Abstract
Objective
To evaluate the reliability of WhatsApp in comparison to the images viewed on a workstation monitor (gold standard) for the identification and interpretation of radiographic images of jaw pathologies.
Methods
150 panoramic radiographs were screened for the assessment of jaw pathologies in the workstation monitor. The radiographs were sent to two observers (Observer A and B) via WhatsApp® Messenger which were viewed independently on smartphones. A structured proforma was prepared to evaluate the radiographs for the presence or absence of various radiographic pathological characteristics.
Results
The reliability of WhatsApp for observers A and B concerning various characteristics like vital structures, pathological fractures, periodontal ligament widening, and root resorption indicated almost perfect agreement (0.8–0.97). The Kappa coefficients for WhatsApp for observers A and B for pre-categorized radiographic impressions were 0.95 and 0.97 which indicated almost perfect agreement.
Conclusion
WhatsApp based expert teleradiology consultation can be a suitable and effective alternative for radiographic interpretations.
Keywords: Teledentistry, Teleradiology, WhatsApp, Instant messaging applications
1. Introduction
The internet was introduced to the world at the end of the 20th century. Roughly twenty years later, the internet has been embraced by every section of our society.1 Worldwide, more than 1,65,000 health-linked applications have been made and approximately 62% of smartphone users have been using their mobile phones to gain health-related advice.2 Telemedicine and teledentistry are indispensable in the present world. It is cost-effective and is readily available even in remote rural areas. With increasing availability and access to smartphones, this can improve the reliability and quality of healthcare set-up due to a quick response time. Social media and instant messaging (IM) platforms have become very popular these days. Applications like WhatsApp Messenger and other instant messaging services have become common methods of communication between doctors and patients as well as between health care professionals.2,3
IM platforms are applications through which the technology allows smartphone users to effectively communicate using a variety of files (text messages, photographs, videos, and audio files). These applications support Internet-based synchronous chat, through point-to-point communication between consumers on similar systems.4 High-quality clinical images, radiographs, images obtained from computed tomography (CT) ultrasound, magnetic resonance imaging, cone-beam CT, histopathology slides can be transmitted.3
Several studies have reported the reliability and efficacy of teledentistry using IM platforms for the diagnosis and treatment planning of dental caries in children with mixed dentition,5 testing the use of smartphone camera for dental caries screening,6 improvement of oral hygiene,7 videoconferencing as a modality for treatment planning for patients requiring prosthetic rehabilitation8 and real-time telemedicine for exchanging expert opinions in the area of pre-implant dental assessment.9
However, there were no studies that have evaluated the role of IM platforms in the evaluation of jaw pathologies on digital radiographs. With this background, we aimed to evaluate the inter-observer reliability in the interpretation of radiographs transmitted by WhatsApp in comparison to the gold standard images. The null hypothesis was that there would be no significant differences in observer agreement between instant messaging application and the gold standard desktop monitor at the workstation.
2. Materials and methods
This was a retrospective study was carried out on the extraoral radiographs from the archives of Oral Radiology (September 2018 to August 2019). Permission to conduct the study was obtained from the Institutional Ethics Committee (IEC: 655/2019). The sample size was estimated to be 149, considering a power of 80%, expected kappa of 0.9, attest value of 0.6, and an expected observed proportion of lesions of 10%.10
A total of 150 panoramic radiographs were screened systematically by a senior maxillofacial radiologist for the assessment of jaw pathologies using a workstation monitor. This was considered as the gold standard for all the findings and interpretation of the radiographic images. Only panoramic images were exclusively considered since it is the most easily available, affordable, and employed modality of imaging for routine screening for jaw pathologies.
Radiographs with and without any evidence of jaw pathologies were included in the study. Radiographs that had technical errors and artefacts that obscured the region of the pathology were excluded from the study. Panoramic radiographs taken for the assessment of common dental pathologies namely progression of caries, assessment of periodontal status, evaluation of impacted third molars were not included in the study as these are seldom sent for second opinion and consultations. Of the 150 panoramic radiographs that were screened, 84 radiographs showed evidence of pathology, and 66 were normal.
