Spring 2025 QI Presentations

“Fasthead” MRI scans for children with headache: A review of two years practice

 

Background
Following a successful pilot, a “Fasthead” MRI protocol was introduced in February 2023. This reduces scan time from 13 to 4 minutes accompanied by a modest loss of image quality, considered acceptable in this clinical setting with agreed clinical criteria/ exclusions. Following a two year period of use, this project sought to determine if scans are of diagnostic quality and whether this protocol is being adhered to.

Methodology
MRI scans performed over a two year period in which “headache” was included in the clinical information were identified via EPR and retrospectively reviewed to determine if the “Fasthead” or conventional sequences were performed. Adherence to the agreed clinical criteria was determined for each examination. Findings of the examination and requirement for further neuroimaging was established.

Results
399 examinations were identified. 172 were appropriate conventional scans (43%)
72 were appropriate Fastheads (18%). 133 were conventional scans which met criteria for a Fasthead scan (33.3%). 22 were Fasthead scans in which criteria had not been met (5.5%).
4 of 94 Fasthead scan patients underwent further imaging. MRI in 3 cases (one space-occupying lesion; features of raised intracranial pressure; callosal lipoma), CT was subsequently performed in one case, but considered a separate presentation.
Only one case was identified in which the patient may have benefitted (modestly) from a conventional scan rather than Fasthead.

Conclusion
Study demonstrates scope for improvement in vetting requests but Fasthead is considered safe and efficient. Estimated 23.5 hours scan time saved during this period.

Inconsistency in CT Head PACS data storage: Opportunities for carbon savings?

 

Background
With 2024 confirmed to be both the hottest year ever recorded and exceeding 1.5 C above pre-industrial levels, sustainable imaging practice is of increasing urgency. Data storage of imaging examinations is a significant (and growing) source of carbon emissions. While judicious storage of archived examinations is essential for good quality care and clinical governance, anecdotal reports suggest wide variation in the volume of data stored for similar examinations at different centres. This study sought to demonstrate inconsistency in data storage practice, with a view to implementing lean storage practices as a carbon-saving strategy.

Method
Utilising the Scottish National PACS system, a CT head (non-contrast) examination performed within working hours was selected for every registered radiology centre on 7th November 2024. Each examination was interrogated to determine the total number of images stored and the nature of individual folders analysed.

Results
166 CT head examinations from across Scotland were included, with a range of 161 to 1913 images stored per examination. Median of 525 and an interquartile range (IQR) of 286. Co-efficient of variation was 49%. Identified opportunities for savings include: elimination of duplicate scanograms and non-patient images; deletion of thick-slice reconstructions after reporting; elimination of redundant non-diagnostic images. Routine storage of fine bone reconstructions should also be scrutinised according to clinical context.

Conclusion
This study highlights marked variation in data storage practice, and opportunities for energy savings. Robust guidelines to ensure good practice are advocated.

Does linking XR CHEST with XR Shoulder or XR Pelvis increase ED plain film turnaround times?

 

Background
Plain film skills mix including reporting radiographers, consultants and registrars. Only consultants and post FRCR registrars can verify chest radiographs (CXR). Preliminary reporting of CXR is an important part of training but is often limited by preference and availability of supervisors.
CXR often linked and inseparable from appendicular e.g. CXR and Pelvis / hip in the case of neck of femur fractures or additional CXR when dealing with shoulder trauma. Anecdotally these linked accessions are those that go unreported the longest.
NHS GGC targets are to report 95% of ED films within 24 hours. RCR guidance for England suggest all ED films should be reported within 12 hours.

Aims
To investigate potential cause and remedy of delay to plain film reporting times.

Methods
CRIS data requested for ED radiographs from November 2023 including XSHRL, XSHRR, XCHES and XPELV and data highlighted for multiple exam codes performed concurrently.
Time to final report was primary outcome.

Intervention
Radiographers instructed to un-link requests prior to acquisition. Repeat cycle carried out November 2024.

Results
Linking studies led to delayed reporting times. Good uptake with unlinking. Residual linked studies still have a longer turn around time. Still below targets for turn around times.

