|
|
||||||||
|
J Am Dent Assoc, Vol 140, No 4, 415-423.
© 2009 American Dental Association |
COVER STORY |
| ABSTRACT |
|---|
|
|
|---|
Methods. The authors searched the population of studies used in five reviews of sealant effectiveness as established in split-mouth design studies involving resin-based sealants with no reapplication of lost sealant. They required included studies to contain sufficient data to estimate the risk of caries in FS teeth relative to that in NS teeth
and its 95 percent confidence interval (CI). To estimate the mean RR by year since sealant placement, they used a weighted bivariate model and tested for heterogeneity using the quantity I2.
Results. The weighted mean RR was 0.998 (95 percent CI, 0.817–1.220) one year after placement (four studies, 345 tooth pairs) and 0.936 (95 percent CI, 0.896–0.978) at four years (five studies, 1,423 tooth pairs).
Conclusions. Teeth with fully or partially lost sealant were not at a higher risk of developing caries than were teeth that had never been sealed.
Clinical Implications. Inability to provide a retention-check examination to all children participating in school sealant programs because of loss to follow-up should not disqualify a child from receiving sealants.
Key Words: Dental sealants; pit-and-fissure sealants; retention; caries
Abbreviations: CDC: Centers for Disease Control and Prevention. CWF: Community water fluoridation. FMR: Fluoride mouthrinse. FS: Formerly sealed. M: Permanent molar. NA: Not applicable. NR: Not reported. NS: Never sealed. PM: Permanent premolar. RB1: Ultraviolet light–polymerized resin-based sealant. RB2: Autopolymerized resin-based sealant. VT: Visual/tactile. Y1: Year 1. Y1.5: Year 1.5. Y2: Year 2. Y3: Year 3. Y4: Year 4. Y4.5: Year 4.5.
Almost 70 percent of youth have experienced dental caries by late adolescence.1 Available data show that children and youth from low-income families (those with an income of less than 200 percent of the federal poverty guidelines) are more than twice as likely to have untreated caries in their permanent teeth as are their higher-income counterparts.1 Overall, about 90 percent of carious lesions are found in the pits and fissures of permanent posterior teeth,2 with molars being the most susceptible to caries in comparison with other tooth types.3
Researchers have shown that dental sealants delivered in clinical or school settings are highly effective in preventing dental caries, reducing caries in the pits and fissures by 60 percent from two to five years after placement.4 Sealant effectiveness is linked to sealant retention, and a retained sealant has been shown to be 100 percent effective.5 Although systematic reviews4,6 have demonstrated the effectiveness of dental sealants, recent national data indicate that sealant prevalence among children and youth—30 percent1—is well below the national Healthy People 20107 target of 50 percent. Disparities exist according to income, with children from lower-income families about one-half as likely to have received a sealant as their counterparts from higher-income families.1
School programs providing dental sealants are an important intervention to increase childrens receipt of sealants. On the basis of strong evidence of effectiveness, the independent, nongovernmental Task Force on Community Preventive Services8—a volunteer body of public health and prevention experts whose members are appointed by the director of the Centers for Disease Control and Prevention (CDC), Atlanta—issued a strong recommendation that school-based sealant programs be part of a comprehensive community strategy to prevent dental caries. The task force also acknowledged that these programs typically deliver services to children unlikely to receive them otherwise (such as children from lower-income families). School-based sealant programs also have the potential to link students with treatment services in the community.
One potential barrier to delivering sealants is the concern that a tooth with a partially lost sealant may be at a higher risk of developing caries than it would be if it never had been sealed. The theoretical rationale is that food particles could become trapped under a partially retained sealant, thus increasing the availability of nutrients for cariogenic bacteria. Because school-based programs typically deliver sealants to children who are more likely to move during or between school years than are higher-income children,9 follow-up examinations for all children receiving sealants may not be possible. This concern about risks associated with sealant loss led a CDC-sponsored Expert Work Group that was developing guidelines for school-based sealant programs to request an analysis of relevant clinical studies. We were charged with carrying out this analysis. Therefore, the objective of our research was to determine if the risk of developing caries in a formerly sealed (FS) tooth with fully or partially lost sealant exceeds the risk in a never-sealed (NS) tooth.
Data abstraction and quality assessment.
