Validation of Size Guides for Fitting Molar Bands to Stainless Steel Crowned Primary Molars

Abstract

Purpose:

To create size guides for fitting Denovo^® (Baldwin Park, CA) plain molar bands and bands with tubes to 3M^™ ESPE^™ (St. Paul, MN) stainless steel crowned primary molars and validate the guides.

Methods:

Two size guides were created for fitting these bands to corresponding crown sizes, with an additional pre-existing guide available for comparison. Eight volunteer dentists chose their best fit band sizes for stainless steel crowned typodont teeth, with each seeing a subset of clinical conditions as determined by a randomized factorial experiment. Comparisons of dentists’ best fits and each guide individually were evaluated over 96 fits for robustness of clinical condition in an analysis of variance and agreement using intraclass correlation. Savings in attempted fits using size guides were estimated.

Results:

No practical significant differences over clinical conditions were observed when evaluating size guides and dentists’ best fits. Intraclass correlations between plain band, plain band with tubes, and the pre-existing plain band with tubes guides and observed fits were 0.997, 0.998, and 0.998, with a corresponding expected savings of 1.1, 0.7, and 0.5 attempts per fit.

Conclusion:

Size guides created for fitting plain bands and bands with tubes to stainless steel crowned primary molars were validated by volunteer dentists and showed promise for decreasing waste and increasing efficiency during dental visits.

Introduction

A posterior primary molar lost prematurely may require a space maintainer to hold the space until the permanent tooth erupts. A fixed unilateral appliance called a band and loop is often utilized to help negate the mesial shift that occurs as a result of eruption of the permanent first molar, as well as distal drift from the canine.¹ One way to fabricate a band and loop space maintainer is to fit a plain molar band to the abutment molar, and then take an impression for future laboratory fabrication. An extension of stainless steel is soldered to the existing band during laboratory fabrication, and usually results in an additional appointment for the patient. This type of fixed appliance is considered a laboratory fabricated band and loop. The alternative is to fit a molar band that contains pre-soldered tubes. A separate prefabricated extension arm is fit to size, and inserted into the tubes. This type of space maintainer requires only a single appointment, and is considered a chair-side band and loop. Fitting bands for space maintainers, regardless of system used, is generally done by trial-and-error.¹ Often, the molars used as abutments have been restored with stainless steel crowns (SSC), and because prefabricated crowns come in sizes, the decision was made to create a size guide that will give the practitioner the band/band with tube size to fit each respective primary molar crown size.

The focus of this study was to evaluate sizing of plain molar bands as well as bands with pre-soldered tubes when fit to stainless steel crowned primary molars. Each time a space maintainer band is tried on and determined to not be the adequate size, it must either be reprocessed or discarded. Many clinical settings will repurpose these rather than discard, as Denovo^® (Baldwin Park, CA) provides processing instructions, with the exception of bands that have been crimped or altered in any way.² The Nemours Children’s Health, Delaware Valley hospital-based program, however, recognizes the bands as single-use. This has become more common in hospitals, as reprocessing can be time consuming and labor intensive for centralized processing staff, especially as many hospitals are currently short staffed and experiencing high staffing turnover.³ Each plain molar band costs approximately $4.35⁴ and each band with arm costs approximately $7.75.⁵ Multiple band trials can result in not only material lost, but also chair time lost. For these reasons, this study set out to create size guides to help decrease waste and increase efficiency as there were no published size guides available to the authors’ knowledge. After the data collection phase, however, the research team learned that a private practitioner, E. LaRee Johnson, DDS, MS, had created a size guide specific for fitting chair-side bands with tubes to primary molar crowns. This author’s data were included in this study. The purpose of this study was to not only create size guides, but to also validate or invalidate them to assess clinical applicability.

Methods

This research project was reviewed by the Institutional Review Board of Nemours Children’s Health and received exempt status (IRB #1809038). Two band fitting size guides were created. The first guide, band size guide (BSG), consisted of plain general purpose Denovo^® molar bands that would fit to respective 3M^™ ESPE^™ (St. Paul, MN) primary molar SSC. The second guide, band with tubes size guide (BTSG), consisted of Denovo^® bands with pre-soldered tubes that would fit to respective 3M^™ ESPE^™ primary-sized SSC. All sizes were mutually agreed upon by two dentists. If there were instances where two sizes were considered to be clinically adequate, the smaller band was chosen. Eight volunteer dentists were then recruited from within the institution to participate in fitting and choosing their best fit band sizes for both band types, referred to as the dentist’s best fit (DBF). Following informed consent, volunteer dentists fit bands to a randomized subset of different sizes of stainless steel crowned typodont teeth on a benchtop typodont. They had band seaters available if needed. The number of band fit attempts for both systems was also tracked. The subjects did not have access to the authors’ size guides. The data collected were then analyzed against the size guides in an attempt to either validate or invalidate the guides. Fit was also analyzed against a pre-existing size guide, Johnson’s band with tubes size guide (JBTSG).

