myBSPR.org

Website of the BSPR

BSPR 2008 Abstracts

Abstract for the BSPR Birmingham November 2008

 

 

Multiple unexplained fractures in infants : non-accidential injury or osteogenesis imperfecta?

Sudigali V, Dimond D, Grier D, Smithson S, Burren C, King S.

 Department of radiology, orthopaedics, endocrinology and clinical genetics, Bristol Children’s Hospital and Department of Radiology, Weston General Hospital, Weston Super Mare, Somerset.

 

Aims

To review the history, clinical and radiological findings of non-accidental injury and osteogenesis imperfecta with a view to differentiating between these two conditions.

 Materials and Methods

Retrospective case note review of all infants who underwent a skeletal survey over a period of 5 years at Bristol Children’s Hospital.  Clinical history, signs and symptoms and radiological findings were noted.

 Results

There were a total of 109 skeletal surveys.  The final diagnosis was osteogenesis imperfecta in 5 infants and non-accidental injury in 14.

 1/5 children with osteogenesis imperfecta had no family history of the disease.  All of the 14 children with non-accidental injury had a history incompatible with the clinical findings.

There was overlap of radiological features in a minority of cases.  One child was difficult to diagnose due to overlapping clinical and radiological features and was eventually diagnosed as osteogenesis imperfecta by genetic testing.

 Conclusion

Differentiating osteogenesis imperfecta from non-accidental injury is important and can sometimes be difficult.  Key to differentiation is the correlation of history, examination and radiological findings.  Genetic testing may be conclusive.

 

 

Histories given in unexplained long-bone fractures:  work in progress

 

Background:  Fractures of diaphyses of long bones are the commonest single type of non-accidental injury which produced radiological changes.  If the history given at presentation is judged “incompatible”, it is routinely ignored except as a risk factor for non-accidental injury.  However, lengthy and highly detailed accounts are often given of circumstances said by the child’s family to be associated with the fracture.

 

Methods:  Up to 15 cases have been identified from my own archive so far.  Modified Yale criteria were applied, independent of the history itself, to identify high-risk and unknown-risk cases.  The records were trawled for first-hand account of circumstances surrounding the fracture, including medical notes, police interviews, solicitors’ statements, etc.  Specialised text-analysis software (Nvivo) was used to identify key features of the histories and to cross-correlate them.

 

Results:  Commonly occurring features included multiple histories, observation that the child was suffering on more than one occasion before presentation to the health services, attribution of the history to accidents during routine care of the child, and self-doubt about the veracity of the accounts.

 

Conclusion:  There were many stereotypical features of these histories.  These features may be used in future to help identify children at high risk of non-accidental injury.

 

 

Rib Fractures Identified at Postmortem Examination in Sudden Unexpected Deaths in Infancy  (SUDI)

Weber MA, Risdon RA, Offiah AC*, Malone M, Sebire NJ

Department of Histopathology & *Department of Radiology

Great Ormond Street Hospital for Children

 

Introduction:  Rib fractures are associated with physical abuse;  the significance of fresh fractures in relation to resuscitation remains undetermined.  This study aimed to establish the characteristics of rib fractures identified at autopsy in cases of sudden unexpected death in infancy (SUDI)

 

Methods:  Anonymised records of SUDI cases were searched to identify those with rib fractures.

 

Results:  Over a 10 year period, 546 SUDI post-mortem examinations were performed, including 94 forensic autopsies.  Rib fractures were identified in 24 (4%).  In nine (2% of SUDI) there were only fresh fractures with no histological evidence of healing;  7 (78%) of these, all situated in the anterolateral chest, were classified as resuscitation – related fractures – there being no other injuries.  15 (3%) infants demonstrated healing fractures, and of these, 10 (67%) demonstrated additional features suggestive of abuse.  Abuse-associated fractures were posterior and costochrondral junction, not seen in apparent resuscitation – related cases.  Healing rib fractures were apparent on 93% of skeletal surveys;  fresh rib fractures only on 22%.

 

Summary:  Rib fractures in infancy indicate abuse, particularly when healed or healing, posterior or costochondral.  Isolated fresh anterolateral fractures are likely to be resuscitation – associated.  Fresh rib fractures are often missed on skeletal survey, but are reliably detected at postmortem.

