the influence of different factors on the survival of root

Upload: jihanjohari8053

Post on 03-Jun-2018

213 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    1/8

    The Influence of Different Factors on the Survival of RootCanal Fillings: A 10-Year Retrospective StudyRichard Stoll, PD, Dr.med.dent.habil.,* Kirsten Betke, Dr.med.dent., and

    Vitus Stachniss, Prof.Dr.med.dent.*

    Abstract

    The aim of the present study was to determine thesurvival time of root canal fillings performed in 1990and 1991 at the Dental School, Philipps University ofMarburg, Germany. Data were collected retrospectivelyfrom the dental records and control radiographs wereevaluated. The survival probability was determined byKaplan-Meier analysis. Intergroup differences were ver-ified with the log-rank test. The 914 evaluated rootcanal fillings yielded an overall 10-yr cumulative sur-vival probability of 0.74. The factors baseline periapicalcondition (p 0.001), length (p 0.001), condensa-tion (p 0.001), vitality (p 0.001), and pain symp-toms (p 0.005) were found to have a significantinfluence on the long-term success of root canal fillings.Higher survival rates were recorded for teeth withhealthy periapical conditions, root canal fillings of thecorrect length, homogeneously condensed root canalfillings, root canal fillings in previously vital teeth, andteeth that had been asymptomatic during treatment. Anonsignificant influence was recorded for the parame-ters operator (p 0.606) and retreatment (p 0.196).

    Key Words

    Root canal treatment, survival analysis

    From the *Department of Operative Dentistry and End-odontics, Dental School, Philipps-University of Marburg, Mar-burg, Germany; Private practice.

    Address requests for reprint to Dr. Richard Stoll, Philipps-University of Marburg, Dental School, Department of OperativeDentistry and Endodontics, Georg Voigt Str. 3, D-35033 Marburg,Germany. E-mail address: [email protected].

    Copyright 2005 by the American Association ofEndodontists

    In recent years, opportunities for endodontic therapy have undergone a marked fur-ther development coincidingwith an increase in thedemandfor endodontic treatment(1) and for endodontic restoration. This is due in part to a rise in the proportion ofelderly persons with a relatively complete dentition and to increased dental awareness(1). Patients expectations concerning the success of endodontic restorations are ac-cordingly high (2).

    Varied long-term success rates of root canal fillings have been reported by differ-ent authors; figures range from 70% (3) to more than 90% (4, 5). A survey of therelevant literature was provided by Friedman and co-workers (6), and an introduction

    to meta-analysis by Lewsey et al. (7). In most published studies, the analyzed collectiveconsisted of patients who had been treated either by specially trained endodontists or bysupervised dental students (4 6, 8).

    Most authors reported that the long-term success rate was notably higher in teethdisplaying no apical radiolucency at the start of treatment (4, 6, 912). Moreover, theprospects of success were given a lower rating for retreatments than for initial treat-ments (10).

    The aim of the present study was to evaluate all root canal fillings performed in1990 and 1991 at the Department of Operative Dentistry, Dental School, Philipps Uni-versity of Marburg, Germany, and to calculate their survival functions in dependence onvarious baseline parameters, filling quality parameters and operator-related parame-ters.

    Materials and MethodsAll root canal fillings performed in 1990 and 1991 at the Department of Operative

    Dentistry, Dental School, Philipps University of Marburg, Germany, were followed upwith reference to the dental records. The patients name and date of birth were takenfrom the radiographic records of the respective year with reference to the controlradiograph made on completion of any endodontic treatment.

    The unit tooth was specified as the smallest study unit. In cases where more thanone tooth had undergone endodontic treatment, each tooth was evaluated individually.Only permanent teeth were investigated.

    Following endometric and radiographic length determination, the root canals hadbeen prepared with standardized hand instruments (reamer, K-files and hedstroemfiles) and a step-back technique.All root canalshad been filled with gutta-percha points

    and sealer (Seal-Apex, Kerr, Karlsruhe, Germany), using the lateral condensation tech-nique. A control radiograph had been made in each case on completion of the rootcanal filling.

    The date of birth, gender and data on the affected tooth/teeth of all patients wereobtained from the records. Data recorded in addition to the date of accesswere whetherthetooth reacted positively to cold on that date, whether it was painful, and whether theroot canal filling was being retreated. The date of the root canal filling was also takenfrom the records together with data on whether the treatment had been performed by adental student or a qualified dentist.

    The criteria condensation grade, length, and periapical status according to the PRI(Table 3) (13) were assessed by one calibrated investigator with reference to thepostfilling control radiograph (Tables 3, 5, and 6). In the case of multi-rooted teeth, theworst value of condensation grade, the worst PRI value and the worst value of length

    Clinical Research

    JOE Volume 31, Number 11, November 2005 Survival of Root Canal Fillings 783

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    2/8

    (overfilled was considered poorer than short) was registered. All ra-diographs were assessed individually in a radiograph viewer.