All the panoramic radiographs were acquired with a Planmeca 2D S3 unit (Planmeca Oy, Helsinki, Finland). Images were viewed on the workstation monitor using Planmeca Romexis dental imaging software. The radiographs were delinked and anonymized. The senior maxillofacial radiologist sent the 150 radiographs to observers A and B via WhatsApp® Messenger (Version 2.20.71 WhatsApp Inc, Mountain View, CA). Both the observers were trained oral and maxillofacial radiologists. The images on WhatsApp were viewed by the observers on their smartphones (Fig. 1). Observer A viewed the images on a smartphone equipped with the Android version 9 PKQ1.190616.001, MIUI version 11.0.5.0(PCOINXM) operating system having 6.3 inches IPS LCD capacitive touchscreen display, 1080 × 2340 pixels resolution at ~ 409 PPI density, 19.5:9 contrast ratio with 50% display brightness in a room with ambient light. Observer B viewed the images on a smartphone equipped with the iOS operating system with software version 13.5.1 having a 6.1-inch (diagonal) all-screen Liquid Crystal Display (LCD), 1792 × 828-pixel resolution at 326 Pixels per inch (PPI), 1400:1 contrast ratio (typical) with 50% display brightness in a room with ambient light. Both the observers viewed the panoramic images in landscape mode for better visualization. They also used the zoom feature for better clarity whenever it was necessary.
Fig. 1.
Screenshot of the radiographic images sent to the two observers on the WhatsApp® platform.
The observers independently viewed the images and recorded the interpretation in a specially designed proforma. The pathologies were assessed for parameters like the anatomical location and quadrant involved, the type of margins and the type of lesion, presence or the absence of the effect of the pathology on the adjacent vital structures namely pathological fracture, periodontal ligament space widening, and root resorption. Each of the images was further classified as “Benign odontogenic cyst/tumor” or “Benign non-odontogenic cyst/tumor, reactive” or “Malignant non-odontogenic cyst/tumor”; “Others” or “Normal”.
2.1. Statistical analysis
Data were analyzed using SPSS version 20 (IBM Corp. Released 2011. IBM SPSS Statistics for Windows, Version 20.0. Armonk, NY). A p-value of <0.05 was considered statistically significant. Kappa Coefficient was used to evaluate the reliability of observers (Whatsapp images) with the gold standard (work station monitor).
3. Results
Of the 150 panoramic radiographs, 84 showed evidence of pathology, and 66 were normal. Benign odontogenic cysts like odontogenic keratocyst and dentigerous cyst, benign odontogenic tumors like odontoma and ameloblastoma, malignancies of the jaws, and developmental disturbances of the teeth like dentin dysplasia and concrescence were the pathologies noted in the images. Several other findings were also noted like altered condylar morphology, maxillary sinus polyp, multiple impacted supernumerary teeth, tooth floating in space appearance, dense bone islands, fractures, developmental abnormalities of the teeth, and elongated styloid process.
The Kappa coefficient for both the observers that evaluated the status of the teeth and radiograph indicated almost perfect agreement for variables. The reliability of WhatsApp-based radiographic evaluation for identification of quadrants for both the observers was 0.98 which indicated almost perfect agreement. The Kappa coefficients for observers A and B for the assessment of the location of the pathology indicated almost perfect agreement (0.97–0.99.)
The reliability of WhatsApp-based radiographic evaluation for observers A and B in comparison to the gold standard for variables that evaluated the effect on vital structures namely pathological fractures, periodontal ligament widening, and root resorption ranged from 0.8 to 0.97 which indicated almost perfect agreement (Table 1). The reliability of WhatsApp based observations for the assessment of margins and the type of the lesion have been presented in Table 2, Table 3. The Kappa coefficients for WhatsApp based radiographic evaluation for observers A and B for pre-categorized radiographic impressions were 0.95 and 0.97 which indicated almost perfect agreement (Table 4).
Table 1.
The reliability of WhatsApp based observations in comparison to the gold standard for the assessment of various effects on the teeth and surrounding vital structures.
| Gold standard |
Agreement (%) | Kappa (SE) | |||
|---|---|---|---|---|---|
| Absent | Present | ||||
| Effect on vital structures | Observer A | ||||
| Absent | 103 | 2 | 98.67 | 0.97 (0.02) | |
| Present | 0 | 45 | |||
| Observer B | |||||
| Absent | 103 | 2 | 98.67 | 0.97 (0.02) | |
| Present | 0 | 45 | |||
| Pathological fracture | Observer A | ||||
| Absent | 147 | 0 | 99.33 | 0.8 (0.2) | |
| Present | 1 | 2 | |||
| Observer B | |||||
| Absent | 148 | 0 | 100 | 1 (0) | |
| Present | 0 | 2 | |||
| Periodontal widening | Observer A | ||||
| Absent | 112 | 3 | 98 | 0.95 (0.03) | |
| Present | 0 | 35 | |||
| Observer B | |||||
| Absent | 112 | 2 | 98.67 | 0.96 (0.03) | |
| Present | 0 | 36 | |||
| Root resorption | Observer A | ||||
| Absent | 105 | 2 | 98.67 | 0.97 (0.02) | |
| Present | 0 | 43 | |||
| Observer B | |||||
| Absent | 105 | 2 | 98.67 | 0.97 (0.02) | |
| Present | 0 | 43 | |||
Table 2.
