Nuclear medicine survey statistics

This page contains usage statistics and a summary of the data submitted to the NDRLS nuclear medicine survey.

If you are conducting a publicly funded research project and would like to gain access to an anonymised version of the NDRLS MDCT dataset, please contact the NDRLS team.

Data sources and calculation methodology

The DRLs were derived from a survey, conducted in 2021/22, of nuclear medicine facilities. We asked participants to provide the prescribed activity for all procedures offered at their facility and to report every dose (from both radiopharmaceutical and CT sources) delivered within their facility over a two-week period.

Results from the 2021 nuclear medicine survey have been used to formulate the national DRLs for procedures on adult patients. In all cases, the DRLs have been based on the 75^th percentile of the distribution of Facility Reference Levels (FRLs), where the FRLs are doses/activities indicative of common practice on a particular scanner at a particular facility.

How FRLs were defined varied for different modalities. For general nuclear medicine, the prescribed activity was generally used. A similar approach was used for PET, however facilities that based prescribed activity on patient weight (or a similar characteristic) were treated separately to those that delivered the same activity to all patients. For the CT component of SPECT/CT and PET/CT, an FRL was assigned to each scanner that conducted a particular scan type, with the value of the FRL being the median dose delivered to the patients included in the survey.

The metric chosen to define the DRLs was the 75^th percentile of the FRL distribution, keeping the nuclear medicine DRLs consistent with the CT DRLs already published by ARPANSA. In addition to the DRL, the 50^th and 25^th percentiles of the distributions were also calculated.

Community participation

ARPANSA engaged Services Australia to send invitations to all facilities listed on the Location specific practice number register that are accredited to offer nuclear medicine services. Services Australia reportedly sent approximately 300 invitations. Table 1 contains the number of facilities that engaged with the survey and Table 2 shows the number of facilities and scans included in the final dataset.

Table 1: The number of facilities that engaged with the survey. Note that a single facility could register for both the NM and PET components of the survey, hence the Total column isn’t simply the sum of the other two columns.

	NM facilities	PET facilities	Total
Registered	108	59	127
Submitted survey	92	45	107
Included in analysis	86	44	99

Table 2: The data reported by facilities from which data was used when determining DRLs

	Number of facilities	Number of scans
General nuclear medicine	86	6201
NMCT	82	3784
PET	44	4322
PET/CT	44	4258

Submitted data

Summaries of the number of surveys submitted and the number of scans reported are presented in tables 3 to 7. Also included in the tables are the 25^th, 50^th and 75^th percentiles of the FRL distributions and, where appropriate, the most common activity reported. The FRL distribution for all protocols can be accessed by clicking the entries in the DRL tables on the Current NM DRLs page.

Table 3: The general nuclear medicine protocols reported by NDRLS survey participants. Only the protocols that were used to develop the DRLs are shown.

						Percentiles (MBq)
Category	Scan		Pharmaceuticals	No. FRLs	No. scans	25th	50th	75th	Most common activity (MBq)
Cardiovascular	Gated blood pool scan		Pertechnetate, RBCs	72	180	878	953	1000	1000
	MPI 1-day:	1st phase (rest)	Tetrofosmin, MIBI	52	411	300	305	350	300
		2nd phase (stress)	Tetrofosmin, MIBI	51	371	900	1000	1150	1000
	MPI 2-day	1st phase	Tetrofosmin, MIBI	41	163	350	500	600	600
		2nd phase	Tetrofosmin, MIBI	41	163	350	500	600	600
Endocrine	Thyroid		Pertechnetate	84	199	200	200	200	200
	Parathyroid	Without subtraction	MIBI	31	43	750	800	800	800
		With subtraction	MIBI	49	71	750	800	900	800
		Thyroid subtraction	Pertechnetate	50	58	40	60	220	40
Gastrointestinal	Gastric emptying (solid phase)		Colloid, DTPA	81	125	40	40	40	40
	Colonic transit		67Ga Citrate	75	27	10	20	20	20
Hepatobiliary	Hepatobiliary		HIDA, DISIDA, Mebrofenin	81	67	200	250	300	200
Infection	Infection		67Ga Citrate	70	25	200	200	220	200
Lymphatic	Sentinel node (breast)†:	Same day surgery	Colloid	75	229	20	40	40	40
		Delayed	Colloid	32	59	42	80	80	80
	Sentinel node (melanoma)†		Colloid	65	49	20	40	52	40
Nervous system	Brain		ECD, HMPAO	66	170	740	750	800	750
Pulmonary	Lung perfusion		MAA	82	455	200	200	220	200
Skeletal	Bone scan		MDP, HDP	85	1962	800	825	900	800

