With increasing use of computed tomography (CT) in modern medicine, concerns have arisen regarding increasing radiation dose to the community from med

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights. ORIGINAL RESEARCH PEDIATRIC IMAGING Sarabjeet Singh, MBBS Mannudeep K. Kalra, MD Michael A. Moore, MD Randheer Shailam, MD Bob Liu, PhD Thomas L. Toth, BS Ellen Grant, MD Sjirk J. Westra, MD Dose Reduction and Compliance with Pediatric CT Protocols Adapted to Patient Size, Clinical Indication, and Number of Prior Studies 1 Purpose: Materials and Methods: To assess compliance and resultant radiation dose reduction with new pediatric chest and abdominal computed tomographic (CT) protocols based on patient weight, clinical indication, number of prior CT studies, and automatic exposure control. The study was institutional review board approved and HIPAA compliant. Informed consent was waived. The new pediatric CT protocols, which were organized into six color zones based on clinical indications and number of prior CT examinations in a given patient, were retrospectively assessed. Scanning parameters were adjusted on the basis of patient weight. For gradual dose reduction, pediatric CT (n 692) examinations were performed in three phases of incremental stepwise dose reduction during a 17-month period. There were 245 male patients and 193 female patients (mean age, 12.6 years). Two radiologists independently reviewed CT images for image quality. Data were analyzed by using multivariate analysis of variance. 1 From the Department of Radiology, Massachusetts General Hospital, 25 New Chardon St, 4th Floor, Boston, MA 02114 (S.S., M.K.K., M.A.M., R.S., B.L., E.G., S.J.W.); and GE Healthcare, Waukesha, Wis (T.L.T.). From the 2007 RSNA Annual Meeting. Received August 29, 2008; revision requested October 30; revision received December 8; accepted January 29, 2009; final version accepted February 16. S.J.W. supported in part by a grant from the Society for Pediatric Radiology Research and Education Foundation. Address correspondence to S.S. (e-mail: ssingh6@partners.org ). Results: Conclusion: Compliance with the new protocols in the early stage of implementation (chest CT, 58.9%; abdominal CT, 65.2%) was lower than in the later stage (chest CT, 88%; abdominal CT, 82%) (P.001). For chest CT, there was 52.6% (9.1 vs 19.2 mgy) to 85.4% (2.8 vs 19.2 mgy) dose reduction in the early stage of implementation and 73.5% (4.9 vs 18.5 mgy) to 83.2% (3.1 vs 18.5 mgy) dose reduction in the later stages compared with dose at noncompliant examinations (P.001); there was no loss of clinically relevant image quality. For abdominal CT, there was 34.3% (9.0 vs 13.7 mgy) to 80.2% (2.7 vs 13.7 mgy) dose reduction in the early stage of implementation and 62.4% (6.5 vs 17.3) to 83.8% (2.8 vs 17.3 mgy) dose reduction in the later stage (P.001). Substantial dose reduction and high compliance can be obtained with pediatric CT protocols tailored to clinical indications, patient weight, and number of prior studies. RSNA, 2009 Supplemental material: http://radiology.rsnajnls.org/cgi /content/full/2521081554/dc1 RSNA, 2009 200 radiology.rsnajnls.org Radiology: Volume 252: Number 1 July 2009

With increasing use of computed tomography (CT) in modern medicine, concerns have arisen regarding increasing radiation dose to the community from medical imaging and regarding the associated increasing estimated risk for radiation-induced cancer (1,2). With regard to risk of cancer, studies (3) have shown that children are much more susceptible to radiation than adults. Consequently, several strategies have been assessed to reduce radiation dose for children, including that described by Frush et al (4) for radiation dose optimization with weight-based color-coded protocols for CT body scanning in children. These strategies adapt fixed tube current and peak voltage according to children s weight and enable reduction of CT radiation dose for children. Use of automatic exposure control techniques for automatically adapting the radiation dose to the size of children while maintaining desired or specified image quality has also been described (5 7). In addition, studies have shown that reduced-dose CT scanning can be performed for certain clinical indications, such as for evaluation of kidney stones (8), most bone indications (9,10), follow-up head CT for assessing shunt patency in hydrocephalus (11), and follow-up chest CT for cystic fibrosis (12), bronchiectasis (13), and pulmonary nodules (14,15). However, to the best of our knowledge, no systematic method has described pediatric CT protocols that involve patient weight, clinical indication, number of prior studies, and automatic Advances in