Conclusions
Linking appendicular and CXR is associated with increased turn-around time.
Findings presented at governance meeting so re-audit should hopefully see further improvement in numbers of unlinked studies and therefore reporting times.

Tele vs Home : Skipping Priors, Scanning More?

 

Background
While there was almost no private teleradiology presence in Scotland 10 years ago, all but one Scottish health board rely on them to meet exponential increase in imaging demand.
Comparing with prior imaging is a key Royal College of Radiologists recommendation, to improve quality of radiology reports while reducing unnecessary imaging recommendation^1,2.

Objective
How often do radiologists (in-house v tele-radiologists) state comparison with prior imaging?

Method
Following statistical advice, a retrospective audit of 250 consecutive CT Chest, Abdomen and Pelvis (CT CAP) reports was performed to evaluate for group level difference in (1) comparison with previous (2) further imaging recommendation and (3) throughput of imaging.

Results
No significant difference in the sex (71F v 65F) and age (63.6 v 64.9 years) distribution in the patients examined by the two groups of radiologists. Nineteen (15.2%) patients from the outsourced group underwent their first CT CAP, compared to 54 (43.2%) for the in-house group (p < 0.0001).
Teleradiologists explicitly compared with previous imaging (where available) in 87.7% of reports compared with 95.8% of in-house radiologists.
Further imaging was recommended for 3 (4.2%) patients by in-house radiologists compared to 13 (12.3%) by outsourced group.
One hundred and twenty-five CT CAP were reported in 56 days by outsourced versus 273 by in-house radiologists.

Conclusion
Outpatient CT CAP are 5 times as likely to be reported by teleradiologists, who are less likely to compare with previous imaging and more likely to recommend further imaging.

References
1. Royal College of Radiologists. Standards for interpretation and reporting of imaging investigations. Second edition. 2018. https://www.rcr.ac.uk/media/yiglbn35/rcr-publications_standards-for-interpretation-and-reporting-of-imaging-investigations-second-edition_march-2018.pdf. (Accessed 23.3.25)
2. Judith D Akwo, Phuong Trieu, Sarah Lewis, Does the availability of prior mammograms improve radiologists’ observer performance?—a scoping review, BJR|Open, Volume 5, Issue 1, 1 November 2023, 20230038, https://doi.org/10.1259/bjro.20230038

Are Radiology Residents safe to report feeding NG tubes on Chest X-rays?

Authors: Dr. Lucy McGuire (1,2), Prof. Patrick J O’Dwyer (2), Dr. David Young (3), Prof. Cindy Chew (2,4)
1. Queen Elizabeth University Hospital, NHS Greater Glasgow and Clyde, Glasgow, G51 4TF, United Kingdom
2. Undergraduate School of Medicine, University of Glasgow, Glasgow, G12 8QQ, United Kingdom
3. Department of Mathematics and Statistics, University of Strathclyde, Glasgow, G1 1XQ, United Kingdom
4. Department of Radiology, University Hospital Hairmyres, NHS Lanarkshire, East Kilbride, G75 8RG, United Kingdom

Objectives
 The task of issuing reports on whether nasogastric tubes (NGT) are safe to use for feeding (or not) on chest x-ray (CXR) often falls to radiology residents. This study is the first attempt to evaluate the ability and confidence of radiology residents to correctly interpret NGT position on CXR. In addition to quantifying adherence to the standard of formal teaching.

Methods
Radiology residents were invited to participate in an online NGT on CXR study. The CXR images comprised of 20 NGTs, 14 of which were correctly sited, while 4 were in the distal oesophagus, and 2 in the lung.

Results
Despite reminder emails from Training Program Directors and incentives to participate, only 28 (15%) of 185 radiology residents responded. Of those, only 10 (35.7%) correctly identified all NGTs on CXR. The most common error was reporting a correctly sited NGT as unsafe for feeding. This occurred on 46 occasions or 8.2% of the time. Residents who correctly interpreted all 20 NGT CXRs were significantly more confident in their abilities on a 5-point Likert scale than those who got at least 1 NGT CXR wrong [4.4 (0.52) versus 3.8 (0.79), p = 0.02]. Only 12 (42.9%) of residents had received previous training in NGT CXR interpretation, and only 7 (25.0%) received this during radiology training.

