A clinical trial of patient satisfaction and prosthodontic needs with ball and bar attachments for implant-retained complete

overdentures: Three-year results

Michael I. MacEntee, LDS (I), PhD,a Joanne N. Walton, DDS,b and Ned Glick, PhDc University of British Columbia, Vancouver, Canada Statement of problem. Few prospective trials of implant-retained mandibular dentures have evaluated the increase and
duration of patient satisfaction, costs of denture maintenance in relation to different methods of attaching overdentures to implants, or the use of a reinforced framework. Purpose. This report evaluates subjects satisfaction and prosthodontic maintenance during a 3-year randomized clinical trial of implant-retained mandibular complete dentures, whether reinforced or not with a cast framework, and attached by bar-clip or 2.25-mm ball-spring matrices to endosteal dental implants. Materials and methods. One hundred edentulous subjects, each having at least 1 years experience with conventional complete dentures, were selected from respondents to a university dental clinics request for volunteers. Candidates were examined to verify adequate mandibular bone and medical suitability for implants. Subjects then received 2 implants in the anterior mandible before being stratied by mandibular bone height and gender and assigned randomly to 1 of 4 treatment groups. Every subject received a new maxillary complete denture in addition to an implant-supported mandibular complete denture, with or without a reinforcing framework, connected to implants by either a bar-clip or a ball-spring patrix and matrix. The dentures were adjusted and repaired as needed. Subjects indicated on a visual analogue scale (VAS) satisfaction with conventional dentures prior to the study and then with new dentures at 1 month, 1 year, and 2 years. The results reported here are from the rst 68 subjects observed for 3 years after receiving new dentures (19 subjects received new dentures less than 3 years before this analysis, and another 13 subjects were lost to follow-up). VAS scores are presented in simple tables and graphs, and results for different groups were compared using 2-sided nonparametric rank tests and repeated measures ANOVA. With respect to costs and maintenance, t tests were used to compare group means. Sample size and other design considerations used a .05 signicance level. Results. After receiving new dentures with mandibular implant supports, improved satisfaction within subject was prompt, durable, substantial, and statistically signicant, regardless of the attachment mechanism, and with or without a reinforcing framework. In contrast, there were no notable satisfaction differences between the 2 attachment mechanisms, or with the presence or absence of a reinforcing framework, either at specic intervals after receiving the new dentures, or in repeated measures ANOVA. For both attachment groups, most denture adjustments occurred during the rst year. This accounted for 81% of total adjustments during 3 years, when the 34 subjects in the ball-spring group and the 34 in the bar-clip group were combined. The mean numbers of adjustments per subject and associated clinical times did not differ signicantly between the 2 groups. Conversely, denture repairs declined more slowly than adjustments. Almost all repairs (90%) occurred in the ball-spring group to correct problems with the attachments, 39% in the rst year, and tapering off only slightly in the following 2 years. Over 3 years of follow-up, mean numbers of repairs per subject differed signicantly between groups: 6.7 repairs per person in the ball-spring group, compared to 0.8 in the bar-clip group (P,.001), and mean time per appointment was greater for repairs in the ball-spring group: 18.9 minutes compared to 16.9 (P,.01). The cast framework had no inuence on the satisfaction expressed or on adjustments and repairs. Conclusion. Subjects were very satised with the new dentures, although the ball-spring attachment tested in this trial required substantially more repairs. (J Prosthet Dent 2005;93:28-37.)

CLINICAL IMPLICATIONS
Implant-supported mandibular dentures can satisfy edentulous patients, and the visual analogue scale helps to evaluate a patients specic concerns, such as masticating and stability. A cast framework in the denture base does not improve overdenture maintenance.

I
This study was supported by Health Canada, National Health and Research Development Program (Project no. 6610-2061-403) and Nobel Biocare Canada, Inc. a Professor, Department of Oral Health Sciences, Faculty of Dentistry. b Professor, Department of Oral Health Sciences, Faculty of Dentistry. c Professor Emeritus, Department of Health Care and Epidemiology, and Department of Statistics, Faculties of Medicine and Science.

mplant-retained removable complete overdentures offer an effective rehabilitative treatment for edentulous mandibles,1,2 although sometimes difcult to maintain without frequent adjustments and repairs to the attachment mechanisms between the implants and the dentures.3-7 Attachment components are prone to fracture, distortion, and disengagement, consequently, several designs have been suggested to reduce the need for
VOLUME 93 NUMBER 1