Experimental Design.

A factorial experiment using block confounding with replication was implemented for data collection of band fitting, first for the plain molar bands, and then for the bands with pre-soldered tubes. Forty-eight band-fit conditions (maxillary [MX]/mandibular [MD] × primary first molar [PFM]/primary second molar [PSM] × left/right × size)⁶ represent the clinical conditions for fits. Within an individual block only, a subset of 12 clinical conditions were fit in consideration of dentists’ time and the potential for fatigue effects over repetitive trials. A unique block of clinical conditions was determined for each dentist using block confounding. To achieve that structure, size was artificially re-expressed in two factors small/large × level (1, 2, 3) within each of small/large. The confounding scheme determined four conditions in each block, where blocks were confounded with three interaction terms MX/MD × PFM/PSM × left/right, MX/MD × PFM/PSM × small/large, and left/right × small/large. Then, for each of the four conditions within a block, levels 1, 2, and 3 were paired, resulting in 12 conditions within a block performed by one dentist. All 12 conditions within each block were completely randomized. A full replicate, retaining the same confounding scheme but new randomization, was conducted with four additional dentists. This design confounded three interactions with blocks, but also guaranteed that dentist pairs, one from each replicate, would each see the same set of 12 clinical conditions, and that overall, two replications of the 48 conditions would be run.

Responses.

For each condition (assigned SSC crown size), dentists were free to fit bands until obtaining what they determined to be their individual DBF in available increments of 0.5 from a selection of Denovo^® bands ranging in size from 23.5 to 38. This size was recorded, together with condition specifics (e.g., MX/MD, PFM/PSM), the number of attempts made, and the BSG model value that was previously determined by the authors. The BSG, DBF, the difference BSG less DBF, and the number of attempts were each examined during the analysis.

Analysis.

Descriptive statistics summaries for BSG, DBF, and their difference were produced. Through analysis of variance consistent with the factorial model described, clinical condition factors were evaluated for their effects on the difference between BSG and DBF. The artificial construct of level was suppressed in the analysis to increase the degrees of freedom for error. Significant effects for the difference measure were explored further using Tukey’s honestly significant difference post hoc tests for mean comparison. The interpretation of practical significant effects was to indicate disagreement between BSG and DBF and detract from BSG use under clinical conditions where those effects were observed.

Direct association of DBF (Y) and BSG (X) was investigated using simple linear regression over all fit conditions. Model suitability was assessed using explained variation R², residual error diagnostics, slope, and intercept. The intraclass correlation coefficient (ICC) with bias as provided by JMP version 16.0.0 (SAS Institute, Inc., Cary, NC 1989–2022) assessed absolute agreement between DBF and BSG over all conditions and dentists. Regression analysis was repeated for individual blocks of matching fit conditions shared by two dentists randomly assigned to the same block, and the ICC for each was reported separately.

Efficiency gained by employing the BSG first to achieve the DBF was examined. Descriptive statistics were produced to summarize the distribution for the number of attempts to achieve the DBF, and normality was assessed using the Shapiro-Wilk test. Sensitivity of the number of attempts to clinical conditions was first explored using the nonparametric Wilcoxon Rank Sum test or Kruskal Wallace test for two or more groups, respectively, with grouping defined by MX/MD, PFM/PSM, left/right, size, and dentist. Sizes 2–7 were treated categorically. Efficiency was calculated as the savings in number of attempts required when starting with the BSG and proceeding ideally in 1/2-sized steps to the DBF relative to the actual number of attempts recorded by dentists to achieve their DBF. The difference of the average number of attempts required by dentists without use of the BSG and the average anticipated number of attempts supported by BSG determined the savings.

For the band with tube configuration, the BTSG and the JBTSG were compared with the DBF in separate analyses. The experimental design with new randomization scheme and analyses proceeded exactly as described above.

Results

Size guides were created in tabular form, each containing band fittings of 24 sizes of primary molar stainless steel crowns (Table 1, 2, 3).

Table 1.