Conclusions
This study suggests that radiology residents are not adequately trained to interpret the position of feeding NGT on CXRs. These errors could delay or withhold appropriate feeding. Early and compulsory training in this important skill should be instituted urgently.

Objective
Computed tomography pulmonary angiogram (CTPA) is the gold standard imaging investigation for the diagnosis of patients with possible pulmonary embolism (PE). UK national guidelines (NICE) advise that CTPA requests should be guided by pre-test probability (PTP) assessments (e.g. Well’s Criteria) and the use of D-dimer if the 2-level PE Well’s Criteria yields low clinical probability. This Quality Improvement Project (QIP) performed at a tertiary centre in Edinburgh aims to evaluate whether CTPA is used appropriately and identify the diagnostic yield of scans for PE or alternative diagnoses. The Royal College of Radiologists (RCR) suggest that at least 15% of CTPAs should detect PE with an alternative diagnosis in at least a further 50%. These figures were set as the target.

Methods
The clinical details of 50 patients that underwent CTPA between July and October 2023 at the Western General Hospital (WGH), Edinburgh, were retrospectively reviewed. The patients’ baseline characteristics, D-dimer levels, preceding chest X-ray (CXR) results, documented Well’s Criteria, referrer’s details and CTPA results were collected and analysed.

Results
Of the 50 CTPAs, only 4 (8%) confirmed a PE, an alternative diagnosis was reported in 35 scans (70%) with no abnormality being detected in the remaining 11 scans (22%). Within the 4 confirmed PE scans, 50% were large saddle emboli and the other 50% were subsegmental clots. A form of PTP was applied in only 9 cases (18%). Specifically, 2-point Well’s Criteria was used in 8 of the 9 patients and Simplified Geneva Score was used in only 1 patient. By contrast, D-dimer was requested in the majority of patients (86%). 3 patients were found to have a negative D-dimer and still proceeded to CTPA scanning.

Conclusion
Whilst the alternative diagnosis rate (70%) met the RCR recommendation, the PE pickup rate (8%) was below the expected standard. The limited use of PTP in this cohort and related inadequate use of D-dimer testing could have contributed to possible over-requesting of CTPAs.

Diagnostic Accuracy and Radiation Dose in Renal Colic: An Audit of Low Dose CT KUB in Renal Colic

Dr May Chua, Dumfries and Galloway Royal Infirmary
Dr Sohail Iqbal, Dumfries and Galloway Royal Infirmary

Background
CTKUB is the preferred radiological investigation for patient presenting with acute renal colic. It aligns with British Association of Urological Surgeon (BAUS) (1) and NICE guideline for management of acute ureteric colic (2). Low radiation (radiation dose/exam of < 3.5mSV) is also recommended for patients with BMI of less than 30 (3). Patients affected by renal colic are often young adults that require multiple CT scans over their lifetime; thus, we should ensure that there is appropriate diagnostic yield and that no unnecessary repeat CT is performed for the same presentation (1,3), as each scan’s dose is critical for long term safety.

Objectives and Methods
We want to investigate the diagnostic yield of CTKUB in patient with renal colic with a secondary aim of the dose being used for patients with BMI <30.

Patients admitted to Dumfries and Galloway Royal Infirmary from June to August 2024 with unilateral loin pain high suspicious of renal colic are identified. Data were collected on the investigation they had, recent admission with renal colic, time of request, time of execution, CTDI Vol (mGy) and their BMI.

Results
53 patients presented to DGRI and GCH with renal colic between June and August 2024. 72% (n=38) of patients underwent CTKUB as diagnostic investigation. 8% (n=4) of patients had CT AP for possible renal colic. 92% (n=35) of CTKUBs requested were performed within 24 hours. Of the CTKUBs performed, 68% (n=26) had a positive finding of confirmed calculi(s), 8% (n=3) of alternative diagnosis and 16% (n=6) with no radiological findings. 5 patients do not have their weight recorded on the system. 19 patients had a BMI <30 and 79% (n=15) of them had low dose (LD) CTKUB. 3% (n=1) had CTKUB done less than 3 months ago.