The same two reviewers independently abstracted data from the included studies. If there was disagreement on a specific item on the abstraction form, both reviewers re-examined the relevant portion of the study and reached consensus on the appropriate value. Because included studies were randomized controlled split-mouth trials and selected from among published systematic reviews that included explicit quality criteria for inclusion, we did not assign a quality score. However, we collected information on selected aspects of study quality (Table 1Our outcome measure was whether a tooth, when assessed at each annual follow-up examination, had developed caries.
![]()
MATERIALS AND METHODS
TOP
ABSTRACT
MATERIALS AND METHODS
RESULTS
DISCUSSION
CONCLUSION
REFERENCES
Inclusion criteria and identification and selection of studies.
We searched MEDLINE and the Cochrane Library for systematic reviews of sealant effectiveness that were published in English from 1990 through 2005. Five systematic reviews,4,6,10–12 which included 37 unique studies,13–49 met these inclusion criteria. Two reviewers (S.K.G. and S.O.G.) screened these studies. They excluded 30 of the studies for the following reasons: publication in language other than English,47 adult rather than child or youth subjects,34 absence of concurrent comparison group that had not received sealants,19,25,37,40,42,43,46 intervention not involving placement of resin-based sealants on permanent posterior teeth with no reapplication,26,27,35,38,39,44,48 absence of description of caries status by retention status17,18,20,21,23,28,30,31,33,41 and absence of a split-mouth design.16,24,36,45
, page 416), including loss to follow-up and validity (caries assessment method) and reliability (examiner agreement) of caries status determination. Because studies involved randomized controlled trials with a split-mouth design, we determined it to be unlikely that initial assembly and maintenance of comparable groups was an issue. We also should note that it is difficult to blind examiners as to whether a sealant was placed or not placed unless the sealant was removed before follow-up, a scenario that is not typical in most sealant studies.
|
FS teeth included teeth that had fully or partially lost sealant material. We also estimated the 95 percent confidence interval (CI) for the RR reported in each study by assuming that paired teeth were independent (further information about estimation of the variance is available as supplemental data to the online version of this article, found at "http://jada.ada.org"). We also collected data on the percentage of FS teeth on which the sealants were partially lost. Synthesis of findings. To estimate the mean RR, we used a weighted bivariate model in which we weighted each study by the reciprocal of its squared standard error. We also calculated the median RR across studies. To determine if the weighted bivariate analysis was heterogeneous, we calculated the quantity I2.50
| RESULTS |
|---|
|
|
|---|
Subjects ages ranged from 5 through 14 years. In three studies, investigators reported that subjects were exposed to fluoride via community water systems or mouthrinse program participation,13,14,22 two studies reported no fluoride exposure29,32 and two studies did not report background fluoride exposure.15,49 Caries incidence among NS teeth at the first-year follow-up examination ranged from 24 to 47 percent. Researchers in all studies used visual or tactile methods or both to assess caries; however, those in one study also used radiographs.49 For studies with more than one examiner, reported agreement among examiners (one study did not report agreement32) was greater than 90 percent. Loss to follow-up ranged from 5 to 21 percent for the five studies in which researchers conducted their first follow-up examination one to 1.5 years after placement13–15,22,29 and from 19 to 37.5 percent for the two studies in which investigators conducted their follow-up examinations four years after placement.32,49
Our findings suggest that heightened concern about partially lost sealants trapping food and thus increasing the risk of caries development may be unfounded.
For studies in which researchers reported sealant loss at the tooth level versus the site level,13–15,29,49 the mean percentage of FS teeth accounted for by partially lost sealants was at least 60 percent, up to and including three years after placement (Table 2
). The mean percentage of FS teeth accounted for by partially lost sealants declined over time, and there did not appear to be a difference according to generation of sealant material. Two studies reported retention at the site level (pit and fissure; data not shown)22,32; in one of them,22 the proportion of FS teeth accounted for by partially lost sealants was 27 percent one year after placement and 32 percent two years after placement, and in the other study,32 it was 32 percent four years after placement.
|
|
| DISCUSSION |
|---|
|
|
|---|
Theoretically, it is possible that partially retained sealants may offer some protection,51 especially if a specific tooth site remains sealed. Indeed, in one study included in our analysis, Horowitz and colleagues32 found that sealant effectiveness increased with the extent of retention. One possible explanation as to why our review did not find an association is that the unit of observation (tooth) used in most studies was not sufficiently sensitive to detect a difference. For example, let us assume that all teeth without sealants develop caries and that 10 teeth, each with two sites, are sealed while their contralateral teeth remain unsealed. If one site on each tooth lost its sealant while the other site remained sealed, then the RR calculated at the tooth level would be 100 percent/100 percent = 1, while the RR at the site level would be 50 percent/100 percent = 0.5. In the two studies that used site as the unit of measurement, both22,32 had an RR of less than 1. However, only one study22 did not include 1 in the 95 percent CI.