28 THE JOURNAL OF PROSTHETIC DENTISTRY

MACENTEE, WALTON, AND GLICK

THE JOURNAL OF PROSTHETIC DENTISTRY

maintenance.8,9 There are 2 dominant designs for the attachment at present.5-8 One uses plastic or metal clips in the denture base to engage a metal-alloy bar connected rigidly to both implants, while the other consists of a metal cap or a rubber ring embedded in the denture base to engage a ball abutment attached independently to each implant. The metal caps usually have either a spring or an adjustable claw to retain the denture on the abutment. Biological consequences of stress induced in the bone supporting the implants by either attachment design are controversial,10-14 and there is no empirical evidence that the prosthodontic contribution of any particular attachment design is superior to the other. The need to reinforce the acrylic resin denture base to improve the strength of implant dentures is not clear. One study15 indicated that 60% of the prosthetic complications observed were related to problems with the acrylic resin matrix, although there was no mention of whether or not a metal framework was used in the prostheses. While some researchers have indicated that a reinforcing framework, typically in stainless steel or a chromium cobalt alloy, should be used to reinforce the denture,10,16 at least 1 group17 has since changed that opinion and stated that the extra costs associated with the reinforcement may not be justied. This randomized clinical trial was conducted to evaluate the 2 common attachment modes and to identify the benets of reinforcing the denture base. Previous reports have been published on patient satisfaction as well as costs of adjustments and repairs up to 1 year after the subjects received new dentures.18 Mandibular dentures supported by 2 endosteal dental implants were retained either by a bar with 2 clips or by 2 ball attachments with metal caps. Fabrication time, number of appointments, and chair time for adjustments were similar for each design, but the ball-attachment dentures required many more repairs, and reinforcement of the denture base had no apparent inuence on treatment outcomes. In all, after 1 year, the subjects in both treatment groups were comparably satised with the dentures, although the bar-clip design had been easier to maintain. This article reports further results from the same clinical trial. The new analyses consider prosthodontic maintenance required over 3 years, and evaluate several aspects of subjective satisfaction over 2 years with the new dentures.

MATERIAL AND METHODS

Subjects were recruited by advertising and sending information about the planned trial to dentists and denturists throughout the province, to part-time faculty at the University, and to various organizations serving elderly persons in the region. The Universitys Clinical Research Ethics Board approved the research protocol, including recruitment procedures, exclusion/ inclusion criteria, and the informed consent. Respondents were
JANUARY 2005

interviewed in telephone conversations by a dental assistant associated with the study, and then mailed written information about the trial. Subsequently, volunteers attended the university Oral Implant Clinic and were screened jointly by a surgeon and a prosthodontist, each a specialist with at least 10 years of clinical experience. The inclusion criteria required that participants be: edentate, with at least 1 year of experience wearing conventional complete dentures, medically and psychologically suited for implant surgery, able to complete study forms and to communicate orally in English, and be willing to commit to 2 years of participation in the trial after receiving new dentures. Volunteers were excluded from the study if they had: received treatment previously with oral implants, need for additional preprosthetic surgery, insufcient bone height for at least an 8.5-mm mandibular implant, or a history of systemic or neurologic disease or head and neck radiation. Some volunteers were not enrolled in the trial after implant surgery because the prosthodontist found that the implants diverged more than 15 degrees from each other, or that the implants were located less than 20 mm or more than 35 mm apart, because of evidence that such an orientation and location of implants could disturb the stability and maintenance of the implant-retained denture.19,20 Each subject paid approximately $1300 (US) for implant placement and denture fabrication, although all follow-up care was free of charge for the rst 2 years, and at cost recovery beyond that period. Enrollment of 100 subjects was calculated for good statistical power (anticipating some dropouts) to detect a differential of 25 percentage points or more between the proportions requiring repairs in the 2 attachment groups. Subjects received either the ball or bar-clip attachment for mandibular dentures, using an acrylic resin base either with or without a cast metal alloy framework embedded within the resin, and the treatments were assigned randomly in blocks of 4 subjects (via random permutations, generated in an Excel spreadsheet), using distinct randomizations within each of the 4 strata based jointly on gender and on ridge resorption (severe or normal). The primary intent of this stratied randomization was to promote overall comparability of the ball and barclip attachment groups. Previous studies indicated a preponderance of women seeking oral implants, so there were concerns about a greater prevalence of osteoporosis and severe ridge resorption in women and, also, about the possibility that men might apply more physical stress to the dentures. Residual ridge resorption of each subjects mandible was classied as severe if either mental foramen on a panoramic radiograph was located at the crest of the residual ridge or as normal if both foramina were below the crest of the ridge. Five surgeons, 1 prosthodontist, and 3 certied dental assistants provided clinical care. The surgeon placing
29

THE JOURNAL OF PROSTHETIC DENTISTRY

MACENTEE, WALTON, AND GLICK

Fig. 1. A, Bar-clip attachment showing metallic clip embedded in denture base. B, Noble alloy bar splinting 2 implants.

