Evaluation of data imputation strategies in complex, deeply-phenotyped data sets: the case of the EU-AIMS Longitudinal European Autism Project

An increasing number of large-scale multi-modal research initiatives has been conducted in the typically developing population, e.g. Dev. Cogn. Neur. 32:43-54, 2018; PLoS Med. 12(3):e1001779, 2015; Elam and Van Essen, Enc. Comp. Neur., 2013, as well as in psychiatric cohorts, e.g. Trans. Psych. 10(1):100, 2020; Mol. Psych. 19:659–667, 2014; Mol. Aut. 8:24, 2017; Eur. Child and Adol. Psych. 24(3):265–281, 2015. Missing data is a common problem in such datasets due to the difficulty of assessing multiple measures on a large number of participants. The consequences of missing data accumulate when researchers aim to integrate relationships across multiple measures. Here we aim to evaluate different imputation strategies to fill in missing values in clinical data from a large (total N = 764) and deeply phenotyped (i.e. range of clinical and cognitive instruments administered) sample of N = 453 autistic individuals and N = 311 control individuals recruited as part of the EU-AIMS Longitudinal European Autism Project (LEAP) consortium. In particular, we consider a total of 160 clinical measures divided in 15 overlapping subsets of participants. We use two simple but common univariate strategies—mean and median imputation—as well as a Round Robin regression approach involving four independent multivariate regression models including Bayesian Ridge regression, as well as several non-linear models: Decision Trees (Extra Trees., and Nearest Neighbours regression. We evaluate the models using the traditional mean square error towards removed available data, and also consider the Kullback–Leibler divergence between the observed and the imputed distributions. We show that all of the multivariate approaches tested provide a substantial improvement compared to typical univariate approaches. Further, our analyses reveal that across all 15 data-subsets tested, an Extra Trees regression approach provided the best global results. This not only allows the selection of a unique model to impute missing data for the LEAP project and delivers a fixed set of imputed clinical data to be used by researchers working with the LEAP dataset in the future, but provides more general guidelines for data imputation in large scale epidemiological studies.

Utility of 3D printed models as adjunct in acetabular fracture teaching for Orthopaedic trainees

OBJECTIVE

To evaluate the use of 3-D printed models as compared to didactic lectures in the teaching of acetabular fractures for Orthopaedic trainees.

METHODS

This was a randomised prospective study conducted in a tertiary hospital setting which consisted of 16 Orthopaedic residents. Ten different cases of acetabular fracture patterns were identified and printed as 3-D models. The baseline knowledge of orthopaedic residents regarding acetabular fracture classification and surgical approach was determined by an x-ray based pre-test. Trainees were then randomly assigned into two groups. Group I received only lectures. Group II were additionally provided with 3-D printed models during the lecture. Participants were then assessed for comprehension and retention of teaching.

RESULTS

Sixteen trainees participated in the trial. Both Group 1 and 2 improved post teaching with a mean score of 2.5 and 1.9 to 4.4 and 6 out of 10 respectively. The post test score for fracture classification and surgical approach were significantly higher for 3-D model group (p < 0.05). Trainees felt that the physical characteristics of the 3-D models were a good representation of acetabular fracture configuration, and should be used routinely for teaching and surgical planning.

CONCLUSION

3-D printed model of real clinical cases have significant educational impact compared to lecture-based learning towards improving young trainees' understanding of complex acetabular fractures.

Care pathway for autistic children and their families in Europe

Introduction

Autism is a lifelong complex neurodevelopmental condition that affects brain development and behaviour with significant consequences for everyday life (WHO, 2018). Despite its personal, familial and societal impact, there is still a European-wide lack of harmonised guidelines about the support needed from early stages, the most sensitive time to gain positive future outcomes (Berajamo-Martin et al, 2019).

Objectives

The objectives were: 1. To analyse autistic children care pathway and patient/carer journey in three European countries: Italy, Spain and U.K. 2. To propose policy recommendations on how to improve this pathway.

Methods

To identify major barriers and treatment gaps, we conducted a rapid literature review of the care pathway in Europe and a survey aimed at parents or carers of autistic children ages 0 to 18 living in the three countries. The survey gathered information on screening, diagnosis, accessibility and support received before, during and after diagnosis. Members of the working group met to discuss results and propose policy recommendations.

