Radiotherapy error and near-miss data report: December ... · Of the RTE reports, 65.8% (n = 10,414) were ‘near miss’ or ‘other non-conformances’ with no impact on patient

Radiotherapy error and near-miss data report: December 2015 to November 2017 Report No. 5

Radiotherapy error and near-miss data report: December 2015 to November 2017

2

About Public Health England

Public Health England exists to protect and improve the nation’s health and wellbeing

and reduce health inequalities. We do this through world-leading science, knowledge

and intelligence, advocacy, partnerships and the delivery of specialist public health

services. We are an executive agency of the Department of Health and Social Care,

and a distinct delivery organisation with operational autonomy. We provide

government, local government, the NHS, Parliament, industry and the public with

evidence-based professional, scientific and delivery expertise and support.

Public Health England

Wellington House

133-155 Waterloo Road

London SE1 8UG

Tel: 020 7654 8000

www.gov.uk/phe

Twitter: @PHE_uk

Facebook: www.facebook.com/PublicHealthEngland

Prepared by: Medical Exposures Group

For queries relating to this document, please contact: [email protected]

© Crown copyright 2019

You may re-use this information (excluding logos) free of charge in any format or

medium, under the terms of the Open Government Licence v3.0. To view this licence,

visit OGL. Where we have identified any third-party copyright information you will need

to obtain permission from the copyright holders concerned.

Published May 2019

PHE publications PHE supports the UN

gateway number: GW-446 Sustainable Development Goals

http://www.gov.uk/phe

https://twitter.com/PHE_uk

http://www.facebook.com/PublicHealthEngland

mailto:[email protected]

https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/

https://www.gov.uk/government/collections/sustainability-and-public-health-a-guide-to-good-practice


3

Contents

About Public Health England 2

Executive summary 4

1. Introduction 7

2. Background 8

3. Data 10

3.1 Obtaining the data 10 3.2 Number of reports 11 3.3 Lag time for reporting 13 3.4 Organisation of the database 13

3.5 Quality assurance of the data 13

4. Results 15

4.1 Main themes of RTE 15 4.2 Classification level of RTE 18

4.3 Breakdown of classification by process code 19 4.5 Causative factors 28

4.6 Brachytherapy errors 30 4.7 Inspectorate data 31

5. Discussion 34

5.1 Increase in RTE reporting 34 5.2 Main themes 35

5.3 Classification level of RTE 36 5.4 Safety barriers 40

5.5 Causative factors 41 5.6 Inspectorate data 41

6. Conclusion 42

7. Recommendations 43

8. Acknowledgements and steering group 44

PSRT Steering Group membership 44

10. References 45

Appendix A: Radiotherapy error classification grid 47

Appendix B: Refined radiotherapy pathway coding (including safety barrier taxonomy) 48

Appendix C: Causative Factor Taxonomy 63


4

Executive summary

The value of near-miss and error reporting and learning processes is well appreciated

in the UK radiotherapy (RT) community with 100% of NHS RT providers subscribing to

a national voluntary reporting of RT error and near-miss events system.

This report is the fifth in a series of 2-year reports1, providing an overview of

Radiotherapy Error (RTE) data reported voluntarily to the National Reporting and

Learning System (NRLS) and directly to Public Health England (PHE) between

December 2015 and November 2017. The report also contains aggregate data from

December 2009 to November 2015 and compares the new data with that from

previously published reports1. This fifth report includes analysis using the new

taxonomies2 including the refined pathway coding, safety barrier and causative factor

taxonomies. The analysis within this report facilitates comparisons of national RTE

reporting with both local and network trends.

A total of 15,830 RTE reports were included in this analysis. Consistent with previous

reports1 RTEs were spread across all 21 categories of process codes, treatment unit

process codes were the most frequently reported RTE (41.2%, n = 6,526). This is to be

expected as treatment processes are reliant on the accurate interpretation of the

planning information, and RT treatments usually span several visits, providing many

opportunities during the treatment unit process for RTE to occur.

Of the RTE reports, 65.8% (n = 10,414) were ‘near miss’ or ‘other non-conformances’

with no impact on patient outcome. In total, 31.8% (n = 5,035) of the RTEs reported

were not clinically significant and were classified as ‘minor radiation incidents’. Of the

remaining 2.4% (n = 381) RTE reports only 1.0% were reportable. When compared

with the results from the previous 2-year report there has been a reduction in the

percentage of reportable radiation incidents from 1.8% to 1.0%.

Safety barriers (SB) are embedded across the RT pathway coding. The SB taxonomy

can be used to separately identify both failed and effective SB. For this reporting

period, 5,191 subcodes were identified as failed SB, and ‘on-set imaging approval

process’ was the most frequently occurring failed SB (15.3%, n = 796). Future reports

will also contain effective SB or method of detection. There were 7,778 RTE reported

between January 2017 and November 2017 of which 35.9% (n = 2,793) contained

causative factors. The most frequently reported primary CF was ‘slips and lapses’

(32.1%, n = 897).

After 10 years of reporting, the proportion of significant incidents has started to reduce

slightly. RT is delivered in a complex and rapidly evolving landscape. Therefore, it is

imperative that RTE trends continue to be reported, analysed and monitored on a


5

cyclical basis, to inform ongoing safe and effective RT practice. This is especially

pertinent as new techniques and technologies are implemented and as new clinical RT

providers are established. This work supports a risk-based approach to improving

safety both locally, regionally and nationally and indicates a culture that is open,

transparent and already present in the UK RT community. Providers are encouraged to

contact PHE if they have any queries about reporting or would like further advice.

Local provider recommendations are:

• all UK RT providers should continue to use Towards Safer RT3 and Development of

Learning2 to classify and code all levels of RTEs for local analysis to inform local

learning, policy and practice

• local learning should be compared with the national picture and used to inform local

and network level learning

• UK NHS RT providers should continue to submit coded and classified RTE reports

to the national voluntary reporting system using the mechanisms identified within

this report

• UK NHS RT providers should ensure all levels of RTE are reported locally and

nationally

• providers are encouraged to report monthly to ensure timeliness of learning

• UK RT providers should adopt all codes, including safety barriers and causative

factors when coding RTE and the use of secondary or additional subcodes

• independent providers should consider submitting all levels of RTE to the national

voluntary reporting system

• UK RT providers should review effectiveness of local end of process checks across

the entire pathway

National recommendations are:

• the patient safety in radiotherapy steering group (PSRT) should continue to develop

analysis of the reports, with regular dissemination of findings to the RT community

for group learning

• the data should be used both by the PSRT and by individual RT providers as part of

a risk-based approach to allocating resources for improving patient safety in RT and

to inform audit and research

• the mechanism to enable providers in Scotland and Northern Ireland to submit RTE

reports to PHE for collation and analysis as part of a UK dataset should be further

refined

• PSRT should review the effectiveness of corrective actions and preventative

measures (safety barriers)


6

• vendors should consider developments to reduce the rate of RTE related to

equipment failure1

1 PSRT. Radiotherapy errors and near misses: biennial report. Nos 1-4. Available at

www.gov.uk/government/publications/radiotherapy-errors-and-near-misses-data-report 2 PSRT. Development of learning from radiotherapy errors. Available at

www.gov.uk/government/publications/development-of-learning-from-radiotherapy-errors 3 RCR, SCoR, IPEM, NPSA, BIR. Towards Safer Radiotherapy. Royal College of Radiologists, London (2008).

Available at www.rcr.ac.uk/towards-safer-radiotherapy

http://www.gov.uk/government/publications/radiotherapy-errors-and-near-misses-data-report

http://www.gov.uk/government/publications/development-of-learning-from-radiotherapy-errors

http://www.rcr.ac.uk/towards-safer-radiotherapy


7

1. Introduction

The fundamental role of reporting and learning reporting systems is to enhance patient

safety by learning from failures of the healthcare system1. It is known that most

problems are not just a series of random, unconnected one-off events; they are

provoked by weaknesses in systems and processes and often have common root

causes which can be generalised and corrected. Although each event is unique, there

are likely to be similarities and patterns in sources of risk that may go unnoticed if

incidents are not reported and analysed. To maintain or improve patient safety, error

must be prevented, or minimised.

The value of near miss and error reporting and learning processes is well appreciated

in the UK radiotherapy (RT) community with 100% of NHS RT providers subscribing to

a system of national voluntary reporting of RT error and near miss events.

The 20062 report of the Chief Medical Officer for England and Towards Safer

Radiotherapy (TSRT), published in 20083, were seminal documents in the field of RT

safety; both contained practical recommendations for the RT community aimed at

improving safety and reducing errors. These recommendations have been adopted by

UK RT providers. In 2016, the Development of Learning (DoL) from Radiotherapy

Errors4, a guidance document supporting the enhancement of learning from RTE and

their analysis, was published.

This report is the fifth in a series of 2-year reports5, providing an overview of

Radiotherapy Error (RTE) data reported voluntarily to the National Reporting and

Learning System (NRLS) and directly to PHE between December 2015 and November

2017. The report also contains aggregate data from December 2009 to November 2015

and compares the new data with that from previously published reports5.


8

2. Background

TSRT3 provides definitions for the terminology to be used in discussing radiotherapy

(RT) errors that include near misses (RTE) and proposed 2 taxonomies for use in

describing RTEs. The ‘Classification of radiotherapy errors grid’ (Appendix A) describes

the severity of the error and the ‘Radiotherapy pathway coding’ described where in the

RT pathway the error occurred. In December 2016, a guidance document containing

the refinement of the RT pathway coding to include safety barriers (SB) (Appendix B)

and a proposed causative factor (CF) taxonomy (Appendix C) was published4. The

document also contains definitions and examples on the application of the taxonomies.

To aid the RT community in the adoption of the taxonomies, a workshop was held in

October 2017. The workshop also provided opportunities for sharing between the RT

community and the Patient Safety in Radiotherapy Steering Group (PSRT) on the

analysis of RTE. The new taxonomies have been adopted by NHS RT providers across

the UK.

In 2008, the PSRT was tasked with monitoring the implementation of the

recommendations from TSRT, through a collaborative programme of work with the RT

community. PSRT membership includes representatives from Public Health England

(PHE), Royal College of Radiologists, Society and College of Radiographers, Institute

of Physics and Engineering in Medicine and a lay representative. The Group is chaired

by PHE where detailed analysis of RTE reports is undertaken, reported to the PSRT

and disseminated to the RT community to provide learning opportunities.

The National Reporting and Learning System (NRLS) is an anonymised voluntary

reporting system to collect and learn from patient safety incidents for England and

Wales. The NRLS is managed by the Patient Safety Team who currently sit within NHS

England and Improvement. PHE has a data sharing agreement with the NRLS and

under this agreement continues to facilitate extraction of RTE data from the NRLS, and

share learning from these events to make services safer for patients. This collaboration

led to the publication in July 20105 of the first 2-year data report on a back catalogue of

patient safety incidents reported to the NRLS between August 2007 and November

2009.

The first report aimed to provide learning from incidents already reported, to trial the

application of the taxonomies from TSRT and to test the mechanism for reporting.

Subsequently the PSRT asked RT provider staff to classify and code all errors

according to the taxonomies presented in TSRT within their local incident reporting

systems for subsequent submission for analysis by PHE staff as part of a voluntary

reporting scheme. This publication was closely followed in September 2010 by the

introduction of a series of newsletters entitled Safer Radiotherapy6 providing regular

updates on the analysis of RTE reports for professionals working in the RT community.


9

In 2012, a second report was published5. In November 2013, a mechanism for

providers in Northern Ireland and Scotland to contribute to this voluntary reporting

scheme was introduced. Subsequently, data from across the UK, including data

received from the inspectorates for IR(M)ER for England, Wales, Northern Ireland and

Scotland was published within the third and the fourth5 2 yearly reports.

This fifth report includes analysis using the new taxonomies4 including the refined

pathway coding, safety barrier and causative factor taxonomies.


10

3. Data

The data presented in this report is anonymised and received as part of a voluntary

reporting scheme. As with any voluntary reporting system, the data will only reflect

those incidents that are reported and may not necessarily be representative of the

actual levels of occurrence, as such, this data needs interpreting with care.