 Standards for Radiological Investigations of Suspected Non-accidental Injury (NAI)  (March 2008) suggest that ‘consideration should be given to developing local and regional networks to facilitate a second opinion’

 At our institution between Jan 2005 and June 2008, over 80% (39/48) of skeletal surveys for suspected NAI were double reported by radiologists with a paediatric subspecialty interest.  There was inter-observer discrepancy in 4/39 (~10%).  All of these discrepancies were consider ‘major’, based upon their impact over the final radiological report.  In 2 cases, a third opinion was required before consensus was reached.

 Using double reporting as the standard, single reporting alone would have identified all the abnormalities in 89.7% of cases with a 95% confidence interval of 75.8-97.1%.

 These findings support the recommendation that double reporting should be carried out for all skeletal surveys with suspected NAI.  Ideally a consensus report should follow discussion between the radiologists.  There may be considerable logistical difficulties when examinations are sent away for a second opinion, and therefore we would endorse the generation of reporting networks.

 


MR Imaging of intra and extra cranial haemorrhage in the neonate:  birth injury or non-accidental injury.

 L L Wallis, S V Gandhi, M F Smith, M N J Paley, P d Griffiths, E H Whitby.

 Academic Unit of Radiology, The University of Sheffield, C Floor, Royal Hallamshire Hospital, Glossop Road, Sheffield S10 2 JF

 

Purpose

Cranial haematomas in infants are associated with non-accidental head injury.  We sought to establish (12) the frequency and natural history of subdural haemorrhages in asymptomatic term neonates (2) whether the presence of intracranial bleeds are relected in an increased number of extracranial bleeds (3) the influence of the mode of delivery.

 

Methods

Term neonates were imaged within 48 hours of delivery using a 0.2T MRI scanner.  Obstetric details were recorded retrospectively.  A neonatal radiologist recorded incidence of subdural or subgaleal haemorrhage and cephalohaematoma.

 

Results

Imaging from 484 neonates revealed 38 subdural haemorrhages (7.7%).  Of these, 19 were NVD, 3 were forceps deliveries (odd’s ratio = 1.23), 6 were forceps after attempted ventouse delivery (OR=3.4; p=0.02), 8 followed ventouse delivery (OR=1.96) and one followed emergency caesarean section (OR=0.38).  All had resolved by the 4 week rescan with no reoccurrence to date.  The distribution of subgaleal haemorrhage (n=18) and cephalohaematoma (n=11) followed that of subdural haemorrhage.

 

Conclusion

High numbers of clinically silent intracranial and extracranial haemorrhages occur frequently in instrumental deliveries and resolve spontaneously without detrimental effects within 4 weeks.  Subdural haemorrhages after 4 weeks of age are therefore not related to birth injury and traumatic causes must be considered.

 


Early Experience of Hip Screening in Cerebral Palsy

 K A Duncan & S L Barker

 

Background

Hip displacement occurs in approximately one third of children with cerebral palsy.  There is a broadly linear relationship between the severity of the disease the the likelihood of hip problems.  Based on this we have introduced a local protocol to guide clinical and radiological monitoring of hips at risk in children with cerebral palsy.  The children are clinically monitored by community paediatricians and referred at intervals for radiographs depending on their age and level of disability.

 

Method

An AP pelvic film is taken in a standardised radiographic position with the lumbar lordosis abolished.  The acetabular indices and migration percentages are calculated using software on our PACS equipment and a report issued.  If there is evidence of migration of 30% or above, children are referred to one of the Orthopaed Surgeons for further assessment and management.

 

Results

Since introducing this screening policy we have screened 20 children and to date have detected 12 hips in 7 children requiring orthopaedic referral.

 

Conclusion

In this relatively small group of patients we have detected abnormality in 35% which had not been previously suspected clinically.  By earlier detection and referral for management it is hoped that the degree of bony incongruity can be minimised, facilitating easier surgical management and preserving mobility.

 

OCT for the Detection of Complete Tracheal Rings in Children with Suspected LSCTS

McLaren C, Elliot M, Roebuck DJ

Great Ormond Street Hospital for Children

Purpose:  LSCTS  with complete (0-shaped) tracheal cartilage rings is a rate life-threatening condition that usually presents in infancy.  Because accurate preoperative assessment is critically important, we evaluated the feasibility of demonstrating complete rings with OCT.