Denovo^® Plain Molar Band Sizes for Corresponding 3M^™ ESPE^™ Stainless Steel Crown Sizes^*

Arch**	Primary Molar Stainless Steel Crown Sizes**
	D2	D3	D4	D5	D6	D7	E2	E3	E4	E5	E6	E7
UR/UL (maxillary)	24	25	25.5	27.5	28.5	30	31.5	33	34	35	36.5	38
LR/LL (mandibular)	23.5	25	26	27	28	29.5	30.5	32	33	34	35.5	36.5

^*Size guide developed for fitting Denovo^® plain molar bands to corresponding 3M^™ ESPE^™ stainless steel crowned primary molars.

^**Abbreviations based on 3M^™ ESPE^™ manufacturing labeling system: D=first primary molar, E=second primary molar, UR=upper right, UL=upper left.

Table 2.

Denovo^® Molar Band with Tube Sizes for Corresponding 3M^™ ESPE^™ Stainless Steel Crown Sizes^*

Arch**	Primary Molar Stainless Steel Crown Sizes**
Size	D2	D3	D4	D5	D6	D7	E2	E3	E4	E5	E6	E7
UR/UL	24	25.5	26	27.5	29	30.5	31.5	33	34.5	35	36.5	38
LR/LL	23.5	25	26	27	28.5	29.5	30.5	32.5	33.5	34	35.5	36.5

^*Size guide developed for fitting Denovo^® plain molar bands with tubes to corresponding stainless steel crowned primary molars.

^**Abbreviations based on 3M^™ ESPE^™ manufacturing labeling system: D=first primary molar, E=second primary molar, UR=upper right, UL=upper left.

Table 3.

Johnson Denovo^® Molar Band with Tube Sizes for Corresponding 3M^™ ESPE^™ Stainless Steel Crown Sizes^*

Arch**	Primary Molar Stainless Steel Crown Sizes**
Size	D2	D3	D4	D5	D6	D7	E2	E3	E4	E5	E6	E7
UR/UL	24.5	25.5	27	28	29	30.5	31.5	33.5	34.5	35	37	38
LR/LL	24	25.5	26.5	27.5	28.5	30	31	32	33.5	34.5	35.5	36.5

^*Preexisting size guide developed for fitting Denovo^® molar bands with tubes to corresponding stainless steel crowned primary molars to use as a comparator to the band with tubes size guide by Dr. E. LaRee Johnson, DDS, MS.

^**Abbreviations based on 3M^™ ESPE^™ manufacturing labeling system: D=first primary molar, E=second primary molar, UR=upper right, UL=upper left.

Band Size Guide.

The distributions for DBF and BSG were symmetric with mean ± standard deviation [SD] (30.28 ± 4.33) and (30.38 ± 4.32), respectively. No outliers were present. The difference measure (0.10 ± 0.41) was also symmetric, but with one suspected outlier (−1) for dentist 2 at condition MD/PFM/left/size=6. The value was judged not sufficiently extreme to remove from the analysis. An error in following the design led to a missing value declared for dentist 5. Suppressing left/right, the value was imputed based on the identical DBF values for the three other instances of the MD/PSM/size=6 condition.

Analysis of variance found a single significant main effect: MX/MD (P=0.001). The MD condition resulted in slightly higher estimates for the BSG relative to the DBF, with mean difference ± standard error [SE] (0.219 ± 0.079), while the MX condition showed similar performance between the two fits BSG and DBF (−0.021 ± 0.079). The effect size, effect ± SE (0.240 ± 0.072), though statistically significant, represented a difference of less than 1/2 the possible sizing sensitivity to make a fit. Thus, relative to the range of sizes (23, 38.5) observed for DBF, the effect was considered noise with respect to the utility of the BSG.

The regression of DBF on BSG showed strong agreement between measures, with an intercept of −0.04, a slope of 1.00, and explained variation R² = 99.1%. The BSG appeared to be a good proxy for the DBF achieved by dentists with unlimited trials. Residuals showed no problematic structure, and the pure variation estimate for the SD was 0.42, which was less than the fit steps of 0.5. The ICC was 0.997 for absolute measure agreement over all data collected. Individual fit lines for each dentist showed close model agreement (Fig. 1). The ICC for absolute DBF and BSG agreement isolating on conditions shared by dentist pairs (blocks) (1, 5), (2, 6), (3, 7), and (4, 8) were 0.996, 0.998, 0.997, and 0.999, respectively.

Figure 1.

Similar individual dentist regression for DBF (Y) vs BSG (X) showing guide applicability among dentists and limited data variability about lines with slope approximately equal to one illustrating absolute agreement between BSG and DBF sizing among clinical conditions. BSG, band size guide; DBF, dentist’s best fit; ICC, intraclass correlation coefficient.