The data in this report is represented in three timeframes, the current analysis (for

reporting period December 2015 to November 2017), the previous 2016 analysis (for

reporting period December 2013 to November 2015), and the aggregate data (for

reporting period December 2007 to November 2015).

3.1 Obtaining the data

The voluntary data was obtained through 2 distinct routes: from the NRLS for providers

in England and Wales and directly from providers in Northern Ireland and Scotland.

These routes are described in detail below. An anonymised closed synopsis of

reportable radiation incidents was shared directly from the inspectorates for IR(ME)R

for England, Wales, Northern Ireland and Scotland with PHE for inclusion within the

database.

3.1.1 National voluntary reporting system

The clear majority of reports came through the NRLS7 at NHS England and

Improvement, which collates reports for England and Wales. The NRLS operates a

voluntary reporting system to collect and learn from patient safety incidents. A patient

safety incident (PSI) is defined as:

‘Any unintended or unexpected incident which could have, or did, lead to harm for one

or more patients receiving healthcare’7.

PSIs are primarily reported by NHS organisations in England and Wales through bulk

upload via local trust risk management systems and web-based forms to the NRLS.

Independent providers are also starting to report using bulk upload. Patients, carers

and independent providers can also report directly through an open access form8. The

NRLS offers a unique dataset to help understand harm associated with healthcare. It

was established in 2003 and now has over 19.5 million PSI reports9, from many areas

of healthcare, in the database. The NRLS is now under redevelopment as part of the

patient safety incident management systems (DPSIMS) project10, the new system will

support learning, so the NHS can continue to improve safety.


11

The NRLS can be interrogated for relevant incidents by searching the free text field of

any incident report using key words or search terms. During the development of this

work, a system was created to extract targeted data from the NRLS using a trigger

code ‘TSRT9’. This was proposed and described in ‘Implementing Towards Safer

Radiotherapy: guidance on reporting radiotherapy (RT) errors and near misses

effectively’11. This code is searched for in the free text field rather than using search

terms that were found to be less determinant. PSIs that are not RTEs, such as a report

of a patient falling in ‘radiotherapy’, are not included in the RTE dataset. An RTE is

defined in TSRT as:

‘a non-conformance where there is an unintended divergence between a RT treatment

delivered or a RT process followed and that defined as correct by local protocol’3.

3.1.2 Northern Ireland and Scotland

A mechanism was developed to enable providers in Northern Ireland and Scotland to

participate in this scheme in 2013. Once agreements for data sharing were achieved

with health boards and hospital trusts, predefined criteria consistent with those

employed for the NRLS data were shared with RT providers in Northern Ireland and

Scotland for inclusion in reports.

Anonymised data has been accepted from providers on Microsoft Excel spreadsheets

for direct upload into the PHE RTE incident database to minimise the possibility of

transcription error and to ensure the anonymity of the data.

PHE is working to further streamline the reporting mechanism for providers in Northern

Ireland and Scotland while continuing to receive and include this data in the tri-annual

analysis6.

3.2 Number of reports

A total of 15,979 RTE reports were submitted to the voluntary reporting scheme

between December 2015 and November 2017, with an average 666 reports per month.

There has been an increase in reporting since the last 2-year reporting period, where a

total of 12,800 RTE reports were submitted between December 2013 and November

2015, with an average 533 reports per month. Figure 1 indicates a peak in reporting in

October. This spike in October and peak in May could be related to the 6 monthly

NRLS reporting cut-off dates for their annual reports. RT providers may be more likely

to bulk upload in these 2 months of the year to hit this deadline9. The data also

indicated that 2 separate providers reported a backlog of data during October. This

variation highlights that not all providers report monthly as indicated in figure 1.


12

Figure 1. Average number of RTE reports submitted to the national voluntary reporting system by month

For this 2-year period, reports were received from the clear majority of NHS RT

providers, 59 (96.7%). This is a slight decrease on the previous analysis where 60 NHS

providers reported to the national analysis. It is clear there is some variance in the

number of reports provided, one provider reported 2,160 RTE across all levels of RTE

over the 2-year period and 62.3% (n = 38) providers reported less than the mean

average of 260 RTE over the 2-year period (Figure 2). There is a wide range in

provider size and capacity, and the number of reports per provider has not been

normalised to account for this variation. It should be noted that those providers

reporting higher numbers of RTE represent providers with mature reporting cultures

and should be encouraged to continue reporting.

Figure 2. Number of RTE reported per provider, Dec 2015 to Nov 2017

0

200

400

600

800

1000

1200

Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov

Dec 15-Nov 17averageDec 13-Nov 15averageAggregateaverage

260

0

500

1000

1500

2000

2500

Num

ber

of R

TE

rep

orts

RT provider


13

3.3 Lag time for reporting

A lag time between the date of the RTE and the date on which it was reported to the

NRLS or PHE was calculated for each report included in the dataset. This measures

the time from the date of RTE or discovery of the RTE through local reporting on to

PHE.

A minimum reporting lag of 0 days and a maximum 871 days was found for the

individual RTE, with an average of 66 days and a median of 48 days across providers.

This lag time is like the previous 2-year analysis where the minimum reporting lag was

1 day and maximum 864 days for each individual RTE and an average of 58.8 days

across providers. There were 4 outliers of more than 2 years. Two reports did not

contain enough information to explain the extended time lag. One report was a level 1

RTE and may have been reported once the full investigation was completed. The other

outlier was identified when preparing a patient for a second course of treatment 861

days after the original error. This was a level 5 error and was found to be an error in the

recording of the patient summary.

3.4 Organisation of the database

An established SQL database acts as a repository for the data. Searches within the

database can include review of data by taxonomies, date and keyword searches. It also

supports the analysis of the data using scripted reports on frequency and trends

analysis. The database has access restricted to named individuals directly involved

with the quality assurance and analysis of the data. All data was submitted for upload

by PHE staff to the database by import of Microsoft Excel spreadsheets. For the

purposes of the analysis each RTE report submitted is included as a single report.

3.5 Quality assurance of the data

All providers were asked to include a trigger code, classification and coding in RTE

reports. In December 2016, the DoL4 taxonomies were published and providers were

asked to include a trigger code, classification, pathway coding, including failed safety

barriers, causative factor and, where applicable, effective safety barriers (detection

methods) to their RTE reports to facilitate both local and national analysis.

The data supplied was initially scrutinised by PHE staff and consistency checking of the

application of the coding and classification undertaken. Reports run in the database

were checked against manual calculations to verify the expected outcomes.

On receipt of the reports, PHE staff with clinical RT expertise performed consistency

checking of the local application of the classification and coding. During consistency

checking the coding is reviewed for all RTE classified as reportable through to near


14

miss (levels 1-4) and 10% of non-conformances (level 5) RTE are audited. This formed

part of the data quality assurance process completed prior to analysis of the reports.

Consistency checking of the appropriate application of the coding and classification

taxonomies remained high, with 82.4% agreed in this dataset. However, this is lower

than 87.1% agreement achieved in the 2016 report5. Data received between December

2015 and December 2016 indicated a 79.7% agreement rate, 85.4% consistency rate

was seen in the months January 2017 to November 2017. This increase in consistency

may be due to the DoL document which was published in December 2016, with an aim

to provide guidance on the application of the taxonomies4. The PSRT held a workshop

in October 2017 on the application of the refined pathway coding and new taxonomies

from the DoL4 one of the aims of the workshop was to allow consistency in the uptake

of the new and amended taxonomies.

Reports were categorised into complete, incomplete or non-RTE. Complete reports

contain the classification and coding, complete fixed reports are defined as complete

reports which have had the classification and/or the pathway coding amended for

consistency reasons. Incomplete reports are defined as reports without the

classification and coding being applied locally prior to submission, incomplete fixed

reports are reports which had sufficient text descriptors to assign the classification

and/or pathway coding.

Of the 15,979 RTE reports received, a total of 15,830 reports were included in the

analysis 14,412 had been classified and coded by local RT providers (Figure 3). There

were 1,418 incomplete reports, all of which contained sufficient information to assign

classification and coding. The remaining 149 non-RTE or PSI reports, were excluded

from the analysis. A total of 99.1% (90.2% complete and 8.9% incomplete) of the data

submitted was included for analysis in this report, this is consistent with reported data in

20165.

Figure 3. Data quality Dec 2015 to Nov 2017 (n = 15,979)

11873

2539

01418 149

Complete Report

Complete - Fixed Report

Incomplete Report

Incomplete - Fixed Report

Non-RTE Report


15

4. Results

4.1 Main themes of RTE

The 15,830 RTE reports were categorised by classification, process code or process

subcode including failed and effective safety barriers and causative factors, so that the

main themes could be derived.

4.1.1 Breakdown of process codes

The entire dataset was broken down by process code and classification level. The RTE

reported associated with ‘treatment unit processes’ comprised 41.2% (n = 6,526) of the

data and 15.4% (n = 2,431) occurred during ‘pre-treatment planning process’ (Figure

4). RTEs were spread across all 21 categories of process code. In this subset of data, it

may be seen that the majority (64.6%, n = 9,698) of reports comprise of ‘near misses’

(Level 4) and ‘other non-conformances’ (Level 5). ‘Minor radiation incidents’ (Level 3)

made up 32.9% (n = 4,940) of these reports. The remaining 2.5% (n = 368) reports

were ‘non-reportable radiation incidents’ (level 2) and ‘reportable radiation incidents’

(Level 1).

Figure 4. Breakdown of RTE main activity Dec 2015 - Nov 2017 (15,006/15,830 subset of RTE)

The main activity was then reviewed for this 2-year period, the 2016 analysis and

aggregate data as shown in figure 5. The RTE reported associated with ‘treatment unit

6526

2431

1819

1130

997

745

509

333

264

252

0 1000 2000 3000 4000 5000 6000 7000

Treatment unit process

Pretreatment planning process

Pretreatment activities

Treatment data entry process

Booking process

Communication of intent

Referral for treatment

On-treatment review process

Process prior to first appointment

Timing

Number of RTE Reports

Level 1

Level 2

Level 3

Level 4

Level 5


16

processes’ and ‘pre-treatment activities’ has been consistent across this 2-year period

and the previous 2-year period. There has been a decrease in ‘pre-treatment planning

process’ activity from 17.3% across the aggregate data to 15.4% in this 2-year

reporting period. Both ‘booking process’ and ‘treatment data entry process’ has

increased since the last 2-year period and in the aggregate data.

Figure 5. Breakdown of RTE main activity as percentage of RTE reports

4.1.2 Breakdown of process subcodes

The most frequently reported process subcode was ‘on-set imaging: production

process’, making up 11.5% (n = 1,827) of all RTEs reported. This was followed by

‘accuracy of data entry’ 5.1% (n = 805) and ‘on-set imaging: approval process’ 5.0% (n

= 796). Of note, on-set imaging processes represented 4 of the most frequently

reported subcodes making up 24.1% (n = 3,813) of all RTE. The most frequently

reported subcodes and their classification levels are presented in figure 6. It can be

seen that most of the main themes are made up of ‘near misses’ (level 4) at 26.3% (n =

1,921) and ‘other non-conformance’ (level 5) at 29.1% (n = 2,127). In reality, the

incidence of level 4 and 5 reports is likely to be much higher as not all these events are

currently reported under the national voluntary reporting system12. ‘Minor radiation

incidents’ (Level 3) made up 43.2% (n = 3,155) of this subset of reports. The remaining

1.4% (n = 103) reports were ‘non-reportable radiation incidents’ (level 2) and

‘reportable radiation incidents’ (Level 1).