The reliability of WhatsApp based observations in comparison to the gold standard for the assessment of margins of the lesion.
| Margins | Gold standard |
Agreement | Kappa (SE) | ||
|---|---|---|---|---|---|
| None | Well defined | Ill defined | |||
| Observer A | |||||
| None | 112 | 3 | 1 | 0.9 (0.04) | |
| Well defined | 0 | 27 | 0 | ||
| Ill defined | 0 | 2 | 5 | ||
| Observer B | |||||
| None | 112 | 2 | 1 | 0.93 (0.03) | |
| Well defined | 0 | 29 | 0 | ||
| Ill defined | 0 | 1 | 5 | ||
Table 3.
The Kappa coefficient for observers A and B for the variable that evaluated the type of lesion.
| Type of lesion | Gold standard |
Agreement | Kappa (SE) | ||||
|---|---|---|---|---|---|---|---|
| None | Radiolucent lesion | Radiopaque lesion | Mixed radiolucent radiopaque lesion | Impacted teeth | |||
| Observer A | |||||||
| None | 94 | 0 | 0 | 0 | 0 | 97.33 | 0.95 (0.02) |
| Radiolucent lesion | 0 | 27 | 0 | 0 | 0 | ||
| Radiopaque lesion | 0 | 0 | 4 | 0 | 0 | ||
| Mixed radiolucent radiopaque lesion | 0 | 0 | 0 | 3 | 0 | ||
| Impacted teeth | 0 | 3 | 0 | 1 | 18 | ||
| Observer B | |||||||
| None | 94 | 0 | 0 | 0 | 0 | 98 | 0.96 (0.02) |
| Radiolucent lesion | 0 | 27 | 0 | 0 | 0 | ||
| Radiopaque lesion | 0 | 0 | 4 | 0 | 0 | ||
| Mixed radiolucent radiopaque lesion | 0 | 0 | 0 | 4 | 0 | ||
| Impacted teeth | 0 | 3 | 0 | 0 | 18 | ||
Table 4.
The Kappa coefficients for WhatsApp based radiographic evaluation for observers A and B for pre-categorized radiographic impression.
| Gold standard |
Agreement (%) | Kappa (SE) | |||||
|---|---|---|---|---|---|---|---|
| benign odontogenic cyst/tumor | benign nonodontogenic cyst/tumor, reactive | malignant nonodontogenic cyst/tumor | Others | Normal | |||
| Observer A | |||||||
| Benign odontogenic cyst/tumor | 26 | 0 | 0 | 0 | 0 | 96.67 | 0.95 (0.02) |
| Benign nonodontogenic cyst/tumor, reactive | 0 | 4 | 0 | 0 | 0 | ||
| Malignant nonodontogenic cyst/tumor | 0 | 0 | 3 | 0 | 0 | ||
| Others | 5 | 0 | 0 | 48 | 0 | ||
| Normal | 0 | 0 | 0 | 0 | 64 | ||
| Observer B | |||||||
| Benign odontogenic cyst/tumor | 28 | 0 | 0 | 0 | 0 | 98 | 0.97 (0.02) |
| Benign nonodontogenic cyst/tumor, reactive | 0 | 4 | 0 | 0 | 0 | ||
| Malignant nonodontogenic cyst/tumor | 0 | 0 | 3 | 0 | 0 | ||
| Others | 3 | 0 | 0 | 48 | 0 | ||
| Normal | 0 | 0 | 0 | 0 | 64 | ||
4. Discussion
Instant messaging applications began as a predominantly text-based one-to-one communication. Today, messaging applications have advanced to the extent of providing audio, video, documents, photograph sharing facilities.4 The adoption of instant messaging applications in teledentistry and radiology would require the assessment of reliability.5
Several studies have tested the efficacy of IM applications in the field of dentistry all of which were performed by sharing clinical images of patients. A recent study showed that IM and e-mail platforms can be used as reliable tools for remote consultations for the diagnosis of maxillofacial fractures.11 Also, mobile phone-based teledentistry in the diagnosis and treatment planning of dental caries in mixed dentition was shown to be reliable.5 A teledentistry model utilizing a smartphone camera for screening dental caries concluded that users were satisfied.6 Real-time telemedicine for exchanging expert opinions on pre-implant dental assessment was found to provide satisfactory preoperative evaluation.9 Sarode et al.12 reported that the WhatsApp application was a suitable and reliable tool for seeking expert consultation for histopathological diagnosis. Vinayagamoorthy et al. demonstrated the feasibility of using a remote screening of oral potentially malignant disorders using IM application.13 Very few studies have been performed in the field of teleradiology for testing the effectiveness of free messaging services in the interpretation of radiographs of which all of them evaluated the fractures.14,15
Our study is evaluated the inter-observer reliability of IM applications in obtaining a radiographic interpretation and impression of jaw pathologies. Details like the margins of the lesion, effect on the vital structures, widening of the periodontal ligament space, and root resorption have also been assessed and evaluated by two observers. Oral and maxillofacial radiologists receive a lot of teleconsultations for a second opinion. IM platforms like the WhastApp are preferred for such consultation due to the ease at which those platforms can be operated. Moreover, many clinicians would have the provisions for acquiring an image may not have ready access to the radiologist for interpretation.