† Quoted activities are the total delivered, not per injection. Table 4: The CT scans reportedly conducted as part of a SPECT/CT scan. Note that brain scans are not included here as the brain NMCT DRL was not updated for the 2023 DRLs. Only the protocols that were used to develop the DRLs are shown.         CTDIvol (mGy) DLP (mGy.cm) Category Region No. FRLs No. scans 25th percentile 50th percentile 75th percentile 25th percentile 50th percentile  75th percentile Cardiac Chest (heart) 61 1069 1.51 1.73 2.02 31 36 48 Lymphatic (breast ca.) Chest 39 148 2.25 2.92 3.77 69 84 132 Parathyroid Neck/chest 58 146 3.47 4.62 7.12 106 147 236 Pulmonary Chest (lung) 63 253 2.46 3.06 4.56 83 104 154 Skeletal Single width 102 1155 2.83 3.55 4.78 112 142 195 Double width 64 483 2.64 3.69 4.77 207 281 361 Table 5: The variable prescribed activity PET protocols reported by NDRLS survey participants. Only the protocols that were used to develop the DRLs are shown.         Variable prescribed activity (MBq/kg) Scan Pharmaceutical No. FRLs No. scans 25th percentile 50th percentile 75th percentile Most common Whole body 18F FDG 44 3322 3.0 3.2 3.5 3.0 NETs 68Ga DOTAT-TATE 10 75 2.0 2.0 2.15 2.0 Prostate cancer 68Ga PSMA 10 106 2.0 2.0 2.15 2.0 18F DCFPyL 7 66 3.15 3.55 3.68 3.0 Table 6: The fixed prescribed activity PET protocols reported by NDRLS survey participants. Only the protocols that were used to develop the DRLs are shown.         Fixed prescribed activity (MBq) Scan Pharmaceutical No. FRLs No. scans 25th percentile 50th percentile 75th percentile Most common activity Whole body 18F FDG 44 3322 229 249 267 265 Parkinsonian/ Alzheimer's 18F FDG 24 107 189 200 228 200 NETs 68Ga DOTAT-TATE 17 64 150 185 200 200 Prostate cancer 68Ga PSMA 26 278 160 170 198 160 18F DCFPyL 13 120 240 252 265 252 Table 7: The CT scans reportedly conducted as part of a PET/CT scan. Note that brain scans are not included here as the brain PETCT DRL was not updated for the 2023 DRLs. Only the protocols that were used to develop the DRLs are shown.         CTDIvol (mGy) DLP (mGy.cm) Region Arm position No. FRLs No. scans 25th percentile 50th percentile 75th percentile 25th percentile 50th percentile 75th percentile Brain vertex to prox./mid thighs Up 49 2363 2.61 3.21 4.13 281 336 427 Down 51 857 3.00 3.77 5.25 339 412 553 Brain vertex to toes Up 16 91 2.27 2.74 3.82 428 488 674 Down 46 680 2.62 3.01 4.59 479 570 823 Whole body PET FDG statistics The 2021/22 NDRLS nuclear medicine survey collected data relating to over 3000 whole body 18F FDG scans conducted on 52 different scanners, believed to represent approximately 50% of all PET cameras in use in Australia at the time of the survey. The number of data, and the detail provided, allowed for analysis beyond that used to derive the DRLs. The analysis presented on this page is done so in the hope of providing facilities conducting their DRL comparisons greater context. Administered Activity The variable DRL for whole body FDG was set based on the 75th percentile of the distribution of prescribed activity reported by each facility. However, facility reference levels can also be expressed as the median dose delivered to the patient sample and calculated for each scanner rather than each facility. Doing so results in the FRL distribution shown in Figure 1.

Note that the 75^th percentile of this FRL distribution is slightly higher than the DRL (3.55 MBq/kg compared to 3.5 MBq/kg). This is fairly typical of the data submitted to the survey for all procedure types, facilities tend to administer slightly higher doses than they prescribe. Table 8 provides additional summary statistics of the FRL distribution.

Table 8 Summary statistics from the scanner level FRL distribution shown in Figure 1. With the exception of the number of scanners, all values are expressed in terms of MBq/kg.