Knowledge Pediatric CT protocols based on the clinical indications for CT and the number of prior CT examinations can have up to 88% compliance. Such protocols are associated with 16% 89.5% radiation dose reduction for pediatric chest and abdominal CT studies. Lesion conspicuity and diagnostic acceptability were acceptable with low radiation dose chest and abdominal CT performed in compliance with these new protocols. exposure control for optimizing radiation dose. Therefore, the purpose of our study was to assess compliance and radiation dose reduction with new pediatric chest and abdominal CT protocols based on patient weight, clinical indication, number of prior CT studies, and combined modulation automatic exposure control. Materials and Methods An author (M.K.K.) received research funding from GE Healthcare (Waukesha, Wis). An author (T.L.T.) is an employee of GE Healthcare. The remaining authors (S.S., M.A.M., R.S., B.L., E.G., S.J.W.) have no financial disclosures and had complete unrestricted access to data at all stages of the study. Patients The study was approved by the human research committee of the institutional review board and was compliant with Health Insurance Portability and Accountability Act guidelines. The requirement for informed consent was waived. To reduce radiation dose in a graded manner to allow the radiologists and the technologists to get used to noisier images associated with CT examinations with lower radiation dose, the division of pediatric radiology at our institution introduced new protocols in three phases for gradual incremental stepwise reduction in CT radiation dose for children (phase 1: January 1, 2007 to October 31, 2007; phase 2: November 1, 2007 to February 28, 2008; phase 3: March 1, 2008 to May 31, 2008) (Appendix E1, http: //radiology.rsnajnls.org/cgi/content/full /2521081554/DC1). Consecutive pediatric patients ( 18 years) undergoing chest or abdominal CT studies performed between January 1, Implication for Patient Care Weight-based protocols for radiation dose reduction in children can be structured to adapt radiation dose to clinical indications, presence of prior CT, and body region. 2007 and May 31, 2008 were included in our study. The study cohort consisted of 692 CT examinations (chest CT, n 328; abdominal CT, n 364) performed in 438 children (245 male patients, 193 female patients; age range, 0 18 years; mean age, 12.6 years) (Fig 1). Scanning Protocols All pediatric patients were scanned with an eight-section (LightSpeed Ultra; GE Healthcare), 16-section (LightSpeed 16; GE Healthcare), or 64-section (Light- Speed VCT; GE Healthcare) multidetector CT scanner. Color zones. For each phase, the new CT protocols were divided into six color zones based on clinical indication and presence of one or more prior CT examinations performed at our institution. These zones included pink (most routine indications or rule-out situations; eg, to rule out appendicitis), green (follow-up CT with one prior CT examination performed at our institution; eg, first follow-up of liver abscess), red (index CT for bone evaluation or multiple prior CT examinations performed for the same abnormality and body region; eg, bone fractures or deformity or multiple prior CT examinations performed for pancreatic pseudocyst or pneumonia follow-up), yellow (kidney stone), blue (subtle or small lesions suspected or identified with other imaging results or with clinical examination results), and gray (vascular assessment with CT angiography). The decision Published online before print 10.1148/radiol.2521081554 Radiology 2009; 252:200 208 Author contributions: Guarantors of integrity of entire study, M.K.K., M.A.M., S.J.W.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, S.S., M.K.K., M.A.M., B.L., E.G., S.J.W.; clinical studies, S.S., M.K.K., M.A.M., R.S., E.G., S.J.W.; statistical analysis, S.S., M.K.K.; and manuscript editing, M.K.K., M.A.M., B.L., T.L.T., E.G., S.J.W. See Materials and Methods for pertinent disclosures. See also the editorial by Frush in this issue. Radiology: Volume 252: Number 1 July 2009 radiology.rsnajnls.org 201