Fig. 2. A, Cap-spring attachments with C spring within titanium alloy cap-embedded in denture base. B, To engage 2.25-mm ball-spring attachment.

implants did not know the subjects future prosthodontic treatment, but blinding was not possible for either the patients or the prosthodontist. Each surgeon placed the endosteal dental implants (Nobel Biocare Canada Inc, Richmond Hill, Ontario, Canada) bilaterally in the lateral incisor or canine area of each subjects mandible. Soft tissue healing and implant stability were monitored to ensure good healing and support for the dentures. In essence, 4 months of healing were allowed before loading the implants with a removable implantretained denture and constructing a new maxillary denture. The mandibular denture frame was waxed on the residual ridge of the denitive cast, and cast in a chrome-cobalt alloy (Vitallium; Dentsply Int, York, Pa) before packing the acrylic resin. The denture bases were processed with a heat-polymerized acrylic resin (Lucitone 199; Dentsply) and supported acrylic resin teeth (Bioform; Dentsply) arranged in lingualized occlusion.21 A noble alloy bar (Round Gold Bar System; Nobel Biocare Canada, Inc) splinted the 2 implants together in the bar-clip treatment group to retain and stabilize the overdenture by engaging a gold alloy matrix clip (Nobel Biocare Canada, Inc) embedded in the denture base (Fig. 1). In the ball-spring group, a 2.25mm noble alloy matrix (Nobel Biocare Canada, Inc)
30

Table I. Baseline characteristics of subjects followed in the clinical trial for 3 years
Treatment group Characteristic Bar-clip Ball-spring (n = 34) (n = 34) Chi-square test (1 df ) to compare groups v2 P value

Women* With severe ridge resorption* Married In excellent vs. good or fair general health (self-reported) Monthly income > $1500 US Edentate 4 years or more

65% 73% 59% 38% 59% 85%

62% 68% 77% 18% 53% 91%

0.06 0.28 2.42 3.58 0.24 0.57

.50 .40 .10 .05 .40 .36

*Gender (female or male) and ridge resorption (severe or normal) were stratication variables.

placed onto each implant supported and stabilized the mandibular overdenture by engaging a metal C spring within a titanium alloy matrix (Nobel Biocare Canada, Inc) embedded in the denture base (Fig. 2). The maintenance required for each prosthesis was recorded either as an adjustment or a repair. Adjustment indicated a modication that did not add new material
VOLUME 93 NUMBER 1

MACENTEE, WALTON, AND GLICK

THE JOURNAL OF PROSTHETIC DENTISTRY

Fig. 3. A, Self-reported overall satisfaction with mandibular complete denture (CD) at baseline, and with implant-supported denture (ISP) at 1 month, 1 year, and 2 years, reported by subjects on visual analogue scale (VAS). B, Self-reported overall satisfaction with maxillary complete denture reported by subjects on visual analogue scale (VAS) at baseline, and at 1 month, 1 year, and 2 years with mandibular implants. Table II. Median satisfaction scores, reported by subjects using visual analogue scales (VAS), comparing baseline scores (with previous dentures) to scores after 1 month with new dentures, for subjects within each of 2 treatment groups
Median VAS score (0 = least satised; 100 = most satised) Bar-clip group (n = 34) Baseline with previous denture 1 month with new denture Change* P value Baseline with previous denture Ball-spring group (n = 34) 1 month with new denture Change* P value

Satisfaction relating to: Pain: maxilla mandible Comfort: maxilla mandible Appearance: maxilla mandible Mastication: maxilla mandible Stability: maxilla mandible Speech Hygiene: maxilla mandible Overall: maxilla mandible

91 41 88 25 81 45 79 18 84 7 67 89 89 87 13

94 91 94 88 94 93 92 84 89 92 95 96 95 95 93

.19 ,.001 .01 ,.001 .001 ,.001 .02 ,.001 .05 ,.001 ,.001 .001 .13 .007 ,.001

94 51 86 23 79 64 78 34 80 14 75 90 85 82 13

95 93 92 86 95 95 87 91 92 94 93 96 95 95 94

.68 .001 .12 ,.001 .003 ,.001 .02 ,.001 .12 ,.001 .002 .001 .001 .001 ,.001

*Changes for scores within each group were tested using 2-sided Wilcoxon signed-rank tests.