    The follow-up data were obtained from the patients records at theDepartment of Operative Dentistry and the Department of MaxillofacialSurgery, allowing both new endodontic and surgical interventions on

    the root-filled teeth to be registered. The criteria for functional rootcanal filling and loss were stipulated before baseline (Table 1).

    In the event of a functional root canal filling, the most recent dateon which the root filling had been recorded as in situ was noted. In theevent of loss, the date on which the root canal filling had been lostthrough retreatment, root resection or extraction was recorded.

    If a root canal filling had been retreated and replaced within thestudy period, the retreatment was included as a separate case.

    The interval between access and definitive root canal filling wascalculated in days, and the total observation period in months. Thecalculations were based on the date of the definitive root canal fillingand the date of the loss or last documented success.

    All data were first noted on a record sheet and entered into the

    SPSS 10.0 statistics program for evaluation by the Kaplan-Meier method(13). This method calculates the cumulative survival probability (csp)as a function of time. Csp-values and standard error were reported forthe time of the last loss. The log-rank test at a significance level of p 0.05 was used for intergroup comparisons.

    ResultsFollowed up were 965 teeth that had undergone endodontic treat-

    ment in 1990 and 1991 at the Department of Operative Dentistry, Phil-ipps University of Marburg, Germany. Of these, 441 root canal fillings(45.7%) dated from 1990,and 524 (54.3%) from1991. There were51root canal fillings (5.3%) that had been placed presurgically forplanned root resections not included in the study, so that 914rootcanal

    fillings were available for further evaluation.The gender distribution was well balanced with 474 female(49.2%) and 490 male patients (50.8%). At the time of the root canalfilling, the patients ranged in age from 10 to 82 yr. Table 2 shows thedistribution of the teeth according to their location. Table 3 showsthe distribution of the apical findings in dependence on vitality. Aclassification based on the parameter interval between access androot canal filling relative to pulpal status is presented in Table 4. Thelengths and condensation grades achieved in the root canal fillingare shown in Tables 5 and 6.

    The mean observation period was 33.7 months (range: 0 to 124months). A histogram of the observation period is presented in Fig. 1.The Kaplan-Meier survival analysis revealed 105 losses. The mean sur-vival time calculated was 104 months (95% confidence interval: 101

    108 months). At the time of the last loss after 106 months, the cumula-tive survival probability wascsp0.74(standard error: 0.03; Table 7).

    The 914 cases were divided into subgroups according to the pa-

    rameters vitality, type of treatment, operator, symptoms, radiographicfindings, length, and condensation grade. Each of these subgroups wasthen tested for verification of the survival function with the log-rank test.

    Before trepanation, 419 teeth had been classified as vital and 495as nonvital. At the time of the last loss after 95 months, this yielded acumulative survival probability of csp 0.81 for the group of teethoriginally classified as vital. At the time of the last loss after 106 months,the teeth originally classified as nonvitalhad a cumulative survival prob-ability of csp 0.68. The intergroup comparison revealed a significantdifference (p 0.0001) in the survival function (Table 7 and Fig. 2).

    There had been 508 teeth classified as pain-free before treatment,and 406 as painful. The pain-free teeth yielded a cumulative survivalprobability of csp 0.79 after 99 months, and the painful teeth a

    cumulative survival probability of csp 0.67 after 106 months. Thelog-rank test revealed a significant intergroup difference (p 0.0047)in the survival functions (Table 7 and Fig. 3).

    PRI classes 1 and2 (definitely or probably no apical lesion) and 4and 5 (definitely or probably apical periodontitis) were combined (Ta-ble 7 and Fig. 4). Whereas the cumulative survival probability calculatedafter 95 months for the group with PRI classes 1 and 2 was csp 0.88,that calculated after 106 months for the group with PRI classes 4 and 5was csp 0.64. Here too, the log-rank test revealed a significant dif-ference (p 0.0001).

    Another factor influencing the survival rate of the root canal fillingis its quality. To assess the influence of length, length classes 2 (rootcanal filling extending to within 12 mm of the apex) and 3 (root canal

    filling extending to within 01 mm of the apex) were combined (n 595, cumulative 106-month survival probability csp 0.85) and com-pared with classes 4 (too long, n 104, cumulative 99-month survivalprobability csp 0.49) and 1 (too short, n 116, cumulative 96-month survival probability csp 0.41) (Table 7 and Fig. 5). Thisrevealed significant differences in the log-rank test between the com-bined group and thetwo remaining classes (p 0.0001, respectively).

    The comparison of the homogeneous (n 690, cumulative 96-month survival probability csp 0.84) with the inhomogeneous rootfillings (n 222, cumulative 106-month survival probability csp 0.45) also revealed a statistically significant difference (p 0.0001).