37.6%

17.3%

11.8%

6.9%

4.9%

41.3%

16.2%

11.4%

6.4%

5.2%

41.2%

15.4%

11.5%

7.1%

6.3%

0 5 10 15 20 25 30 35 40 45


Pretreatment planningprocess



Booking process

Percentage of RTE reports

2018, data for 2015-2017(n = 15,830)2016, data for 2013-2015(n = 12,691)Aggregate, 2009-2015(n = 23,662)


17

Figure 6. Breakdown of RTE main themes by classification Dec 2015 - Nov 2017 (7,306/15,830 subset of RTE)

The main themes were then reviewed for this 2-year period, the 2016 analysis and the

aggregate data as shown in figure 7. The RTE reported associated with ‘on-set

imaging: production process’ has increased from 6.5% of the aggregate data to 8.4% in

the last 2-year period and 11.5% in this 2-year period. ‘Accuracy of data entry’ and

‘documentation of instructions’ have also increased. Whilst ‘on-set imaging: approval

process’ and ‘use of on-set imaging’ has decreased.

Figure 7. Breakdown of RTE main themes 2018/2016 and aggregate for top 5

1827

805

796

758

646

544

443

398

391

363

335

0 200 400 600 800 1000 1200 1400 1600 1800 2000

(13z) On-set imaging: production process

(12f) Accuracy of data entry

(13aa) On-set imaging: approval process

(10j) Documentation of instructions/information

(13i) Use of on-set imaging

(13bb) On-set imaging: recording process

(11j) Generation of plan for approval

(11o) Management of process flow within planning

(11n) Recording of patient specific instructions

(6a) Bookings made according to protocol

(13l) Movements from reference marks

Total number of RTE

Level 1

Level 2

Level 3

Level 4

Level 5

6.5%

4.7%

5.4%

4.0%

5.0%

8.4%

4.3%

5.8%

4.2%

6.5%

11.5%

5.1%

5.0%

4.8%

4.1%

0 2 4 6 8 10 12




(10j) Documentation of instructions



2018, data for2015-2017 (n = 15,830)

2016, data for2013-2015 (n = 12,691)

Aggregate, data for2009-2015 (n = 23,662)


18

4.2 Classification level of RTE

Each of the reports were classified as ‘other non-conformances’, ‘near miss’, ‘minor

radiation incident’, ‘non-reportable radiation incident’ and ‘reportable radiation incident’.

Figure 8 includes data for this 2-year period, the 2016 analysis and aggregate data.

Of the RTE reports, 65.8% (n = 10,414) were ‘near miss’ or ‘other non-conformances’

with no impact on patient outcome. In total, 31.8% (n = 5,035) of the RTEs reported

were not clinically significant and were classified as ‘minor radiation incidents’. Of the

remaining 2.4% (n = 381) of RTE reports only 1.0% were reportable under either of 2

statutory instruments, IR(ME)R13 or IRR14, to the appropriate authority. This is reflective

of the previous 2-year analysis. Of note, across both 2-year analyses and aggregate

data the ‘non-reportable radiation incidents’ have stayed the same at 1.4%, however

the ‘reportable radiation incidents’ have reduced from 1.8% during the last 2-year

analysis and 1.7% within the aggregate data to 1.0% within this 2-year analysis.

Figure 8. Classification levels as a percentage of RTE reports

4.2.1 Providers reporting per classification

It is clear there is some variance in the classification levels reported as part of the

national voluntary scheme. Figure 9 indicates that not all providers report all levels of

errors. All providers submitting to this system reported ‘near miss’ data and the majority

98.3% (n = 58) reported ‘minor radiation incidents’, however only 91.5% (n = 54)

reported ‘other non-conformances’ this may be due to lack of resource and non-

electronic reporting systems15. Providers are required to report all ‘reportable radiation

incidents’ to the appropriate authority13,14. Only 74.6% (n = 44) of providers reported

reportable and non-reportable radiation incidents; this may be due to reporting burden

38.2%

29.0%

29.7%

1.4%

1.7%

36.6%

28.7%

31.5%

1.4%

1.8%

40.2%

25.6%

31.8%

1.4%

1.0%

0 5 10 15 20 25 30 35 40 45

5 Other non-conformance

4 Near miss

3 Minor radiation incident

2 Non-reportable radiation incident

1 Reportable radiation incident


2018, data for 2015-2017(n = 15,830)2016, data for 2013-2015(n = 12,691)Aggregate, data for 2009-2015(n = 23,662)


19

and not reporting incidents twice. A national survey of providers undertaken in

November 2018 by PHE staff confirmed this assessment12.

Figure 9. Classification level reported by providers Dec 2015 to Nov 2017

4.3 Breakdown of classification by process code

In this section, the RTE reports are broken down by classification into their attributed

process codes and subcodes.

4.3.1 Breakdown of Level 1 (reportable radiation incident) RTE

‘Reportable radiation incidents’ fall into the category of incidents reportable under either

of the statutory instruments, IR(ME)R13 or IRR14. Clearly, reporting to the national

voluntary reporting scheme does not negate regulatory requirements to report events to

the appropriate authority.

Most level 1 events reported affected only a single fraction of treatment and thus were

correctable over the remaining fractions with no significant impact on the patient or

outcome of treatment. There were 52 different subcodes associated with the 159 Level

1 RTE (Figure 10). The most frequently reported occurred during ‘localisation of

intended volume’ comprising of 10.1% (n = 16) of all level 1 RTE. This was followed by

‘setting of couch position/angle’ (7.5%, n = 12), ‘on-set imaging: approval process’

(6.9%, n = 11) and ‘authorisation to irradiate’ (6.3%, n = 10). Examples of RTE

associated with ‘localisation of intended volume’ include when the incorrect anatomical

area of the patient is CT scanned for planning and requires multiple re-scans.

Examples of RTE associated with ‘setting of couch position/angle’ include when the

incorrect couch height is set for a post FSD patient. Examples of RTE associated with

‘on-set imaging: approval process’ includes when an image has been mismatched and

treated to the incorrect vertebrae leading to a clinically significant difference in

treatment.

54

59

58

44

44

0 10 20 30 40 50 60

5 Other non-conformance

4 Near miss

3 Minor radiation incident

2 Non-reportable radiation incident

1 Reportable radiation incident

Number of providers


20

Figure 10. Breakdown of most frequently reported reportable radiation incidents (level 1) Dec 2015 to Nov 2017 (n = 88/159 subset of RTE)

The ‘reportable radiation incidents’ were then reviewed for this 2-year period, the 2016

analysis and aggregate data as seen in figure 11. There are similarities in percentage

with the previous 2-year reporting period and the aggregate data. However, there is an

increase across all the process subcodes shown in figure 10 for this reporting period.

There were only 52 out of 206 different subcodes associated with the level 1 RTE,

therefore the comparison graph should be interpreted with care.

Figure 11. Percentage of level 1 most frequently reported RTE for this 2-year reporting period compared with the 2016 analysis and aggregate

16

12

11

10

8

7

7

6

6

5

0 2 4 6 8 10 12 14 16

(10c) Localisation of intended volume

(13q) Setting of couch position/angle


(5k) Authorisation to irradiate (IR(ME)R)

(4i) Choice of other concurrent treatment


(4b) Verification of diagnosis/extent/stage

(10f) Production of images demonstrating correct detail

(13g) Patient positioning

(13c) Patient data ID process

Number of RTE reports

7.8%

2.3%

4.3%

2.3%

2.5%

7.3%

2.2%

4.7%

3.9%

3.4%

10.1%

7.5%

6.9%

6.3%

5.0%

0 2 4 6 8 10 12




(5k) Authorisation to irradiate

(4i) Choice of other current treatment or interventions


2018, data for2015-2017(n = 159)2016, data for2013-2015(n = 232)Aggregate, datafor 2009-20015(n = 398)


21

4.3.2 Breakdown of Level 2 (non-reportable radiation incident) RTE

A ‘non-reportable radiation incident’ is defined by TSRT as a radiation incident ‘not

reportable, but of potential or actual clinical significance’3.

There were 67 different subcodes used to report the Level 2 reports, and the most

frequently reported can be seen in figure 12. Of the level 2 reports, 14.9% (n = 33)

occurred during the ‘on-set imaging: approval process’. Examples of RTE reports

associated with ‘on-set imaging: approval process’ included the incorrect matching of

reference image and verification image and several reports indicated this was due to

matching the incorrect vertebrae level. This was followed by ‘localisation of intended

volume’ (7.7%, n = 17) which was also the most commonly reported subcode within the

level 1 reports. The 2017 Department of Health guidance16 contains guideline factors

for concomitant imaging; the reports classified as level 2 did not reach these guideline

factors. Examples of RTE associated with ‘localisation of intended volume’ include the

incorrect area of the patient CT scanned for planning and requires additional re-scan.

‘Production of images demonstrating correct detail’ made up 5.0% (n = 11) of all level 2

RTE. Examples of these types of RTE include when the image required for planning

does not contain previous permanent marks or patient scars.

Figure 12. Breakdown of most frequently occurring non-reportable radiation incidents (level 2) Dec 2015 to Nov 2017 (n = 125/ 222 subset of RTE)

The ‘non-reportable radiation incidents’ were then compared with the previous 2-year

analysis and the aggregate data as seen in figure 13. There has been a decrease in

the percentage of ‘on-set imaging: approval process’ and ‘movements from reference

marks’ associated RTE within this subcategory. An increase has been seen within the

level 2 RTE associated with ‘localisation of intended volume’, ‘production of imaging

33

17

11

10

10

7

6

6

5

5

5

5

5

0 5 10 15 20 25 30 35





(13cc) Management of variations/unexpected events/errors


(13k) ID of reference marks




(13r) Use of immobilisation devices





22

demonstrating correct detail’ and ‘management of variations’ within the treatment unit

process area.

Figure 13. Percentage of level 2 most frequently reported RTE for this 2-year reporting period compared with 2016 analysis and aggregate

4.3.3 Breakdown of Level 3 (minor radiation incident) RTE

A ‘minor radiation incident’ is defined by TSRT as a radiation incident ‘in the technical

sense but one of no potential or actual clinical significance’3.

There were 133 different subcodes used to report the Level 3 reports, and the most

frequently reported can be seen in figure 14. The most frequently occurring subcode

was ‘on-set imaging: production process’ making up 32.2% (n = 1,620) of all Level 3

RTE. This was followed by ‘on-set imaging: approval process’ (8.0%, n = 404) and ‘use

of on-set imaging’ (6.5%, n = 330). Examples of RTE associated with ‘on-set imaging:

approval process’ includes the mismatch of verification images causing slight

inaccuracies in treatment. Examples of RTE associated with ‘use of on-set imaging’

includes the verification image being taken when not required leading to additional

radiation dose. Of note, all but one subcode from the top 10 most frequently reported

level 3 RTE and 48.4% (n = 2,435) was associated with the treatment unit process.

A review of the ‘on-set imaging: production process’ subcode of level 3 RTE revealed

that 49.3% (n = 799) related to procedural failures, examples of which included

‘incorrect imaging parameters selected’, ‘wrong image acquisition image mode

selected’, ‘incorrect blade moved for image capture’, and ‘imager not extended or

appropriately positioned’. The remaining 50.7% (n = 821) related to equipment failure.

Examples of these types of reports included ‘failure of the image device during image

acquisition’, ‘image not captured after exposure’, flooded image’, or ‘image unavailable

offline’. This resulted in additional imaging exposure being undertaken. RT providers

13.7%

3.0%

2.7%

1.5%

8.8%

16.4%

2.3%

2.8%

1.7%

6.8%

14.9%

7.7%

5.0%

4.5%

4.5%

0 2 4 6 8 10 12 14 16 18




(13cc) Management of variations



2018, data for2015-2017(n = 222)2016, data for2013-2015(n = 177)Aggregate, data for2009-2015(n = 329)


23

are encouraged to audit and report these events locally so appropriate and timely

preventative measures might be implemented. In addition, the Medicines and

Healthcare products Regulatory Agency (MHRA)17 should be advised of all equipment

failures.

Figure 14. Breakdown of most frequently occurring minor radiation incidents (level 3) Dec 2015 to Nov 2017 (n = 3,602/ 5,035 subset of RTE)

The ‘minor radiation incidents’ were then reviewed for this 2-year period; the 2016

analysis and the aggregate data as seen in figure 15. There has been an increase in

percentage of ‘on-set imaging: production process and ‘patient positioning’ from the

previous 2-year report and the aggregate data. A decrease in percentage of RTE

associated with ‘use of on-set imaging’, ‘movements from reference marks’ and ‘on-set

imaging: approval process’ has been seen.