Results

1. Current care pathway analysis showed the following treatment gaps: Long waiting time from first concerns until screening visit and confirmed diagnosis. Delayed or no access to intervention once diagnosis has been confirmed. Overall limited information about autism and how to access early detection services. Overall deficient support to families. 2. Please see Box 1 for our proposed policy recommendations.

Box 1. Policy recommendations

Conclusions

Our findings and recommendations will inform policy harmonisation in Europe to shorten long waiting times, diagnosis process and intervention, and therefore, improve autistic people and their families’ journey experience and quality of life.

Disclosure

No significant relationships.

Perceived symptoms as the primary indicators for 30-day heart failure readmission

Background

To identify 30-day rehospitalizations in patients discharged with heart failure (HF) based on clinical indications, physiologic measures and symptoms.

Methods

Fifty-six patients with heart failure participated. After discharge to home, clinical indicators of dyspnea, fatigue, orthopnea, dyspnea with exertion, daily weight, edema, heart rate, blood pressure, mental condition, medication adherence, and overall well-being were reported by participants daily for up to 30 days.

Results

Joint modeling of longitudinal and time-to-event approach was applied to assess the association of readmission with longitudinal measurements. There was no association between demographic, physiological, or laboratory variables and re-hospitalization within 30 days post discharge. Perceptions of dyspnea (p = .012) and feeling unwell (p < .001) were associated with rehospitalization. Patients struggling to breath were 10.7 times more likely to be readmitted than those not struggling to breath.

Conclusion

Perceived measures, particularly dyspnea and feeling unwell were more important factors than demographic, physiological, or laboratory parameters in predicting 30-day rehospitalizations in this racially diverse cohort. The symptomatic experience of heart failure is an important indicator of rehospitalization.

Case Study: Soft Tissue Infection with Raoultella ornithinolytica

Raoultella ornithinolytica is a rare encapsulated Gram-negative aerobic and facultative anaerobic rod belonging to the Enterobacteriaceae family. It tends to inhabit water and soil environments and can be found on insects, fish, ticks, and termites, but can also found in the hospital environment.^1,2R ornithinolytica has been documented in respiratory, urinary, gastrointestinal, and biliary tract infections as well as bacteraemia and systemic infections but has rarely been documented in soft tissue infections.² This case study describes a recurrent paronychia infection secondary to R ornithinolytica in a young woman not responding to antibiotics and successfully treated with surgical management.

Analysis of an adaptive lead weighted ResNet for multiclass classification of 12-lead ECGs

Background. Twelve lead ECGs are a core diagnostic tool for cardiovascular diseases. Here, we describe and analyse an ensemble deep neural network architecture to classify 24 cardiac abnormalities from 12 lead ECGs.Method. We proposed a squeeze and excite ResNet to automatically learn deep features from 12-lead ECGs, in order to identify 24 cardiac conditions. The deep features were augmented with age and gender features in the final fully connected layers. Output thresholds for each class were set using a constrained grid search. To determine why the model made incorrect predictions, two expert clinicians independently interpreted a random set of 100 misclassified ECGs concerning left axis deviation.Results. Using the bespoke weighted accuracy metric, we achieved a 5-fold cross-validation score of 0.684, and sensitivity and specificity of 0.758 and 0.969, respectively. We scored 0.520 on the full test data, and ranked 2nd out of 41 in the official challenge rankings. On a random set of misclassified ECGs, agreement between two clinicians and training labels was poor (clinician 1:κ= -0.057, clinician 2:κ= -0.159). In contrast, agreement between the clinicians was very high (κ= 0.92).Discussion. The proposed prediction model performed well on the validation and hidden test data in comparison to models trained on the same data. We also discovered considerable inconsistency in training labels, which is likely to hinder development of more accurate models.

Outcomes in Nonagenarians with Hip Fractures Treated Conservatively and Surgically

Introduction

We aimed to assess the clinical outcomes in nonagenarians following a hip fracture. We also further investigated the factors that influence these outcomes, such as method of treatment (operative versus conservative), co-morbidities, and pre-morbid function.

Materials and methods

We studied 65 nonagenarians that were identifiable from our hospital hip fracture database. We reviewed various parameters of these patients admitted after sustaining a hip fracture (neck of femur or intertrochanteric) and investigated how these parameters affected patient outcomes. The main outcomes studied were: inpatient morbidity, and mortality at one year.