    The kind of treatment, divided into initial treatment (n 793,cumulative 106-month survival probability csp 0.74) and retreat-

    TABLE 1. Definition of Success

    Census (functional) The tooth including the original root canal filling dating from 1990 or 1991 is in situLoss The root canal fi lling has been completely or partially retreated, the tooth has been

    extracted, hemisected or resected at one or more root tips

    TABLE 2. Classification of Individual Groups of Teeth Into Nonvital and Vital

    Tooth Group Vital Nonvital

    Upper incisors 86 91Upper premolars 71 95Upper molars 76 93Lower incisors 60 35Lower premolars 68 65Lower molars 58 116Total 419 495

    TABLE 3. Distribution of Apical Findings in Nonvital and Vital Teeth

    PRI Designation Vital Nonvital

    1 Definitely no osseousinvolvement

    132 20

    2 Probably no osseousinvolvement

    163 34

    3 non-assessable 104 734 Probably osseous involvement 16 1765 Definitely osseous involvement 4 191Total 419 495

    PRI (Probability Index) according to Reit and Grndahl (13).

    Clinical Research

    784 Stoll et al. JOE Volume 31, Number 11, November 2005

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    3/8

    ment (n 121, cumulative 96-month survival probability csp 0.72)showed no significant difference (p 0.196) in the log-rank test (Ta-ble 7 and Fig. 7).

    Whether the treatment was carried out by dental students (n 693, cumulative 106-month survival probability csp 0.72) or byqualified dentists (n 221, cumulative 96-month survival probabilitycsp 0.78) had no influence (p 0.606) on the survival function(Table 7 and Fig. 8).

    DiscussionThere were 914 teeth that had undergone root canal treatment and

    were followed up in the present study. The number of teeth investigatedin comparable studies ranges between 22 and 2298 (7), one study hadaccess to 110,766 cases using a data-warehousing approach at an in-surance company database (15). By contrast, some authors specifiedthe number of roots; this ranged between 501 (8) and 1277 (16).

    In the present study, the tooth was selected as the smallest studyunit. A subdivision based on roots (4, 5, 9, 16) is unfavorable in that afailure cannot always be allocated to one specific root. A failure at onlyone root moreover leads to loss of the entire tooth in the worst-casescenario. The highproportion of multi-rooted teeth with successful rootcanal fillings in the above-stated studies may have had a very positiveimpact on the results. A direct comparison of studies based on the toothas the smallest unit with those based on the root as the smallest unit is

    thus open to criticism.Becaues of the retrospective design of the present study, the fol-

    low-up period ranged from 0 months to 10 yr. The time span of com-parable studies is between 1 yr (17) and 10 yr (18).

    Apart from the exclusion of deciduous teeth and fillings in prepa-ration for a surgical intervention,which wasdefined in the present studyas a failure of the root canal filling, no exclusion criteria were formu-lated. The material accordingly included complex cases, e.g. teeth weresevere root canal curvature, third molars, or retreatments, and thus arelatively high proportion of cases that would have fallen short of theexclusion criteria set in other studies(3, 10, 19). Nordid a sporadicallyreduced quality of the radiographs lead to exclusion from the study, ashad been the case with Grahnen and Hansson, for example (16).

    The criterion of success adopted for the present study was thecontinuedpresenceofthetoothwiththeoriginalrootcanalfillingdatingfrom 1990 or 1991. By contrast, any intervention at the root canal filling(full or partial retreatment)or at thetooth (extraction, root amputation,or root resection) was regarded as a failure. Because of differences inthe definition of success and failure as reported in the literature, resultsare notreadily comparable and have to be discussed in thesettingof thestudy design. The method used in the present study is comparable to adata warehousing approach used by studies that access large insurancecompany databases (15).

    As classification into success or failure in the present retrospectivestudy wasbasedsolelyon thepatientsrecords, clinical parameters suchas lack of pain or a radiographic evaluation of the periapical status

    could not be applied. These clinical parameters were previously re-garded as reliable criteria for determining success (20). However, theextent to which these criteria lead to different results from the moreglobal parameters used in the present study is a matter of speculation(18). The difference between a strict criterion (complete healing of theperiapex)andamoreliberalone(incompletehealingisalsocountedasa success) for the probability of success is 0.08 (p-value) according toa meta-analysis by Lewsey et al. (7).

    At least when clinical symptoms occur, therapy leading to a failurerating can be expected to be initiated. If the observation period is suf-ficiently long, the results of the study should not be affected in suchcases.

