From Volume 8, Issue 1, January 2015 | Pages 24-27
An audit was conducted within a district general hospital's orthodontic department to improve the quality of digital lateral cephalometric radiographs. The results of the audit show improvements in the quality of lateral cephalometric radiographs, following the introduction of a new picture archiving and communications system and charged coupling device cephalometric machine.
Lateral cephalometric radiography is a practical, quantitative method providing information on the morphology and relationship of skeletal, dental and soft tissues. Exposures for lateral cephalometric radiographs result in lower effective doses when compared to three-dimensional imaging.1 Antero-posterior relationships remain some of the most challenging aspects of orthodontic treatment. As such, the cephalogram still represents a dose-efficient diagnostic tool.
Cephalometric analyses vary in measurements, but a common set of anatomical landmarks is generally utilized.2 As well as an aid to diagnosis, cephalograms also form a pre-treatment record, monitor the progress of treatment and monitor growth. All cephalograms produced should be of acceptable diagnostic quality and should be clinically justified prior to exposure. Good quality radiographs are essential for efficient and safe healthcare delivery.
A Quality Assurance programme is one of the essential legal requirements of the Ionizing Radiations Regulations 1999 (IRR99). Clinical audit ensures that quality assurance mechanisms are satisfactory and lead to relevant change and improvement.1 The Health Protection Agency (HPA) has developed subjective ratings of radiographic quality to be used for audit purposes. They class Grade 1 as excellent, Grade 2 as diagnostically acceptable and Grade 3 as unacceptable. The HPA minimum standards for radiological image quality are:
This study is a repeat of an audit completed in 2011, which evaluated the quality of 20 digital lateral cephalometric radiographs requested by the Department of Orthodontics, Victoria Hospital, Fife.4 The previous audit met HPA standards, with 85% radiographs rated as Grade 1, 15% radiographs rated as Grade 2 and no radiographs rated as Grade 3. The radiographs which received a rating of Grade 2 had problems with head positioning, identification of adenoids, identification of lower incisor root apices and tips, and measurement of incisor angulation.
Following the original audit, the results were presented at a local departmental meeting. Recommendations were made to radiology staff on correct head positioning procedures. Improved communication between the radiology and orthodontic departments was also established so that discussion of the causes of poor quality cephalograms would occur.
Two significant changes have occurred since the previous audit. One, a new picture archiving and communications system (PACS) Carestream Vue PACS (Carestream Health Inc, Rochester, New York, USA) has been introduced. The new software provides improved functionality with automated 3D multi-planar reconstruction. Two, a new charged coupling device (CCD) cephalomatric machine, OC200 D Instrumentarium (Instrumentarium Dental, Tuusula, Finland) was installed, replacing the existing storage phosphor plate (SPP) system. CCDs allow instantaneous imaging, avoiding the need to scan storage phosphor plates. Storage phosphor plates are sensitive to ambient light, lose information within minutes of exposure and must be exposed to high intensity light to remove latent images.5 All these factors may lead to a reduction in image quality.
The standard was determined from the results and recommendations of the previous audit cycle which advised that a minimum of 85% cephalograms should be rated as Grade 1, 10% as Grade 2, 5% as Grade 2* and 0% as Grade 3. The extra grade 2* was added to the standard HPA ratings owing to the complexity of cephalograms.
This was a retrospective study which assessed the quality of digital lateral cephalograms. Thirty-five consecutive radiographs requested by the Orthodontic Department at Victoria Hospital, starting from the 5 June 2013, were analysed. The cephalograms were viewed by a single observer on a 17” monitor with Carestream Vue PACS software, allowing visual enhancement. The radiographs were viewed in blocks of 10 units at a time to avoid observer fatigue.6,7
A proforma used for data collection was adapted from methodology set out by the Royal College of Surgeons of England,8 to include five extra variables (italicized), in addition to those outlined by the Royal College.
The 19 variables included:
‘Date of birth’ was included as it was relevant to the cephalogram's clinical context. A ‘Scale of at least 30 mm in the mid sagittal plane’ provides a measure of the magnification of the cephalogram. ‘Airway visible’ and ‘Adenoidal area visible’ were added as the orthodontic department receives referrals for sleep apnoea patients, and ‘No need to repeat’ was added to allow greater correlation with the HPA grading system (Figure 1).
From the proformas, each image was assigned a grade from 1 to 3:
As part of quality assurance, to ensure standards have been maintained since the last audit, the following additional aspects were assessed: cephalometric machine variables, archiving and retrieval, size of monitor, resolution of monitor, PACS file format and provision of image enhancement.