Results

Inpatient morbidity was more likely in patients with an ASA grade of 3 to 5. Urinary tract infection was the most common medical complication. The 1-year mortality was 15.4% and was significantly influenced by advancing age. Surgically managed patients had a 1-year mortality rate (14.3%) slightly less than non-operative patients (17.4%). Post injury mobility was significantly better in those who received operative treatment with 63% of surgical cases regaining ambulatory status versus 7% of conservatively managed patients.

Conclusions

We presented the outcomes of hip fractures in an extreme age group in the population. In nonagenarians with hip fractures surgery was associated with a 1-year mortality rate of 14.3% which is comparable to the general hip fracture population and less than the mortality rate of conservatively managed patients (17.4%). The primary advantage of surgery would be that two-thirds of patients return to ambulatory status. This information is useful to counsel patients and their families especially since the elderly are often more fearful of surgical intervention.

Preference for biological motion is reduced in ASD: implications for clinical trials and the search for biomarkers

BACKGROUND

The neurocognitive mechanisms underlying autism spectrum disorder (ASD) remain unclear. Progress has been largely hampered by small sample sizes, variable age ranges and resulting inconsistent findings. There is a pressing need for large definitive studies to delineate the nature and extent of key case/control differences to direct research towards fruitful areas for future investigation. Here we focus on perception of biological motion, a promising index of social brain function which may be altered in ASD. In a large sample ranging from childhood to adulthood, we assess whether biological motion preference differs in ASD compared to neurotypical participants (NT), how differences are modulated by age and sex and whether they are associated with dimensional variation in concurrent or later symptomatology.

METHODS

Eye-tracking data were collected from 486 6-to-30-year-old autistic (N = 282) and non-autistic control (N = 204) participants whilst they viewed 28 trials pairing biological (BM) and control (non-biological, CTRL) motion. Preference for the biological motion stimulus was calculated as (1) proportion looking time difference (BM-CTRL) and (2) peak look duration difference (BM-CTRL).

RESULTS

The ASD group showed a present but weaker preference for biological motion than the NT group. The nature of the control stimulus modulated preference for biological motion in both groups. Biological motion preference did not vary with age, gender, or concurrent or prospective social communicative skill within the ASD group, although a lack of clear preference for either stimulus was associated with higher social-communicative symptoms at baseline.

LIMITATIONS

The paired visual preference we used may underestimate preference for a stimulus in younger and lower IQ individuals. Our ASD group had a lower average IQ by approximately seven points. 18% of our sample was not analysed for various technical and behavioural reasons.

CONCLUSIONS

Biological motion preference elicits small-to-medium-sized case-control effects, but individual differences do not strongly relate to core social autism associated symptomatology. We interpret this as an autistic difference (as opposed to a deficit) likely manifest in social brain regions. The extent to which this is an innate difference present from birth and central to the autistic phenotype, or the consequence of a life lived with ASD, is unclear.

Evaluation of the impact of human immunodeficiency virus pre-exposure prophylaxis on new human immunodeficiency virus diagnoses during the COVID-19 pandemic

AIMS

The national PrEP programme launched in Ireland in November 2019 with tenofovir/emtricitabine free to those meeting eligibility criteria. We assessed the impact of the first year of the PrEP programme on new HIV diagnoses in the largest sexual health and HIV service in Ireland.

METHODS

A free PrEP service was established in November 2019. We reviewed the number of new diagnoses of HIV between November 2018-2019, before the introduction of the national PrEP programme and compared this with the number of new HIV diagnosis between November 2019-2020.

RESULTS

There were 95 new HIV diagnoses (63.3% MSM) between November 2018 and 2019 and 73 new HIV diagnoses (65.7% MSM) between November 2019 and 2020. There was a statistically significant decline in new HIV diagnoses between the 2 years (P = 0.0003). 546 patients were prescribed PrEP as of December 2020.106 patients (19.4%) changed their PrEP dosing regimen due to lockdown. 178 individuals (32.6%) had a rectal infection diagnosed.

CONCLUSION

There has been a reduction in new HIV diagnoses in our cohort (although this has occurred during a global pandemic). It is too early to say if PrEP reduces late presentations of HIV based on our findings. A significant number of rectal infections were identified in the PrEP clinic suggesting ongoing risk despite pandemic restrictions. Further research into sexual practices during COVID-19 is needed to assess if this had an impact on the lower rates of HIV acquisition.