1620

404

330

246

232

200

173

150

131

116

0 200 400 600 800 1000 1200 1400 1600 1800







(13u) Use of compensators



(13r) Use of immobilisation devices


18.0%

8.4%

9.3%

6.4%

2.4%

23.1%

9.0%

12.4%

5.3%

2.8%

32.2%

8.0%

6.6%

4.9%

4.6%

0 5 10 15 20 25 30 35







2018, data for 2015-2017 (n = 5,035)

2016, data for 2013-2015 (n = 3,994)

Aggregate, data for 2009-2015 (n = 7,036)


24

4.3.4 Breakdown of Level 4 (near miss) RTE

A ‘near miss’ is defined in TSRT as a potential radiation incident ‘that was detected and

prevented before treatment delivery’3.

There were 166 different subcodes used to report the level 4 RTE, the most frequently

reported can be seen in figure 16. The most frequently occurring subcode was

‘accuracy of data entry’ making up 9.9% (n = 400) of all level 4 RTE. Examples of RTE

associated with ‘accuracy of data entry’ includes the incorrect entry of information into

the oncology management system and other transcription inaccuracies. This was

followed by ‘documentation of instructions/information’ (8.0%, n = 323) and ‘on-set

imaging: approval process’ (5.0%, n = 204). Examples of RTE associated with

‘documentation of instructions/information’ include the incorrect documentation of

instructions including patient set-up or immobilisation, which was identified at treatment

set-up before commencement of treatment. Examples of RTE associated with ‘on-set

imaging: approval process’ includes the incorrect off-line image match leading to

incorrect reference moves which are picked up before treatment commences.

Figure 16. Breakdown of most frequently occurring near misses (level 4) Dec 2015 to Nov 2017 (n = 2,009/ 4,058 subset of RTE)

The ‘near miss’ RTE were then reviewed for this 2-year period, the 2016 analysis and

the aggregate data as seen in figure 17. There has been an increase in the percentage

of RTE associated with ‘accuracy of data entry’ and ‘documentation of instruction’

within the level 4 RTE. There has been a decrease in the percentage of ‘on-set

imaging: approval process’ within this 2-year reporting period when compared to the

aggregate and previous 2-year period. The prevalence of level 4 RTE associated with

400

323

204

196

191

181

174

121

111

108

0 50 100 150 200 250 300 350 400








(11i) Target and organ at risk delineation


(5a) Completion of request for treatment (paper/electronic)



25

‘generation of plan for approval’ and ‘on-set imaging: recording process’ is shown to be

consistent with the 2016 analysis and the aggregate data.


4.3.5 Breakdown of Level 5 (other non-conformances) RTE

‘Other non-conformances’ is defined by TSRT as ‘non-compliance with some other

aspect of a documented procedure but not directly affecting RT delivery’3.

There were 182 different subcodes used to report the level 5 RTE, and the most

frequently reported can be seen in figure 18. The most frequently occurring subcode

was ‘management of process flow within planning’ making up 5.9% (n = 372) of all level

5 RTE. Examples of level 5 RTE associated with ‘management of process flow within

planning’ include plan being delayed and not ready for treatment time. This was

followed by ‘documentation of instructions/information’ (5.1%, n = 323) and bookings

made according to protocol’ (4.7%, n = 298). Of note, only 2 of the most frequently

occurring level 5 RTE subcodes were associated with the treatment unit process.

Examples of RTE associated with ‘documentation of instruction/ information’ includes

the incorrect documentation of patient set-up or immobilisation which was identified at

the data entry checking stage before treatment. Examples of RTE associated with

‘bookings made according to protocol’ includes booking patient treatment course onto

the incorrect machine, require movements of appointments at data entry checking

stage.

8.1%

6.5%

6.6%

4.3%

4.3%

7.2%

7.2%

6.6%

4.6%

4.1%

9.9%

8.0%

5.0%

4.8%

4.7%

0 1 2 3 4 5 6 7 8 9 10







2018, data for2015-2017 (n = 4,058)

2016, data for2013-2015 (n = 3,646)

Aggregate, data for2009-2015 (n = 6,858)


26

The ‘other non-conformances’ RTE were then reviewed for this 2-year period, 2016

analysis and the aggregate data as seen in figure 19. Non-conformances associated

with ‘management of process flow within planning’ and ‘documentation of instructions’

had a similar percentage within the aggregate data and the 2016 data, however they

have decreased and increased respectively within this 2-year reporting period.

Figure 18. Breakdown of most frequently occurring non-conformances (level 5) Dec 2015 to Nov 2017 (n = 2,465/ 6,356 subset of RTE)


372

323

298

266

219

206

194

154

145

144

144

0 50 100 150 200 250 300 350 400





(6d) Communication of appointments to patient


(6b) Bookings made according to request details

(5a) Completion of request for treatment (paper/electronic)





7.3%

3.5%

4.2%

3.3%

3.5%

7.4%

3.6%

5.0%

3.7%

4.1%

5.9%

5.1%

4.7%

4.2%

3.4%

0 1 2 3 4 5 6 7 8





(6d) Communication of appointments to patient


2018 data for2015-2017(n = 6,356)2016, data for2013-2015(n = 4,642)Aggregate, datafor 2009-2015(n = 9,041)


27

4.4 Safety Barriers

Critical control points, detection methods or defence in depth, are any process steps

whose primary function is to prevent errors occurring or propagating through the RT

workflow18.

The new safety barrier (SB) taxonomy was published in the DoL in December 2016 and

the application of these codes was actively encouraged from then onwards. As SBs

form part of the pathway coding, analysis of data before the publication of the was

possible. There are 86 safety barriers (SB) embedded across the 206 pathway codes4.

Each RTE report can contain multiple SB codes to identify all the points in the pathway

where the error was not detected. All subcodes were analysed across the 15,830 RTE

reports for the reporting period December 2015 to November 2017, and a total of 5,191

subcodes were identified as failed SB. ‘On-set imaging: approval process’ was the

most frequently reported failed SB (15.3%, n = 796) followed by ‘use of on-set imaging’

(12.4%, n = 646). Only 2.9% (n =150) of reported failed SB led to a level 1 or 2 RTE as

represented in figure 20.

The SB taxonomy can also be utilised to identify effective SB or methods of detection,

these will be included in analysis in future publications.

Owing to the complex and multi-faceted nature of RT it is usual to have checking

processes in place at points in the pathway where patient data is handed over to a new

part of the pathway. Therefore the ‘end of process check’ SB subcode is repeated

across the RT pathway. Failed SB associated with ‘end of process checks’ made up

8.7% (n = 451) of all reported failed SB for this reporting period. This is a vast decrease

from 20.4% (n = 2,584) since the previous 2-year reporting period5. The most

frequently reported ‘end of process check’ was at pre-treatment planning (33.5%, n =

151) whereas in the last 2-year reporting period5 the most frequently occurring ‘end of

process check’ subcode occurred at the treatment unit process (34.1%, n = 883).


28

Figure 20. Breakdown of failed primary safety barriers by classification level reported, Dec 2015 to Nov 2017 (n = 2,916/ 5,191 subset of RTE)

4.5 Causative factors

Use of causative factor taxonomy enables identification of system problems or root

causes that could precipitate a range of different incidents19.

The new causative factor (CF) taxonomy was published in the DoL in December 2016

and the first CF code was allocated to RTE reported in January 2017. Therefore, only

11 months of data is presented within this 2-year report. There were 7,778 RTE

reported between January 2017 and November 2017 of which 35.9% (n = 2,793)

included a CF. Figure 21 illustrates the most frequently reported primary CF which can

be seen as the root cause of an incident. A root cause can be defined as an identified

event that leads to anticipated operational occurrences or accident conditions20. The

most frequency occurring primary CF was ‘slips and lapses’ (32.1%, n = 897), followed

by ‘adherence to procedures/ protocols’ (18.7%, n = 521). ‘Slips and lapses’ was most

frequently attributed to ‘on-set imaging: production process’ (8.7%, n = 78) and

‘adherence to procedures/protocols’ was most frequently attributed to ‘accuracy of data

entry’ (8.8%, n = 46).

796

646

391

266

185

181

157

151

143

0 100 200 300 400 500 600 700 800





(11m) Recording of definitive treatment prescription

(4j) Consent process and documentation

(14c) On-treatment review of notes

(11t) End of process checks

(12g) End of process checks

Number of RTE

Level 1

Level 2

Level 3

Level 4

Level 5


29

Figure 21. Breakdown of most frequent primary causative factors by classification level, Jan to Nov 2017 (n = 2,686/ 2,793 subset of RTE)

Several CF codes can be attributed to each individual RTE. A review of the second to

fifth CF codes indicates the contributory factors associated with each RTE. A contributory

factor is defined as the latent weakness that allows or causes the observed cause of an

initiating event to happen, including the reasons for the latent weakness20. Contributory

factors were indicated across 607 reports; of these 92 contained multiple contributory

factors leading to 699 contributory factors in total. Figure 22 shows the most frequently

reported contributory factors. The most frequently occurring was ‘adherence to

procedures/protocols’ (34.8%, n = 243); these were most frequently attributed to ‘on-set

imaging: approval status’ (8.2%, n = 20) and ‘use of on-set imaging’ (7.8%, n = 19).

Figure 22. Breakdown of most frequently occurring contributory factors, Jan to Nov 2017 (n = 618/ 699 subset of RTE)

897

521

468

351

127

103

96

55

34

34

0 100 200 300 400 500 600 700 800 900

(CF 1c) Slips and lapses

(CF 2c) Adherence to procedures / protocols

(CF 1d) Communication

(CF 3a) Equipment or IT network failure

(CF 5d) Inadequate staffing

(CF 1a) Failure to recognise hazard

(CF 1b) Decision making process

(CF 7a) Other

(CF 2b) Inadequate procedures / protocols

(CF 2d) Process design


Level 1

Level 2

Level 3

Level 4

Level 5

243

115

78

70

35

25

18

17

17

0 50 100 150 200 250

(CF 2c) Adherence to procedures / protocols

(CF 1d) Communication

(CF 1c) Slips and lapses

(CF 2d) Process design

(CF 5d) Inadequate staffing

(CF 2b) Inadequate procedures / protocols

(CF 5e) Inadequate training

(CF 1a) Failure to recognise hazard

(CF 1b) Decision making process

Number of times contributory factor reported


30

4.6 Brachytherapy errors

Errors coded with brachytherapy process codes as the primary code account for 0.5%

(n = 80) of RT errors for the reporting period December 2015 to November 2017, which

is a slight increase since the previous reporting period (0.4%, n = 46)5. Brachytherapy

is a small specialised practice within RT, therefore the number of brachytherapy-

associated RTE would be expected to be low. Most of the brachytherapy RTE reported

were ‘near misses’ or ‘non-conformances’ (77.5%, n =62), none were reported as

reportable incidents and only 2.5% (n =2) were classified as a non-reportable incident

(Figure 23). ‘Initial positioning of applicators/ sources’ was the most frequently reported

brachytherapy associated RTE (22.5%, n = 18). An example of this type of RTE

includes the incorrect grid positioning of needles for prostate brachytherapy, or the

incorrect positioning of the intrauterine tube for gynaecological patients.

A total of 85 subcodes were identified across the 80 brachytherapy associated RTE

reports; only 10 different subcodes were identified as failed SB (12.9%, n = 11).

‘Correct applicators/sources’ was the most commonly reported failed SB within this

subset of data.

Figure 23. Breakdown of brachytherapy RTE coded 15 by classification level, Dec 2015 to Nov 2017 (n = 80)

Brachytherapy RTE were then reviewed for this 2-year period. The new data, the 2016

analysis and the aggregate data can be seen in figure 24. Both ‘initial position of

applicators/sources’ and ‘planning of treatment’ have a similar frequency within the

aggregate data and this 2-year reporting period; there has also been a slight decrease

in both since the previous 2-year reporting period.