about the use of the pink, green, or red zone in patients with known malignancy undergoing follow-up CT scanning was left to the individual radiologist. Weight categories. To adapt radiation dose to patient weight, each zone was further divided into four (for phases 1 and 2) or five (for phase 3) categories on the basis of patient weight (Table 1). These weight categories were set arbitrarily to enable selection of specific voltage for different-sized patients, as well as noise indexes used in our study. Scanning parameters pertaining to the color zones and their weight categories were archived on each scanner and were named (eg, pink 0 9 kg, pink 10 27 kg, and so on for each of the four or five weight categories belonging to the six color zones). Archiving and naming these protocols in the scanner allowed technologists to select and scan by using the preloaded protocols, without making any changes to scanning parameters. Automatic exposure control. All studies were performed by using a combined modulation type (AutomA 3D; GE Healthcare) of automatic exposure control. The noise indexes for the automatic exposure control technique used in the new protocols were selected empirically on the basis of prior studies in adult patients of different sizes (16,17). These patient studies showed that the use of a lower noise index was necessary for smaller adult patients ( 27.2 kg) compared with larger adults. In the study of Kalra et al (16), the authors also described the use of a cone-shaped phantom to document the fact that lower noise indexes should be used for smaller objects compared with average-sized or larger objects. Thus, a noise index of 10 used in this study for smaller patients was selected for larger children ( 45.4 kg) for the new pink protocol. For smaller children, the noise indexes were decreased empirically. Furthermore, the use of different noise indexes for different clinical indications, such as kidney stone protocol, has also been reported for achieving higher dose reduction (8). CT workflow with new protocols. On receiving a request for CT examination in a child, a pediatric radiologist specified the zone on the basis of clinical indication, body region, and images from prior CT studies available in the picture archiving and communication system and/or radiology information system. The color zone was specified on an online protocoling system (IDxRAD; GE Healthcare) available at our institution. For example, for ruling out appendicitis, the radiologist specified pink in the protocoling system, whereas for a kidney stone, yellow was specified. Figure 1 Figure 1: Flowchart depicts patient demographics in the three phases of incremental dose reduction for implementation of new pediatric CT protocols. MDCT multidetector CT scanner, n number of CT examinations. 202 radiology.rsnajnls.org Radiology: Volume 252: Number 1 July 2009

On the basis of the patient weight at the time of scanning, a weight category was selected for each color zone recommended by the radiologist. When radiologists were unavailable prior to scanning or when the scanning indication was emergent, technologists were instructed to use the pink zone, the default zone of radiation dose, and guess Table 1 New Pediatric CT Protocols for Chest and Abdominal Examinations Zone and Weight Category Voltage (kvp) Tube Current Range (ma) the weight of the child to determine the appropriate weight category. Phase 1 Noise Index Phase 2 Noise Index Pink 0 9 kg 80 65 130 4 5 5 10 26 kg 100 80 160 6 7 7 27 45 kg 120 95 190 8 10 10 46 100 kg 120 110 220 10 12 12 101 kg 120 125 300 NA NA 15 Green 0 9 kg 80 50 100 6 7 7 10 26 kg 100 60 120 8 9 9 27 45 kg 120 70 140 10 11 11 46 100 kg 120 80 160 12 13 13 101 kg 120 90 240 NA NA 16 Red 0 9 kg 80 50 100 7 8 8 10 26 kg 100 50 100 9 10 10 27 45 kg 100 50 100 11 12 12 46 100 kg 120 50 100 13 14 14 101 kg 120 50 150 NA NA 18 Yellow 0 9 kg 80 65 130 5 6 6 10 26 kg 100 80 160 7 8 8 27 45 kg 120 95 190 9 12 12 46 100 kg 120 110 220 11 14 14 101 kg 120 125 250 NA NA 17 Blue 0 9 kg 80 80 160 4 5 5 10 26 kg 100 100 200 6 7 7 27 45 kg 120 120 240 8 9 9 46 100 kg 120 140 280 10 11 11 101 kg 120 160 320 NA NA 14 Gray* 0 9 kg 80 100 200 4 5 5 10 26 kg 100 120 240 6 7 7 27 45 kg 120 95 190 8 10 10 46 100 kg 120 120 240 10 12 12 101 kg 120 120 300 NA NA 15 Phase 3 Noise Index Note. Radiation dose for color zones and weight categories was adapted by using appropriate noise index, tube current range, and voltage. For phase 3, a fifth weight category was added for heavier patients to improve compliance, and pitch was increased for all weight categories to lower radiation dose. NA not applicable. * Despite lower tube currents for 27 45-kg patients in the gray zone compared with those of the 10 26-kg group, dose was greater for the former category because of higher voltage. Assessment of Compliance with New Protocols Patient age, sex, and weight, type of examination (chest or abdominal), color zone and weight category used for scanning, and date of examination were recorded. In addition, we also recorded all scanning parameters, as well as the CT dose descriptors CT dose index volume (in milligrays) and dose-length product (in milligray-centimeters). For each CT examination, two authors (S.S. and M.K.K., with 1 and 9 years of experience, respectively) matched