160 TIME TO BIN ‘NEXT OF KIN’?—JUNIOR DOCTORS DEMONSTRATE POOR UNDERSTANDING OF THE TERM IN IRELAND

Background

‘Next of kin’ (NOK) is a term widely used in the healthcare setting in Ireland to indicate an individual that can be contacted in the event of an emergency when a patient is in hospital. While it has its origins in Irish inheritance law, the term confers no legal or decision-making authority to any other individuals on behalf of a patient in hospital. However, anecdotally, the term is commonly misunderstood with a false belief among healthcare staff that consent should be sought from the NOK when a patient is unable to do so. In this study we sought to assess current understanding of the term among Irish non-consultant hospital doctors (NCHDs).

Methods

Data was collected via an online survey, completed by NCHDs across a number of medical and surgical disciplines to investigate their understanding of the term. Descriptive analysis of data was performed in Excel.

Results

118 NCHDs completed the survey. While 110 (93.2%) believed a NOK was an emergency contact, 53 (44.9%) believed they assisted in decision making; 58 (49.2%) thought they could provide medical information about the patient. Forty-four doctors (37.3%) believed a NOK was permitted to make medical decisions while 59 (50%) felt they could provide consent on behalf of an incapacitated patient. Sixty-three (53.4%) believed a NOK had legal entitlement to information over and above that of another family member or friend.

Conclusion

This study highlights several misconceptions surrounding the term ‘next of kin’ in the Irish healthcare setting, particularly around the area of decision-making authority and consent. The term is confusing and unhelpful and should not be used in healthcare. ‘Contact person’ or ‘emergency contact’ are appropriate alternatives. Further education to NCHDs around consent and maintenance of patient autonomy is necessary.

Magnetic resonance imaging in the evaluation of idiopathic frequent premature ventricular complexes with normal ventricular function

Background

The role of cardiac magnetic resonance (CMR) imaging in the diagnostic evaluation of patients with frequent premature ventricular complexes (PVCs) and normal left ventricular ejection fraction (LVEF) remains unclear. Existing data has been predominantly derived from highly selected populations, that may overestimate the true prevalence of abnormal findings on CMR in this patient population.

Purpose

The aim of this study was 2-fold: 1) to establish the prevalence of CMR imaging abnormalities in a cohort with normal LVEF and high PVC burden; 2) to identify predictors of CMR imaging abnormalities in patients with frequent PVCs and normal LVEF.

Methods

In this cohort study, 211 patients (age 53.2±19 years; 41% male) with frequent PVCs (≥5%/24 h), of normal LVEF (≥50% by echocardiography) and no known underlying structural heart disease were prospectively enrolled from 2016–2020. Of these, 166 (79%) patients were symptomatic from their PVCs in the form of palpitations, fatigue, chest pain, dizziness, and/or dyspnea. Patients underwent CMR imaging (1.5 Tesla) with a late gadolinium enhancement (LGE) protocol at the time of enrollment for the detection of scarring and/or fibrosis.

Results

Patients had a median baseline echocardiographic LVEF of 60% (± 5%) with 195 (92%) of patients having a normal native QRS morphology. Median PVC burden of the study cohort was 16% (± 14%). CMR LGE abnormalities were found in 19 (9%) patients including 17 scans with non-ischemic LGE and 2 with ischemic LGE. Age &gt;60 (odds ratio [OR]: 3.20, 95% confidence interval [CI]: 1.20–8.51, p=0.020), male sex (OR: 4.67, 95% CI: 1.61–13.50, p=0.004), history of hypertension (OR: 3.43, 95% CI: 1.31–8.97, p=0.012), native QRS duration (OR: 1.03, 95% CI: 1.00–1.05, p=0.031), and history of non-sustained ventricular tachycardia (OR: 2.81, 95% CI: 1.03–7.68, p=0.044) were significantly associated with the presence of imaging abnormalities on CMR. Dominant PVC origin from the left ventricle had a positive trend (OR: 2.60, 95% CI: 0.99–7.66, p=0.083) to association with CMR imaging abnormalities. On multivariate analysis, male sex (OR: 4.10, 95% CI: 1.40–12.04, p=0.010) and history of hypertension (OR: 2.89, 95% CI: 1.08–7.75, p=0.035) remained significantly associated with the presence of CMR abnormalities. There was no association between CMR imaging abnormalities and the burden of PVCs or the number of PVC morphologies.

