Incidence of file separation in an endodontic post-graduate program: A retrospective cohort study

Drs. Kirk Brown, Stephen J. Clark, Joseph Morelli and Ricardo Caicedo explore the factors contributing to file separation

Abstract
In a retrospective cohort study, 4,965 teeth with 13,055 canals receiving root canal therapy in a post-graduate endodontic clinic were evaluated for file separation. 119 instruments separated with 111 separations during initial treatment and 8 during retreatment. The incidence of file separation for number of teeth treated was 2.4% (2.5% for initial treatment, 1.6% for retreatment), and the incidence of file separation for number of canals treated was 0.91% (.93% for initial treatment, .71% for retreatment). There was no significant difference in incidence of file separation between initial treatment and retreatment (p = .279). Fisher’s exact test indicated a significant difference for incidence of separation between individual residents. No significant difference was found for the group of residents in incidence of file separation between first year of training (3.12%) and second year of training (2.69%) (p = 0.479). Files were found to separate more often in the apical third of the canal (58.3%) than the middle third (34%) or coronal third (7.8%) (p = .0001). Grouped by tooth type, there is a significant trend for file separation in posterior teeth (p < 0.00001). A higher incidence of file separation was found in maxillary teeth than in mandibular teeth (p = 0.0142).
The results of this study reveal an incidence of file separation and a trend for location of file separation that is consistent with previous reports in the literature.

Introduction
The endodontic community has moved toward greater use of nickel-titanium files in a motorized handpiece. While nickel-titanium files have been shown to provide great benefits in canal instrumentation, some limitations still exist. The most prominent concern is that of instrument separation during use in the root canal. A significant amount of research has been conducted to evaluate factors contributing to instrument breakage during root canal preparation, including operator experience1,2, canal morphology and curvature3-5, forces on the file such as torque and flexing6,7, rotational speed4,5,8, file design9,10, file size and taper3, 11-13, glide path14,15, manufacturing process and defects16-18, electropolishing18-21, number of file uses10,22,23, and the sterilization process24-26. Some previous studies have been conducted to report the incidence of file separation during endodontic therapy27-32.
The primary aim of this study was to determine the incidence of file separation in teeth treated by first and second year endodontic residents in a university post-graduate endodontic program. A secondary aim of this study was to evaluate the incidence of file separation in the first year of resident training and compare it to the incidence for the second year of the resident training. Other variables that were investigated were the effect of instrumenting different tooth types on instrument separation and location of separated instruments within the canal.

Materials and methods
The University of Louisville Human Studies Committee (Institutional Review Board) approved the investigation protocol.
This was a retrospective cohort study that included 4,965 teeth that received endodontic therapy in the post-graduate endodontic clinic at the University of Louisville School of Dentistry. Each case record and radiograph was inspected and evaluated to record the following variables: tooth number; treating resident; number of canals in the treated tooth; if endodontic therapy was completed; initial treatment or retreatment; occurrence of separated instrument; if separated instrument was removed, and location of separated instrument in the canal. For the purposes of this study, the level at which the most coronal portion of the fractured file was located determined its location in the canal. For those teeth in which a separated instrument occurred, the following information was recorded: first or second year resident operator and the location of the separated instrument. This data was recorded with no patient identifiers but with a unique identifier for each case.
Fisher’s exact test was used to determine any statistical difference for file separation between residents. A chi-square test was used to determine if canal location was significant for file separation. The Pearson’s correlation was then used to determine if a trend existed for file separation in certain canal locations. A Mantel-Haenszel chi-square test was used to determine whether a significant difference existed between file separation rates for residents during their first and second year of training. Using logistic regression, significance of file separation in different tooth types was tested.

Results
Of the total 4,965 treated teeth and 13,055 treated canals, the number of separated files was 119; an incidence of 2.4% for treated teeth and 0.91% for treated canals. Of the 119 separated instruments, 111 were separated in initially treated teeth and 8 separated in retreated teeth. Incidence of file separation for initially treated teeth was 2.5% and for retreatment was 1.6%. There was no significant difference between the two (p = 0.279; 95% CI 0.762 – 3.744) (Table 1).
The incidence of file separation was also calculated and compared to the number of canals treated. The incidence of file separation for total number of canals treated was 0.91%, for initially treated canals it was 0.93%, and for retreated canals, the incidence was 0.71%. There was no significant difference between overall incidence of file separation in initially treated canals and retreated canals (p = 0.620; 95% CI 0.637 – 3.106) (Table 1). The incidence of file separation in the first and second year of training for the six residents for which there were two full years of data was 3.12% (n=36) for the first year and 2.69% (n=43) for the second year (Figure 1). Using the Mantel-Haenszel chi-square test, there was no significant difference in overall file separation (Table 2) (p = 0.479; df = 1).
Of the 119 separated instruments, four were not included because either there was no radiograph of the separated instrument and the file was removed without noting the location of separation in the patient record or after canal obturation, the file could not be visualized. For the remaining 115 separated instruments, 9 (7.8%) were separated in the coronal one-third of the root, 39 (34%) were separated in the middle one-third of the root, and 67 (58.3%) were separated in the apical one-third of the root (p = .0001) (Figure 2). A trend toward files separating in the apical area > middle area > coronal area of the canal was found (Pearson’s correlation = 0.9998, p = 0.013). Three (33%) of the files that separated in the coronal third were retrieved, 2 (5.1%) of the files in the middle third were retrieved, and only 1 (1.5%) of the files separated in the apical third was retrieved (two cases were not included due to removal prior to documenting location). A significant difference was found for file retrieval between apical and middle thirds (p = 0.0282; std. error 1.0134) and between apical and coronal thirds (p = 0.0045; std. error 1.2309). Also, a trend exists for success rate of file removal; coronal > middle > apical thirds (p = 0.00485; std. error 0.6522).
Incidence of file separation by tooth type was as follows: 4.1% (52 of 1,277) of maxillary molars had separation occur compared to 2.9% (54 of 1,893) of mandibular molars, 1.6% (9 of 573) of maxillary premolars, zero mandibular premolars or mandibular anteriors had file separations, and 0.5% (4 of 789) of maxillary anteriors had separated instrument occur (Table 3, Figure 3). A significant trend was noted in file separation with more files separating in molars than premolars than anterior teeth (p < 0.00001; std. error 0.2199). Also, a trend of files separating in maxillary teeth more often than in mandibular teeth was found (p = 0.0142; std. error 0.1904). When maxillary molars were compared to mandibular molars, there was a tendency for increased file separation in maxillary molars (p = 0.0624; std. error 0.1978). Teeth with only 1 canal had file separations 0.4% (4 of 1,120) of the time compared to 1.7% (9 of 529) of teeth with 2 canals, 3.1% (48 of 1525) of teeth with 3 canals, and 3.9% (50 of 1,273) of teeth with 4 canals (Figure 4). Using logistic regression, teeth with a higher number of canals were found to have a higher rate of file separation (p < 0.00001; std. error 0.09914).

Discussion
Incidence of nickel-titanium file breakage in root canals has been reported in previous studies27-32. The method by which the incidence is reported varies from reporting the total number of canals instrumented or the total number of cases instrumented. In the literature, the reported incidence of file separation by number of teeth treated ranges from 1.68% to 3.3%; the reported incidence of file separation by number of canals ranges from 0.39% to 1.3%. In these studies, more specific variables were examined such as the area in the canal in which the instrument separated, the tooth type involved, severity of curvature of the canal in which the instrument separated, type of file involved, and number of uses before the instrument separated. Of the studies that recorded the location of the separated fragment, all of them reported that the apical one-third was the most frequent location27,29,31,32. Iqbal et al29 reported the odds of instrument separation were three times greater in molars than premolars, but that no difference was found between maxillary and mandibular teeth of the same tooth type. Di Fiore et al27 found no difference in file breakage between different types of files used while Wolcott et al30 concluded that ProTaper® (Dentsply Tulsa Dental Specialties) files can be used up to four times without significantly increasing the chance of separation.
If a nickel-titanium file separates in the root canal, it may hinder the ability of the clinician to properly clean and obturate the canal. For this reason, a separated file is an undesirable event during root canal therapy. Some studies reported the effect on outcomes that a retained separated instrument has on the healing of the treated tooth. Spili et al32 conducted a study in which more than 8,000 cases were analyzed, and those cases that had separated instruments were followed. The overall healing rates between cases with and without separated instruments were not significantly different with an overall healed rate of 91.8% for cases with a separated instrument and 94.5% for the matched controls. Crump and Natkin33 also reported that separated files in root canals had no significant impact on healing with a success rate in cases with separated files being 81.2% while the control group had a 73.6% success rate.
In this study, the incidence of files separated for teeth treated was 2.4% (119 of 4,965), and the overall incidence of files separated for canals treated was 0.91% (119 of 13,055). This incidence is consistent with the range of 1.68% to 3.3%, which has been reported in the literature by past investigators27-32. Although this study reports a decrease in file separation incidence from the first year of training (3.12%) to the second (2.69%), there was no statistical significance between the two groups. Perhaps, if the sample size were larger, a statistical difference would have been noted. A trend toward less file separation as experience of the operator increases might be expected.
Perhaps more than 1 year of training is needed to see a statistical difference in file separation. It could also be proposed that the incidence of file separation in the second year of training could be greater than the first year. Residents in their second year of training may be treating more difficult cases or attempting to increase their clinical efficiency while inadvertently increasing their risk of separating an instrument. Studies comparing experienced endodontists and inexperienced clinicians have been carried out1,2; however, the studies were either on acrylic blocks, extracted teeth, or had very small numbers.
Yared et al1 compared file separation between an experienced endodontist and recent dental school graduates who were either trained or untrained in using a specified instrumentation technique on extracted teeth. They found that the untrained group separated and deformed significantly more files than did the trained recent graduate or experienced endodontist. Vieira et al2 conducted a clinical study in which an experienced endodontist treated 80 teeth endodontically without separating a file, and inexperienced dental students treated 50 teeth that resulted in deformed and separated files. It would be helpful to compare the incidence of file separations between studies that include large case numbers and compare experienced endodontists with endodontic residents.
Iqbal et al29 reported that file separation was 33.5 times more likely in the apical one-third of the canal than in the coronal one-third of the canal. The findings of this study are in agreement and found that instruments had a greater likelihood of separating in the apical portion of the root than in the middle or coronal area of the root. Cunningham et al34 reported that secondary curvatures in the root occurred primarily in the apical one-third of the root. This could also be a factor in instrument separation in the apical one-third of the canal.
In this study, a significant trend was found for increased file separation in molars as compared to incisors. Molars can be more difficult to reach, and precise control over the rotary instrument may be more difficult to maintain. Straight line access can be more difficult to obtain than in an anterior tooth. Molars may also have more calcified canals from being more heavily restored than are anterior teeth. Perhaps more importantly, molars have more canals to be treated than the teeth anterior to them; therefore, the instruments used receive more stress and use than anterior teeth and have a greater chance of instrument separation.
Teeth with more canals require more instrumentation than teeth with fewer canals. Analysis was performed using the number of canals in the tooth to determine if that had a significant effect on file separation. Using logistic regression analysis, it was found that the more canals a tooth had in it, the more likely a file was to separate. This is similar to the analysis of tooth type; however, the same types of teeth may have a different number of canals. For example, a lower molar or a maxillary molar could have one, two, three, or four canals, and if put in the same category, could skew the results. Therefore, statistical analysis of teeth categorized by canal number rather than tooth type can better describe the likelihood of file separation with a greater number of canals.
Iqbal et al29 found no statistical difference in file separation between maxillary and mandibular teeth of the same tooth type. In this study, files separated in the maxillary arch (65 of 2639) more often than in the mandibular arch (54 of 2326) with statistical significance (p = 0.0142). When maxillary and mandibular molars were compared, a tendency for a greater incidence of file separation in maxillary molars than mandibular molars existed, however, it was not statistically significant. This may be due in part to the fact that maxillary first molars have a high rate of second mesiobuccal canals that can be calcified and difficult to negotiate.
The results of this study report an incidence of file separation that is consistent with what has been previously reported in the literature. File separation between first and second years of training did not significantly differ. Nickel-titanium rotary files were found to be more likely to separate in maxillary teeth, in teeth with greater numbers of canals, and in the apical one-third of the canal.