 

Method

Complete rings can usually be seen at bronchoscopy.  OCT was therefore performed only on selected children in whom the bronchoscopic diagnosis was not absolutely certain.  A 150-µm (0.006-in) optical imaging probe (Lightlab Imaging, Westford, MA) was introduced through a 1.0mm bronchographic catheter (n=5) or a 2.8mm flexible bronchoscopy (n=5).  Surgical correlation was available in five children.

 

Results:

Complete rings were easily identified in the four children aged <1 year, and confirmed at surgery.  In the older children (aged 1 to 6 years) complete rings were progressively more difficult to identify with age, because it become increasingly more difficult to capture the entire tracheal cross-section on a single image.  This is because the currently available equipment has a fixed field of view of 8mm.

 

Conclusion:

Detection of complete tracheal cartilage rings with OCT is feasible in infants with LSCTS.  Technical modifications may increase the age range over which this technique is useful.

 

Clinical Relevance

OCT is useful when the diagnosis of LSCTS in uncertain.

 


Title of Paper:  GP referred UTIs:  Should we be NICE?

 Authors:  Dr Kirsteen McDonald, Dr Ian Kenney

 Institution:  Royal Alexandra Children’s Hospital, Brighton.

 

Introduction:

In August 2007 the National Institute for Clinical Excellence 9NICE) published new guidelines for ther management of children with suspected urinary tract infection 9UTI).  Only a selected group of high risk patients will now be investigated with imaging.  The efficacy of the new guidelines in diagnosing relevant urinary tract abnormalities is unclear.

 

Methods:

We applied the new imaging strategy to a historic cohort of 934 patients with UTI referred by General Practitioners (GO) to a specialist children’s hospital between 1996 and 2002.  We evaluated specificity, sensitivity, positive, and negative predictive values of the new strategy.

 

Results:

Of the 934 patients referred, 218 would have been investigated according to NICE guidelines.  In total there were 105 patients with abnormal results, and 44 of these belonged to the group investigated by NICE.  The positive predictive value of the new imaging strategy was 20.6% (C.I 15.1-26.1%), the negative predictive value was 91.5% (89.2-93.4%).  Specificity was 79% and sensitivity 42%).

 

Conclusion:

With the implementation of the new guidelines, less patients presenting with UTI will require imaging.  In our study about 10% had a significant abnormality (reflux nephropathy, or conditions that may require surgical intervention), but only half of them would be identified with the new protocol.  The relevance of these urinary tract abnormalities on the development of long term renal impairment or hypertension is still debated.  Whilst following the new imaging guidelines may result in less abnormalities being identified at an early stage, the long term effect for patients is unclear.

 


Imaging Children:  do we provide a child-centred service in adult hospitals?

 Mathers, S,: Anderson, H: McDonald S,:  NHS Grampian

 

Background: Children’s National Service Framework (2003 DoH) emphasised health services for children should be child-centred to meet their needs, no matter where they live.  Children are major users of healthcare but not all children have access to a paediatric hospital, the majority of children being imaged in facilities primary design for adults.  The main aims of this study were to:  i)  investigate the imaging service for children in adult hospitals and  ii) establish children’s involvement.

 

Methods:  A questionnaire was developed and distributed to superintendent radiographers in all English and Welsh hospitals with imaging facilities.  Quantitative data were entered into SPSS-PC.

 

Results:  A response rate of 70% (289/411) of adult hospitals was achieved and 81% (236/289) imaged children.  From respondents’ estimates, 1.3 million children are imaged in adult hospitals.  76% of departments reported no separate amenities such as waiting rooms and toilers for children.  84% stated they had no protocols in place for children with special needs.  Children’s views on the provision of imaging services were seldom sought in a systematic way.

 

Conclusions:  There is a need for child-centred facilities and policies for imaging children in adult departments to be developed and implemented.

 

 Survey was funded by College of Radiographers (UK) Research Award.


 

The role of Doppler Ultrasound in the Diagnosis of Hepatic Veno-Occlusive Disease.