Conclusion

In this cohort, only 9% of patients with apparently idiopathic frequent PVCs and normal LVEF had concealed myocardial abnormalities on CMR imaging. Male sex and history of hypertension were associated with a higher rate of CMR abnormalities.

Funding Acknowledgement

Type of funding sources: None.

FFRCT and Invasive Coronary Angiography – assessing concordance in an unselected UK real world population

Background/Introduction

CT coronary angiography (CTCA) Fractional Flow Reserve (FFRCT) is a key investigation in chronic coronary syndrome (CCS) guidelines. FFR calculated from CTCA may help improve CTCA specificity for flow limiting disease, preventing unnecessary invasive coronary angiography (ICA).

Purpose

To (1) assess the treatment planning potential of FFRCT by determining the concordance of FFRCT with an ICA +/− invasive physiological assessment for the detection of flow limiting CAD in a real world NHS setting and (2) concordance sub-analysis of concordance of invasive iFR and FFRCT.

Methods

A single-centre retrospective analysis of a prospectively maintained clinical CTCA database. We identified patients with CCS who had CTCA FFRCT and subsequent ICA from August 2018 to January 2021. Concordance was assessed on a per patient and per vessel basis (major epicardial vessels: left main stem [LMS], left anterior descending artery [LAD], circumflex [LCx], right coronary artery [RCA]). Two non-invasive ischaemia thresholds were examined an FFRCT ≤0.8 and &lt;0.75 (2 cm distal to stenosis). This was compared to a clinical ICA assessment where a flow limiting lesion was defined as a stenosis &gt;70% (or &gt;50% in the LMS) and/or an end vessel iFR ≤0.89 / FFR ≤0.8. All vessels that underwent an iFR, at the interventionists discretion, were then assessed relative to their end-vessel FFRCT.

Results

565 patients had a CTCA with FFRCT and 164 patients had a subsequent ICA and were suitable for analysis. On a per patient basis 69% of those referred to ICA with an FFRCT ≤0.8 of at least one major epicardial vessel had any flow limiting CAD at ICA. With an FFRCT &lt;0.75 this was 73%. Table 1 illustrates the per vessel concordance.

A total of 120 vessels were included in the iFR subsection analysis. The mean FFRCT was 0.71 (±0.13) and mean iFR was 0.89 (±0.1). Accuracy was 54% (95% CI 45–63%) with a sensitivity of 89% (95% CI 76–96%), specificity 32% (95% CI 22–44%), positive predictive value 45% (95% CI 40–50%) and negative predictive value of 83% (95% CI 66–92%). A Pearson's correlation coefficient of 0.23 was found.

Conclusion

This study demonstrated that the negative predictive value of FFRCT was excellent, including importantly for LMS analysis. The specificity on a per vessel basis was good with the exception of the LAD assessment. This may have implications for interventional planning with this imaging modality. FFRCT correlated poorly with invasive iFR in this subsection analysis although selection bias may be contributing. There remains a significant proportion of patients referred for an ICA where no flow limiting CAD is found.

Funding Acknowledgement

Type of funding sources: None. Table 1. Per vessel concordance analysis of potential flow limiting FFRCT relative to clinical ICA findings. Analysed with an ischaemia threshold of FFRCT ≤0.80 and &lt;0.75. PPV = Positive Predictive Value. NPV = Negative Predictive Value.

Optimal reconstructive strategies in the setting of post-mastectomy radiotherapy - A systematic review and network meta-analysis

BACKGROUND

A third of breast cancer patients require mastectomy. In some high-risk cases postmastectomy radiotherapy (PMRT) is indicated, threatening reconstructive complications. Several PMRT and reconstruction combinations are used. Autologous flap (AF) reconstruction may be immediate (AF→PMRT), delayed-immediate with tissue expander (TE [TE→PMRT→AF]) or delayed (PMRT→AF). Implant-based breast reconstruction (IBBR) includes immediate TE followed by PMRT and conversion to permanent implant (PI [TE→PMRT→PI]), delayed TE insertion (PMRT→TE→PI), and prosthetic implant conversion prior to PMRT (TE→PI→PMRT).