or replace broken or missing components, whereas repair indicated the addition or replacement of material or teeth. The time required to complete an adjustment or repair was recorded on each occasion. Subjects were asked to commit to the study for 2 years after receiving new dentures, during which time the investigators could record the prosthetic maintenance, assess the health of the implants and related soft tissues, and record subjective satisfaction. However, the incidences and types of
JANUARY 2005

prosthetic maintenance were monitored further for those subjects who continued to attend the clinic after 2 years. Based on the results of previous research,2,3,5 baseline information was collected to assess comparability of treatment groups and to evaluate the possible effects of demographic variables, such as gender, age, selfreported general health, marital status, and income. At baseline (with the previous dentures) and then at 1
31

THE JOURNAL OF PROSTHETIC DENTISTRY

MACENTEE, WALTON, AND GLICK

Table III. Median satisfaction scores, reported by subjects using visual analogue scales (VAS), comparing satisfaction at 1 month and at 2 years after receiving new dentures for subjects within each of 2 treatment groups
Median VAS score (0 = least satised; 100 = most satised) Bar-clip group (n = 34) 1 month with new dentures 2 years with new dentures Change* P value 1 month with new dentures Ball-spring group (n = 34) 2 years with new dentures Change* P value

Satisfaction relating to: Pain: maxilla mandible Comfort: maxilla mandible Appearance: maxilla mandible Mastication: maxilla mandible Stability: maxilla mandible Speech Hygiene: maxilla mandible Overall: maxilla mandible

94 91 94 88 94 93 92 84 89 92 95 96 95 95 93

96 96 95 94 95 95 95 94 94 97 97 97 97 96 96

.45 .05 .54 .02 .79 .59 .13 .004 .10 .06 .01 .36 .07 .39 .15

95 93 92 86 95 95 87 91 92 94 93 96 95 95 94

96 95 94 92 96 95 95 94 94 88 94 96 97 96 93

.03 .16 .04 .27 .10 .41 .003 .25 .23 .24 .16 .30 .04 .06 .76

*Changes for scores within each group were tested using 2-sided Wilcoxon signed-rank tests.

Table IV. Numbers of adjustments and repairs to implant-supported dentures in 2 treatment groups, each with 34 subjects
Treatment group Ball-spring (n = 34) Year 1 Year 2 Year 3 Year 1 Bar-clip (n = 34) Year 2 Year 3

Adjustments to: Denture base Occlusion Tighten implant abutment Ball-spring Retentive clip Miscellaneous components Total adjustments Repairs: Denture reline Broken denture Replace denture tooth Ball-spring Ball-cap Dislodged or broken bar-clip Fractured bar Miscellaneous components Total repairs

62 23 17 23 6 131 4 2 1 44 35

9 1 3 1 1 15 3 1 53 27

6 0 2 1 3 12 2 2 27 8

86 18 2 45 6 157 3

7 1 1 12 1 22 2

12 1 17

3 89

2 86

15 54

4 2 2 11

month, 1 year, and 2 years after receiving the new dentures, each subject used a visual analogue scale (VAS) to express degrees of satisfaction with the dentures.22,23 The intent and operation of the VAS was explained or demonstrated to each subject. The scales assessed global or overall satisfaction, and also 7 specic (although correlated) components of satisfaction, regarding pain,
32

comfort, appearance, mastication, speech, stability, and oral hygiene. Each scale used a 100-mm uninterrupted line to represent a continuum of feelings, with complete dissatisfaction at one end of the line and complete satisfaction at the other, but no marks or indicators in-between. VAS scores are ordinal and continuous, but ratios may not be meaningful and scores may not be
VOLUME 93 NUMBER 1

MACENTEE, WALTON, AND GLICK

THE JOURNAL OF PROSTHETIC DENTISTRY

Table V. Distributions of denture adjustments and repairs and associated costs for subjects in 2 treatment groups over 3 years
Treatment group Bar-clip (n = 34) Ball-spring (n = 34) Difference t test* P value

Adjustments: Mean number per subject Mean clinical time in minutes per appointment Mean clinical time in minutes per subject Repairs: Mean number per subject Mean clinical time in minutes per appointment Cost of repairs in US dollars: Mean dental laboratory fee per subject Dental laboratory fee per subject who had repairs Dental laboratory fee per subject who paid for repairs
*2-sided t tests to compare treatment group means.

6.0 16.9 76 (n = 9) 0.8 16.9 $24 $101 $135

4.9 16.2 98 (n = 32) 6.7 18.9 $42 $44 $142

.20 .46 .21 ,.001 ,.01 .30 .09 .88

normally distributed; consequently, the statistical analyses of VAS scores used nonparametric tests rather than t tests. The Wilcoxon signed-rank test was used to assess changes in satisfaction within each group, and comparisons between groups used the Mann-WhitneyWilcoxon rank-sum test (2-sided testing at the .05 signicance level). Results reported here relate primarily to the 68 subjects (34 in each attachment group) for whom data were available for at least 3 years after new complete dentures with mandibular implant supports were received. Of the 100 subjects enrolled at baseline, 2 died, 5 withdrew, 6 were otherwise lost to follow-up, and 19 subjects had less than 3 years in the trial since receiving new dentures. The 2 attachment groups not only provided equal numbers of subjects, but also similar 3-year completion rates (ball-spring group: 67%; bar-clip group: 69%).