    No assessment was undertaken of the periapical region in com-parison with the findings immediately after root canal filling, as in some

    TABLE 4. Time Between Access and Root Canal Filling, Depending on Vitality

    Weeks 0 02 24 46 68 812 1216 1620 2024 2428 2853

    Vital 86 122 75 35 15 29 18 13 11 3 11Nonvital 55 123 87 58 34 47 25 29 13 7 18

    TABLE 5. Radiographically Detectable Condensation Grade of Root CanalFillings

    CondensationGrade

    Designation Number Percentage

    1 Homogeneous 690 75.5Without radiologically

    detectable faults2 Nonassessable 123 13.43 Inhomogeneous 101 11.1

    Radiologicallydetectable faults

    Total 914 100

    TABLE 6. Radiographic Length Control of Root Canal Fillings

    Length Designation Number Percentage

    1 2 mm short of apex 116 12.72 12 mm short of

    apex279 30.5

    3 01 mm short ofapex

    316 34.6

    4 overfilled 104 11.45 Nonassessable 99 10.8Total 914 100

    Figure 1.Distribution of observation times (root filling up to last census orloss).

    Clinical Research

    JOE Volume 31, Number 11, November 2005 Survival of Root Canal Fillings 785

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    4/8

    other studies(6, 8, 10, 21). This meant that any negative changes in theperiapical region were not recorded unless they were clinically evident.In addition, no distinction was made in the event of an extraction as towhether it had been performed for endodontic reasons or for otherreasons such as periodontal problems or trauma. The fact that any

    extracted tooth was evaluated as a failure, irrespective of the reason forextraction, may have had a negative impact on the result of the study.In contrast to studies with follow-ups at specific timepoints (3, 4, 8),

    no fixed follow-up timepoints were required in the present study (10). Theevaluationofsuccessorfailurewasundertakenateachcontrolappointmentkept by the patients after the definitive root canal filling.

    The total observation period for the individual root canal fillingswas calculated retrospectively from the interval between the date of theroot canal filling and the last follow-up appointment at which successwas recorded. In the event of a failure, the date on which the root canalfilling was lost was regarded as the endpoint of the observation. Thedemand for the longest possible observation periods (22) was met witha maximum possible observation period of 10 yr. While the mean ob-servation period of 33.7 months seems short compared with the 10-yr

    total observation period, it conforms with the results of similarly de-signed studies (23).

    In the present study, the recall rate recorded in the first year afterdefinitiveroot canal fillingwas relatively high. However, this rate wasnot

    maintained. A study by Selden (24) showed that the number of dropoutsincreases with the increasing length of a study. A recall rate of approx.50% after 6 months but of only 11% after 18 months was reported. Theproblem of patient compliance in clinical studies has long been known.The reasons for the increasing dropout rate in follow-up studies arecertainly varied (4, 25).

    Survival studies have the advantage of utilizing all collected infor-mation until the endpoint of the study (26). As no fixed examinationintervalswere specified in the present study,the procedure proposed byKaplan and Meier (14) was the statistical method of choice (23). Thegain in the utilization of censored survival data depends on the numberof probands, the relative completeness of the further examinations, andthe relative size of the initial loss rate (26). It has to be borne in mindthat the initial loss rate in dental studies tends to be small (23). Espe-

    TABLE 7. Results of Kaplan-Meier Survival Analysis in the Total Group and the Individual Subgroups Together with the Result of the Log-Rank Test for the SameSurvival Functions

    GroupMean Survival Time with

    Confidence Interval(mo)

    Cases Losses SuccessesSurvival

    Rate(last loss)

    StandardError

    LastLoss

    (in mo)

    Log-Rankp

    All 104 (101108) 914 105 809 0.74 0.03 106Initial filling 105 (101109) 793 87 706 0.74 0.03 106 0.196Retreatment 94 (85104) 121 18 103 0.72 0.07 96Vital 111 (107115) 419 31 388 0.81 0.04 95 0.0001Nonvital 94 (8999) 495 74 421 0.68 0.04 106Students 101 (97105) 693 78 615 0.72 0.04 106 0.606Dentists 105 (99112) 221 27 194 0.78 0.04 96Pain-free 108 (104113) 508 47 461 0.79 0.03 99 0.0047Painful 98 (92104) 406 58 348 0.67 0.05 106Condensation good 110 (107114) 690 52 638 0.84 0.03 96 0.0000Condensation poor 84 (7593) 222 51 171 0.45 0.08 106Length short 77 (6490) 116 30 86 0.41 0.10 96 0.0000

    0.0000Length correct 113 (110116) 595 35 560 0.85 0.03 106Length above apex 76 (6487) 104 27 77 0.49 0.10 99PRI 12 116 (113120) 349 15 334 0.88 0.03 95 0.0000PRI 45 93 (8699) 387 69 318 0.64 0.04 106

    Figure 2. Cumulative survival functions in dependence on the factor vitality(vitaln 419, nonvitaln 495); the vertical lines represent censor points.

    Figure 3. Cumulative survival functions in dependence on thefactor pain symp-toms (no pain n 508, pain n 406); the vertical lines represent censorpoints.