AIM

Perform a network metanalysis (NMA) assessing optimal sequencing of PMRT and reconstructive type.

METHODS

A systematic review and NMA was performed according to PRISMA-NMA guidelines. NMA was conducted using R packages netmeta and Shiny.

RESULTS

16 studies from 4182 identified, involving 2322 reconstructions over three decades, met predefined inclusion criteria. Studies demonstrated moderate heterogeneity. Multiple comparisons combining direct and indirect evidence established AF-PMRT as the optimal approach to avoid reconstructive failure, compared with IBBR strategies (versus PMRT→TE→PI; OR [odds ratio] 0.10, CrI [95% credible interval] 0.02 to 0.55; versus TE→PMRT→PI; OR 0.13, CrI 0.02 to 0.75; versus TE→PI→PMRT OR 0.24, CrI 0.05 to 1.05). PMRT→AF best avoided infection, demonstrating significant improvement versus PMRT→TE→PI alone (OR 0.12, CrI 0.02 to 0.88). Subgroup analysis of IBBR found TE→PI→PMRT reduced failure rates (OR 0.35, CrI 0.15-0.81) compared to other IBBR strategies but increased capsular contracture.

CONCLUSION

Immediate AF reconstruction is associated with reduced failure in the setting of PMRT. However, optimal reconstructive strategy depends on patient, surgeon and institutional factors. If IBBR is chosen, complication rates decrease if performed prior to PMRT.

PROSPERO REGISTRATION

CRD 42020157077.

AB0913-PARE METHOTREXATE FOR RHEUMATOID ARTHRITIS: PATIENT PERSPECTIVES ON MONITORING IN PRIMARY CARE

Background:

Monotherapy with methotrexate (MTX) is one first-line option for newly diagnosed Rheumatoid Arthritis (RA)1. Although treatment is usually commenced by a specialist, repeat prescribing and monitoring responsibilities often lie with primary care. While inadequate monitoring is a safety concern, unnecessary duplication of monitoring invariably has cost implications for General Practice (GP).

Objectives:

To ensure that patients in one South-West GP practice who are taking oral MTX for RA are appropriately monitored in line with current guidance from the British Society of Rheumatology (BSR). Current guidance recommends that once a stable MTX dose is maintained for six weeks, followed by monthly bloods for the next three months, that at least twelve weekly blood monitoring is sufficient from then on1.

Methods:

A randomised sample of 50 patients registered in one South-West GP practice that were taking a stable dose oral MTX for RA; for at least 6 months; was collected. The length of time between the patient’s two most recent blood tests was recorded. A random selection of ten patients with more frequent monitoring than BSR guidance suggest were asked a series of questions to determine patient perspectives on reasons for monitoring frequency.

Results:

58% of patients had more frequent monitoring bloods than current BSR guidelines recommend. Within this group, the mode frequency of monitoring was four weeks, in line with the previous National guidance which was superseded in 20172.The most common themes in patient’s perspectives on monitoring frequency were:

a.Previous abnormal blood results requiring close monitoring

b.Multiple disease modifying anti-rheumatoid drug (DMARD) regimes that included MTX

c.Patient preference

d.Unclear or unknown reason

The purported increased monitoring for multiple DMARD regimes were not in line with current national guidelines1.

Conclusion:

This audit demonstrates that over half of patients taking MTX for RA in one GP practice are having more frequent blood monitoring than BSR guidance suggests is necessary. Furthermore, it demonstrates that patient’s understanding of their perceived need for increased monitoring is largely inaccurate or unclear to them. There is a clear deficit in patient education surrounding monitoring frequency which must be addressed in order to empower patients, as well as reducing unnecessary duplication of blood tests. As a result of this audit, patient education on MTX monitoring frequency was formally introduced as part of the RA annual patient review. An information sheet with these monitoring requirements was produced to aid practitioners in the education process.

References:

[1]J, Ledingham et al | BSR/ BHPR guideline for the prescription and monitoring of non-biologic disease-modifying anti-rheumatic drugs | Rheumatology 2017; 56: 865-8.

[2]Chakravarty K etc al | BSR/BHPR guideline for disease-modifying anti-rheumatic drug (DMARD) therapy in consultation with the British Association of Dermatologists | Rheumatology 2008; 47: 924–5.

Disclosure of Interests:

None declared