Of the 35 radiographs analysed, 89% were found to be Grade 1, this represents a slight improvement from 85% of radiographs rated as Grade 1 in the initial cycle of the audit. The percentage of radiographs rated as Grade 2 increased slightly from 10% in the initial audit to 11% in the re-audit. No radiographs were rated as Grade 2* in the re-audit, compared to 5% of radiographs in the original audit. No radiographs received a rating of Grade 3 in either audit (Table 1 and Figure 2). The standard set from the previous audit was achieved.
| Grade 1 | Grade 2 | Grade 2* | Grade 3 | |
|---|---|---|---|---|
| First Audit 2011 | 85% (n=17) | 10% (n=2) | 5% (n=1) | 0% (n=0) |
| Re-audit 2013 | 89% (n=31) | 11% (n=4) | 0% (n=0) | 0% (n=0) |
Transverse head tilt causing a lack of superimposition of the lower borders of the mandible was classified as ‘incorrect head position’. This was identified on three cephalograms. The entire cranial vault was visible on two cephalograms and this meant that important structures were not centred on the film. One cephalogram showed both incorrect head positioning and important structures not centred on the film. Figure 3 shows the number of cephalometric imaging errors in the first audit and re-audit.
The standard cephalometric machine variables for a single cephalogram in Victoria Hospital is 77 kV; 8.32 mA; 0.52s. No images were incorrectly archived and all were retrieved quickly (subjective observation). 17” TFT monitor screens with resolutions of 1280 x 1024 (1.3 megapixels): pixel size 0.264 mm were available for viewing cephalograms throughout the orthodontic department. DICOM (digital imaging and communications in medicine) standard PACS file format is still used with Carestream Vue PACS software.9Carestream Vue PACS software allows enhancement of digital images.
Improvements in the quality of digital lateral cephalometric radiographs, following changes implemented since the previous audit, have been demonstrated.
A number of cephalograms in the re-audit displayed transverse head tilt, causing a lapse in the superimposition of the lower border of the mandible. Averaging out two images of the lower border of the mandible is unlikely to yield a valid, traced anatomical feature, and may lead to inaccuracies in cephalometric measurements.10 Projection errors and facial asymmetry may also contribute to the production of double images of bilateral structures.11
The entire cranial vault was captured on two of the cephalograms. This unnecessary irradiation of the cranium and neck can be avoided with appropriate use of paediatric settings and collimation on the OC200 D Instrumentarium machine. This would provide greater compliance with the ALARP (As Low As Reasonably Practical) principle.1
A local audit meeting was organized between the radiology and orthodontic departments at Victoria Hospital to discuss the findings of the re-audit. Correct head positioning procedures were discussed with the aid of light beam markers. The importance of appropriate use of paediatric settings and collimation has also been advised when taking cephalograms for younger patients.
The subjective nature of landmark identification is an issue in all studies of cephalometric radiographs. Ambiguity in identifying cephalometric landmarks can arise due to a lack of sharpness of the image, the radiographic complexity and superimposition of some anatomical regions on cephalograms.11 Greater experience in cephalometric analysis has been shown to increase the reliability of cephalometric landmark identification.12
One observer was used to assess the cephalograms in this re-audit as there is evidence of low variance in intra-observer cephalometric landmark identification.7,13 Inter-observer errors of landmark identification are smaller for monitor-viewed cephalograms compared to hard copies of cephalograms.14 Although the observer has changed between the original audit and re-audit, cephalograms were viewed on monitors with the same specifications in both studies, increasing the validity of the comparison of results.
The tube potential settings for cephalograms at Victoria Hospital was within the recommended 60–90 kV for extra-oral x-rays in the British Orthodontic Guidelines for radiographs. 17” TFT monitors with resolutions of 1280 x 1024 (1.3 megapixels) are used throughout the Orthodontic Department in Victoria Hospital and this complies with the quality assurance recommendations made in the initial audit.4 17” monitors help prevent the loss of visual information which has been identified with 15” monitors. This is due to pixel size mismatch between the CCD array and the monitor screen.15 17” monitors have proven to provide good image quality following introduction of a PACS in a university hospital.16 It is important to use TFT flat screen monitors to prevent parallax errors and distortion.15
Carestream Vue PACS includes features for image enhancement which can increase the reliability of cephalometric landmark identification17 and should be used to avoid repeat exposures.18 When utilizing image enhancement, the minimum amount of adjustment should be applied to images to prevent invalid landmark identification and the introduction of systematic error.19Carestream Vue PACS continues to comply with quality assurance recommendations made in the previous audit that viewing systems should offer image manipulation, especially for contrast adjustment. The pan-manufacturer quality assurance specification ‘DICOM’ was developed in the USA20 and this optimal file format is used with Carestream Vue PACS.
The standard of cephalograms rated as 85% as Grade 1, 15% as Grade 2 and 0% as Grade 3 was met in this second audit, although only a slight improvement in the quality of cephalograms was found when compared to the first audit. A high standard of digital lateral cephalometric radiographic quality is consistently being achieved.