the scanning parameters used by technologists with those specified for the color zone and weight category. A CT examination was labeled as compliant with the new protocol when all scanning parameters were correctly used by the technologist for that specific color zone and weight category. CT examinations performed with the use of incorrect scanning parameters were labeled as noncompliant examinations. On CT images, an author (S.S.) measured the maximum skin-to-skin transverse diameter of the chest (just above the diaphragm for all chest CT studies) and the abdomen (above the anterior superior iliac spine for all abdominal CT studies). These parameters were also recorded for prior CT examinations, if available, to compare radiation dose between the new and the prior protocols. Image Quality Assessment Two radiologists (S.J.W., with 22 years of subspecialty experience; R.S., with 6 years of experience) reviewed representative consecutive chest (n 30 for phase 1, n 33 for phase 2, n 50 for phase 3) and abdominal (n 41 for phase 1, n 45 for phase 2, n 50 for phase 3) CT examination results included in our study on a picture archiving and communication system diagnostic workstation (Impax ES; AGFA Technical Imaging Systems, Ridgefield Park, NJ). All consecutive chest and abdominal CT studies performed within 2 weeks of implementation of each phase were reviewed, regardless of study indication, color zone, or weight category. Both radiologists independently evaluated image quality of randomized images obtained with and without the use of the new protocols on a picture archiving and communication system workstation. The radiologists were informed about the phase of study to be reviewed but were not aware of the identity of compliant and noncompliant exam- Radiology: Volume 252: Number 1 July 2009 radiology.rsnajnls.org 203

Table 2 Distribution of Patients Scanned according to Color Zones during Three Phases of Implementation of New CT Protocols Parameter inations, patient demographics, and the date of examination when evaluating for image quality. Image quality was assessed in terms of subjective image noise, beam hardening, visibility of small structures, lesion size, lesion conspicuity, lesion attenuation, and diagnostic acceptability. If there were multiple lesions, the readers were asked to grade the smallest lesion. In case of a discrepancy between the two readers, a study investigator (S.S.) informed the readers about the lesions that should be evaluated. Subjective image noise was graded with a five-point scale based on presence and amount of image mottle or graininess (1 minimal image noise, 2 less than average noise, 3 average image noise, 4 above average noise, 5 unacceptable image noise). The visibility of small structures was also graded by using a fivepoint scale, with 1 indicating excellent visualization and 5 indicating unacceptable visualization of small structures. Beam hardening or streak artifacts were graded with a five-point scale (1 artifacts affecting the interpretation of a lesion or an organ of interest, 2 pronounced artifacts interfering with diagnosis although a diagnosis can be made, 3 moderate artifacts slightly interfering with diagnostic decision making, 4 mild artifacts not interfering with diagnostic decision making, and 5 complete absence of artifacts). Diagnostic acceptability for clinical interpretation of the study was evaluated with a four-point scale, which included fully acceptable (grade 1), probably acceptable (grade 2), acceptable only in limited conditions (grade 3), and diagnostically unacceptable (grade 4). Lesion conspicuity was ranked with a five-point scale, with a score of 5 indicating a wellseen lesion with well-visualized margins and a score of 1 indicating a definite artifact mimicking a lesion. Most assessed image quality attributes used in our study have been described in the European Guidelines on Quality Criteria for Computerized Tomography document (18). In addition to the subjective assessment of image quality, objective image noise and CT numbers (in Hounsfield units) were measured (S.S.) for CT examinations included in the present study. For all chest CT studies (n 113) evaluated by the radiologists, a circular region of interest (10 20 mm 2 based on the size of aorta) was drawn in the descending thoracic aorta at the level of the carina. For all abdominal CT studies (n 126) assessed by the radiologists, a region of interest was drawn in the homogeneous area of the right lobe of liver (20 30 mm 2 ) or in the paraspinal muscles (20 30 mm 2 ) in the presence of heterogeneous liver parenchyma. Data and Statistical Analysis Data were analyzed by using software (SAS, version 9.1.3, SAS