    Clinical Research

    786 Stoll et al. JOE Volume 31, Number 11, November 2005

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    5/8

    cially when censored data are used, the predictive validity of survivalstudies carried out over a short period is low; such studies are thusbetter suited to long follow-up periods (23). A large quantity of earlycensored data results in long-term survival figures having a high stan-dard error and in survival probabilities being underestimated (27).

    The log-rank test is suitable for comparing different groups withpredicted differences in loss rate. Unlike the Breslow test, for example,all losses are taken equally into account, irrespective of the timepoint.Especially in conjunction with a large quantity of censored data, thelog-rank test has a higher predictive potential than the Breslow test(28). For this study, we manually modeled the test groups for eachcovariate. We considered this approach as advantageous compared to ageneral multivariate model like Cox regression.

    A meta-analysis of published studies (7) calculated from the re-

    sulting pooled data a success probability of csp

    0.84 (with a 95%confidence interval of 0.80 0.87). The 124-month cumulative survivalprobability for all cases investigated in that analysis was csp 0.74.Similar results were reported by Lost et al. (10), Fritz and Kerschbaum(29), and Rocke et al. (18), who also calculated survival probabilities.Table 8 shows a comparison of the results of some selected studies. Anumber of studies (4, 22, 30) reported considerably higher successrates in part. However, the distinction was confined to success or fail-ure, without the partially substantial fluctuations in the length of thefollow-up period being taken into account.

    Some studies classify reasons of failure into categories. It wasshown that the most common reason for failure was caries because ofpoororal hygieneand loss of therestoration sealing theroot canal (31).Teeth without subsequent restorations had a four times higher inci-

    Figure 4.Cumulative survival functions in dependence on the factor baselinesituation of the periapex (PRI 1&2 group without chronic apical periodon-titis (n 349), PRI 4&5 group with chronic apical periodontitis (n

    387)); the vertical lines represent censor points.

    Figure 5.Cumulative survival functions in dependence on the factor length(02 mm to apexn 595, too shortn 116, overfilledn 104; nonas-sessable root canal fillings are not shown); the vertical lines represent censorpoints.

    Figure 6.Cumulative survival functions in dependence on the factor condensa-tion (homogeneous n 690, inhomogeneous n 224); the vertical linesrepresent censor points.

    Figure7.Cumulative survival functions in dependenceon the factor retreatment(initial treatmentn 793, retreatmentn 121); the vertical lines representcensor points.

    Clinical Research

    JOE Volume 31, Number 11, November 2005 Survival of Root Canal Fillings 787

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    6/8

    dence of extraction than teeth with single unit restorations (15). Therewere 59.4% of extractions following nonsurgical root canal treatmentcaused by prosthetic failure, 32% were caused by periodontal failure,and only 8.6% were caused by endodontic failure (32).

    In the present study, significant differences were registered in thesurvival probability of teeth classified as vital (csp 0.81) or nonvital(csp 0.68) at the time of access. Teeth classified preoperatively asvital thus had a higher survival probability.

    The cryogenic method of sensitivity testing was used in contrast toother studies where an electric pulp testerwas preferred(31). Althoughthis method is in widespread use, it has to be pointed out that it is only

    thesensitivity but not the vitality of the respective tooththat is tested, i.e.it is primarily the conduction response to a given stimulus. The vitality ofthe tooth is dependent upon an efficient flow of blood to the vessels. Asit is fundamentally possible for a tooth to be vital despite a lack ofsensitivity, some of the teeth classified as nonvital must be assumed tohave been merely nonsensitive.

    Similar conclusionsare drawn in various other studies(3, 4, 6, 11,25, 30), although the distinction between vital and nonvital teeth is notequallyclearinallcases.Someauthorsreportednodifferencesbetweenthe success probabilities of vital and nonvital teeth (9, 13, 25).

    In the present study, teeth that were asymptomatic at the time ofaccess were found to have longer survival times than those that werepainful (p 0.005), i.e. the factor pain symptoms must be assumed to

    have an impact on the success of root canal treatment. Contrary to thefrequent claim that pain has no impact on the success of root canaltreatments (4, 9, 17), there is agreement in this point with findings ofFriedman et al. (6), who also reported a highly significant differencebetween asymptomatic and painful teeth.

    In the group with periapical findings (PRI 4&5), the loss rate wasmore than four times higher than in the group without such findings(PRI 1&2). At the time of the last loss, the survival probability of thegroup without periapical findings was csp 0.88, while that of thegroup with periapical findings was only csp 0,64; this difference wasstatistically significant (p 0.0001).