The distribution of the number of attempts was determined as non-normal (P<0.0001) and was slightly skewed right with median (3), minimum (1), maximum (6), mean (2.7), and SD (1.2). The distribution of the number of attempts compared between or among groups led to insignificant findings for MX/MD (P=0.164), PFM/PSM (P=0.745), left/right (P=0.628), size (P=0.983), and dentist (P=0.096). Thus, no special considerations for the number of attempts expected by a dentist were tied to a specific set of fit conditions. Had the BSG been used as a starting point to get to the ideal fit determined by the DBF, only 57 additional attempts would have been required for the 96 fittings, an average of 1.6 attempts per fit. This contrasted with the observed average number of fits of 2.7 for a savings of resources that would be expended to perform an additional 1.1 attempts per fit.

Band with Tubes Size Guide.

The distributions for DBF and BTSG were symmetric with mean ± SD (30.46 ± 4.27) and (30.54 ± 4.29), respectively, with no outliers. The difference measure (0.08 ± 0.38) was also symmetric with no outliers. Analysis of variance on BTSG-DBF found three significant main effects: MX/MD (P=0.015), PFM/PSM (P=0.007), and small/large (P=0.007). The effect sizes for the difference between conditions, effect ± SE, were for MX/MD (0.167 ± 0.067), PFM/PSM (0.188 ± 0.067), and small/large (0.188 ± 0.067), with conditions MD, PSM, and small having slightly larger BTSG values relative to DBF on average. Again, though statistically significant, clinical relevance was not observed due to the small effect sizes relative to fit adjustments of 0.5.

The regression of DBF on BTSG showed strong agreement between measures, with an intercept of 0.19, slope of 0.99, and explained variation R² = 99.2%. The BTSG appeared to be a good proxy for the DBF. Residuals were consistent with assumptions and the pure variation estimate for the SD was 0.38, and less than the fit steps of 0.5. The ICC was 0.998 for measure agreement over all conditions. Individual fit lines for each dentist showed close model agreement (Fig. 2). The ICC for absolute DBF and BTSG agreement isolating on conditions shared by dentist pairs (blocks) (1, 5), (2, 6), (3, 7), and (4, 8) were 0.998, 0.996, 0.999, and 0.996, respectively.

Figure 2.

Similar individual dentist regression for DBF (Y) vs BTSG (X) showing guide applicability among dentists and limited data variability about lines with slope approximately equal to one illustrating absolute agreement between BTSG and DBF sizing among clinical conditions. BSG, band size guide; BTSG, band with tubes size guide; DBF, dentist’s best fit; ICC, intraclass correlation coefficient.

The distribution of the number of attempts was determined as non-normal (P<0.0001) and was slightly skewed right with median (2), minimum (1), maximum (5), mean (2.3), and SD (0.8). The distribution of the number of attempts compared between or among groups led to insignificant findings for MX/MD (P=0.476), PFM/PSM (P=0.994), left/right (P=0.297), size (P=0.770), and dentist (P=0.057). Thus, no adjustment to the number of attempts expected by a dentist were tied to a specific set of fit conditions. With the BTSG used as a starting point to achieve the DBF, only 52 additional attempts would have been required for the 96 fittings, an average of 1.5 attempts per fit. This contrasted with the observed average number of fits of 2.3 for a savings of 0.8 attempts per fit.

Johnson Band with Tubes Size Guide.

The distribution for the Johnson guide was symmetric with mean ± SD (30.79 ± 4.15) and no outliers. The distribution of DBF was reported under band with tubes. The difference measure (0.33 ± 0.42) was symmetric with no outliers. Analysis of variance on JBTSG-DBF found two significant main effects: MX/MD (P=0.038) and small/large (P=0.001). The effect sizes for the difference between conditions, effect ± SE, were for MX/MD (0.167 ± 0.079) and small/large (0.271 ± 0.079), with conditions MD and small having slightly larger JBTSG average values relative to DBF. Though statistically significant, small effect sizes relative to possible fit adjustments indicated no clinical relevance.

The regression of DBF on JBTSG showed strong agreement between measures, with an intercept of −1.07, slope of 1.02, and explained variation R² = 99.1% but with a slight positive bias. The JBTSG appeared to be a good proxy for the DBF. Residuals were consistent with assumptions and the pure variation estimate for the SD was 0.41, which was less than the fit steps of 0.5. The ICC was 0.998 for measure agreement over all conditions. Individual fit lines for each dentist showed close model agreement (Fig. 3). The ICC for absolute DBF and JBTSG agreement isolating on conditions shared by dentist pairs (blocks) (1, 5), (2, 6), (3, 7), and (4, 8) were 0.998, 0.998, 0.996, and 0.996, respectively.

