improved self-care; furthermore, it has been suggested that there is little value to the professional intervention if hygiene advice is not given [6].

Provision of single-visit scale and polish incurs costs for privately-paying patients and for tax-payers in publicly-funded healthcare systems. It is unclear what proportion of this treatment is clinically necessary [7], yet in 2009/10, 12 million (44.1% of total) courses of treatment carried out on adult National Health Service (NHS) patients in England included a scale and polish [1].

Despite its routine provision, there is a debate regarding the clinical effectiveness of single-visit scale and polish, and the frequency at which it should be provided. A systematic literature review [3] was unable to reach firm conclusions about the beneficial effects on periodontal health and recommended (practice-based) randomised controlled trials (RCTs) to investigate the effectiveness of the intervention.

This paper reports the findings of a preliminary RCT which aimed to compare gingival health outcomes of single-visit scale and polish, performed at 6-, 12- or 24-month intervals, in healthy adults, with no significant periodontal disease who were regular attenders at 'family' dental practices. The objectives were to compare presence of gingival bleeding, dental plaque, and amount of calculus between groups receiving single-visit scale and polish at these intervals.

The trial protocol responded to recommendations [3] and was reviewed and approved by Cheshire Local Research Ethics Committee (reference: Q/1506/100.) The trial was registered with UKCRN: (ID5101); and ISRCTN (ISRCTN56889016). The Oral Health Unit at The University of Manchester funded the trial. Research support costs were met by Cheshire and Merseyside Comprehensive Research Network (funded by the National Institute for Health Research.)

The study was a randomised, 3-arm, parallel clinical trial with an allocation ratio of 1. The follow-up period was 24 months; the maximum period advised between dental check-ups by national guidelines [8].

The CONSORT flow diagram [20] is presented in Figures 2, 3, 4. There were 40 dedicated recruitment sessions 02/2006 to 09/2007. Of the 826 patients approached, 44.7% (N = 369) consented, and were randomly allocated to a trial group. Of the 369 participants commencing the trial, 3 were found to have been recruited in error and did not fulfil the inclusion criteria. Seventeen participants chose to discontinue the trial (5 from 6-month group, 4 from 12-month group, and 8 from 24-month group.) Two participants were withdrawn by their family dentist and a further fifteen were withdrawn from the trial by the independent trial examiners due to concerns that they had a BPE code of 3 (Total 17: 6 from 6-month group; 4 from 12-month group; 9 from 24-month group.). Follow-up data were collected for 83.5% of the original participants; 76.2% attended all 5 trial appointments. Baseline demographic and clinical characteristics of trial participants are presented in Table 1.

Prevalence of gingival bleeding increased in all groups between baseline and follow-up (Table 2). There were no significant differences between groups at follow-up (P = 0.746). Odds ratios, adjusted for baseline bleeding, showed no significant association between frequency of scale and polish and bleeding prevalence.

There were no significant differences between groups at follow-up with respect to prevalence of plaque (P = 0.183) and calculus (P = 0.615) (Table 3).

The mean amount (height) of calculus present on the lower anterior teeth at follow-up was: 6-month group 0.71 mm (SD 1.00); 12-month group 0.89 mm (SD 0.99); 24-month group 0.95 mm (SD 0.97). Adjustment of follow-up calculus for baseline measures showed this difference to be statistically significant (ANCOVA p = 0.001).

The results of the multiple imputation analysis were similar to the complete dataset analysis: no additional statistically significant differences were identified.

Ancillary analyses examined the proportion of index teeth with bleeding or plaque. All groups demonstrated increased prevalence in bleeding from baseline to follow-up. Follow up proportions of teeth with bleeding (Table 2), were 37.9% (SD 30.3) in 6-month group; 38.8% (SD 30.7) in 12-month group; 39.8% (SD 30.2) in 24-month group. These proportions were not significantly different between groups (ANCOVA p = 0.979)

Follow-up proportions of teeth with plaque were 39.4% (SD 34.1), 43.0% (SD 34.9), and 43.7% (SD 32.4) in the 6-month, 12-month and 24-month groups respectively. Differences between groups were not significant (ANCOVA P = 0.597.)