We compared the caries in FS teeth with that in NS teeth at the individual tooth level. It is important to note, however, that at the community level, the relevant question is not a direct comparison of caries rates in FS and NS teeth but rather a comparison of the caries rate in the group with sealed teeth (FS teeth plus fully retained sealants) versus the caries rate in the group with NS teeth. It must be remembered that the caries rate in the group with sealed teeth is based on the sealant loss rate and the caries rate in teeth that lost sealants (that is, FS teeth). Researchers conducting a systematic review that included only studies in which lost sealants were not reapplied found that sealants reduced caries by more than 70 percent.10 This finding indicates that the sealant loss rate multiplied by the caries rate in the group with FS teeth is less than the caries rate in the group with NS teeth or, equivalently, that the benefits of delivering sealants to children for whom follow-up cannot be ensured exceed the potential risks. Additionally, the findings of our study indicate that at the individual tooth level, the risk of caries development in FS teeth does not exceed that in NS teeth.
Because current guidance recommends sealant placement only when there is a risk of caries development52 and because sealant effectiveness is linked directly to retention,5 the maximum protection against caries can be achieved when a sealant is fully retained. Our findings do not suggest that practitioners can be any less careful in their sealant-application technique or in the evaluation or maintenance of sealants after placement in clinical practice. Our findings, however, do suggest that a child should not be deprived of the benefits of a sealant even when follow-up care cannot be ensured.
If we consider Cochrane inclusion/exclusion criteria for study design6 as the gold standard, then the overall quality of studies included in this review was good. Of the four studies included in this review that were not in the Cochrane review,6 three22,29,49 were randomized controlled trials and had dropout rates meeting the Cochrane criteria. Of these three studies, two were excluded from the Cochrane review because they did not meet the intervention criteria of RB2 sealant material,29,49 and one was excluded because the children in the study participated in a biweekly mouthrinse program.22 One additional study had a four-year dropout rate of 37.5 percent.32 The Cochrane review excluded studies with three-year dropout rates exceeding 30 percent and did not specify a threshold for four years after sealant placement.
One limitation of this analysis was the finding of heterogeneity for pooled results two to four years after sealant placement. The presence of heterogeneity suggests that there were significant differences between studies. These differences may not be as important in this study, in which our primary purpose was to determine if the preponderance of evidence indicated that FS teeth were at greater risk of developing caries than were NS teeth. We were not trying to obtain a precise point estimate of effect. For four13,22,32,49 of the seven studies included in this review, the point estimate of the RR for each year since sealant placement was always less than 1. In only one14 of the remaining three studies was the RR consistently above 1, and in that study the highest point estimate of the RR was 1.186.
Finally, we limited our search to studies included in systematic reviews of sealant effectiveness. For this analysis, we chose a less resource-intensive method to identify and screen potential studies. This approach is attractive because it provides an efficient method of collecting data from well-conducted studies. The studies included in systematic reviews have met rules of study design, conduct and measurement. In addition, we minimized bias in selecting studies for this analysis because the authors of the original systematic reviews determined the universe of studies. Inclusion and exclusion criteria in this analysis were explicit, and we specified them before screening available studies.
All but one14 of the studies included in this analysis were published in the 1970s, when flouride exposure was lower. Furthermore, in some of the studies we included,14,29,32,49 researchers used a generation of sealant material (RB1) that no longer is commercially available in the United States. It is unlikely, however, that these factors influenced our findings. Among this group of studies, the RR did not appear to vary according to background fluoride exposure or generation of sealant material.
| CONCLUSION |
|---|
|
|
|---|
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
B. F. Gooch, S. O. Griffin, S. K. Gray, W. G. Kohn, R. G. Rozier, M. Siegal, M. Fontana, D. Brunson, N. Carter, D. K. Curtis, et al. Preventing Dental Caries Through School-Based Sealant Programs: Updated Recommendations and Reviews of Evidence J Am Dent Assoc, November 1, 2009; 140(11): 1356 - 1365. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. H. Beaumont SEALANTS REVISITED J Am Dent Assoc, August 1, 2009; 140(8): 969 - 970. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |