Working length (WL) determination is an important step in endodontic therapy, because it defines the point at which root canal preparation and obturation should terminate [1]. A root canal preparation short of the accurate WL can result in incomplete elimination of bacteria from the apical third of the root [2], whereas instrumentation beyond the apex results in overfilling and periapical inflammation [3]. The anatomy of the apical foramen has been suggested as presenting an apical constriction (AC) that apically enlarges toward the major foramen (MF) [4]. However, some studies have shown that frequently this configuration is not present [56].

Radiographic measurements have been used for WL determination. However, variations in the position of the foramen can lead to incorrect measurements when only radiographs are used. The introduction of electronic apex locators (EALs) for use in endodontic therapy has enabled more predictable and accurate measurements in this regard [1]. Although a recent review suggested that at least one radiograph should be acquired to determine WL [7], some studies have demonstrated that EALs are more reliable than radiographic measurements [8]. An increase in the use of EALs may decrease the excessive use of radiographs [9].

Different methodologies have been used to evaluate the accuracy of EALs. Some studies have tested the ability of EALs to locate AC, while some have tested their ability to locate MF. However, it has been reported that MF is a more reliable landmark [6]. Moreover, some studies have used extracted teeth and some have obtained in vivo measurements, followed by comparisons with radiographic measurements [7]. Some in vitro studies have shown that the accuracy of EAL devices may vary when tested in vivo. Although several studies have evaluated EALs, the majority have been conducted in vitro [1].

Third-generation EALs use multiple frequencies, which provide more accurate measurements. Propex II (Dentsply Maillefer, Ballaigues, Switerzland) and Root ZX II (J. Morita, Tokyo, Japan) are 2 examples of third-generation EALs. Propex II works by calculating the ratio of the impedances measured simultaneously at frequencies of 0.5 and 8.0 kHz, while Root ZX II uses the mean square roots of the impedances measured at frequencies of 0.4 and 8.0 kHz [10]. The aim of the present in vivo study was to investigate the accuracy of Propex II and Root ZX II and radiographic measurements for locating MF during endodontic treatment.

In total, 32 maxillary and mandibular premolars with single root canals that required extraction for orthodontic reasons were included in this study. All patients provided written informed consent for participation, and the study was approved by the Institutional Review Board of Faculdade de Odontologia São Leopoldo Mandic (2010/0062). The inclusion criteria were as follows: completely formed apices and a positive response to cold sensitivity tests (Endofrost, Coltene/Whaledent AG, Altstätten, Switzerland). Teeth presenting open apices, signs of root canal calcification, caries, and/or previous restorations were not included. Two radiographs in orthogonal and distal angulations were taken and observed to exclude teeth presenting 2 roots or curvatures of > 5° according to Schneider's technique [11].

Following the induction of local anesthesia and rubber dam isolation, the access cavity was prepared using diamond burs and the presence of vital tissue was confirmed. The pulp chamber was irrigated with 5 mL of 2.5% sodium hypochlorite (NaOCl), and a size 10 K-file was used for initial canal negotiation. The tooth was then evaluated under the magnification of a dental operating microscope (DOM) to confirm the presence of a single canal. The pulp tissue was extirpated using size 10 and 15 K-files in a sequential order, with continuous irrigation using 2 mL of 2.5% NaOCl. S1 and SX rotary ProTaper files (Dentsply Maillefer) were then used at 300 rpm and 3 N to prepare the cervical third of the teeth. Finally, the root canal space was irrigated with 2.5% NaOCl and aspirated to eliminate excess irrigant solution.

Propex II and Root ZX II were used with a size 15 K-file to locate MF. The point at which each EAL maintained the 0.0 measure for 5 seconds was confirmed to be MF. This step was repeated 3 times with both Propex II and Root ZX II. If the same value was obtained with all three measurements, we registered that value. If 2 same and 1 different values were obtained, we used the value that appeared twice. If 3 different values were obtained, we determined the median value. The files were then measured using a digital caliper with a ± 0.01 mm resolution (FNCL, Worker Gage, Esteio, Brazil), and the obtained values were registered in a spreadsheet. Subsequently, radiographic measurements using Ingle's technique were obtained to determine the length of each canal [12]. Based on the initial image a radiograph was taken with a size 15 K-file 2 mm short of the radiographic apex. The difference between the tip of the file and the radiographic apex was added and a new radiograph was taken; the results were recorded. Thus, each tooth was measured a total of 7 times: once with Ingle's technique, 3 times with Propex II, and 3 times with Root ZX. The same experienced operator performed all measurement procedures.

Following in vivo measurements, the teeth were extracted and stored in 2.5% NaOCl. The teeth were then held in a manner that the MF could be directly visualized from a lateral view. A size 15 K-file was introduced into the canal until the tip reached MF. The obtained length was registered in the same digital caliper. A DOM under a × 25 magnification was used for accurate measurements.

All results were analyzed using the Biostat 4.0 program (AnalystSoft Inc., Chapel Hill, NC, USA). Pearson's correlation analyses were used to correlate the in vivo and in vitro measurements. Descriptive statistics using analysis of variance (ANOVA, Tukey's post hoc test) were used to evaluate discrepancies among measurements obtained using radiographic technique and the 2 EALs.

Measurements obtained using Root ZX II and Propex II correlated with the extracted tooth measurements for 75% and 59.37% of the cases, respectively. According to Pearson's correlation analyses, the actual measurements obtained using the visual method exhibited the strongest correlation with the measurements obtained using Root ZX II (r = 0.94), followed by those obtained using Propex II (r = 0.90) and Ingle's technique (r = 0.81; p < 0.001; Table 1).

Descriptive statistics using ANOVA (Tukey's post hoc test) showed significant differences between the radiographic measurements and the 2 EAL measurements (p < 0.05; Table 2).

This study aimed to evaluate the accuracy of 2 third-generation EALs, namely Propex II and Root ZX II, and a radiographic technique for locating MF during endodontic treatment. Several recent studies have evaluated the accuracy of different EALs in vitro [13] and in vivo [1415]. An in vivo study presents a great challenge for evaluation; however, it simulates an actual clinical situation more closely. One challenge is the standardization of specimens. We selected maxillary and mandibular premolars with a single canal and comparable estimated lengths for our study. In addition, the patients were of a similar age, thus ensuring similar root development and avoiding anatomical variations due to cement deposition in older patients. Each specimen was measured by both EALs and radiographic technique, which eliminated intergroup differences in apical sizes. A recent meta-analysis showed that EALs are not influenced by the pulp status [1], whereas another study showed that the results are influenced by the pulp tissue status [16]. To avoid any possible bias, we only included teeth with vital pulp tissue.

We chose MF as a reference point in the present study on the basis of a report showing greater anatomical consistency for this landmark than for AC [6] and a report showing the higher accuracy of EAL in locating MF [17]. Following in vivo location of MF, we compared the measurements with those obtained in vitro using direct visualization. In the present study, both EALs presented better results compared with radiographic technique. A recent study comparing 2 different EALs and a radiographic technique, but found no difference in measurements [18], while another study showed that EAL was more accurate [8]. Kim et al. [19] conducted an in vivo study and showed that the combined use of EALs and radiographs provide the most accurate measurements.

In the present study, measurements obtained using Root ZX II and Propex II precisely correlated with the extracted tooth measurements for 75% and 59.37% cases, respectively. These values were lower than those obtained by Plotino et al. [20] (84.22%) and Real et al. [18] (91.9%) for Root ZX. However, it is important to emphasize that these previous studies were in vitro studies that selected AC as the reference landmark. In addition, one study considered a variation of 0.5 mm as acceptable, while another study allowed a deviation of 1.0 mm within an accurate measurement. In our in vivo study, we used MF as the reference landmark and strictly followed an error-free protocol; by selecting these points, which may have influenced the results, we can consider our correlation rate to be accurate.

Parente et al. [21] have shown a 100% accuracy for MF location by both devices. Our results are in agreement with their finding that there is no significant difference in accuracy between the 2 devices, although our accuracy values were lower for both. Vasconcelos et al. [17] presented an accuracy of 73.5% for Root ZX and 67.6% for Propex II with regard to MF location; these values are similar to ours. One in vivo study [22], however, has presented less accurate results compared with those obtained in vitro. Specifically, only 40% of MF measurements obtained using Propex II were accurate.

In the present study, the accuracy of both third-generation EALs were superior to that of Ingle's radiographic technique. When interpreting and comparing the results of different studies, clinicians should consider the design (in vivo or in vitro), the adopted reference landmark, and the error tolerance for measurements.

Both third-generation electronic apical locators assessed in this study located MF with a similar accuracy that were higher than that of the radiographic measurements. Our results suggest that the use of these EALs for MF location is more accurate than the use of radiographic measurements.