National Diagnostic Reference Level Service statistics

Introduction

ARPANSA's Medical Imaging Section is conducting an ongoing national dosimetry service for common Multi Detector Computed Tomography (MDCT) protocols from which the first Australian diagnostic reference levels (DRLs) for MDCT have been calculated. The National Diagnostic Reference Level Service (NDRLS) began operation in August 2011 and data is collated on an annual (calendar year) basis. The service is a collaborative project conducted in partnership with the Commonwealth Department of Health (Health), the Royal Australian and New Zealand College of Radiologists (RANZCR), the Australasian College of Physical Scientists and Engineers in Medicine (ACPSEM), the Australian and New Zealand Society of Nuclear Medicine (ANZSNM) and the Australian Society of Medical Imaging and Radiation Therapy (ASMIRT).

Multi detector computed tomography data (MDCT)

The current set of Australian Adult MDCT DRLs are calculated from data collected in 2011. The currents sets of Australian Paediatric (baby and child) MDCT DRLs were calculated from data collected by an independent RANZCR MDCT survey.

A technical report covering the first three years of the National Diagnostic Reference Level Service (2011-2013) is available at ARPANSA/TR171, February 2015.

An overview of the data required per acquisition protocol is available on page 6 of the User Guide:

PDF iconNational Diagnostic Reference Level Survey User Guide - 2016
FileNational Diagnostic Reference Level Survey User Guide - 2016

 

Successful use of the service requires facility registration, appropriate staff contacts, logging of MDCT platforms and completion of individual surveys for specific protocols with sufficient patient dosimetry data to generate a facility reference level (FRL). The FRL is the median value of the CTDIvol and DLP dose metrics submitted per compliant survey.

The following graphs and tables provide an overview of the NDRLS results to date:

  1. Overview
    1. Data Table
  2. Facility Registrations
    1. By State/Territory & Year
    2. By Type & Year
  3. Compliant Surveys
    1. Survey Submissions by Year
    2. Paediatric (Baby & Child) Compliant Surveys by State/Territory and Protocol
  4. FRL Data
    1. Cumulative Distributions
  5. Data Variation
    1. DLP 95% Confidence Intervals
    2. CTDIvol 95% Confidence Intervals

MDCT iterative reconstruction

Information from facilities concerning their use of iterative reconstruction (IR) has been collected since May 2013. Significant dose savings are apparent.

The following graphs provide an overview of the dose savings per protocol by applying IR technology:

  1. CTDIv variation
  2. DLP variation
  3. Percentage Reduction per Protocol

Image guided interventional procedures: draft survey results

In October 2014 the NDRLS released the DRL survey for Image

In October 2014 the NDRLS released the DRL survey for Image Guided Interventional Procedures (IGIP) in an Excel spreadsheet format. Facilities need to register for participation and are sent a workbook with a choice of 7 procedures; normal coronary angiogram, 3 vessel or less cerebral angiogram with/without 3D rotational angiography, 4 or more vessel cerebral angiogram with/without 3D rotiational angiography, abdominal angiogram and ERCP. The NDRLS requests 30 patients to be logged per procedure type along with accompanying details of equipment.

The following tables and graphs provide an overview of the survey results to date for the normal coronary angiogram procedure:

Guided Interventional Procedures (IGIP) in an Excel spreadsheet format. Facilities need to register for participation and are sent a workbook with a choice of 7 procedures; normal coronary angiogram, 3 vessel or less cerebral angiogram with/without 3D rotational angiography, 4 or more vessel cerebral angiogram with/without 3D rotiational angiography, abdominal angiogram and ERCP. The NDRLS requests 30 patients to be logged per procedure type along with accompanying details of equipment.

The following tables and graphs provide an overview of the survey results to date for the normal coronary angiogram procedure:

  1. IGIP Facility Survey Data Input Table
  2. IGIP KAP Meter Calibration Survey
  3. IGIP Draft Survey Data Distributions 1
  4. IGIP Draft Survey Data Distributions 2

NDRLS data overview

To date, on a yearly and monthly current year basis, a brief overview of submitted data is as follows:

Table 1: NDRLS MDCT FRL data overview
  2011 2012 2013 2014 2015 2016 2017 TOTAL
Facility Registrations ACT 5 0 0 0 2 3 0 10
NSW 13 44 12 7 11 36 0 123
NT 0 0 0 1 0 0 0 1
QLD 12 21 6 3 9 14 0 65
SA 3 7 0 1 3 1 0 15
TAS 3 6 1 0 1 0 0 11
VIC 30 7 20 21 40 8 0 123
WA 16 4 1 0 1 2 0 24
Total 82 89 40 33 67 64 0 372
Adult Facility Reference Level Reports (FRL) AbdomenPelvis 51 100 150 128 194 274 904 904
Chest 44 78 112 113 202 258 788 788
ChestAbdomenPelvis 40 68 100 93 156 200 643 643
Head 56 113 166 147 177 284 977 977
LumbarSpine 34 75 116 106 135 237 734 734
Neck 30 57 80 76 141 192 582 582
Baby (0-4 yr)
FRL
AbdomenPelvis 0 0 2 2 1 2 0 7
Chest 0 3 5 4 17 4 0 19
Head 0 20 23 18 3 20 1 100
Child (5-14 yr)
FRL
AbdomenPelvis 0 3 6 4 3 5 0 24
Chest 0 7 8 7 22 9 0 39
Head 0 28 26 19 8 22 0 117
Survey Submissions Started 378 977 1041 896 1242 1784 82 6399
Compliant (FRL) 255 552 794 717 1062 1507 47 4933
Non-compliant 123 425 247 179 180 277 35 1466

Facility registration

A facility is a site performing radiology and/or nuclear medicine and/or IGIP as defined by the allocation of a Medicare Location Specific Practice Number (LSPN). Previously there has been some confusion using the term practice to mean both (i) a site of practice and (ii) the performance of a radiological procedure. Recently the IAEA has suggested that meaning (i) should be changed to the term facility. It is the intention of the NDRLS to take up and apply this to our nomenclature and documentation.