Conclusion
Patients with renal colic as differentials do not always get CTKUB as first line of investigation. We suggest that BMI should be placed on the referral card to allow radiographers to choose appropriate dose for the patients rather than relying on estimated body weight. Referrer should review differentials and consider other methods of investigation if there are uncertainties, recent imaging, or recent admission with similar presentation. This can be improved by providing education to the respective teams and re-audit in 3 months’ time with view of 100% standard in low dose CTKUB for BMI<30 and 80% in diagnostic yield.

References
1. Tsiotras, A., Smith, R.D., Pearce, I., O’Flynn, K. and Wiseman, O. (2017). British Association of Urological Surgeons standards for management of acute ureteric colic. Journal of Clinical Urology, 11(1), pp.58–61. doi:https://doi.org/10.1177/2051415817740492.
2‌. Renal and ureteric stones: assessment and management. NICE guideline NG118
3. Rob, S., Bryant, T., Wilson, I. and Somani, B.K. (2017). Ultra-low-dose, low-dose, and standard-dose CT of the kidney, ureters, and bladder: is there a difference? Results from a systematic review of the literature. Clinical Radiology, 72(1), pp.11–15. doi:https://doi.org/10.1016/j.crad.2016.10.005.

RCDS : What is happening in NHS Lanarkshire?

 

Background
The RCDS 2023 report highlighted NHS Lanarkshire had the lowest cancer detection rate and highest non-cancer diagnosis rate compared amongst 4 other health boards.1 This may have reflected its infancy, therefore the service’s current effectiveness remains unclear. 20-40% CTCAP’s can reveal actionable incidental findings, however, true incidence within RCDS and impact on patients and the health system is unknown.2

Question: What is the outcome and cancer pick-up rate from RCDS in Lanarkshire?

Method
A retrospective audit of 100 consecutive patients undergoing RCDS CTCAP between 1/10/2024 – 15/11/2024 across NHS Lanarkshire were identified from the RIS. Information collected included gender, age, findings, further imaging recommendation and final diagnosis. Clinical history was interrogated to determine adherence to RCDS Lanarkshire inclusion/exclusion criteria.

Results
59 female and 41 male patients, with median age of 68.4 years were scanned. All patients met referral indications. 19 patients were scanned despite meeting exclusion criteria, of which 16 were eligible for site-specific USOC pathways. Forty-two additional imaging examinations were generated for 31 patients from the 100 CTs, of which 9 (for 7 patients) including 2 PETCTs were not radiology recommended. 5 cancers were diagnosed from the primary 100 CTs, with a further cancer (breast) confirmed after mammography and biopsy. Overall cancer rate was 6 from 142 imaging examinations (4.2%). A mean time of 40.84 days between RCDS referral to report outcome was found.

Conclusion
A fifth of patients were scanned under the RCDS pathway despite not meeting inclusion criteria. Significant additional imaging was generated, with implications on patients and health system costs.

References
1. Meer PRMPRVD. Final Report of the Evaluation of Rapid Cancer Diagnostic Services. Scotland: University of Strathclyde; 2023.
2. Lumbreras B, Donat L, Hernández-Aguado I. Incidental findings in imaging diagnostic tests: a systematic review. Br J Radiol. 2010;83(988):276-89.

Are Chest X-Rays a Clinically Significant Step in the Neck of Femur Fracture Pathway?

Authors: Dr. Catherine Cobburn, Dr. Nicole Harley
Institution: Queen Elizabeth University Hospital (QEUH), 2024

Background
Neck of femur (NOF) fractures represent a growing public health burden due to an aging population. Care standards for the NOF fracture management are set and audited at a national level in Scotland. Despite their common use, chest radiographs are not mandated in national NOF management guidelines. This study evaluates the clinical utility and cost-effectiveness of routine admission CXRs, in particular screening radiographs requested by radiographers, in NOF patient management at QEUH.