Figure 3.

Similar individual dentist regression for DBF (Y) vs JBTSG (X) showing guide applicability among dentists and limited data variability about lines with slope approximately equal to one illustrating absolute agreement between JBTSG and DBF sizing among clinical conditions. BSG, band size guide; DBF, dentist’s best fit; ICC, intraclass correlation coefficient; JBTSG, Johnson’s band with tubes size guide.

The distribution of the number of attempts for bands with tubes was again leveraged in assessing the benefit of the JBTSG. Using JTBSG as a starting point to reach the ideal fit determined by the DBF, 82 additional attempts would have been required for the 96 fittings, an average of 1.9 attempts per fit. This contrasted with the observed average number of fits of 2.3 for a savings of 0.4 attempts per fit.

Discussion

Molar bands have many uses in pediatric dentistry and orthodontics. This study’s focus was limited to two different types of Denovo^® molar bands fit to 3M^™ ESPE^™ primary molar stainless steel crowned typodont teeth. The plain bands and bands with soldered arms are both commonly used for fixed unilateral space maintenance. The authors created size guides for fitting these bands to crowned teeth as a way to decrease waste and increase efficiency and attempted to validate or invalidate them by testing their potential clinical acceptability among volunteer dentists. The authors became aware of a previous existing size guide for fitting bands with tubes to crowned primary molar teeth. With that dentist’s permission, this guide was also tested against the data collected from the dentist fittings. Although a size guide for bands with tubes did exist, a plain band size guide did not, and no validation attempts for guides had previously been characterized.

In relation to the DBF, a limited number of significant effects (six), based on condition factors or their interactions, were identified over the three guides, BSG (Table 1), BTSG (Table 2), and JBTSG (Table 3). However, in each case, the effect size differences were sufficiently small with respect to feasible fitting sensitivity that they were deemed not to have clinical significance. All three size guides were determined to be proxies for the DBF over an unlimited number of attempts.

When number of fit attempts observed were considered, utilization of the guides in relation to DBF could have resulted in a savings of an average of 1.1 attempts per fit with the BSG, 0.8 attempts per fit with the BTSG, and 0.4 attempts per fit with the JBTSG. Decreased fit attempts can decrease chair-time, potentially contributing to patient experience/satisfaction while also contributing to provider efficiency. Decreased procedure times involving oral rehabilitation specifically could also lend to decreased exposure of sedation medications or general anesthetics. Furthermore, when considering the above savings in fit attempts, if this hospital-based volunteer group were to utilize the BSG, BTSG, and/or JBTSG, a potential cost savings per molar fitting of $5.27⁴, $6.20⁵, and $3.10⁵, respectively, could be appreciated as this hospital system recognizes them as single-use items. Discarding trial fits lends to direct waste and requires staffing resources for restocking and repurchasing. Meanwhile, reprocessing can be tedious² and time consuming, especially in offices that are currently affected by the national staffing shortages.³ Minimizing the number of discarded and/or reprocessed bands is a cost saving measure.

Limitations

Molar band retention relies heavily on the elasticity of the band material and it’s fit to the tooth.⁶ Retention is also supplemented by the luting agent; however, sole reliance on cement-dependent retention is not advised.⁶ Additionally, malleability of the bands allows for crimping and contouring to help with retention and minimization of cement wash out. Comparison of band fit in relation to retention rates has not been characterized to our knowledge. During size guide creation, when there were two band sizes considered to be clinically acceptable for a respective crown size, the authors of BSG and BTSG erred toward the smaller fit. On the other hand, the author of the JBTSG erred toward the larger fit, with the expectation that crimping and contouring would be applied to aid in retention. Further studies are needed to evaluate retention rates based on the 0.5 mm size differentials. Further studies could also explore potential application with different band and crown manufacturers.

Conclusions

Based on this study’s results, the following conclusions can be made:

1. Nemours Children’s Health, Delaware Valley size guides for fitting plain molar band and bands with tubes could be utilized as a chair-side adjunct starting point for initial band size selection.

2. The Johnson size guide for fitting bands with tubes could also be utilized as a chair-side adjunct starting point for initial band size selection.

3. Size guide utilization has the potential to minimize the number of band trial fits, decreasing reprocessing/waste and increasing provider efficiency.

4. Future studies are needed to determine retention rates when more than one band size might be considered clinically acceptable by clinicians to help determine selection.