Kirk Brown graduated with a DDS degree from Virginia Commonwealth University in 2004 and with an MS and Endodontic Certificate from the University of Louisville in 2010. He is in private practice in Flower Mound, Texas.

Stephen J. Clark earned his DMD degree and Endodontic Certificate from the University of Kentucky, and he is a professor at the University of Louisville.

Joseph M. Morelli received his DDS degree from Loyola University and Endodontic Certificate from Tufts University. He currently is an associate professor of endodontics at the University of Louisville.

Ricardo Caicedo received his Dr Odont degree from Colegio Odontologico Colombiana, Bogota, Colombia and his Endodontic Certificate from the University of Louisville. He currently is associate professor of endodontics at the University of Louisville.

References

1. Yared G, Bou Dagher F, Kulkarni K (2003). Influence of torque control motors and the operator’s proficiency on ProTaper failures. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 96(2):229-33.

2. Vieira EP, Franca EC, Martins RC, Buono VT, Bahia MG (2008). Influence of multiple clinical use on fatigue resistance of ProTaper rotary nickel-titanium instruments. Int Endod J. 41(2):163-72.

3. Pruett JP, Clement DJ, Carnes DL, Jr (1997). Cyclic fatigue testing of nickel-titanium endodontic instruments. J Endod. 23(2):77-85.

4. Kitchens GG, Jr., Liewehr FR, Moon PC (2007). The effect of operational speed on the fracture of nickel-titanium rotary instruments. J Endod. 33(1):52-4.

5. Zelada G, Varela P, Martin B, Bahillo JG, Magan F, Ahn S (2002). The effect of rotational speed and the curvature of root canals on the breakage of rotary endodontic instruments. J Endod. 28(7):540-2.

6. Parashos P, Messer HH (2006). Rotary NiTi instrument fracture and its consequences. J Endod. 32(11):1031-43.

7. Parashos P, Gordon I, Messer HH (2004). Factors influencing defects of rotary nickel-titanium endodontic instruments after clinical use. J Endod. 30(10):722-5.

8. Gabel WP, Hoen M, Steiman HR, Pink FE, Dietz R (1999). Effect of rotational speed on nickel-titanium file distortion. J Endod. 25(11):752-4.

9. Schafer E, Florek H (2003). Efficiency of rotary nickel-titanium K3 instruments compared with stainless steel hand K-Flexofile. Part 1. Shaping ability in simulated curved canals. Int Endod J. 36(3):199-207.

10. Yared G, Kulkarni GK, Ghossayn F (2003). An in vitro study of the torsional properties of new and used K3 instruments. Int Endod J. 36(11):764-9.

11. Ullmann CJ, Peters OA (2005). Effect of cyclic fatigue on static fracture loads in ProTaper nickel-titanium rotary instruments. J Endod. 31(3):183-6.

12. Grande NM, Plotino G, Pecci R, Bedini R, Malagnino VA, Somma F (2006). Cyclic fatigue resistance and three-dimensional analysis of instruments from two nickel-titanium rotary systems. Int Endod J. 39(10):755-63.

13. Ounsi HF, Salameh Z, Al-Shalan T, Ferrari M, Grandini S, Pashley DH, et al (2007). Effect of clinical use on the cyclic fatigue resistance of ProTaper nickel-titanium rotary instruments. J Endod. 33(6):737-41.

14. Berutti E, Negro AR, Lendini M, Pasqualini D (2004). Influence of manual preflaring and torque on the failure rate of ProTaper rotary instruments. J Endod. 30(4):228-30.