KJ Taylor-Robinson, S Wilkins, R Keenan and LJ Abernethy

 

Introduction:  Hepatic Veno-Occlusive Disease (VOD) is a serious complication of Haemopoetic Stem Cell Transplantation (HSCT) with mortality up to 28%.  Diagnosis of VOD following HSCT is primarily based on clinical and biochemical (Seattle) criteria, but ultrasound is frequently requested to help confirm the diagnosis.  The recognised ultrasound features of VOD are hepatomegally, gallbladder wall thickening, blood flow reversal in the portal vein, and ascites.

 

Method:  Retrospective review of 66 children receiving HSCT over 6 years, and also 8 children with Acute Lymphoblastic Leukaemia (ALL) who had VOL.

 

Results:  9 (14%) of the HSCT group were treated for VOD.  Ultrasound showed hepatomegally and ascites in 6 cases (67%), but gall bladder wall thickening and portal vein reversal were found in only 2 (22%).  Portal vein reversal and gall bladder wall thickening were specific but not sensitive for VOD.  Ascites was specific and moderately sensitive.  In the ALL group 4 (50%) had hepatomegally;  2 (25%) gall bladder wall thickening;  3 (38%) portal vein reversal;  2 (25%) hepatic vein flow attenuation;  4 (50%) ascites.

 

Conclusion:  Our results suggest that ultrasound does not alter management of VOLD in HSCT because treatment is initiated early on clinical criteria.  Ultrasound may be more valuable in ALL, in which VOD is less common.

 

Ultrasound Guided Drainage of Empyeme in Children – 5 year experience –

 S Ramamurthy, H Evans, I Doull, A Evans

 

Purpose:  To review our practice of ultrasound guided empyeme drainage in children.

 

Background:  Pneumonia in children may be complicated by parapneumonic effusion and less commonly by empyema.  Those unsuccessfully treated by antibiotics alone may require drainage,.e  Although this has traditionally been performed by surgically placed large-bore drain or thoracotomy, there has been a recent move to less invasive techniques including radiological placement of small-bore pigtail catheters and video assisted thoracoscopic techniques.  At our centre we have used ultrasound guided placement and urokinase instillation as an alternative to surgical management over the last 8 years and almost exclusively over the last 5 years.

 

Methodology and Results:  A 5 year retrospective analysis was made of children with empyema treated with ultrasound guided radiological drainage.  Information about the patient demographics, prior medical treatment, ultrasound findings, plain radiograph findings, complications of catheter placement, duration of drain, hospital stay and any long term complications was made.  The vast majority of the patients responded well to the procedure with little post procedure complication.

 

Conclusion:  Imaging-guided pigtail catheter drainage of empyema in children is a safe and effective procedure and an effective alternative to traditional surgical therapy.  Combined with fibrinolytic use, the failure rate and complications are minimal.

 

Audit of Radiological Investigations to Delineate Perineal Anatomy in Males with High An-rectal Malformations

K Elmalik, G Murthi, A Sprigg


Aim of the Study:  In males with high anorectal malformation, following colostomy formation, accurate delineation of perineal anatomy is essential prior to reconstruction.  The aim of this audit was to review the results of radiological investigations and their correlation with operative findings in our practice.

 

Methods:  Radiological data collected included:  accuracy of information on request card, technical quality of studies, whether anatomy was demonstrated to satisfaction, whether studies repeated, number of radiologists performing study, whether study attended by consultant surgeon.  Surgeon’s operation notes were reviewed for documentation of level of fistula.

 

Results:  Data on request cards for these investigations was accurate in 3/11 patients only.  Technically 7/11 studies were considered satisfactory.  Both loopogram and MCUG were performed on the same visit for 7/11 studies only.  Studies were repeated on 4 occasions.  The recto-urethral fistula was clearly demonstrated in 7/11 cases.  These investigations were performed by eight different radiologists.  Data regarding attendance by surgeon was unavailable.  Radiological diagnosis correlated with operative findings in 3 cases only due to either inadequate radiology 4/11.

 

Conclusion:  Performance of this investigation by an assigned Consultant Radiologist with attendance by the Surgeon will enhance the quality and accuracy of the results.

 

Sonography for Gastro-Oesophageal Reflux in the Paediatric Age Group

 Dr PFC Lung, Dr A D’sa and Dr S Maheshwaran

 

Introduction:   Last year’s audit of paediatric barium swallows at Mayday Hospital, confirmed most (71%) of the requests were to look for gastro-oesophageal reflux (GOR), which is implicated in conditions such as asthma and failure to thrive.  According to published literature, ultrasound is more sensitive than barium studies in the diagnosis of reflux, but is rarely performed in hospitals in England.  From May 2007, we offered ultrasound as the first line investigation to look for GOR.  This audit examined the outcome of this change.