1817

1110

44

3

3

22

2

2

11

0 2 4 6 8 10 12 14 16 18

(15g) Initial positioning of applicators / sources

(15h) Planning of treatment

(15e) Correct applicators /sources

(15n) Management of variations

(15i) Maintenance of position of applicators /sources

(15c) Source calibration

(15f) Correct theatre equipment

(15p) On-set imaging: production process

(15j) Removing of applicators / sources

(15m) Authorisation of plan

(15o) Use of on-set imaging

(15l) Validation of applicator/ source position

(15d) Sterility of sources

(15b )Delivery of sources


Level 1

Level 2

Level 3

Level 4

Level 5


31

Figure 24. Percentage of brachytherapy most frequently reported RTE for this 2-year reporting period compared with 2016 analysis and aggregate

4.7 Inspectorate data

There is a requirement that the appropriate authority is notified of all level 1

incidents13,14. The inspectorates for IR(ME)R for England, Wales, Northern Ireland and

Scotland were approached and asked to share their anonymised synopsis of closed

reportable radiation incidents from December 2015 to November 2017.

A total of 288 reports were shared; this is a decrease in the numbers since the last

report where 437 reports were shared for the same time period5. This decrease may be

due to the Department of Health guidance on what constitutes a reportable event,

published in January 201716. There were 159 level 1 reports received through the

voluntary scheme; this highlights a difference in the number of incidents reported to the

appropriate authorities and the voluntary reporting and learning scheme. The

classification and coding from the DoL document4 was applied to the inspectorate

incidents.

On review of the inspectorate data it became clear that there was wide variation in the

locally applied classification of events. It was found that 44.4% (n = 128) of the reported

events could have also been reported as level 2 (non-reportable radiation incidents) or

level 3 (minor radiation incidents). Examples of these include single repeat on-set

verification images which were reported to the inspectorate but could be classified as

level 3 (minor radiation incidents). The 2017 DH guidance16 sought to address this

disparity and contains guideline factors for concomitant imaging.

Of the 288 Level 1 inspectorate RTE reports shared, 45.1% (n = 130) occurred during

‘treatment unit process’. A comparison of inspectorate and voluntary data (Figure 25)

revealed a similar pattern of spread in the higher frequency reports. The inspectorate

23.6%

19.8%

7.5%

0.9%

10.4%

1.9%

26.5%

24.5%

10.2%

2.0%

10.2%

2.0%

22.5%

21.3%

13.8%

12.5%

5.0%

5.0%

0 5 10 15 20 25 30

(15g) Initial positioning of applicators / sources

(15h) Planning of treatment

(15e) Correct applicators /sources

(15n) Management of variations/unexpected events/errors

(15i) Maintenance of position of applicators /sources

(15c) Source calibration


2018, data for2015-2017 (n = 80)2016, data for2013-2015 (n = 49)Aggregate, data for2009-2015 (n = 106)


32

data indicated that 7.3% (n = 21) of incidents shared were associated with ‘referral for

treatment’, however this was higher in the voluntary data (11.3%, n = 18). This trend is

similar with reportable RTE associated with ‘communication of intent’ where the

inspectorate data indicates 6.9% (n =20) and the voluntary data is 13.8% (n =22). The

inspectorate data contains 2.1% (n = 6) brachytherapy associated RTE, however the

voluntary data did not contain any reportable radiation incidents associated with

brachytherapy. The text descriptors for the inspectorate data were more detailed than

those of the voluntary data, which may explain differences in coding.

Figure 25. Percentage frequency of process codes found in reportable radiation incidents from the inspectorate and voluntary datasets, Dec 2015 to Nov 2017

The inspectorate data was also broken down by most frequently occurring process

subcode. This revealed pre-treatment ‘localisation of intended volume’ (8.3%, n = 24)

and treatment ‘on-set imaging: production process’ (8.0%, n = 23) were the most

frequently occurring process subcodes within the inspectorate data (Figure 26). There

are similarities in subcode frequency of the inspectorate and voluntary data, however

there is a significant difference in the treatment ‘on-set imaging: production process’

associated reportable radiation incidents. This difference may be due to the guidance

issued in 201716 and a large proportion of these types of events being classified as

level 3, see figure 7. The most frequently reported subcodes for this reporting period

are similar to the previous reporting period (December 2013 to November 2015) where

all the subcodes featured in the previous most frequently reported inspectorate data

apart from ‘verification of diagnosis/extent/stage’ and ‘patient data ID process’5.

Due to the detail contained within the inspectorate data CF could be applied to all

reports. CF for the inspectorate data was analysed and indicated that slips and lapses

were most frequently reported as the primary CF (46.2%, n = 133). This is consistent

with the voluntary data where the most frequent level 1 CF was slips and lapses

(36.8%, n = 7).

45.1%

21.9%

9.4%

7.3%

6.9%

2.4%

2.1%

2.1%

2.7%

40.3%

20.1%

7.5%

11.3%

13.8%

1.9%

0.0%

0.6%

4.6%

0 5 10 15 20 25 30 35 40 45 50



Pretreatment planning process

Referral for treatment

Communication of intent


Brachytherapy

Booking process

Miscellaneous


Voluntary data (n = 159)

Inspectorate data (n = 288)


33

Figure 26. Breakdown of inspectorate and voluntary data as a percentage of most frequently occurring process subcodes, from Dec 2015 to Nov 2017

8.3%

8.0%

6.3%

5.6%

5.2%

4.9%

3.8%

3.8

3.1%

10.1%

0.6%

6.9%

7.5%

3.8%

4.4%

3.1%

4.4%

2.5%

0 2 4 6 8 10 12






(4b) Verification of diagnosis/extent/stage



(5a) Completion of request for treatment

Percentage of Level 1 RTE reports

Voluntary data (n = 159)

Inspectorate data (n = 288)


34

5. Discussion

The analysis within this report facilitates comparisons of national RTE reporting with

both local and network trends. Aggregate data and previous reports have been used for

comparison to this 2-year period, this representation of the data allows local providers

to compare their RTE over time. As local analysis and learning from RTE is developed

consideration should be given to the appropriateness of corrective actions and

preventative measures.

5.1 Increase in RTE reporting

During this 2-year reporting period, there has been an increase in the numbers of RTE

reported to the national voluntary reporting system. As seen in section 3.2 a total of

15,979 RTE reports were submitted to the voluntary reporting scheme between

December 2015 and November 2017, compared with the last 2-year reporting period,

where a total of 12,800 RTE reports were submitted, 7,655 for the 2014 report and

3,316 for the 2012 report5. This is an increase in reporting levels of 24.8% since the

2016 report and 381.8% increase since the 2012 report5.

The increase in reported RTE reflects a maturing reporting culture. Further increases in

the number of RTE reports submitted to the national voluntary reporting system are

expected as service provision increases and as providers develop full electronic

reporting solutions and reporting mechanisms are further streamlined.

The number of providers contributing to this initiative was 59 (96.7%), this is a decrease

from the previous reporting period where 100% (n = 60) NHS providers reported to the

national analysis5. This decrease in number of providers contributing to the national

scheme may be due to resources and changes in local procedures. Providers are

encouraged to contact PHE if they have any queries about reporting or would like

further advice.

The Euratom/European Union Basic Safety Standards Directive 2013 (BSSD) sets out

standards for radiation protection21; additional requirements included recording of

analyses of incidents involving or potentially involving accidental and unintended

exposures. This was transposed into UK legislation in February 201813. Additionally,

recommendations of the Francis report22 into failings at the Mid-Staffordshire NHS

Foundation Trust included a requirement for openness, transparency and candour

throughout the NHS to support a culture of protecting patients and removing poor

practice. The ethos of these requirements has mainly been adopted across UK RT

providers reflecting a mature reporting culture.


35

There is still some disparity in the number of reports received per provider, reporting

ranged from a minimum of 1 incident to 2,160 incidents reported per provider over the

2-year period. This variation is also reflected in the lag time for reporting, as seen in

section 3.3, which ranged from 0 days to a maximum of 871 days. All providers are

encouraged to report monthly to ensure timeliness of learning.

Data received from the National Cancer Registration and Analysis Service (NCRAS)23

revealed an estimated 335,354 RT prescriptions were delivered across 3,668,047

attendances in England during the period from December 2015 until November 2017.

This data was extrapolated for the UK population to an estimated 396,876 prescriptions

across 4,340,962 attendances in the UK for the same period. This is a slight decrease

from the estimated 402,998 prescriptions (delivered across 4,518,571 attendances)

within the previous 2-year period. This reflects the adoption of hypofractionation of

breast, prostate and bony metastases treatments and the introduction of stereotactic

regimes.

To establish a reported error rate, it was accepted the clear majority of RTEs reported

affected a single attendance as part of a prescription. With this caveat an estimated

reported error rate per prescription of 4.0% was calculated for the current reporting

period. This compares with an estimated reported error rate of 3.2% for the previous 2-

year reporting period, marking a slight increase in the reported error rate. However,

using the same premise an estimated reported error rate of level 1 events to the

inspectorates for 0.07% can be calculated, for the current reporting period marking a

decrease from 0.11% for the previous 2-year reporting period 5. This suggests that

whilst the reporting culture continues to grow a slight decrease in level 1 events is

being seen.

This decrease in percentage of reported level 1 RTEs in relation to prescriptions should

be interpreted carefully. Also, it should be noted that the majority of reported events

(65.8%) did not impact on the patient’s treatment and opportunities to correct for

detected RTE was possible in an additional 33.2%.

5.2 Main themes

Analysis of the RTE reports demonstrates that generation of error is not confined to

one professional group or to any particular point in the pathway. It also indicates that

the pattern of errors is replicated across service providers in the UK.

5.2.1 Breakdown of process codes

Consistent with previous reports5 RTEs were spread across all 21 categories of

process codes (as seen in section 4.1.1), treatment unit process codes were the most

frequently reported RTE (41.2%, n = 6,526). RTEs attributed to treatment unit


36

processes are consistent with the previous 2-year report (41.3%, n = 5,254). This is to

be expected as treatment processes are reliant on the accurate interpretation of the

planning information and RT treatments usually span a number of visits, providing

many opportunities during the treatment unit process for RTE to occur.

5.2.2 Breakdown of process subcodes

The areas in the RT pathway where RTEs frequently occur can be seen in table 1.

Guidance has already been published on how to minimise all these frequently occurring

RTEs in the periodic newsletter Safer Radiotherapy through the ‘error of the month’

column as some of these trends were noted in the regular analysis6.

Frequently occurring process subcodes Safer Radiotherapy guidance6

On-set imaging: production process (13z) Issue 7, 16, 25

Accuracy of data entry (12f) Issue 2

On-set imaging: approval process (13aa) Issue 3, 7

Documentation of instructions/information (10j) Issue 8

Use of on-set imaging (13i) Issue 7

On-set imaging: recording process (13bb) Issue 7

Generation of plan for approval (11j) Issue 15

Management of process flow within planning (11o) Issue 5

Recording of patient specific instructions (11n) Issue 10

Bookings made according to protocol (6a) Issue 17

Movements from reference marks (13l) Issue 1, 22

Table 1. Frequently occurring process subcodes and associated newsletter guidance

All the main themes within this reporting period were also the most frequently occurring

process subcodes within the previous 2-year report5, as seen in section 4.1.2. The

percentage of RTE associated with ‘on-set imaging: production process has increased

from 6.5% within the aggregate data to 8.4% in the last 2-year period and 11.5 % in this

reporting period. This notable increase may be associated with the increase in the use

of on-set verification images due to its further availability and the recognition of its

preventative role in minimising geographical displacement during treatment24.

5.3 Classification level of RTE

The clear majority of reports were of lower level events, thus not affecting the outcome

of patient care. Of the level 1 and 2 incidents reported, it is known most them affected

only one fraction of a course of treatment. This meant that corrective action could be

taken over the remaining treatment fractions, so the incident did not have a significant


37

impact on the patient or the outcome of their treatment. This was like the trends seen in

each of the previous 2-year reports.

A small number of higher level incidents and a much greater number of lower level

incidents are consistent with findings in the literature. It is known that for every level 1

‘reportable radiation incident’ that occurs, many lower level incidents are also seen.