15. Patino PV, Biedma BM, Liebana CR, Cantatore G, Bahillo JG (2005). The influence of a manual glide path on the separation rate of NiTi rotary instruments. J Endod. 31(2):114-6.

16. Gambarini G, Grande NM, Plotino G, Somma F, Garala M, De Luca M, et al (2008). Fatigue resistance of engine-driven rotary nickel-titanium instruments produced by new manufacturing methods. J Endod. 34(8):1003-5.

17. Zinelis S, Darabara M, Takase T, Ogane K, Papadimitriou GD (2007). The effect of thermal treatment on the resistance of nickel-titanium rotary files in cyclic fatigue. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 103(6):843-7.

18. Anderson ME, Price JW, Parashos P (2007). Fracture resistance of electropolished rotary nickel-titanium endodontic instruments. J Endod. 33(10):1212-6.

19. Herold KS, Johnson BR, Wenckus CS (2007). A scanning electron microscopy evaluation of microfractures, deformation and separation in EndoSequence and Profile nickel-titanium rotary files using an extracted molar tooth model. J Endod. 33(6):712-4.

20. Bui TB, Mitchell JC, Baumgartner JC (2008). Effect of electropolishing ProFile nickel-titanium rotary instruments on cyclic fatigue resistance, torsional resistance, and cutting efficiency. J Endod. 34(2):190-3.

21. Cheung GS, Shen Y, Darvell BW (2007). Does electropolishing improve the low-cycle fatigue behavior of a nickel-titanium rotary instrument in hypochlorite? J Endod. 33(10):1217-21.

22. Gambarini G (2001). Cyclic fatigue of nickel-titanium rotary instruments after clinical use with low- and high-torque endodontic motors. J Endod. 27(12):772-4.

23. Yared G (2004). In vitro study of the torsional properties of new and used ProFile nickel titanium rotary files. J Endod. 30(6):410-2.

24. Silvaggio J, Hicks ML (1997). Effect of heat sterilization on the torsional properties of rotary nickel-titanium endodontic files. J Endod. 23(12):731-4.

25. Peters OA, Roehlike JO, Baumann MA (2007). Effect of immersion in sodium hypochlorite on torque and fatigue resistance of nickel-titanium instruments. J Endod. 33(5):589-93.

26. Cheung GS, Shen Y, Darvell BW (2007). Effect of environment on low-cycle fatigue of a nickel-titanium instrument. J Endod. 33(12):1433-7.

27. Di Fiore PM, Genov KA, Komaroff E, Li Y, Lin L (2006). Nickel-titanium rotary instrument fracture: a clinical practice assessment. Int Endod J. 39(9):700-8.

28. Knowles KI, Hammond NB, Biggs SG, Ibarrola JL (2006). Incidence of instrument separation using LightSpeed rotary instruments. J Endod. 32(1):14-6.

29. Iqbal MK, Kohli MR, Kim JS (2006). A retrospective clinical study of incidence of root canal instrument separation in an endodontics graduate program: a PennEndo database study. J Endod. 32(11):1048-52.

30. Wolcott S, Wolcott J, Ishley D, Kennedy W, Johnson S, Minnich S, et al (2006). Separation incidence of ProTaper rotary instruments: a large cohort clinical evaluation. J Endod. 32(12):1139-41.

31. Tzanetakis GN, Kontakiotis EG, Maurikou DV, Marzelou MP (2008). Prevalence and management of instrument fracture in the postgraduate endodontic program at the Dental School of Athens: a five-year retrospective clinical study. J Endod. 34(6):675-8.

32. Spili P, Parashos P, Messer HH (2005). The impact of instrument fracture on outcome of endodontic treatment. J Endod. 31(12):845-50.

33. Crump MC, Natkin E (1970). Relationship of broken root canal instruments to endodontic case prognosis: a clinical investigation. J Am Dent Assoc. 80(6):1341-7.

34. Cunningham CJ, Senia ES (1992). A three-dimensional study of canal curvatures in the mesial roots of mandibular molars. J Endod.18(6):294-300