Graph of facility registrations by each State/Territory covering 2011-2015 Figure 1: Facility Registrations by Year and State/Territory
Graph of Facility Registrations by Practice Type -including private clinic, private clinic in a private hospital, private clinic in a public hospital and public clinic in a public hospital Figure 2: Facility Registrations by Facility Type and Year

Compliant surveys

NDRLS strongly suggests that 20 patients be logged per protocol for sufficient statistical validation. Facilities are unable to submit surveys with less than 10 patients acquired.

graph of adult  surveys by year Figure 3: Surveys Submissions by Year

 

Paediatric (Baby & Child) Compliant Surveys by State/Territory Figure 4: Paediatric (Baby & Child) Compliant Surveys by State/Territory and Protocol

FRL data

The curves below show the cumulative distributions for the DLP and CTDIvol from adult DRL surveys submitted to ARPANSA in 2016. The plots display the percentage of surveys that achieved a FRL below the dose specified on the x-axis..

The blue curve represents all of the submitted data, the green curve represents the surveys where iterative reconstruction (IR) was used and the red curve represents the surveys where IR was not used. The data portrayed is more recent than the data used to specify the current DRLs and, consequently, the DRLs do not match the 75th percentile of the plotted distributions – in general, the 75th percentile is considerably lower than the DRL.

Curves showing CTDI and DLP cumulative distributions by with/without IR for Abdomen Pelvis, Adult, Single phase

Abdomen Pelvis, Adult, Single phase Figure 5: Abdomen Pelvis, Adult, Single phase

The distributions for all protocols can be found at FRL data.

DRL variation 2011 – 2015

The following charts track the variation in the third quartile of the distribution of FRLs over the survey periods.  The surveys in each calendar year have been further categorised into those utilising iterative reconstruction (IR) and those not.  In earlier years the use of IR was not coded.  Non-parametric statistical confidence intervals for the third quartile in each category were calculated using the method of Conover. (Conover, W.J. (1980) Practical Nonparametric Statistics John Wiley and Sons, New York.)

*95% CI = 0.75N ± 1.96 √0.19N where N = No. FRLs


It is also important to note that since the data were not obtained from a random sample the distributions and their third quartiles are not necessarily statistically representative of the whole population. The 95% CI are purely based on sample size.

95% Confidence Intervals for DLP (mGy.cm) DRLs Figure 6: 95% Confidence Intervals for DLP (mGy.cm) DRLs - AbdoPelvis Scans - Adult Single Phase

 

For the full set of graphs for 95% Confidence Intervals for DLP (mGy.cm) DRLs per protocol, see NDRLS - 95% Confidence Intervals for DLP DRLs per protocol.

Top of Page

95% Confidence Intervals for DLP (mGy.cm) DRLs Figure 7: 95% Confidence Intervals for CTDIvol DRLs - AbdoPelvis Scans - Adult Single Phase

 

For the full set of graphs for 95% Confidence Intervals for CTDIvol DRLs per protocol, see NDRLS - 95% Confidence Intervals for CTDIvol DRL per protocol.

Top of Page

Impact of iterative reconstruction

In April 2013 the MDCT survey form was modified to ask the participants to log their use of iterative reconstruction (IR) with a yes/no check box. Analysis of the subsequent data strongly indicates that the application of IR to all protocols substantially reduces the dose delivered to the patient.

Variation in CTDIvol (mGy) with and without the application of iterative reconstruction algorithms = head, nect, chest, abdopelvis, CAP and Lumbar spine Figure 8: Overall dose saving with the application of IR algorithms.

Image guided interventional procedures

Data submitted for the IGIP survey in 2015-16 are summarised below.

Submissions were received from 23 facilities incorporating 42 interventional rooms. A total of 52 surveys were submitted, with most including data for the requested 30 patients. Facility Reference Levels (FRLs) were computed for the 44 surveys that included at least 20 patients. The data for coronary angiography are presented below as this was the only protocol with sufficient data.

Survey format

Variation in DLP (mGy.cm) with and without the application of iterative reconstruction algorithms - May 2013 to Dec 2015 Figure 9: IGIP survey input data form

A total of 305 patient cases were submitted, giving a reasonably well sized data set for analysis.

Kerma air product and calibration accuracy

It was recognised that a potential source of significant error was the accuracy of the kerma-air product (KAP, Gy.cm2) meters. Subsequent survey submissions reinforced this as a source of error.

Table image - Last KAP calibration dates and estimated accuracy Figure 10: Last KAP calibration dates and estimated accuracy

six Graphs - Survey distributions of Age, Weight, Fluoroscopy Time, DA Time, Frames and KAP Figure 11: Survey distributions of Age, Weight, Fluoroscopy Time, DA Time, Frames and KAP

four graphs -Survey distributions of DA Time vs Frames, KAP vs Frame Rate, KAP vs Make and KAP vs Reference Dose Figure 12: Survey distributions of DA Time vs Frames, KAP vs Frame Rate, KAP vs Make and KAP vs Reference Dose

Further information on International Best Practice for Radiation Protection of Patients can be found on the International Atomic Energy Agency Radiation Protection of Patients website, which can be accessed via the following link:

IAEA Radiation Protection of Patients (RPOP)

IAEA Radiation Protection of Patients Logo