Heinrich illustrated this point in 19313 (Figure 27), it may be seen that as the severity of

an incident decreases, the probability of its occurrence increases.

Figure 27: Heinrich’s triangle (reproduced from Towards Safer Radiotherapy3)

As seen in section 4.2.1 the ‘non-reportable radiation incident ‘(level 2) and ‘minor

radiation incident’ (level 3) have stayed consistent across the 2018, 2016 and

aggregate analyses, as seen in section 4.2.1, however the level 1 ‘reportable radiation

incidents’ have reduced over time. A comparison of the 2018/2016 and aggregate data

with Heinrich’s illustration can be seen in figure 28. This indicates that there are a

smaller number of higher level incidents and a larger number of lower level incidents.

However, the ratios of the RTE data are different to the ratio of Heinrich’s triangle.


38

Figure 28: Comparison of Heinrich’s triangle and 2018, 2016 and aggregate data.

When compared with the results from the previous 2-year report there has been a

change in the percentage of ‘reportable radiation incidents’ from 1.8% to 1.0%. This

reduction may be due to further guidance on reporting much greater than intended

exposures published in January 201716. As seen in section 4.2.1 all providers do not

report all classification of incidents, this may be due to resources or local protocol. With

this in mind the analysis should be interpreted carefully.

Experience has shown that as an organisation’s reporting culture matures, staff

become more likely to report incidents. High reporters aim to learn from incident

reporting to make patient care safer. With this in mind an increase in incident reporting

should not necessarily be taken as an indication of worsening patient safety. It may

indicate an increasing level of awareness of safety issues among healthcare

professionals and a more open and transparent culture across the organisation.

5.3.1 Level 1 (reportable radiation incident) RTE

There were 52 different subcodes associated with the 159 level 1 RTE. With the move

towards hypofractionated treatments there is less opportunity to correct, therefore there

may be opportunity for an increase in level 1 RTE in future analysis. As seen in section

4.3.1 the most frequently reported Level 1 RTE occurred during ‘localisation of intended

0

10

20

30

40

50

60

70

80

90

100

Heinrich (n = 330) 2018 (n = 15,830) 2016 (n = 12,691) Aggregate (n =23,662)

90.9%

65.8% 65.3% 67.2%

8.8%

33.2% 32.9% 31.1%

0.3% 1.0% 1.8% 1.7%

Inci

dent

per

cent

age

Level 1

Level 2+3

Level 4+5


39

volume’ (10.1%, n = 16). The comparison with previous data could be difficult to

interpret due to only 52 out of 206 different subcodes associated with the level 1 RTE.

From these 52 different subcodes only 36 were present in the previous 2-year analysis5

which included 55 different subcodes.

5.3.2 Level 2 (non-reportable radiation incident) RTE

There were 62 different subcodes associated with the 222 level 2 RTE, as seen in

section 4.3.2. ‘On-set imaging: approval process’ was the most frequently occurring

level 2 RTE (14.9%, n = 33). The frequency of this subcode within the level 2 RTE has

decreased since the 2016 analysis5 (16.4%, n = 29) but is an increase from the

aggregate data (13.7%, n = 45). There were similarities in the percentage of Level 2

RTE within the aggregate data and the 2016 analysis5 associated with ‘localisation of

intended volume’, ‘production of images demonstrating correct detail’ and

‘management of variations’. However, there has been an increase in percentage seen

within this reporting period.

5.3.3 Level 3 (minor radiation incident) RTE

There were 133 different subcodes associated with the 5,035 level 3 RTE, as seen in

section 4.3.3. The most frequently occurring subcode was ‘on-set imaging: production

process’ (32.2%, n = 1,620), this is a vast increase from the aggregate data (18.0%, n

= 1,265) and an increase since the 2016 analysis5 (23.1%, n = 922). A review of the

‘on-set imaging: production process’ subcode of level 3 RTE revealed that 49.3% (n =

799) related to procedural failures and the remaining 50.7% (n = 821) were due to

equipment malfunction, this is very similar to the 2016 analysis4.

Of note, when compared to the 2016 report4, all the most frequently occurring

subcodes within this classification were the same, apart from ‘use of immobilisation

devices’ which replaces ‘generation of plan for approval’.

It should be noted that equipment failure reports should be reported to local engineers,

the manufacturers and the MHRA, as appropriate. Equipment failure adverse incidents

above the threshold levels set out in PM7725 (until further guidance is available) should

be reported to the appropriate enforcing authorities as specified within IR(ME)R13. It

would seem the errors reported in this subset are spread across manufacturers and

varied in type. Consideration should be given to undertaking a risk assessment in the

case of a frequently recurring fault, especially where no resultant image acquisition is

achievable to inform the treatment process as to whether imaging on that device should

continue.


40

5.3.4 Level 4 (near miss) RTE


section 4.3.4. The spread of reported level 4 RTEs across the process codes was

consistent with the 2016 analysis5. The most frequently occurring level 4 subcode was

‘accuracy of data entry’ (9.9%, n = 400). This has increased since the 2016 analysis5.

Electronic systems are in use across the majority of RT providers; however, it is seen

from the frequency of level 4 RTE that information is still being incorrectly input into the

oncology management system. Further guidance on reducing this type of error can be

seen in Safer RT6.

5.3.5 Level 5 (other non-conformances) RTE


section 4.3.5. The most frequently reported level 5 subcode was ‘management of

process flow within planning’ (5.9%, n = 372), this was also the most frequently

occurring level 5 subcode in the previous analysis5, however the percentage has

decreased from 7.4% (n = 344).

5.4 Safety barriers

Safety barriers (SB) are embedded across the RT pathway coding. The DoL document4

contains an update to the pathway taxonomy and indicated SB. Several original

pathway codes were identified as SB. There are 206 different subcodes, 86 of which

are identified as SB. The SB taxonomy can be used to identify both failed and effective

SB. As seen in section 4.4 for this reporting period 5,191 subcodes were identified as

failed SB, and ‘on-set imaging approval process’ was the most frequently occurring

failed SB (15.3%, n = 796). Future reports will contain details of effective SB or

methods of detection.

‘End of process checks’ are one form of safety barrier in error detection and, as such

should include safety-critical elements of the pathway. The frequency and effectiveness

of checking and verifying procedures should be audited to ensure they are of value.

Failed SB associated with ‘end of process’ checks comprised 8.7% (n = 451) of all

reported failed SB for this reporting period. This is a vast decrease from the 2016

analysis4 (20.4%, n = 2,584). This reduction may be due to the ‘pause and check’

posters26. Interestingly, the most frequently occurring end of process check was at pre-

treatment planning (33.5%, n = 151) whereas in the last 2-year reporting period the

most frequently occurring end of process check subcode occurred at the treatment unit

process (34.1%, n = 833). Providers should consider reviewing their minimum criteria

for checking across the entire pathway.


41

5.5 Causative factors

The causative factor (CF) taxonomy was published in December 2016; the first CF was

allocated to a RTE reported in January 2017. There were 7,778 RTE reported between

January 2017 and November 2017 of which 35.9% (n = 2,793) contained CF, as seen

in section 4.5. The primary CF is the root cause of an incident and for this reporting

period the most frequently reported CF was ‘slips and lapses’ (32.1%, n = 897). The

contributory factors are the secondary CF indicated in a report. There were 699

contributory factors within this analysis and the most frequent was ‘adherence to

procedures/protocols’ (34.8%, n = 243).

The benefit of the use of causative factor taxonomy is that it enables identification of

system problems or root causes that could precipitate a range of different incidents. If

the root causes are addressed, it can be expected that overall system safety is

enhanced and not just a weakness associated with a particular incident19. For example,

if the most frequently occurring equipment malfunction issues associated with CBCT

failures are addressed, the number of level 3 RTE associated with ‘on-set imaging:

production process’, 821 level 3 RTE, would be significantly reduced.

5.6 Inspectorate data

A total of 288 reports were shared by the inspectorates compared to 159 received

through the voluntary reporting system. The incongruity between the numbers of level 1

reports received by the inspectorates and the voluntary reporting system may be

explained by several factors. Firstly, only anonymised and closed events were shared

by the inspectorates, which may create different time lags. Secondly, the voluntary data

does not contain RTE from the independent sector. Finally, as seen in section 4.7 there

is some variation on the locally applied classification of events, and 44.4% (n = 128) of

the reportable radiation incidents reviewed within the inspectorate data could have

been reported as other classification of incidents within the voluntary data. Like the

level 1 ‘reportable radiation incident’ reports received through the voluntary reporting

scheme, ‘localisation of intended volume’ emerged as the most common process

subcode.


42

6. Conclusion

Although RTEs are rare, when they do occur the consequences can be significant, so it

is essential the RT community does not become complacent about the associated

risks, which may be even greater as new approaches and technologies are utilised.

The overall increase in voluntary reporting indicates a maturing reporting culture. It also

demonstrates the RT community’s commitment to maximise learning from these events

and to minimise the frequency of their occurrence in the future. A further increase in the

number of RTE reports submitted to the national voluntary reporting scheme is

expected, as more clinical providers are encouraged to submit all levels of RTEs

monthly.

The use of all taxonomies as described within the DoL document4 allows wide-ranging

learning from RTE at both a local and national level. This includes the classification,

pathway coding and causative factor taxonomy. The pathway coding can be used to

indicate where in the RT pathway the error first occurred, failed safety barriers and

effective safety barriers shown as method of detection. The implementation of the

national voluntary reporting system described within this report allows clinical providers

to compare their local analysis within networks and with the national picture. Further

improvements can be attained when using the learning from RTE to incorporate

corrective and preventative measures into the RT process.

After 10 years of reporting, the proportion of significant incidents has started to reduce

slightly. RT is delivered in a complex and rapidly evolving landscape. Therefore, it is

imperative that RTE trends continue to be reported, analysed and monitored on a

cyclical basis, to inform ongoing safe and effective RT practice. This is especially

pertinent as new techniques and technologies are implemented and as new clinical RT

providers are established. This work supports a risk-based approach to improving

safety both locally, regionally and nationally and indicates a culture that is open,

transparent and already present in the UK RT community.


43

7. Recommendations

Local provider recommendations are:

• all UK Radiotherapy (RT) providers should continue to use Towards Safer RT3 and

Development of Learning2 to classify and code all levels of RTEs for local analysis

to inform local learning, policy and practice.