Institute, Cary, NC; SPSS, version 13.0, SPSS, Chicago, Ill). Means standard deviations of patient age, weight, and transverse diameter, CT dose index volume, and doselength product were estimated for each color zone and weight category and were compared by using multivariate analysis of variance (SAS statistical software). Image quality scores as described previously were also assessed for each zone and category to determine the number of suboptimal studies by using Wilcoxon signed rank tests (SPSS software). Compliance with the new protocols in each phase of dose reduction was determined by assessing the recorded scanning parameters for each patient and counting the number of compliant and noncompliant examinations during the study period. Radiation dose (CT dose index volume and dose-length product), patient age, No. of Phase 1 Patients No. of Phase 2 Patients No. of Phase 3 Patients Zone Pink 40 (89) 67 (95) 53 (53) Green 75 (108) 28 (67) 24 (24) Red 4 (4) 6 (6) 7 (7) Yellow 13 (13) 4 (6)... Blue 6 (6)...... Gray 1 (1) 3 (3) 1 (1) No. of patients with compliant study 139 (221) 108 (177) 85 (85) No. of patients with noncompliant study 64 (134) 27 (60) 15 (15) Total no. of patients 203 (355) 135 (237) 100 (100) Note. Data in parentheses are numbers of CT examinations performed. No patient was scanned with blue zone protocols during phase 2 or with yellow and blue zone protocols during phase 3. and patient size (weight and transverse diameter) for patients scanned with and those scanned without compliance to the new protocols for different phases of dose reduction were also assessed by using multivariate analysis of variance. Results Patient distribution in individual color zones for individual study phases is summarized in Table 2. Compliance with New Protocols Compared with CT examinations performed in compliance with the new protocol, noncompliant examinations were more often performed in patients who were older, heavier, and larger in size (according to transverse diameter) at all three phases of dose reduction for chest and abdominal CT examinations (P.001). During phase 1, only 58.9% (99 of 168) of chest CT studies and 65.2% (122 of 187) of abdominal CT studies were performed according to the new protocols. In addition, the majority of the noncompliant CT examinations had an inappropriate noise index (32.0% [43 of 134]), tube current range (26.8% [36 of 134]), or combination of voltage, tube current, and noise index (41.0% [55 of 134]). During phase 2, compliance rates for chest and abdominal CT were 76.3% (84 204 radiology.rsnajnls.org Radiology: Volume 252: Number 1 July 2009

of 110) and 73.2% (93 of 127), respectively. Most patients who underwent noncompliant examinations were either scanned at different noise indexes (40% [24 of 60]), tube current ranges (53% [32 of 60]), or combinations of voltage, tube current, and noise index (7% [four of 60]). Phase 3 showed a further increase in compliance for chest CT examination to 88% (44 of 50) from 76.3% in phase 2, while abdominal CT examinations compliance increased to 82% (41 of 50) from 73.2% in phase 2. The majority of noncompliant examinations were performed either at different noise indexes (33% [five of 15]), tube current ranges (60% [nine of 15]), or combinations of voltage, tube current, and noise index (7% [one of 15]). Radiation Dose Compared with the noncompliant examinations, there were significant dose reductions for chest and abdominal CT examinations in all phases (P.001). Radiation dose reductions were noted for chest (Table 3) and abdominal (Table 4) CT examinations performed by using all color zones. There also were dose reductions within each weight category. Regardless of the color zone, there was less variability in radiation dose for CT examinations performed with compliance to the protocols than for examinations performed in the weight- and transverse diameter matched patients scanned without compliance to the protocol (P.001). For example, radiation dose for noncompliant CT examinations performed in patients weighing more than 45 kg ranged from 6 to 55 mgy, whereas radiation dose for the corresponding weight group scanned with compliance to the new protocols was 4 18 mgy. Image Quality Detailed image quality and lesion conspicuity scores for chest and abdominal CT are summarized in Table 5. Despite substantial dose reduction with the use of the new protocols for all age groups, both readers gave similar scores for subjective image quality and lesion conspicuity to the compliant and noncompliant CT studies. Both readers rated image noise, visibility of small structures, lesion conspicuity, diagnostic acceptability, and presence and extent of artifacts as acceptable or above