Aims
Identify how many patients with neck of femur fractures also received a chest radiographs
Determine how many of these CXR’s are requested due a specific clinical indication and many of these are requested by radiographers due to observing a fracture
To assess if the CXR’s performed with no clinical indication had any impact on initial clinical management.

Methods
A retrospective analysis was conducted using CRIS over a two-month period (Jan–Feb 2024) reviewing 587 pelvic/hip radiographs, identifying 123 confirmed NOF fractures. Of these, 102 patients also received a CXR. Reports and clinical records were reviewed to determine the requesting party (radiographer vs. clinician), indication for the CXR, findings, and whether results influenced clinical management.

Results
• 67 CXRs were radiographer-initiated; 36 were clinician-requested.
• CXRs with clinical indication had a 25% rate of positive findings and influenced management in 100% of cases.
• Routine CXRs had only a 5% positive finding rate and affected management in 5% of cases.
• Routine CXRs were often of suboptimal quality and not always reviewed by clinicians.
• Estimated cost of routine CXRs at QEUH exceeded £26,000 annually.

Conclusion
CXRs requested for a clinical indication rather than screening are more likely to influence practice.
These findings do not support the routine request of chest radiographs in patients >50 with NOF fractures.
There may be a significant cost saving benefit if NOF screening CXRs were not performed.

Influence on Practice:
After taking these findings to the local relevant clinicians who attend the Hip Fracture meetings, and the GGC Governance meeting, screening chest radiographs have been removed from the local protocol in GGC.

The impact of incidental findings in silver trauma.

Authors: Emma Curran, Rachael Kirkbride, Hui Yen Teh, James Bott, Charlotte Algeo, Judith Anderson, Gregor J A Stenhouse, Mark T. Macmillan.

Introduction
Whole-trauma body computed tomography (CT) scanning is increasingly used to assess patients who have suffered trauma. In addition to assessing traumatic injuries, incidental findings are frequently detected, which are often challenging to manage and costly to follow-up. This is particularly the case in the elderly, so called silver trauma. This study aims to assess the burden of incidental findings produced by WBCT scanning in elderly trauma patients.

Methods
We retrospectively reviewed all WBCTs within a single healthcare trust between 2021 and 2022 performed on people aged 70 and over (n=465). The electronic clinical notes system was used to determine the follow up appointments, imaging and treatments that resulted from each incidental finding. Incidental findings were then divided into; those that underwent treatment, those that underwent follow-up only, those in which follow-up was intended but not carried out, and those that resulted in no clinical action.

Results
Across the 465 WBCT scans, 39.1% (n=182) reported at least one incidental finding and 227 incidental findings were reported in total. Of the incidental findings, pleural effusions and pneumonia were the most common (n=23). This was followed by lung masses and pulmonary nodules (n=21), biliary duct dilatation (n=14) and adnexal lesions (n=13). Treatment was provided for 15.9% (n=36) of the incidental findings. 27.3% (n=62) of incidental findings underwent follow-up, resulting in 67 imaging studies, 36 clinic appointments, and 11 other investigations, but ultimately did not undergo treatment.

Of all the incidental findings, 28.6% (n=65) would not have been detected had only the area(s) of clinical concern been scanned. 30.6% (n=11) of findings that received treatment would not have been found. These included five cases of lung consolidation, two cases of obstructing kidney stones, two cases of new brain infarcts, one case of metastatic prostate malignancy, and one case of urinary retention. 29.0% (n=18) of findings that resulted in follow-up but no treatment could have been avoided if targeted scanning was used. This would have resulted in 19 less imaging studies, six less clinic appointments, and five less other investigations.

Conclusion
Our work has found that approximately two-fifths of WBCT scans in people over 70 have incidental findings, of which only 16% resulted in treatment. A targeted CT approach would have resulted in reduced follow-up imaging and clinical appointments.

Further work
We plan to run a cost-analysis to show the potential savings that could be made by employing a targeted imaging approach in silver trauma patients. We also intend to carry out a more detailed sub-group analysis to see if demographics (such as age and mechanism of injury) correlate with an increase in the likelihood of finding clinically insignificant findings.