• local learning should be compared with the national picture and used to inform local

and network level learning

• UK NHS RT providers should continue to submit coded and classified RTE reports

to the national voluntary reporting system using the mechanisms identified within

this report

• UK NHS RT providers should ensure all levels of RTE are reported locally and

nationally

• providers are encouraged to report monthly to ensure timeliness of learning

• UK RT providers should adopt all codes including safety barriers and causative

factors when coding RTE and the use of secondary or additional subcodes

• independent providers should consider submitting all levels of RTE to the national

voluntary reporting system

• UK RT providers should review effectiveness of local end of process checks across

the entire pathway

National recommendations are:

• PSRT should continue to develop analysis of the reports, with regular dissemination

of findings to the RT community for group learning

• the data should be used both by the PSRT and by individual RT providers as part of

a risk-based approach to allocating resources for improving patient safety in RT and

to inform audit and research

• the mechanism to enable providers in Scotland and Northern Ireland to submit RTE

reports to PHE for collation and analysis as part of a UK dataset should be further

refined

• PSRT should review the appropriateness of corrective actions and preventative

measures

• vendors should consider developments to reduce the rate of RTE related to

equipment failure


44

8. Acknowledgements and steering group

We would like to acknowledge:

• NHS Radiotherapy Providers across the UK

• National Reporting and Learning System (NRLS) at NHS England and Improvement

• Inspectorates for IR(ME)R

• National Cancer Registration and Analysis Service (NCRAS)

PSRT Steering Group membership

The membership of this group is reviewed on an annual basis. The members contributing to this report were:

• Julia Abernethy (Patient Safety Team, NHS England and Improvement) (Member

March 2019 onwards)

• Helen Best (Public Health England)

• Martin Duxbury (Society and College of Radiographer’s Clinical Representative –

Deputy Head of Radiotherapy, St James Institute of Oncology, Leeds)

• Úna Findlay (Public Health England and Group Chair)

• Marianne Illsley (Royal College of Radiologists – Consultant Clinical Oncologist and

Medical Director, St Luke’s Cancer Centre, Royal Surrey County Hospital

Foundation Trust) (Member April 2018 onwards)

• Maria Murray (Society and College of Radiographers – Professional Officer for

Scotland and UK Radiation Protection Lead)

• Tony Murphy (Lay Representative)

• Tom Roques (Royal College of Radiologists – Consultant Clinical Oncologist and

Clinical Director for Oncology and Haematology, Norfolk and Norwich University

Hospital NHS Foundation Trust) (Member 2010-April 2018)

• Carl Rowbottom (Institute of Physics and Engineering in Medicine – Head of

Physics, The Clatterbridge Cancer Centre NHS Foundation Trust)


45

10. References

1. World Health Organization. Reporting and learning for patient safety. Available at www.who.int/patientsafety/implementation/reporting_and_learning/en/

2. Radiotherapy: Hidden Dangers. Chapter 5. Chief Medical Officer’s Annual Report 2006 (2007). Available at webarchive.nationalarchives.gov.uk/20130107105354/http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/AnnualReports/DH_076817

3. RCR, SCoR, IPEM, NPSA, BIR. Towards Safer Radiotherapy. Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine, National Patient Safety Agency, British Institute of Radiology. Royal College of Radiologists, London (2008). Available at www.rcr.ac.uk/towards-safer-radiotherapy

4. PSRT. Patient Safety in Radiotherapy Steering Group Activity, Development of learning from radiotherapy errors. Available at www.gov.uk/government/publications/development-of-learning-from-radiotherapy-errors

5. PSRT. Patient Safety in Radiotherapy Steering Group Activity, Radiotherapy errors and near misses: biennial report. Nos 1-4. Available at www.gov.uk/government/publications/radiotherapy-errors-and-near-misses-data-report

6. PHE. Safer Radiotherapy: Radiotherapy Newsletter of PHE and Supplementary Data Analysis: Full Quarterly Radiotherapy Error Data Analysis. Nos 1–27. Available at www.gov.uk/government/publications/safer-radiotherapy-error-data-analysis-report

7. NHS Improvement, Report a patient safety incident. Available at https://improvement.nhs.uk/resources/report-patient-safety-incident/

8. NRLS reporting tool, Available at https://report.nrls.nhs.uk/nrlsreporting/ 9. NHS Improvement, National patient safety incident reports. Available at

https://improvement.nhs.uk/resources/national-patient-safety-incident-reports-27-march-2019/ 10. NHS Improvement, The future of the patient safety incident reporting: upgrading the NRLS.

Available at https://improvement.nhs.uk/news-alerts/development-patient-safety-incident-management-system-dpsims/

11. NPSA. Implementing Towards Safer Radiotherapy: guidance on reporting radiotherapy errors and near misses effectively. Available at www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/ https://webarchive.nationalarchives.gov.uk/20110318152355/http://www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/?entryid45=75033&p=1

12. PHE. Safer Radiotherapy: Radiotherapy Newsletter of PHE Issue no 27. Available at www.gov.uk/government/publications/safer-radiotherapy-error-data-analysis-report

13. The Ionising Radiation (Medical Exposure) Regulations (2000). The Stationery Office, London. SI 2000/1059. Available at www.opsi.gov.uk/si/si2000/20001059.htm. The Ionising Radiation (Medical Exposure) Regulations (Northern Ireland) (2000) The Stationery office, London NISR 2000/194. Available at www.legislation.gov.uk/nisr/2000/194/contents/made Updated - The Ionising Radiation (Medical Exposure) Regulations (2017). The Stationery Office, London, SI 2017/1322. Available at www.legislation.gov.uk/uksi/2017/1322/pdfs/uksi_20171322_en.pdf The Ionising Radiation (Medical Exposure) (Amendment) Regulations (2018). The Stationery Office, London, SI 2018/121. Available at www.legislation.gov.uk/uksi/2018/121/contents/made The Ionising Radiation (Medical Exposure) Regulations (Northern Ireland) (2018). The Stationery Office, London, SI 2018/17. Available at www.legislation.gov.uk/nisr/2018/17/contents/made

14. The Ionising Radiations Regulations (1999). The Stationery Office, London. SI 1999/3232. Available at www.opsi.gov.uk/si/si1999/19993232.htm Updated - The Ionising Radiations Regulations (2017). The Stationery Office, London SI 2017/1075. Available at www.legislation.gov.uk/uksi/2017/1075/contents/made

http://www.rcr.ac.uk/towards-safer-radiotherapy

http://www.gov.uk/government/publications/development-of-learning-from-radiotherapy-errors

http://www.gov.uk/government/publications/radiotherapy-errors-and-near-misses-data-report

http://www.gov.uk/government/publications/safer-radiotherapy-error-data-analysis-report

https://improvement.nhs.uk/resources/report-patient-safety-incident/

https://report.nrls.nhs.uk/nrlsreporting/

https://improvement.nhs.uk/resources/national-patient-safety-incident-reports-27-march-2019/

https://improvement.nhs.uk/news-alerts/development-patient-safety-incident-management-system-dpsims/

https://improvement.nhs.uk/news-alerts/development-patient-safety-incident-management-system-dpsims/

http://www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/

http://www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/

https://webarchive.nationalarchives.gov.uk/20110318152355/http:/www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/?entryid45=75033&p=1

https://webarchive.nationalarchives.gov.uk/20110318152355/http:/www.nrls.npsa.nhs.uk/resources/clinical-specialty/radiology-and-radiotherapy/?entryid45=75033&p=1

http://www.gov.uk/government/publications/safer-radiotherapy-error-data-analysis-report

http://www.legislation.gov.uk/nisr/2000/194/contents/made

http://www.legislation.gov.uk/uksi/2017/1322/pdfs/uksi_20171322_en.pdf

http://www.legislation.gov.uk/uksi/2018/121/contents/made

http://www.legislation.gov.uk/nisr/2018/17/contents/made

http://www.opsi.gov.uk/si/si1999/19993232.htm

http://www.legislation.gov.uk/uksi/2017/1075/contents/made


46

15. PSRT. Safer radiotherapy: supplementary survey analysis. Report 3 (2014). Available at www.gov.uk/government/publications/safer-radiotherapy-supplementary-survey-analysis

16. DH, IRMER guidance on much greater than intended- updated January 2017. Available at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/583637/MGTI_guidance_Jan_17.pdf

17. Medicines and healthcare products regulatory agency. Available at https://www.gov.uk/government/organisations/medicines-and-healthcare-products-regulatory-agency

18. Ford E. et al. Consensus recommendations for incident learning database structures in radiation oncology. Medical Physics. 2012; December 39 (12); 7272-7290.

19. Clark B et al. The management of radiation treatment error through incident learning. Radiotherapy and Oncology. 2010; Vol.95, pp344-349.

20. Boadu M & Mohan Rehani M. Unintended exposure in radiotherapy: Identification of prominent causes. Radiotherapy and Oncology. 2009; Vol. 93, pp609-617

21. Basic Safety Standards Directive, (2013/59/Euratom) available at https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32013L0059

22. Francis R. Report of the Mid Staffordshire NHS Foundation Trust public enquiry. HMSO, London (2013). Available at www.gov.uk/government/publications/report-of-the-mid-staffordshire-nhs-foundation-trust-public-inquiry

23. National Cancer Registration and Analysis Service (NCRAS), available at https://www.gov.uk/guidance/national-cancer-registration-and-analysis-service-ncras

24. National Radiotherapy Implementation Group. Image Guided Radiotherapy. Guidance for implementation and use. National Cancer Action Team (2012). Available at www.sor.org/sites/default/files/document-versions/National%20Radiotherapy%20Implementation%20Group%20Report%20IGRT%20Final.pdf

25. HSE. Equipment used in connection with medical exposure. Guidance Note PM77 (3rd ed) (2006). Available at www.hse.gov.uk/pubns/guidance/pm77.pdf

26. SCoR, Have you paused and checked? Radiotherapy, Available at www.sor.org/learning/document-library/have-you-paused-and-checked-radiotherapy

http://www.gov.uk/government/publications/safer-radiotherapy-supplementary-survey-analysis

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/583637/MGTI_guidance_Jan_17.pdf

https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/583637/MGTI_guidance_Jan_17.pdf

https://www.gov.uk/government/organisations/medicines-and-healthcare-products-regulatory-agency

https://www.gov.uk/government/organisations/medicines-and-healthcare-products-regulatory-agency

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32013L0059

https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32013L0059

http://www.gov.uk/government/publications/report-of-the-mid-staffordshire-nhs-foundation-trust-public-inquiry

http://www.gov.uk/government/publications/report-of-the-mid-staffordshire-nhs-foundation-trust-public-inquiry

https://www.gov.uk/guidance/national-cancer-registration-and-analysis-service-ncras

http://www.sor.org/sites/default/files/document-versions/National%20Radiotherapy%20Implementation%20Group%20Report%20IGRT%20Final.pdf



http://www.hse.gov.uk/pubns/guidance/pm77.pdf

http://www.sor.org/learning/document-library/have-you-paused-and-checked-radiotherapy


47

Appendix A: Radiotherapy error

classification grid

Reproduced from Towards Safer Radiotherapy3

No

No

Potentially or

actually clinically

significant?

Radiotherapy error

Yes No

Yes

No Yes

Yes

Correctable

radiation

incident?

Level 1

Reportable radiation

incident

Level 2

Non-reportable

radiation incident

Level 3

Minor radiation

incident

Level 4

Near miss

Level 5

Other non-

conformance

Potential for

radiation

incident?

Radiation

incident?

Reportable?

Yes

No


48

Appendix B: Refined radiotherapy pathway coding (including

safety barrier taxonomy)

Reproduced from Development of Learning guidance4.

Text in red denotes additions to pathway coding in terms of descriptors and new codes.

SB denotes pathway codes that are also safety barriers.


49

Safety

barrier

Process

code Activity code

0 Infrastructure

SB 0a Implementation of national and international codes of practice for radiation dosimetry

0b Development of dosimetry algorithms for local application (includes locally developed software/ programs/ tools for clinical use)

0c Development of treatment planning algorithms for local application

0d Other

0e IT infrastructure (includes change in hardware/ software/ upgrades/ network changes/ archive process/ system compatibility/ data transfer)

Equipment-specific activities

1 Room design

SB 1a Patient safety (includes alteration of room design or use)

SB 1b Staff and public safety (includes alteration of room design or use)

SB 1c Environmental controls

SB 1d Access control

1e Other

2 New equipment

2a Installation

SB 2b Manufacturer’s tests


50

SB 2c Acceptance tests

SB 2d Critical examination under IRR99

2e Customisation and configuration of equipment

SB 2f Commissioning

2g Data recording

2h Preparation of data files for planning systems (to include treatment planning systems, virtual simulation, independent dosimetry checking

software etc.)