average for studies performed in compliance with the new protocols (Figs 2, 3). Discussion Our study shows that high compliance (up to 88% in our department) and substantial dose reduction (up to 89.5% compared with noncompliant CT examinations) can be obtained with pediatric CT protocols based on clinical indications, number of prior CT examinations, and weight-adjusted automatic exposure control. Chest CT studies performed in accordance with Table 3 Radiation Dose and Patient Size for Chest CT Examinations Performed by Using Color Zones in All Phases Parameter No. of Examinations CT Dose Index Volume (mgy) Dose Reduction (%)* Dose-Length Product (mgy cm) Patient Weight (kg) Transverse Diameter (cm) Phase 1 Noncompliant CT 69 19.2 10.3... 672.0 372.3 60.3 26.8 31.4 4.9 Overall compliant CT 99 7.0 3.2 63.5 326.1 72.5 35.6 24.0 25.3 7.0 Pink zone 42 9.1 3.2 52.6 308.0 166.7 36.4 25.0 24.8 6.4 Green zone 54 5.1 1.3 73.4 187.6 80.0 35.1 22.7 26.3 6.8 Gray zone 1 10.9 43.2 720.8 0.0 82.0 0.0 29.2 0.0 Red zone 2 2.8 0.7 85.4 88.0 43.5 36.4 25 24.8 6.4 Phase 2 Noncompliant CT 26 19.1 10.4... 635.3 399.0 61.1 26.3 30.6 5.0 Overall compliant CT 84 7.6 3.9 58.5 179.7 91.2 39.8 24.0 25.8 6.2 Pink zone 48 7.7 2.4 59.6 284.4 148.5 38.9 23.3 24.5 6.4 Green zone 32 3.8 1.5 80.1 175.6 141.8 37.3 25.1 28.8 5.6 Gray zone 3 9.2 2.3 51.8 216.3 31.7 43.5 20.5 23.9 0.0 Red zone 1 2.0 89.5 42.7 0.0 15.0 0.0 23.3 0.0 Phase 3 Noncompliant CT 6 18.5 5... 1298.4 696.7 71.2 14.8 32.3 6.1 Overall compliant CT 44 4.6 2.0 75.1 119.8 64.7 32.0 19.7 23.8 7.2 Pink zone 26 4.9 2.2 73.5 154.8 98.5 26.8 17.6 22.6 6.9 Green zone 13 4.5 1.6 75.6 152.6 67.3 37.8 18.4 26.2 6.2 Red zone 5 3.1 1.1 83.2 52.0 28.5 44.2 28.0 26.0 10.6 Note. Unless otherwise indicated, data are means standard deviations. A significant decrease in radiation dose in phase 3 was noted compared with phase 1 for all color zones except for the green and red zones, as larger patients were scanned in phase 3 of these zones (P.001). As noncompliant CT examinations in phase 3 were performed at a lower dose than those in phase 1, relative dose reduction in the red zone was also lower in phase 3 than in phase 1. * Percentage dose reductions for the different color zones are in relation to the noncompliant studies in each phase. No patient was scanned in gray zone in phase 3. Radiology: Volume 252: Number 1 July 2009 radiology.rsnajnls.org 205

Table 4 Radiation Dose and Patient Size for Abdominal CT Examinations Performed by Using Color Zones in All Phases Parameter No. of Examinations CT Dose Index Volume (mgy) Dose Reduction (%)* Dose-Length Product (mgy cm) Patient Weight (kg) Transverse Diameter (cm) Phase 1 Noncompliant CT 65 13.7 6.6... 545.0 251.5 58.8 14.5 27.8 3.8 Overall compliant CT 122 7.8 3.2 40.4 318.7 117.8 37.2 1.0 23.7 5.3 Pink zone 47 9.0 2.6 34.3 345.5 135.8 40.7 20.9 24.3 4.9 Green zone 54 5.5 2.4 59.8 217.9 131.4 26.7 17.0 21.2 5.0 Yellow zone 13 9.0 2.3 34.3 380.7 131.6 52.2 23.0 27.6 5.8 Blue zone 6 11.5 3.9 16.0 562.4 161.1 58.2 7.3 32.7 5.9 Red zone 2 2.7 0.6 80.2 87.4 29.3 21.0 8.6 20.2 2.4 Phase 2 Noncompliant CT 34 14.1 5.4... 670.7 484.6 62.6 21.7 29.4 5.0 Overall compliant CT 93 7.3 3.1 41.1 228.7 122.2 36.7 25.0 23.6 5.4 Pink zone 47 8.0 2.9 43.2 379.7 222.0 35.0 24.4 24.0 5.0 Green zone 35 4.07 1.6 71.6 243.8 155.0 47.1 28.0 26.5 5.9 Yellow zone 6 6.4 2.4 54.6 213.1 86.4 34.7 24.4 19.4 3.9 Red zone 5 2.0 0.7 85.4 78.3 25.6 22.2 5.6 21.4 3.1 Phase 3 Noncompliant CT 9 17.3 4.7... 863.4 343.0 69.1 15.7 32.2 5.9 Overall compliant CT 41 5.7 1.5 67.0 190.5 56.0 43.8 21.0 25.0 5.7 Pink zone 27 6.5 0.9 62.4 296.8 99.0 48.0 19.8 26.0 5.0 Green zone 11 4.3 1.5 75.1 171.7 66.7 33.7 23.2 22.7 6.5 Gray zone 1 4.9 71.6 174.1 0.0 22.0 0.0 20.2 0.0 Red zone 2 2.8 0.8 83.8 119.4 58.6 48.5 40.3 26.6 11.6 Note. Unless otherwise indicated, data are means standard deviations. A significant decrease in radiation dose in phases 2 and 3 was noted compared with that in phase 1 for all color zones. * Percentage dose reduction for the different color zones are in relation to the noncompliant studies in each phase. No patient was scanned in yellow zone for phase 3. Table 5 Modal Subjective Image Quality Scores and Objective Image Noise for Chest and Abdominal CT Examinations Parameter Subjective Image Noise Visibility of Small Structures the new protocols showed an average radiation dose reduction of up to 69.8% in all color zones in the three phases of the study, whereas abdominal CT studies performed with the new protocols showed a 60.6% radiation dose reduction. The color-coded pediatric body CT protocols assessed in our study differ from those described in earlier studies, which primarily described patient size