    Most authors consider that the presence of a periapical lesion hasa negative impact on the prospective success of root canal fillings (4,912, 17, 33, 34). Friedman et al. stated that apical periodontitis is the

    main prognostic factor in initial endodontic treatment (11). In contrastto the general opinion, Peak (22) stated that root canal fillings in teethwith periapical radiographic findings had a higher survival time thanthose in teeth without such findings.

    Root canal fillings ending within 0 to 2 mm of the radiographicapex are classified as correct with respect to their length (1). For thisreason the cases covered in the present study were divided for statisticalanalysis into three groups. This revealed that teeth whose root canalfilling was of the correct length had a significantly (p 0.001) bettersurvival function than those where theroot canal fillingwas too short ortoo long.

    The results recorded in the present study are in agreement withthose of other studies also reporting notablybetterresults for rootcanalfillings of the correct length (3, 12, 19, 30, 34, 35). The hypothesis ofLin et al.(33), according to whichthe length of the root canal filling hasno influence on its success, could not be confirmed with the findings ofthe present study.

    In the present study the proportion of homogeneous root canalfillings was notably higher than that of inhomogeneous fillings. Thesurvival probability of teeth with homogeneous root canal fillings wassignificantly (p 0.0001) higher than that of teeth with inhomoge-neously condensed fillings. As with length, the homogeneity of all root

    canal fillings wasassessedonly from radiographs,as thisis currentlytheonly feasible method in vivo. As this method is based on the two-dimen-sional representation of a three-dimensional structure, it has to beaccepted that the homogeneity cannot be fully assessed with the givenmeans in all planes. However, in the radiological assessment of homo-geneity a false-positive rating is very much more likely than a false-negative one. Similar conclusions were drawn in comparable studies(4, 9). The discrepancy is most marked in the study by Kerekes andTronstad (9), who reported 93% success for homogeneous and 28%success for inhomogeneous root canal fillings. Only Friedman et al. (6)failed to find a significant difference in the success rate of homogeneousand inhomogeneous root canal fillings.

    At 13.2%, the proportion of retreatment in the present study was

    relatively low.In comparison, the proportions reported by Sjogren et al.and Lost et al. in their collectives were more than twice as high (4, 10).A comparison of the survival functions of retreatments and initial treat-ments in the present study yielded no significant difference (p 0.196). Contrary to these results, notably poorer results for retreat-ments are reported in the literature (4, 6, 10, 36). The outcome ofnonsurgical retreatment depends highly significant on the alteration ofroot canal morphologyand thepresence of apicallesions. Teeth withoutalteration of root canal morphology and without apical lesion showed asuccess rate of 91.6%after a 2 yr period. Teeth with alterated morphol-ogy and with apical lesions had a success rate of 40% following retreat-ment (36). Other important factors are preoperative apical periodon-titis (absent 97% healing, present 78%), perforation (absent 89%,

    present 42%), filling length and lack of definitive restoration (37).One possible explanation of the results in the present study mightbe that the patients investigated had all been treated at a teaching hos-pital, where difficult cases are generally delegated to experienced den-tists in view of the training situation. A comparison of the relative per-centage of retreatments in the individual groups, i.e. 12% for dentalstudents and 17.2% for qualified dentists, reveals that this difference isverifiable but not pronounced. Another explanation might be an in-creased willingness under teaching hospital conditions to take specialcare and to invest more time in performing even protracted, less eco-nomical procedures.

    Approximately 75% of the investigated root canal fillings had beenperformed by undergraduates. However, no further subdivision wasmade into different terms or examination levels. Nor were treatments

    Figure 8.Cumulative survival functions in dependence on the factor operator(dentists n 221, students n 693); the vertical lines represent censorpoints.

    Clinical Research

    788 Stoll et al. JOE Volume 31, Number 11, November 2005

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    7/8

    carried out by qualified dentists differentiated according to the length oftheir vocational experience or to their specialty.

    There was no significant intergroup difference in survival function(p 0.61). The concern that fillings performed by students might havepoorer survival rates than those performed by qualified dentists was notconfirmed in the present study. Kerekes and Tronstad (9) and Sjogrenet al. (4) reported a success rate of more than 90% in student work,while Hellwig et al. reported a failure rate of 5% after 3 yr (8).

    Thesuspicion arisesthat this favorable result may be a result of thedistribution of problem cases, which tend to be delegated to qualifieddentists after the initial clinical examinations. A slightly uneven distri-bution was registered in the present study for retreatments. However,this did not apply to the treatment of molars. A comparison of the

    relative percentages of molars treated in the individual groups showsthat the proportion treated by students (39%) was, in fact, slightlyhigher than that treated by qualified dentists (35%).

    A significantly higher patient satisfaction and quality of life out-come assessed in patients treated by endodontists may be explained bythe increased skill and proficiency of specialists (38).

    In the present study, the long-term results of root canal fillingsperformed with the lateral condensation technique were investigated.Theresults were found to be in good conformity with those published inthe literature.