Number of scanners	52
Mean	3.33
Standard deviation	0.40
Minimum	2.39
25th percentile	3.03
Median	3.44
75th percentile	3.55
Maximum	4.38

 

Time Activity Product

Of the 3322 scans reported, 2972 (from 49 scanners) provided the make and model of the scanner used and an indication of the scan time per bed position. Having the scan time per bed allows for the calculation of the time activity product (TAP), which theoretically provides an indication of the achievable quality of a scan.

All else being equal, a larger TAP means more counts in an image and a higher quality image. In practice, differences in bed field-of-view, bed-position overlap, detector efficiency, reconstruction techniques and patient characteristics mean that TAP isn’t a direct comparator of quality, however it is worthy of consideration.

The TAP can be expressed in MBq.min/bed or, if the weight of the patient being imaged is considered, MBq.min/bed.kg (henceforth referred to as TAP_w). The median TAP_w values reported from scanners in ARPANSA’s nuclear medicine survey are shown in Figure 2. If you find that your dose is unusually high or low, it is worth comparing the TAP_w used at your facility with the distribution displayed in Figure 2; it may provide an indication as to whether you are obtaining images with relatively high or low counts or if you simply use a different balance of dose and imaging time.

Figure 2 Histogram of the median weight dependent time activity products reported on each scanner.

TAP_w is used by the European Association of Nuclear Medicine to set minimum recommended doses (links to a pdf), with a value of 7 MBq.min/bed.kg suggested as the minimum TAP_w for PET whole body FDG scans. For comparison, the median TAP_w shown in Figure 2 suggests that over half of the survey cohort routinely use a TAP_w lower than the EANM recommended minimum.

Facility Type

Figure 3 and Table 9 summarise the median dose, time per bed, and TAP_w reported to the nuclear medicine survey, classified by the funding model of the reporting facility. There were 3 facilities that identified themselves as being a private company in a public facility, for the purposes of this analysis these facilities have been reclassified as private.

Private facilities reported using lower doses and scanning for broadly similar time per bed as public facilities. This resulted in private facilities reporting lower TAP_ws.

Table 9 Summary statistics of the FRL distributions of administered dose and TAP_w for the different facility types that participated in the nuclear medicine 2021/22 survey.

		Administered dose (MBq/kg)			TAPw (MBq.bed/min.kg)
	Scanners	25%	50%	75%	25%	50%	75%
Private	30	3.01	3.16	3.48	6.03	6.32	7.33
Public	19	3.25	3.57	3.73	6.38	8.62	9.04

 

Figure 3 Boxplots of median dose and time metrics classified by facility funding type. Note that the overlayed scatter plots have had a “jitter” applied to make visualisation clearer, there is no underlying meaning to the relative x position

Detector type

In recent years, solid state photomultiplier technology has become available on clinical PET systems. Only six silicon photomultiplier (SiPM) scanners submitted data to the 2021/22 NDRLS nuclear medicine survey (including one that was a whole-body system) and only four of those scanners provided time data.

The doses administered to patients undergoing exams using the SiPM scanners was lower, however there are too few such scanners to draw any broad conclusions about their relative performance. Furthermore, it is clear from Table 10 that the SiPM scanners do not have a strong influence on the overall distribution of dose and scan times.

Table 10 Summary statistics of the FRL distributions of administered dose and TAP_w for SiPM based detectors compared to PMTs.

		Administered dose (MBq/kg)				TAPw (MBq.bed/min.kg)
	Scanners	25%	50%	75%	Scanners	25%	50%	75%
PMT	46	3.031	3.468	3.551	45	6.158	6.958	8.7
SiPM	6	2.967	3.177	3.38	4	4.958	5.443	6.5

Most common scanners

Figure 4 and Table 11 summarise the distribution of administered activity and TAP_wFRLs recorded using the three most common scanner models. The information is presented to allow users of the listed scanners to compare their administered activities and TAP_ws with their most similar counterparts, it is not to compare the merits of the three included scanners.

Table 11 Summary statistics of the FRL distributions of administered dose and TAPw for the three most common scanner models within the nuclear medicine 2021/22 survey cohort.

			Activity (MBq/kg)			TAPw (MBq.bed/min.kg)
Make	Model	Scanners	25%	50%	75%	25%	50%	75%
Siemens	Biograph mCT	18	3.03	3.18	3.46	6.029	6.357	7.362
GE	Discovery 710	7	3.48	3.55	3.62	6.668	8.843	9.14
Philips	Ingenuity TF	5	3.63	3.70	3.75	8.268	8.65	8.653