specific protocols by using fixed tube current (4). The new protocols assessed in our study require special attention to the clinical indication, which is the main determinant of radiation dose for each patient. Compliance rates in our study were slightly lower than those reported by Frush et al (4) (97% compliance with their color-coded protocols and 91% compliance with conventional CT protocols). Although technologists were explicitly instructed to use the new protocols saved on the scanners for all children, irrespective of their size and clinical indications, several factors may have contributed to this difference in compliance rate in our study. These factors include the use of more than one scanner (seven multidetector CT scanners) and use of scanners with different detector geometry (eight-, 16-, and 64-section multidetector CT Lesion Conspicuity Diagnostic Acceptability Objective Image Noise* Chest CT Compliant 3 3 4 1 12.5 5.4 Noncompliant 4 3 4 1 11.6 3.0 Abdominal CT Compliant 3 3 4 1 15.8 5.8 Noncompliant 3 3 4 1 13.6 2.5 Note. There was no significant difference between subjective image quality parameters and objective image noise between compliant and noncompliant CT examinations (P.31.99). * Data are means standard deviations. scanners); there were three separate scanning locations operated by more than 70 full-time and part-time CT technologists who worked in at least three different shifts in our department during the course of our study. It is also possible that 206 radiology.rsnajnls.org Radiology: Volume 252: Number 1 July 2009

PEDIATRIC IMAGING: Dose Reduction with New Pediatric CT Protocols some technologists may have accidentally selected adult protocols for scanning some large or older children, without realizing their age. Interestingly, most noncompliant CT examinations did occur in larger and older children. Westra et al (19) have reported that there are considerable variations in the scanning techniques and radiation doses for large children. We believe that this may have also occurred because there was some confusion regarding the cutoff age for pediatric CT. In phase 2, all technologists were again instructed that patients 18 years or younger should be scanned with the new pediatric CT protocols, and additionally, in phase 3, technologists were instructed to scan children weighing more than 100 kg with the newly created weight category. This may have contrib- uted in part to an increase in compliance noted in our study for phase 3 protocols. Another reason for noncompliance in phase 1 and 2 could be lack of consistent guidelines to the technologists on dealing with children who are obese and well above 45 kg. Technologists tended to employ their previous policy of using adult protocols for these patients more often because of concerns that the use of lower Figure 2 Figure 2: Transverse CT images in 16-year-old female adolescent weighing 64 kg who underwent abdominal CT for evaluation of trauma. (a) Noncompliant abdominal CT image (10 mgy, 147 285 ma, 120 kvp) shows renal laceration (arrow). (b) Follow-up abdominal CT image obtained with the use of green zone and the more than 45 kg category (5.7 mgy, 159 ma, 120 kvp) shows residual left interpolar renal laceration (arrow). Figure 3 Figure 3: Transverse CT images in 6-year-old boy weighing 15 kg who underwent follow-up chest CT for the assessment of stability of pulmonary nodules. (a) Initial CT image obtained with the use of pink zone and the 10 27-kg category (4.8 mgy, 149 159 ma, 100 kvp) shows pulmonary nodule in right middle lobe (arrow). (b) Follow-up chest CT image obtained with the use of red zone and the 10 27-kg category (2 mgy, 99 ma, 100 kvp) demonstrates stable right middle lobe nodule. Radiology: Volume 252: Number 1 July 2009 radiology.rsnajnls.org 207

radiation dose with the new protocols would result in noisier CT images. There were limitations to our study. We did not compare the new protocols implemented in our department with previously described color-coded, weight-based, or automatic exposure control based protocols for pediatric radiation dose reduction (4). However, radiation doses observed with the use of the new protocols were well within the reference dose levels described in previous studies from Great Britain (20,21). Radiation dose reductions stated in our study are, however, relative to the doses of scanning protocols used in our institution prior to the implementation of the new protocols. Thus, radiation doses with the new protocols described in our study may be much lower or higher than those based on the existent protocols in other centers. It is therefore important for radiologists, technologists, and CT physicists to employ the concept