Fig. 4. C springs removed from titanium alloy cap, showing fractured spring on right side.

RESULTS
The 34 subjects followed for 3 years in each attachment group were representative of the respective randomized groups at baseline. Moreover, the baseline characteristics of the subjects in each group were similar, although more of the bar-clip group characterized themselves with excellent general health (Table I). There was no signicant difference (P=.42) between the mean ages of the 2 attachment groups (bar-clip group: 61 years; ball-spring group: 63 years). As anticipated, the women outnumbered men, and approximately 70% of subjects had severe residual ridge resorption. Typically, the subjects were married, reasonably afuent, and had been edentate for more than 4 years at baseline (Table I). Median VAS scores for overall satisfaction at baseline with the previous mandibular dentures did not differ signicantly between men and women or between subjects with and without severe ridge resorption and were identical for the 2 attachment groups: 13 (out of
JANUARY 2005

100 maximum) in the bar-clip group and 13 in the ball-spring group (Table II). With respect to the maxillary denture, median VAS scores for overall satisfaction at baseline with the previous dentures were notably higher, 87 in the bar-clip group, 82 in the ball-spring group (Table II), consistent with the authors emphasis placed on mandibular denture problems. Moreover, the median overall satisfaction with maxillary dentures at baseline was not signicantly different between the 2 groups (Table II). Reported median comparisons between groups used WilcoxonMann-Whitney rank sum tests, but comparisons of means using 2-sample t tests also gave similar results. Within each attachment group, at 1 month after receiving new complete dentures with the mandibular implant supports, the median VAS scores for satisfaction had increased compared to baseline with the previous dentures. Overall satisfaction and every measured component of satisfaction (mastication, comfort, and others) increased, both for mandibular and for maxillary dentures (Table II). Regarding the mandibular denture,
33

THE JOURNAL OF PROSTHETIC DENTISTRY

MACENTEE, WALTON, AND GLICK

Fig. 5. Distribution of repairs among subjects within 2 attachment groups.

overall satisfaction improved dramatically, from 13 to 93 in the bar-clip group, and from 13 to 94 in the ballspring group (Table II) (Fig. 3, A). These mandibular improvements in overall satisfaction, and also the improvements in all satisfaction components, were very signicant statistically (P=.001 or less, in 2-sided Wilcoxon signed-rank tests within each group), apart from the insignicant improvement for satisfaction with hygiene within the bar-clip group. Also, satisfaction with speech improved signicantly in each group (P,.001 for the bar-clip group; P=.002 for the ball attachment group) (Table II). Satisfaction scores at 1 month also were high with the new maxillary dentures, both overall and for specic components of satisfaction, although several improvements from baseline (with previous dentures) did not attain statistical significance or clinical importance, because the VAS medians with previous maxillary dentures were very high at the outset (Table II) (Fig. 3, B). Despite high VAS medians at 1 month after receiving new dentures (Table II) (Fig. 3), the comparable VAS medians were almost all even higher at 2 years (Table III) (Fig. 3), although most of the subsequent improvements were not statistically signicant (at level 0.05, using 2-sided Wilcoxon signed rank tests). As Figure 3 also suggests, overall satisfaction did not differ between the 2 attachment groups. Repeated measures analyses of variance found no signicant differences between the 2 attachment groups, with respect to overall satisfaction or any of the satisfaction variables, for either the mandibular or maxillary denture. The repeated measures analyses also produced prole plots that appear similar (Fig. 3) for overall satisfaction with mandibular and maxillary dentures respectively, except that Figure 3 shows median VAS scores rather than means. Also, Wilcoxon-Mann-Whitney rank-sum comparisons found no differences between the 2 attachment
34