This is the first randomised control trial in the literature investigating the effectiveness of scale and polish when it is delivered in a general practice setting. It should therefore be seen as the first stepping stone to improve the evidence-base for this commonly provided intervention rather than providing definitive evidence to inform policy on resource allocation or as a basis for clinical guidelines. This was a pragmatic trial, involving healthy adult participants with no significant periodontal disease, who had a history of regularly visiting their family dental practitioner. It was not an explanatory trial and did not investigate an intervention performed on patients with periodontal disease or under specialist care. Over a 24-month follow-up period, the researchers could not detect a statistically or clinically significant difference in gingival health measures between groups. The difference in the amount of supragingival calculus between groups was statistically significant yet too small to be clinically significant, given that it is unlikely that it could be detected clinically by family dentists using the instruments commonly used in general dental practice. Prevalence of plaque at baseline (and follow-up) was comparable to findings of national epidemiological surveys [26]. This suggests that the trial has reasonable external validity and that even in a motivated population presence of plaque is the 'norm'. The overall increase in the primary outcome measure trial across the whole trial population gives cause for concern. One interpretation of this finding could be that it is due to a real deterioration in gingival health across all three groups. Furthermore this decline could be attributed to sub-optimal delivery of the scale and polish. This could well be the case, but equally the dental care professionals (hygienists and therapists) could have striven to provide a more thorough scale and polish than usual. Knowing that their work would be independently assessed, it could be argued that this latter scenario is more likely than providing sub-optimal care. In pragmatic trials there will be greater variation in the delivery of the intervention under test than is the case for explanatory trials and therefore the intervention is less likely to be effective. In addition, for interventions for which the practitioner cannot be blinded to the allocation, as is the case in many trials of surgical interventions, there will always be the risk of a Hawthorne effect [27] resulting in the intervention under test being not truly representative of the intervention as it is provided in 'real life'. In this trial participants were monitored by their 'family dentist' every 6 months and dental practitioners could refer participants giving cause for concern to an independent examiner who was blind to the allocation. A small number of patients were referred and a smaller number removed from the trial because of clinical concerns. This and the relative stability of plaque scores between baseline and follow up suggest that inter examiner variation in probing force used to assess gingival bleeding is a more likely to account for the increase in bleeding on probing rather than an overall deterioration in gingival health of the trial population. This interpretation is supported by analyses which demonstrate that examiner 1 assessed more patients at baseline and examiner 2 assessed more patients at follow up (Table 4). The overall increase in bleeding could be explained by examiner 2 using a consistently greater probing force than examiner 1. However there were no significant differences in the distribution of participants examined by the two examiners at baseline and follow up across the three arms of the study and therefore inter-examiner variation would have a minimal influence on the between-group findings.

Significant barriers to executing research in general practice have been reported [28, 29], and were experienced in the operational management of this trial. Lack of empirical data on the expected effect size to inform a power calculation meant that the sample size was based on estimates according to clinical expectations and an ability to identify an effect size which would influence clinicians to consider changing their clinical practise, Recruitment of participants was logistically difficult and caused considerable disruption to day-to-day running of the dental practices, potentially effecting practice income; for this reason clinical measurement was limited and recruitment ceased once the trial had achieved its recruitment target. A larger sample size would have resulted in tighter confidence intervals and the authors acknowledge that a longer follow-up period would have been desirable; a 5-year period [3] was initially proposed but a compromise (2-year follow-up) was reached in response to concern expressed by family dentists about withholding treatment for such a long period. On a positive note, participant loss to follow-up was low (less than 20%) and this preliminary trial demonstrates that dental practices and dental patients can be successfully recruited to practice based trials; with significantly greater resources larger sample sizes and longer follow up periods would be possible.

Interpretation of trials with non-significant findings is difficult, particularly for preliminary trials because bias tending towards a non-significant result can stem from different sources. The potential for allocation bias was considered to be small as the trial manager had no knowledge of the patients other than the basic details required for randomization. Baseline imbalances therefore occurred by chance rather than allocation bias; the magnitude of chance imbalance was not considered to be clinically significant. The 24-month group had a lower proportion of males than other groups, but gender is not a significant risk for periodontal disease progression [30]; and similar proportions of males and females were lost to follow up. Whilst smoking is a risk factor for periodontal disease, smokers were included if they fulfilled the eligibility criteria. The randomisation process ensured that smokers were evenly distributed between groups. Subgroup analysis of the smokers was not undertaken due to the small numbers and because inadequately powered post hoc analyses can produce misleading results [31]. Future trials on this topic may wish to confine the trial population to smokers or power the trial to enable well designed sub-group analyses.

Single-visit scale and polish is a treatment historically embedded in general dental practice rather than being a defined treatment for periodontal disease. The question of whether or not routinely providing a 6-monthly single-visit scale and polish is an effective use of professional resources for healthy adults with no significant periodontal disease cannot be answered by the first practice-based trial in the literature. This preliminary trial does not provide sufficient evidence to support or refute the benefit of 6-month scale and polish over 12- or 24-month treatment provision; it does however raise important questions and will inform the design and conduct of further pragmatic practice-based trials which seek to investigate this subject.