of structured use of clinical indications, presence of prior CT imaging, body region, and patient weight for dose optimization described in our study rather than use the actual scanning parameters employed in our study, which may or may not be transferable to CT scanners with different detector geometry and from different vendors. Indeed, the intent of this study was to enhance the value of the weight-based techniques for radiation dose optimization in children with information related to the scanning indication, presence of prior imaging, and automatic exposure control techniques. We did not assess the accuracy of the clinical indications for CT scanning as stated by the physician or the accuracy and compliance of radiologists in recommending correct color zones. In conclusion, weight-based protocols for radiation dose reduction in children can be structured to adapt radiation dose to clinical indications, presence of prior CT imaging data for the same indication, and body region while using an automatic exposure control technique. References 1. Linton OW, Mettler FA. National conference on dose reduction in CT, with an emphasis on pediatric patients. AJR Am J Roentgenol 2003;181:321 329. 2. Kalra MK, Maher MM, Toth TL, et al. Strategies for CT radiation dose optimization. Radiology 2004;230:619 628. 3. One size does not fit all: reducing risks from pediatric CT. ACR Bull 2001;57:20 23. 4. Frush DP, Soden B, Frush KS, Lowry C. Improved pediatric multidetector body CT using a size-based color-coded format. AJR Am J Roentgenol 2002;178:721 726. 5. Rizzo S, Kalra M, Schmidt B, et al. Comparison of angular and combined automatic tube current modulation techniques with constant tube current CT of the abdomen and pelvis. AJR Am J Roentgenol 2006;186:673 679. 6. McCollough CH, Bruesewitz MR, Kofler JM Jr. CT dose reduction and dose management tools: overview of available options. Radio- Graphics 2006;26:503 512. 7. Greess H, Nömayr A, Wolf H, et al. Dose reduction in CT examination of children by an attenuation-based on-line modulation of tube current (CARE Dose). Eur Radiol 2002; 12:1571 1576. 8. Kalra MK, Maher MM, D Souza RV, et al. Detection of urinary tract stones at low-radiation-dose CT with z-axis automatic tube current modulation: phantom and clinical studies. Radiology 2005;235:523 529. 9. Salamipour H, Jimenez RM, Brec SL, et al. Multidetector row CT in pediatric musculoskeletal imaging. Pediatr Radiol 2005;35: 555 564. 10. Chapman VM, Kalra M, Halpern E, et al. 16-MDCT of the posttraumatic pediatric elbow: optimum parameters and associated radiation dose. AJR Am J Roentgenol 2005; 185:516 521. 11. Udayasankar UK, Braithwaite K, Arvaniti M, et al. Low-dose nonenhanced head CT protocol for follow-up evaluation of children with ventriculoperitoneal shunt: reduction of radiation and effect on image quality. AJNR Am J Neuroradiol 2008;29:802 806. 12. de Jong PA, Nakano Y, Lequin MH, Tiddens HA. Dose reduction for CT in children with cystic fibrosis: is it feasible to reduce the number of images per scan? Pediatr Radiol 2006;36:50 53. 13. Yi CA, Lee KS, Kim TS, Han D, Sung YM, Kim S. Multidetector CT of bronchiectasis: effect of radiation dose on image quality. AJR Am J Roentgenol 2003;181:501 505. 14. Remy-Jardin M, Sobaszek A, Duhamel A, et al. Asbestos-related pleuropulmonary diseases: evaluation with low-dose four-detector row spiral CT. Radiology 2004;233:182 190. 15. Gergely I, Neumann C, Reiger F, Dorffner R. Lung nodule detection with ultra-low-dose CT in routine follow-up of cancer patients [in German]. Rofo 2005;177:1077 1083. 16. Kalra MK, Maher MM, Kamath RS, et al. Sixteen-detector row CT of abdomen and pelvis: study for optimization of Z-axis modulation technique performed in 153 patients. Radiology 2004;233:241 249. 17. Kalra MK, Maher MM, Toth TL, et al. Comparison of Z-axis automatic tube current modulation technique with fixed tube current CT scanning of abdomen and pelvis. Radiology 2004;232:347 353. 18. European guidelines on quality criteria for computed tomography. EUR 16262. http: //www.drs.dk/guidelines/ct/quality/index.htm. Accessed April 21, 2008. 19. Westra SJ, Kalra MK, Liu B, Buckley V, Ferris T. Radiation exposure of obese children from body CT: are they appropriately treated as adults? Radiological Society of North America 2007. http://rsna2007.rsna.org/rsna2007 /V2007/conference/event_display.cfm?em _id 5008891. Accessed April 20, 2008. 20. Wall BF. Implementation of DRLs in the UK. Radiat Prot Dosimetry 2005;114:183 187. 21. Shrimpton PC, Hillier MC, Lewis MA, Dunn M. National survey of doses from CT in the UK: 2003. Br J Radiol 2006;79:968 980. 208 radiology.rsnajnls.org Radiology: Volume 252: Number 1 July 2009