types with respect to overall VAS medians, except for higher mandibular satisfaction in the bar-clip group 2 years after receiving the new dentures (Table III shows 96 versus 93, a difference that appears not to be important clinically). Most (81%) of the denture adjustments were made during the rst year, and two thirds to denture bases or occlusal contacts (Table IV). The ball-abutments loosened (n = 22) more frequently than the bar abutments (n = 3) from implants, whereas approximately three fourths (74%) of the adjustments to either retentive mechanism were associated with the clips on the implant bars. The mean number of adjustments, as well as the clinical time per subject for the adjustments, did not differ signicantly between the attachment groups (Table V). In contrast to the adjustments, the need for denture repairs declined more slowly and is one reason why the data were gathered beyond the original 2-year followup period planned originally for the trial. The rst year accounted for only 39% of all repairs during the 3 years after subjects received new dentures, and almost all repairs in each year were made in the ball-spring group to correct problems with attachments (Table IV). This very signicant nding was the outstanding difference between the bar-clip and ball-spring attachment groups. Dislodged or broken clips for the bar-clip group were replaced occasionally, but not nearly as frequently as the matrix with springs in the other group (Fig. 4). The mean numbers of repairs per subject differed very signicantly (P,.001) between the 2 groups (Table V). Also, the mean time per appointment was signicantly greater (P,.01) for repairs in the ball-spring group. Of 34 subjects in the ball-spring group, 11 individuals had 4 repairs each, and 7 others each had 5 or more repairs, with 1 subject needing as many as 10 repairs (Fig. 5). In all, these ndings were the outstanding
VOLUME 93 NUMBER 1

MACENTEE, WALTON, AND GLICK

THE JOURNAL OF PROSTHETIC DENTISTRY

differences between bar-clip and ball-spring attachment groups. Dental laboratory fees for repairs ranged widely (bar-clip group: $75-$310; ball-spring group: $52$433) and tended to be more costly for the bar-clip group as a whole. Not all repairs incurred a dental laboratory fee, however, the mean laboratory fees paid were comparable in both groups (bar-clip group: $135; ballspring group: $142). Among the 68 overdentures considered in this study, a cast framework was embedded in 28 (11 in the bar-clip group; 17 in the ball-spring group). Signicantly more subjects with a framework wore the dentures while sleeping (P=.03) but, otherwise, the gender division, brushing habits, duration of edentulism, and extent of residual ridge resorption of the framework and nonframework groups remained comparable after 3 years. There was no signicant difference between the framework and nonframework groups either with satisfaction or adjustments and repairs. Consequently, this factor was not considered further in the analysis.

DISCUSSION
The VAS is used typically to measure perceptions of subjective phenomena, such as pain, that are difcult to standardize from individual to individual. VAS scores have been used by others22 to measure satisfaction with dentures, particularly relating to comfort. There were reports that some subjects found the scale confusing because they were not sure how or where to mark the line, and difcult to use because they could not relate easily to the line as a measure of feelings.23 Subjects in the present study marked the scale independently, but in the presence of a research assistant who offered explanation or help as needed. In future, subjects might prefer to indicate VAS scores directly on a computer screen rather than on paper, which could reduce both costs and human measurement errors associated with recording VAS data. Most of the subjects in this study were able to express with the VAS a useful variety of feelings, especially concerning their dentures, that predated enrollment in this clinical trial. For example, VAS scores differed for mandibular and maxillary denture satisfaction, and mandibular VAS scores for appearance and hygiene differed notably from scores for mastication and denture stability. Moreover, the similarity between overall satisfaction scores and scores for satisfaction with masticating and stability suggests that patients seeking implantretained dentures are motivated by concerns that are more practical than cosmetic. Additional studies are underway to explore and explain the factors that motivate requests for implant-retained dentures rather than conventional complete dentures. In the present study, implant overdentures led to satisfaction improvements that were strikingly large, which
JANUARY 2005

support the ndings in other studies.2,21 Such improvements, in part, probably reect the greater satisfaction with a new, in contrast to a previous, denture. However, it is likely that improvements indicate the specic benets of mandibular implant attachments. VAS summaries (Tables II and III) show that improvements in satisfaction with the new maxillary complete dentures were less statistically signicant and less clinically dramatic. The high frequency of repairs (Table IV) is a concern. The number of repairs during the rst year of the trial related particularly to implants that were inclined more than 6 degrees lingually or less than 6.5 degrees facially in the sagittal plane.14 This observation supports the manufacturers recommendation that multiple implants supporting individual ball attachments should be aligned nearly parallel to one another and to the path of insertion of the overdenture. An earlier study14 by 1 of the authors and colleagues found that horizontal alignment of the implants in the frontal plane had no apparent inuence on either repairs or adjustments. The generalizability of the results on the attachment design is limited to the 2 attachment designs tested; consequently, it is not appropriate to conclude that all ball or bar-clip attachments for implant-supported dentures will behave in like manner. The large number of denture repairs for the ball-spring attachment studied in this trial was due probably to a structural aw within the mechanism of the stainless steel C spring and the titanium alloy matrix. Both parts of the matrix separated on many of the occasions when the spring was lost. Although maintenance of the dentures during the rst year of another study was similar with each of 3 ballattachment systems tested, the authors of that study reported similar problems with a titanium alloy matrix made by another manufacturer.9 Perhaps frictional abrasion and stress within the smaller design, especially when it involves stainless steel against a titanium alloy, is too great in prolonged clinical use. During the rst year of another study, investigators found that an adjustable gold alloy matrix in the denture base attached to a ball patrix on implants required less maintenance than 2 metal clips on an implant-bar.8 It is possible that the gold alloy matrix, in contrast to a nonadjustable stainless steel spring, might distribute stress more evenly within the attachment and decrease the maintenance required. Although the need for repair was more likely with the 2.25-mm ball-spring attachment, patients with this attachment could take comfort from the fact that repair costs, when required, were comparable to those incurred with the bar-clip design. Each subject agreed to participate in the clinical trial for 2 years after receiving new dentures, but many subjects found it worthwhile to continue contacts with the clinic for 3 years or longer. Potentially, an additional 9 subjects when they complete 3 years with new dentures could supplement the 68 subjects in the present
35