    It could also be shown that good successes can be achieved withretreated root canal fillings. It thus seems perfectly justified to give toothretention through sometimes resource-intensive retreatment prefer-

    enceover extraction with itsconsequences of implantation or prostheticcover. The results of treatment under quality-controlled training condi-tions are no poorer than those of treatment by experienced dentists.

    Both the baseline state of pulpal tissueand periapex and thequalityof the root canal filling itself could be shown to have a crucial influenceon the long-term prognosis of the filling. It is thus possible for theoperator to evaluate the prospects of success in the treatment of indi-vidual teeth by analyzing the baseline situation and adapting the therapyaccordingly. At the same time, the patients expectations can be directedfrom the very outset into realistic channels to avert potential disappoint-ments or misunderstandings. Irrespective of the initial situation, thedentist has a crucial influence on the treatment outcome with the qualityof hiswork. The operator hasto be constantlyawareof hisresponsibilityand to be correspondingly critical of his own work. Meticulous length

    determination and an adequate condensation technique are the crucialfactors shown in this study within the limitation of the study design.

    References1. Schulte A, Pieper K, Charalabidou O, Stoll R, Stachniss V. Prevalence and quality of

    root canal fillings in a German adult population. Clin Oral Invest 1998;2:6772.2. Briggs PFA, Scott BJJ. Evidence-based dentistry: endodontic failurehow should it

    be managed? Br Dent J 1997;183:15964.3. Heling B, Tamshe A. Evaluation of the success of endodontically treated teeth. Oral

    Surg Oral Med Oral Pathol 1970;30:5336.4. Sjogren U, HagglundB, Sundquist G, WingK. Factors affectingthe long-termresultsof

    endodontic treatment. J Endod 1990;16:498504.5. Benenati FW, Khajotia SS. A radiographic recall evaluation of 894 endodontic cases

    treated in a dental school setting. J Endod 2002;28:3915.6. Friedman S, Lost C, Zarrabian M, Trope M. Evaluation of success and failure afterendodontic therapy using a glass ionomer cement sealer. J Endod 1995;21:38490.

    7. Lewsey JD, Gilthgorpe MS, Gulabivala K. An introduction to meta-analysis within theframework of multilevel modelling using the probability of success of root canaltreatment as an illustration. Community Dent Health 2001;18:1317.

    8. Hellwig E, Klimek J, Ahrens G. Dreijahrige Erfolgskontrolle vonWurzelbehandlungenaus studentischen Behandlungskursen. Dtsch Zahnarztl Z 1982;37:949 53.

    9. Kerekes K, Tronstad L. Longterm results of endodontic treatment performed with astandardized technique. J Endod 1979;5:8390.

    10. Lost C, Weiger R, Axmann-Krcmar D. Prognose von Wurzelkanalbehandlungen unterAnwendung der lateralen Kondensationstechnik und eines Glasionomerzementseal-ers. Dtsch Zahnarztl Z 1995;50:897901.

    11. Friedman S, AbitbolS, Lawrence HP.Treatmentoutcomein endodontics:the TorontoStudy. Phase 1: Initial treatment. J Endod 2003;29:78793.

    12. FarzanehM, AbitbolS, LawrenceHP, Friedman S. Treatmentoutcomein endodontics:the Toronto Study. Phase II: Initial treatment. J Endod 2004;30:3029.

    13. Reit C, Grondahl HG. Application of statistical decision theory to radiographic diag-nosis of endodontically treated teeth. Scand J Dent Res 1983;91:2138.

    14. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J AmStat Ass 1958;53:45781.

    15. Lazarski MP, Walker WA, Flores CM, Schindler WG, Hargraeves KM. Epidemiologicalevaluation of the outcome of nonsurgical root canal treatment in a large cohort ofinsured dental patients. J Endod 2001;27:7916.

    16. Grahnen H, Hansson L. The prognosis of pulp and root canal therapy: a clinical andradiographic follow-up examination. Odontol Revy 1961;12:14665.

    17. Pekruhn RB. The incidence of failure following single-visit endodontic therapy.J Endod 1986;12:6872.

    18. Rocke H, Kerschbaum T, Fehn C. Zur Verweildauer wurzelkanalbehandelter Zahne.Dtsch Zahnarztl Z 1997;52:783 6.

    19. Barbakow FH, Cleaton-Jones PE, Friedman D. Endodontic treatment of teeth withperiapical radiolucent areas in a general dental practice. Oral Surg 1981;51:5529.