Our study evaluated panoramic radiographs for various characteristics on the workstation monitor by a senior maxillofacial radiologist (gold standard). Subsequently, the radiographs were via WhatsApp messenger to two observers. Images that are shared through WhatsApp are compressed to enable the users to share larger files easily and download the images faster. Although this can be an advantage for users in rural areas with limited connectivity and internet speed, the quality and the resolution of the shared images would be lower than the original image. Furthermore, the images sent via WhatsApp are viewed on smartphones having a smaller screen in comparison to the desktop. Within these limitations, our study demonstrated excellent reliability of WhatsApp based radiographic evaluation for various characteristics that were relevant to jaw pathologies.
Reliability of variables like effect on vital structures, pathological fractures, periodontal ligament widening, and root resorption had also perfect agreement. Similar reliability coefficients were reported by Kapıcıoğlu et al.16 with the use of WhatsApp in the diagnosis of type 1 and 2 pediatric supracondylar fractures. Stahl et al.17 also reported near-perfect agreement between interpretations of the radiographs on smartphones in pediatric limb fractures. Effect on the surrounding structures is crucial for assessment as it dictates the surgical approach and treatment outcome.
The Kappa coefficients for WhatsApp based radiographic evaluation for observers A and B for pre-categorized radiographic impressions were 0.95 and 0.97 which indicated almost perfect agreement. The findings were in agreement with the previous study5 which included only clinical images.
The present study has shown that free messaging applications like WhatsApp are valuable and effective in obtaining radiographic interpretations and impressions. It can be a valuable tool in situations of pandemics where the patients cannot access healthcare facilities. Tele-radiology can strengthen the primary and secondary healthcare facilities that often do not have the expertise.18 With the rising rate of cybercrimes, the security of data transfer is a concern. To address this aspect, Whatsapp is marketed as end-to-end encryption to ensure data privacy.
Both the observers in this study used smartphones with a very good resolution which could have yielded higher reliability estimates. Also, factors like internet access and connectivity and availability of the experts have a strong influence on the acceptability of these methods. There is a possibility that smartphones with lower resolution screens could yield lower reliability estimates. However, due to the rapid technological advancement in the development of smartphones, these limitations could be non-existent in the future.
5. Conclusion
Radiographs of jaw pathologies sent via free messaging applications can be interpreted with good reliability.
Acknowledgments
None.