THE JOURNAL OF PROSTHETIC DENTISTRY

MACENTEE, WALTON, AND GLICK

analyses. However, it is unlikely that further extension of the study would add much clinically relevant information. The present studys comparisons demonstrate very dramatic differences, for example, between improved satisfaction from conventional to mandibular implantsupported overdentures and huge differences between repair experiences for the 2 attachment groups. For most other comparisons in the study, differences obviously lack practical importance. In Table III, for instance, most satisfaction improvements that are statistically signicant are not important clinically. In Figure 3, the differences between attachment groups are not consistently in favor of either group, and all differences between the 2 groups are small. Indeed, differences in satisfaction between groups at each of 3 times with new maxillary dentures were much smaller than the indisputably random baseline difference between the 2 groups with previous dentures. Therefore, increasing statistical power in this context would offer no clear benet. The 2-year satisfaction data are clinically important to establish that the high median patient satisfaction at 1 month after receiving new dentures did not decline subsequently, and so could not be viewed as a transient placebo effect improvement over baseline. Moreover, improved satisfaction for at least 2 years is relevant to justication of implant costs in clinical practice. Finally, although diverse aspects of denture satisfaction may be correlated with each other and with overall satisfaction (Table II), it is noteworthy that mandibular denture satisfaction overall seems to parallel satisfaction with mastication and denture stability more than with appearance, speech, comfort, or even pain. In deciding when to recommend implants, and in justifying the costs, it may be helpful for clinicians to regard mastication as a dominant concern of patients and potentially a predictor of treatment outcomes.

fect on patient satisfaction, which was consistently high with the new dentures. 3. Nearly all repairs (90%) were made for the 2.25-mm ball-spring overdenture group; hence mean repairs per subject were signicantly greater in the ball-spring group (P,.001). 4. Adjustments to the implant-retained overdentures in both attachment groups occurred most commonly during the rst year of the trial and required, on average, similar clinical time for both groups. 5. Costs to those who paid for repairs were comparable in both attachment groups, although the clinical time per appointment for repairs was slightly higher for the ball-spring group. 6. The cast framework within the denture base had no inuence on adjustments or repairs to the dentures.