    20. Weiger R, Hitzler S, Hermle G, Lost C. Periapical status, quality of root canal fillings

    TABLE 8. Study Design and Results of Comparable Studies

    Author(s) (year) n Unit Duration (yr) Design X-ray Result

    Grahnen & Hansson (1961) 1277 Root 45 F x 82.8%Heling & Tamshe (1970) 213 Tooth 15 F x 70%Kerekes & Tronstad (1979) 501 Root 35 F x 91%Swartz et al. (1983) 1770 Root 1 F x 89.7%Pekruhn (1986) 925 Tooth 1 F x 94.8%Sjgren et al. (1990) 849 Root 810 F x 91%Smith et al. (1993) 821 Tooth min. 5 F x 84.3%

    Peak (1994) 136 Tooth 0.56 F x 84.6%Friedman et al. (1995) 378 Tooth 0.51.5 F x 78.3%Lst et al.(1995) 161 Tooth 0.53.5 S x csp 0.61Rocke et al. (1997) 485 Tooth 10 S csp 0.81Fritz & Kerschbaum (1999) 504 Tooth 9 S 76.5%

    343 78.5%Lewsey et al. (2001) 38 Studies M csp 0.84Lazarski et al. (2001) 110766 Tooth 10 S 94.44%Benenati et al. (2002) 894 Tooth 0.57 F x 91.05%Dammaschke et al. (2003) 190 Tooth 10 S 85.1%Friedman et al. (2003) 450 Tooth 46 F x 81%Present study 914 Tooth 10 S csp 0.74

    The column headed X-ray shows whether the follow-ups were based on radiographs. Studies are classified as a classical follow-up study (F), a survival study (S), or a meta-analysis (M). The column headed Result

    shows the recorded survival rate or the cumulative survival probability (csp).

    Clinical Research

    JOE Volume 31, Number 11, November 2005 Survival of Root Canal Fillings 789

  • 8/12/2019 The Influence of Different Factors on the Survival of Root

    8/8

    and estimated endodontic treatment needs in an urban German population. EndodDent Traumatol 1997;13:6974.

    21. PettietteMD, Delano EO,Trope M. Evaluationof success rateof endodontic treatmentperformed by students with stainless-steel K-Files and Nickel-Titanium hand files.J Endod 2001;27:1247.

    22. PeakJD. Thesuccess of endodontic treatmentin general dental practice:a retrospec-tive clinical and radiographic study. Prim Dent Care 1994;1:913.

    23. Stoll R, Siewecke M, Pieper K, Stachniss V. Longevity of cast gold inlays and partialcrowns: a retrospective study at a dental school clinic. Clin Oral Invest 1999;3:1004.

    24. Selden HS. Pulpoperiapical disease: diagnosis and healing. Oral Surg 1974;37:27183.25. Sobarzo-Navarro V, Rastl B, Quistorp-Promper M, Hadicke WD, Nolden R. Langzeit-

    erfahrung mit systematischer Endodontie im Klinikbetrieb. Dtsch Zahnarztl Z 1988;43:2725.

    26. CutlerSJ, EdererF. Maximum utilizationof thelifetablemethod inanalyzing survival.J Chron Dis 1958;8:699712.

    27. Davies JA. Dental restoration longevity: a critique of the life table method of analysis.Community Dent Oral Epidemiol 1987;15:2024.

    28. Prentice RL, Marek P. A qualitative discrepancy between censored data rank tests.Biometrics 1979;35:8617.

    29. Fritz UB, Kerschbaum T. Langzeitverweildauer wurzelkanalgefullter Zahne. DtschZahnarztl Z 1999;54:2625.

    30. Smith CS, Setchell DJ, Harty FJ. Factors influencing the success of conventional rootcanal therapy: a five-year retrospective study. Int Endod J 1993;26:32133.

    31. Ettinger RL, Quian F. Postprocedural problems in an overdenture population: alongitudinal study. J Endod 2004;30:3104.

    32. Vire DE.Failureof endodontically treated teeth: classificationand evaluation. J Endod1991;17:33842.

    33. Lin LM, Skribner JE, Gaengler P. Factors associated with endodontic treatment fail-ures. J Endod 1992;18:625 67.

    34. Dammaschke T, Steven D, Kaup M, Ott KHR. Long-term survival of root-canal-treatedteeth: a retrospective study over 10 years. J Endod 2003;29:63843.

    35. Swartz DB, Skidmore AE, Griffin JA. Twenty years of endodontic success and failure.J Endod 1983;9:198202.

    36. Gorni FGM, Gagliani MM. The outcome of endodontic retreatment: a 2-yr follow-up.J Endod 2004;30:14.

    37. FarzanehM, AbitbolS, LawrenceHP, Friedman S. Treatmentoutcomein endodontics:The Toronto Study. Phases I&II: Orthograde Retreatment. J Endod 2004;30:62733.

    38. Dugas NN, Lawrence HP, Teplitsky P, Friedman S. Quality of life and satisfactionoutcomes of endodontic treatment. J Endod 2002;28:81927.

    Clinical Research

    790 Stoll et al. JOE Volume 31, Number 11, November 2005