 

Method:  During the ultrasound examination, the gastro oesophageal junction was observed for up to 5 minutes for the presence of reflux and the intra-abdominal oesophageal lengths were measures.

 

Paediatric barium swallows/meals and sonography for reflux results were analysed over a 1 year period.

 

Results:  57 barium studies and 35 ultrasound scans were performed during the past year at Mayday Hospital, with an average age of 3 years, and 1 year & 8 months respectively.  39% (22) of barium studies demonstrated reflux, compared to 66% (23) with ultrasound.  11 patients had both ultrasound and a barium study.  They agreed on 7 occasions,  whilst ultrasound picked up 4 more episodes of reflux compared to contrast.

 

All the ultrasound requests were for reflux, compared to 60% for barium.  Other indications for barium studies included stridor (15%) and malrotation (9%).  No additional findings were apparent on ultrasound, but barium identified 2 hiatus hernias, an oesophageal stricture and partial duodenal obstruction.

 

Intra-abdominal oesophageal lengths were measured on ultrasound to determine its significance in GOR.  The average oesophageal length in children with reflux is 1.57cm, compared to 2.37cm without.

 

Conclusion:  Ultrasound can be used as a first line investigation for reflux in the paediatric population, reducing the need for barium swallows when hte aim of the examination is to only look for gastro-oesophageal reflux (GOR).  This small sample confirms the association between the intro-abdominal oesophageal lengths with reflux, as seen in other studies.  Obtaining normal values for intra-abdominal oesophageal lengths in age matched controls & technique refinement, to grade the severity of gastro-oesophageal reflux, will make this a more acceptable method of examining for GOR.


Thinking about Measuring

 Dr D Horton, Mrs P Parker

Ultrasound Department,  Hull Royal Infirmary,  Anlaby Road,  Hull,  HU3 2JZ

 

A previous audit at this institution demonstrated wide intra-operator and inter-operator variability in measuring the fluid volumes of locally constructed test objects made to mimic the infant and child bladder.  Operators were attempting high precision measurements whilst not assessing the accuracy or relevance of the volumes calculated.  Recommendations were that operators should consider how precise a measurement really needed to be.

 

A second audit was undertaken to evaluate any change in practice.

 

The audit was repeated.  (The construction of the test objects and test methodology will be briefly described.)

 

The measurements taken were evaluated in terms of accuracy of volume measurement and both intra-operator and inter-operator variation.

 

The results demonstrated a reduction in intra-operator and inter-operator variability but also a greater accuracy of measurement despite a reduction in the precision of the measurements.  Operators had reported volumes to the nearest millilitre, rather than 0.1ml or 0.01ml as previously.

 

The results were discussed with the team of operators and there was consensus that operators now think about the relevance of the volumes being measured.  We postulate the simple act of thinking about how precise a measurement has subconsciously improved the attention of the operator and lead to more accurate measurements.

 

 Ultrasound Department, Hull Royal Infirmary, Anlaby Road, Hull

A previous audit at this institution demonstrated wide intra-operator and inter-operator variability in measuring the fluid volumes of locally constructed test objects made to mimic the infant and child bladder.  Operators were attempting high precision measurements whilst not assessing the accuracy or relevance of the volumes calculated.  Recommendations were that operators should consider how precise a measurement really needed to be.

 

A second audit was undertaken to evaluate any change in practice.

 

The audit was repeated.  (The construction of the test objects and test methodology will be briefly described.)

 

The measurements taken were evaluated in terms of accuracy of volume measurement and both intra-operator and inter-operator variation.

 

The results demonstrated a reduction in intra-operator and inter-operator variability but also a greater accuracy of measurement despite a reduction in the precision of the measurements.  Operators had reported volumes to the nearest millilitre, rather than 0.1ml or 0.01ml as previously.

 

The results were discussed with the team of operators and there was consensus that operators now think about the relevance of the volumes being measured.  We postulate the simple act of thinking about how precise a measurement has subconsciously improved the attention of the operator and lead to more accurate measurements.