REFERENCES
1. Schmitt A, Zarb GA. The notion of implant-supported overdentures. J Prosthet Dent 1998;79:60-5. 2. MacEntee MI, Walton JN. The economics of complete dentures and implant-related services: a framework for analysis and preliminary outcomes. J Prosthet Dent 1998;79:24-30. 3. Walton JN, MacEntee MI. A retrospective study on the maintenance and repair of implant-supported prostheses. Int J Prosthodont 1993;6:451-5. 4. Watson RM, Jemt T, Chai J, Harnett J, Heath MR, Hutton JE, et al. Prosthodontic treatment, patient response, and the need for maintenance of complete implant-supported overdentures: an appraisal of 5 years of prospective study. Int J Prosthodont 1997;10:345-54. 5. Walton JN, MacEntee MI. A prospective study on the maintenance of implant prostheses in private practice. Int J Prosthodont 1997;10:453-8. 6. Bergendal T, Engquist B. Implant-supported overdentures: a longitudinal prospective study. Int J Oral Maxillofac Implants 1998;13:253-62. 7. Naert IE, Gizani S, Vuylsteke M, Van Steenberghe D. A 5-year prospective randomized clinical trial on the inuence of splinted and unsplinted oral implants retaining a mandibular overdenture: prosthetic aspects and patient satisfaction. J Oral Rehabil 1999;26:195-202. 8. Gotfredsen K, Holm B. Implant-supported mandibular overdentures retained with ball or bar attachments: a randomized prospective 5-year study. Int J Prosthodont 2000;13:125-30. 9. Watson GK, Payne AG, Purton DG, Thomson WM. Mandibular overdentures: comparative evaluation of prosthodontic maintenance of three different implant systems during the rst year of service. Int J Prosthodont 2002;15:259-66. 10. Naert I, De Clercq M, Theuniers G, Schepers E. Overdentures supported by osseointegrated xtures for the edentulous mandible: a 2.5-year report. Int J Oral Maxillofac Implants 1988;3:191-6. 11. Kirsch A. Overdentures on IMZ implants: Modalities and long-term results. In: Schepers E, Naert I, Theuniers G, editors. Overdentures on oral implants. Leuven: Leuven University Press; 1991. p. 15-7. 12. Engquist B. Six years experience of splinted and non-splinted implants supporting overdentures in upper and lower jaws. In: Schepers E, Naert I, Theuniers G, editors. Overdentures on oral implants. Leuven: Leuven University Press; 1991. p. 27-41. 13. Mericske-Stern R. Overdentures with roots or implants for elderly patients: a comparison. J Prosthet Dent 1994;72:543-50. 14. Walton JN, Huizinga SC, Peck CC. Implant angulation: a measurement technique, implant overdenture maintenance, and the inuence of surgical experience. Int J Prosthodont 2001;14:523-30. 15. Carlson B, Carlsson GE. Prosthodontic complications in osseointegrated dental implant treatment. Int J Oral Maxillofac Implants 1994;9:90-4. 16. Mericske-Stern R. Clinical evaluation of overdenture restorations supported by osseointegrated titanium implants: a retrospective study. Int J Oral Maxillofac Implants 1990;5:375-83. 17. Naert I, Quirynen M, Theuniers G, van Steenberghe D. Prosthetic aspects of osseointegrated xtures supporting overdentures. A 4-year report. J Prosthet Dent 1991;65:671-80.

CONCLUSIONS
This randomized clinical trial evaluated the satisfaction of subjects with mandibular implant-retained complete dentures opposed by maxillary complete dentures and compared the maintenance required for the mandibular dentures attached by bar-clips or by 2.25-mm ball-springs to endosteal dental implants. The following conclusions were drawn: 1. There was low satisfaction generally at baseline with the previous dentures, but the new implantretained dentures in both attachment groups improved the participants overall satisfaction signicantly (P,.05) and dramatically. 2. Neither the attachment mechanism nor presence or absence of a reinforcing framework had a notable ef36

VOLUME 93 NUMBER 1

MACENTEE, WALTON, AND GLICK

THE JOURNAL OF PROSTHETIC DENTISTRY

18. Walton JN, MacEntee MI, Glick N. One-year prosthetic outcomes with implant overdentures: a randomized clinical trial. Int J Oral Maxillofac Implants 2002;17:391-8. 19. Oetterli M, Kiener P, Mericske-Stern R. A longitudinal study on mandibular implants supporting an overdenture: the inuence of retention mechanism and anatomic-prosthetic variables on periimplant parameters. Int J Prosthodont 2001;14:536-42. 20. Bakke M, Holm B, Gotfredsen K. Masticatory function and patient satisfaction with implant-supported mandibular overdentures: a prospective 5-year study. Int J Prosthodont 2002;15:575-81. 21. Fenton AH, Lang BR. Selecting and arranging prosthetic teeth. In: Zarb GA, Bolender CL, Carlsson GE, editors. Bouchers prosthodontic treatment for edentulous patients. St. Louis: Mosby; 1997. p. 254-5. 22. de Grandmont P, Feine JS, Tache R, Boudrias P, Donohue WB, Tanguay R, Lund JP. Within-subject comparisons of implant-supported mandibular prostheses: psychometric evaluation. J Dent Res 1994;73:1096-104. 23. McDowell I, Newell C. Measuring health: a guide to rating scales and questionnaires. 2nd ed. New York: Oxford University Press; 1996. p. 341-5.

Reprint requests to: DR MICHAEL I. MACENTEE UNIVERSITY OF BRITISH COLUMBIA DEPARTMENT OF ORAL HEALTH SCIENCES FACULTY OF DENTISTRY 2199 WESBROOK MALL VANCOUVER, BC CANADA V6T 1Z3 FAX: 604-822 3562 E-MAIL: macentee@interchange.ubc.ca 0022-3913/$30.00 Copyright 2005 by The Editorial Council of The Journal of Prosthetic Dentistry.

doi:10.1016/j.prosdent.2004.10.013

JANUARY 2005

37