2i Other

3 Routine machine QA

SB 3a Daily consistency checks – geometric parameters (including CT, Linac, gated equipment, in vivo dosimetry devices etc. and discrepancies

between reporting and action level)

SB 3b Daily consistency checks – dosimetric calibration (including CT, Linac, gated equipment, in vivo dosimetry devices etc. and discrepancies

between reporting and action level)

SB 3c Daily consistency checks – safety (IRR compliance)

SB 3d Daily verification of accuracy of data transfer between TPS, R&V system and treatment equipment

SB 3e Planned QA programme checks – geometric parameters (including CT, Linac, gated equipment, in vivo dosimetry devices etc. and

discrepancies between reporting and action level)

SB 3f Planned QA programme checks – dosimetric calibration (including CT, Linac, gated equipment, in vivo dosimetry devices etc. and

discrepancies between reporting and action level)

SB 3g Planned QA programme checks – safety (IRR compliance)


51

SB 3h Planned QA programme checks – image quality parameters (including CT, MR, portal, cone-beam, film processor)

SB 3i Regular preventative maintenance and repair programme

3j Handover of radiotherapy equipment after planned QA and maintenance (including handover to other department such as diagnostic

colleagues)

SB 3k Routine radiation safety checks

3l Other

Patient-specific activities

4 Referral for treatment

SB 4a Identification of patient (verification against primary source data)

SB 4b Verification of diagnosis/extent/stage (including laterality)

4c Choice of dose

4d Choice of modality

4e Choice of energy

4f Choice of fractionation

4g Choice of start date

SB 4h Consideration of patient condition/co-morbidities (including ICED or pacemaker status, prosthesis, patient unsuitable for IV contrast and

changing performance status)

SB 4i Choice of other concurrent treatment or interventions and their sequencing or timing (including patient selection criteria not met)


52

SB 4j Consent process and documentation

4k Other (previously 4i)

5 Communication of intent

5a Completion of request for treatment (paper/electronic) (including incomplete requests or insufficient data and failure to handover referral)

5b Recording of patient ID

5c Completion of required demographics

5d Completion of tumour-specific information (including laterality)

5e Completion of radiation-specific information

5f Completion of details of other professionals

5g Completion of administrative data (including documentation of MDT outcomes)

SB 5h Recording of previous radiotherapy treatment details

SB 5i Recording of patient’s specific requirements (includes communication/ handover/ documentation of patient specific information etc.)

SB 5j Recording of non-standard information/protocol variations

SB 5k Authorisation to irradiate (IR(ME)R) (including requests not signed by appropriately entitled practitioner and authorisation of additional

imaging)

5l Other

6 Booking process (pre-treatment, planning, treatment and follow up)

6a Bookings made according to protocol


53

6b Bookings made according to request details (including requested changes following initial booking)

6c Recording of booked appointments (including requested changes following initial booking)

6d Communication of appointments to patient (including requested changes following initial booking)

6e Other

6f Communication of appointment between staff groups (including requested changes following initial booking)

7 Processes prior to first appointment

7a New patient: registration with healthcare organisation’s PAS

7b New patient: registration with department PAS

7c New patient: generation of notes (including their availability as required across the patient pathway)

7d Old patient: location of healthcare organisation’s notes

SB 7e Old patient: location of department notes/previous treatment details (including availability of archived materials)

SB 7f Availability of reports/imaging required by protocol for treatment (including requirements for these at all points on the pathway)

SB 7g Availability of consent documentation

7h Other

8 Pre-treatment: preparation of patient

SB 8a Confirmation of ID

SB 8b Confirmation of consent


54

SB 8c Confirmation of fertility/pregnancy status

SB 8d Advice on procedure (including training on breath hold, bladder or bowel preparation, ICED or pacemaker status, information on pre-

medication, fiducial insertion etc.)

8e Other

9 Mould room/workshop activities


9b Pre mould room diagnostics/interventions

9c Production of immobilisation devices

SB 9d Checking/fitting of immobilisation devices

9e Production of other accessories/personalised beam shaping device

SB 9f Checking of other accessories/personalised beam shaping device

9g Labelling of mould room/workshop outputs

9h Recording of information in patient record (includes communication/ handover/ documentation of patient specific information etc.)

9i Instructions to patient

SB 9k End of process checks

9l Other

10 Pre-treatment activities / imaging (to include CT, simulation, clinical mark-up, reference image production)



55

10b Positioning of patient

10c Localisation of intended volume (including insufficient scan length, incorrect scanning protocol, incorrect laterality)

10d Production of images using correct imaging factors (including production of reference images)

10e Production of images using appropriate field sizes (including production of reference images)

10f Production of images demonstrating correct detail (including incorrect scanning protocol and production of reference images)

10g Labelling of images (including pre-scan data entry e.g. ID format, orientation etc. and production of reference images)

10h Saving of planning geometry data

SB 10i Recording of radiation data

10j Documentation of instructions/information

10k Marking of patient or immobilisation device

SB 10l End of process checks (including timeliness of sending scans to treatment planning)

10m Identification of staff

10n Other

SB 10o Assessment of patient prior to exposure

10p Use of contrast (including unplanned event such as leaking out, extravasation, timing of contrast etc.)

10q Use of gating (including discrepancy between intended treatment technique and pre-treatment scan, scan acquisition, construction of image

sequence or application of gating equipment etc.)

11 Pre-treatment planning process (including virtual simulation and replans)


56

SB 11a Verification of patient ID, orientation and data entry format to include all patient data, imaging etc.

11b Recording of patient ID on plan

11c Importing of data from external and internal administrative sources

11d Importing of data from external and internal imaging sources

11e Choice of data for planning purposes and to inform planning e.g. MRI, PET, angio, contrast, pre-op/post op data etc.

11f Choice of dose and fractionation inputs

SB 11g Availability of source data

11h Choice of technique/modality (including IMRT/ volumetric/ ART/ superficial or protons etc.)

11i Target and organ at risk delineation (including incorrect growing of volume)

11j Generation of plan for approval (to include DVH, incorrect labelling, inappropriate beam arrangement, replans or missing plan information

etc.)

SB 11k Authorisation of plan

SB 11l Verification of plan/identification of responsible staff

SB 11m Recording of definitive treatment prescription

SB 11n Recording of patient specific instructions

11o Management of process flow within planning (including plan export)

11p Management of authorisation process

11q Timeliness of plan production or approval


57

11r Calculation process for non-planned treatments

SB 11s Calculation checking process for non-planned treatments

SB 11t End of process checks

SB 11u Identification of responsible staff

11v Other

12 Treatment data entry process

SB 12a Pre-data entry verification (including OMS data import)

12b Choice of data entry method (input vs. transcription)

SB 12c Use of correct data

SB 12d Correct ID of patient/all patient input data

SB 12e Correct ID of patient output data

12f Accuracy of data entry (including field sequencing and image scheduling and any required amendments)

SB 12g End of process checks (including OMS data import)

SB 12h Identification of responsible staff

12i Other

13 Treatment unit process (including EXBRT, Protons and Superficial)

13a Availability/timeliness of all required documentation


58

SB 13b Patient ID process

SB 13c Patient data ID process

13d Explanation/instructions to patient

SB 13e Confirmation of pregnancy/fertility status

SB 13f Assessment of patient prior to treatment (including pre-medication prior to treatment e.g. analgesia, antiemetics etc., pace-maker or ICED

status)

13g Patient positioning

SB 13h Use of IVD according to local protocol

SB 13i Use of on-set imaging (including imaging according to local protocol)

13j Transfer of marks

13k ID of reference marks

13l Movements from reference marks

13m Setting of treatment machine parameters

13n Setting of collimator angle

13o Setting of jaw position

13p Setting of asymmetry

13q Setting of couch position/angle (incorrect setting of couch following movement to allow gantry clearance)

13r Use of immobilisation devices (including gating equipment)


59

13s Use of beam shaping devices

13t Use of beam direction aids/applicators

13u Use of compensators (including bolus)

13v Use of wedges

13w Availability of treatment accessories

13x Setting of energy

13y Setting of monitor units

13z On-set imaging: production process (including inappropriate exposure used, image not captured, incorrect CBCT filter used or left in for kV

image, incorrect field localisation of exposure, unsuitable positioning of imaging panel)

SB 13aa On-set imaging: approval process (including image review not completed, image review inaccurate, image matched to wrong reference

image, incorrect prioritisation of structures for matching)

13bb On-set imaging: recording process (recording of result of image review not undertaken, resultant actions from image review not undertaken,

documentation and application of systematic correction)

SB 13cc Management of variations/unexpected events/errors (including management of replans, migration of fiducials, transfer between treatment

machines)

SB 13dd Communication between treatment unit and V&R

13ee Recording of patient attendance

13ff Recording of delivered treatment data

13gg Recording of additional information

SB 13hh End of process checks (including checking of clearance for automated set-ups)


60

SB 13ii Identification of responsible staff

13jj Other

14 On-treatment review process

SB 14a On-treatment review of patient according to protocol by RT staff

SB 14b On-treatment review of patient according to protocol by other professional

SB 14c On-treatment review of notes/data to according protocol (including omission of weekly chart checks)

14d Actions following on-treatment review

14e Other

15 Brachytherapy (including Molecular RT and sealed source IORT)

15a Ordering of sources

15b Delivery of sources

SB 15c Source calibration

SB 15d Sterility of sources

SB 15e Correct applicators /sources

SB 15f Correct theatre equipment

15g Initial positioning of applicators / sources

15h Planning of treatment (including replans)


61

15i Maintenance of position of applicators /sources

15j Removing of applicators / sources

15k Other

SB 15l Validation of applicator/ source position

SB 15m Authorisation of plan

SB 15n Management of variations/unexpected events/errors (including management of replans, seed migration or fiducial migration etc.)

SB 15o Use of on-set imaging (including imaging according to local protocol)

15p On-set imaging: production process (including inappropriate exposure used, image not captured, incorrect field localisation of exposure,

unsuitable positioning of imaging panel)

SB 15q On-set imaging: approval process (including image review not completed, image review inaccurate, image matched to wrong reference

image, incorrect prioritisation of structures for matching)

15r On-set imaging: recording process (recording of result of image review not undertaken, resultant actions form image review not undertaken,

documentation and application of systematic correction)

SB 15s End of process checks

16 End of treatment process

16a Communication of appropriate end of treatment information to patient

SB 16b Recording of treatment summary information in notes

16d Communication of information to referring clinician/GP/CNS etc.

16e Organisation of follow-up appointment to protocol


62

16f Communication of follow-up to patient

16g Other

17 Follow-up process

17a Follow-up consultation and documentation

17b Management of non-attendance

17c Archiving of details of treatment

Other activities contributing to protocol violations

18 Timing

SB 18a Timing of chemo/irradiation

18b Transport issues

18c Portering issues

19 Document management

SB 19a Availability of current protocol, procedures, work instructions forms, training and competency documentation

20 Staff management

SB 20a Availability of staff with competency appropriate to procedure (including engineers, IT, medical, nursing, physics, radiographer etc.)


63

Appendix C: Causative Factor Taxonomy

Reproduced from Reproduced from Development of Learning guidance4

Category Code Description

Category CF 1 Individual

Sub-category CF 1a Failure to recognise hazard (knowledge-based etc.)

CF 1b Decision making process (rule-based or old or invalid rule used etc.)

CF 1c Slips and lapses (skill-based, involuntary automaticity etc.)

CF 1d Communication (inaccuracy or omission of verbal, written etc.)

CF 1e Violation (deliberate action, acting outside scope etc.)

Category CF 2 Procedural

Sub-category CF 2a No procedures / protocols (not in place or unavailable etc.)

CF 2b Inadequate procedures / protocols

CF 2c Adherence to procedures / protocols

CF 2d Process design (impractical and inefficient processes etc.)

Category CF 3 Technical

Sub-category CF 3a Equipment or IT network failure (including immobilisation & accessories)

CF 3b Commissioning/ calibration/ maintenance (including immobilisation & accessories)

CF 3c Device / Product design

Category CF 4 Patient Related

Sub-category CF 4a Medical condition (inability to remain still etc.)

CF 4b Communication with the patient (language issues, comprehension etc.)

CF 4c Non-compliance

Category CF 5 Teamwork / Management / Organisational

Sub-category CF 5a Inadequate leadership (inadequate supervision, congruence or consistency etc.)

CF 5b Unclear responsibilities and lines of accountability (across the radiotherapy pathway)

CF 5c Inadequate capital resources (equipment in use no longer fit for purpose etc.)

CF 5d Inadequate staffing (insufficient staffing levels or skill mix necessary to meet the demands of a service etc.)

CF 5e Inadequate training (inadequate or lack of training etc.)

CF 5f Inadequate risk assessment (poor change management etc.)


64

Category Code Description

Category CF 6 Environmental

Sub-category CF 6a Physical (power cut, control area excessively noisy, distractions etc.)

CF 6b Natural factors (fire, flood etc.)

Category CF 7 Other

CF 7a Other

Documents

Radiotherapy error and near-miss data report: December ... · Of the RTE reports, 65.8% (n = 10,414) were ‘near miss’ or ‘other non-conformances’ with no impact on patient