References
- 1.Desai T., Yee J., Soman S. Smartphone apps: a patient’s new best friend? Clin J Am Soc Nephrol. 2016;11(6):935–937. doi: 10.2215/CJN.03650316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nascimento Ijb do, Oliveira JA. de Q., Wolff I.S. Use of smartphone-based instant messaging services in medical practice: a cross-sectional study. Sao Paulo Med J. 2020;138(1):86–92. doi: 10.1590/1516-3180.2020.0010.R1.28032020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sarode S., Sarode G. WhatsApp use in dentistry: future prospects. J Dent Res Rev. 2016;3(1):3. doi: 10.4103/2348-2915.180104. [DOI] [Google Scholar]
- 4.Bakker G De, Sloep P., Jochems W. Students and instant messaging: a survey of current use and demands for higher education. ALT-J. 2007;15(2):143–153. doi: 10.1080/09687760701470973. [DOI] [Google Scholar]
- 5.Alshaya M.S., Assery M.K., Pani S.C. Reliability of mobile phone teledentistry in dental diagnosis and treatment planning in mixed dentition. J Telemed Telecare. 2020;26(1-2):45–52. doi: 10.1177/1357633X18793767. [DOI] [PubMed] [Google Scholar]
- 6.Estai M., Kanagasingam Y., Xiao D. End-user acceptance of a cloud-based teledentistry system and Android phone app for remote screening for oral diseases. J Telemed Telecare. 2017;23(1):44–52. doi: 10.1177/1357633X15621847. [DOI] [PubMed] [Google Scholar]
- 7.Patil S., Hedad I.A., Jafer A.A. Effectiveness of mobile phone applications in improving oral hygiene care and outcomes in orthodontic patients. J Oral Biol Craniofacial Res. 2020;11(1):26–32. doi: 10.1016/j.jobcr.2020.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ignatius E., Perãlãt S., Mãkelã K. Use of videoconferencing for consultation in dental prosthetics and oral rehabilitation. J Telemed Telecare. 2010;16(8):467–470. doi: 10.1258/jtt.2010.100303. [DOI] [PubMed] [Google Scholar]
- 9.Nickenig H.-J., Wichmann M., Schlegel A., Eitner S. Use of telemedicine for pre-implant dental assessment – a comparative study. J Telemed Telecare. 2008;14(2):93–97. doi: 10.1258/jtt.2007.070806. [DOI] [PubMed] [Google Scholar]
- 10.Kraemer H.C., Bloch D.A., Donner A., Eliasziw M. A goodness-of-fit approach to inference procedures for the kappa statistic: confidence interval construction, significance-testing and sample size estimation. Stat Med. 1994;13(8):876–880. doi: 10.1002/sim.4780130809. [DOI] [PubMed] [Google Scholar]
- 11.Madi M., Kumar M., Pentapati K.C., Vineetha R., Chhaparwal Y. Email and instant messaging applications as platforms for remote oral radiology consultation in maxillofacial fractures. Oral Surg Oral Med Oral Pathol Oral Radiol. 2020:1–6. doi: 10.1016/j.oooo.2020.07.007. 00(00) [DOI] [PubMed] [Google Scholar]
- 12.Sarode S.C., Sarode G.S., Anand R., Patil S., Unadkat H. WhatsApp is an effective tool for obtaining second opinion in oral pathology practice. J Oral Pathol Med. 2017;46(7):513–519. doi: 10.1111/jop.12515. [DOI] [PubMed] [Google Scholar]
- 13.Vinayagamoorthy K., Acharya S., Kumar M., Pentapati K.C., Acharya S. Efficacy of a remote screening model for oral potentially malignant disorders using a free messaging application: a diagnostic test for accuracy study. Aust J Rural Health. 2019;27(2):170–176. doi: 10.1111/ajr.12496. [DOI] [PubMed] [Google Scholar]
- 14.Giordano V., Koch H.A., Mendes C.H., Bergamin A., de Souza F.S., do Amaral N.P. WhatsApp messenger is useful and reproducible in the assessment of tibial plateau fractures: inter- and intra-observer agreement study. Int J Med Inf. 2015;84(2):141–148. doi: 10.1016/j.ijmedinf.2014.11.002. [DOI] [PubMed] [Google Scholar]
- 15.Stahl I., Dreyfuss D., Ofir D. Reliability of smartphone-based teleradiology for evaluating thoracolumbar spine fractures. Spine J. 2017;17(2):161–167. doi: 10.1016/j.spinee.2016.08.021. [DOI] [PubMed] [Google Scholar]
- 16.Kapıcıoğlu M. The reliability of use of WhatsApp in type 1 and type 2 pediatric supracondylar fractures. Jt Dis Relat Surg. 2019;30(2):149–154. doi: 10.5606/ehc.2019.66166. [DOI] [PubMed] [Google Scholar]
- 17.Stahl I., Katsman A., Zaidman M., Keshet D., Sigal A., Eidelman M. Reliability of smartphone-based instant messaging application for diagnosis, classification, and decision-making in pediatric orthopedic trauma. Pediatr Emerg Care. 2019;35(6):403–406. doi: 10.1097/PEC.0000000000001211. [DOI] [PubMed] [Google Scholar]
- 18.Ohannessian R. Telemedicine: potential applications in epidemic situations. Eur Res Telemed/La Rech Eur en Télémédecine. 2015;4(3):95–98. doi: 10.1016/j.eurtel.2015.08.002. [DOI] [Google Scholar]