Pedi-R-MAPP | the development, testing, validation, and refinement of a digital nutrition awareness tool.

BACKGROUND & AIMS: The Remote Malnutrition Application (R-MAPP) was developed during the COVID-19 pandemic to provide support for health care professionals (HCPs) working in the community to complete remote nutritional assessments and provide practical guidance for nutritional care. R-MAPP was adapted into Pediatric Remote Malnutrition Application (Pedi-R-MAPP) using a modified Delphi consensus, with the goal of providing a structured approach to completing a nutrition focused assessment as part of a technology enabled care service (TECS) consultation. The aim of this study was to develop and validate a digital version of Pedi-R-MAPP using the IDEAS framework (Integrate, Design, Assess and Share). METHODS: A ten-step process was completed using the IDEAS framework. This involved the four concept processes; Stage-1, Integrate (Step 1-3) identify the problem, specify the goal, and use an evidence-based approach. Stage-2, (Step 4-7) design iteratively and rapidly with user feedback. Stage 3, (Step 8-9) Assess rigorously, and Stage 4 (Step 9-10) publish and launch of the tool. RESULTS: Stage 1:Evidence-based development, Pedi-R-MAPP was developed using Delphi consensus methodology. Stage 2:Iteration & design, HCPs (n = 22) from UK, Europe, South Africa, and North America were involved four workshops to further develop a paper prototype of the tool and complete small-scale testing of a beta version of the tool which resulted in eight iterations. Stage 3:Assess rigorously, Small scale retrospective testing of the tool on children with congenital heart disease (n = 80) was completed by a single researcher, with iterative changes made to improve agreement with summary advice. Large scale testing amongst (n = 745) children in different settings was completed by specialist paediatric dietitians (n = 15) advice who recorded agreement with the summary advice compared with their own clinical assessment. Paediatric dietitians were in overall agreement with the summary advice in the tool 86% (n = 640), compared to their own clinical practice. The main reasons for disagreement were i) frequency of planned review 57.1% (n = 60/105), ii) need for ongoing dietetic review due to chronic condition 20.0% (n = 21/105), iii) disagreement with recommendation for discharge 16.2% (n = 17/105) and iv) concerns with faltering growth and/or need for condition specific growth charts 6.7% (7/105). Iterative changes were made to the algorithm, leading to an improvement in agreement of the summary advice on re-evaluation to 98% (p=<0.0001). CONCLUSION: A digital version of the Pedi-R-MAPP nutrition awareness tool was developed using the IDEAS framework. The summary advice provided by the tool achieved a high level of agreement when compared to paediatric dietetic assessment, by providing a structured approach to completing a remote nutrition focused assessment, along with identifying the frequency of follow-up or an in-person assessment.

Pedi-R-MAPP: The development of a nutritional awareness tool for use in remote paediatric consultations using a modified Delphi consensus.

BACKGROUND & AIMS: The Remote Malnutrition Application (R-MAPP) was developed during the COVID-19 pandemic to provide support for health care professionals (HCPs) working in the community to complete remote nutritional assessments, and provide practical guidance for nutritional care. The aim of this study was to modify the R-MAPP into a version suitable for children, Pediatric Remote Malnutrition Application (Pedi-R-MAPP), and provide a structured approach to completing a nutrition focused assessment as part of a technology enabled care service (TECS) consultation. METHODS: A ten-step process was completed: 1) permission to modify adult R-MAPP, 2) literature search to inform the Pedi-R-MAPP content, 3) Pedi-R-MAPP draft, 4) international survey of HCP practice using TECS, 5) nutrition experts invited to participate in a modified Delphi process, 6) first stakeholder meeting to agree purpose/draft of the tool, 7) round-one online survey, 8) statements with consensus removed from survey, 9) round-two online survey for statements with no consensus and 10) second stakeholder meeting with finalisation of the Pedi-R-MAPP nutrition awareness tool. RESULTS: The international survey completed by 463 HCPs, 55% paediatricians, 38% dietitians, 7% nurses/others. When HCPs were asked to look back over the last 12 months, dietitians (n = 110) reported that 5.7 ± 10.6 out of every 10 appointments were completed in person; compared to paediatricians (n = 182) who reported 7.5 ± 7.0 out of every 10 appointments to be in person (p < 0.0001), with the remainder completed as TECS consultations. Overall, 74 articles were identified and used to develop the Pedi-R-MAPP which included colour-coded advice using a traffic light system; green, amber, red and purple. Eighteen participants agreed to participate in the Delphi consensus and completed both rounds of the modified Delphi survey. Agreement was reached at the first meeting on the purpose and draft sections of the proposed tool. In round-one of the online survey, 86% (n = 89/104) of statements reached consensus, whereas in round-two 12.5% (n = 13/104) of statements reached no consensus. At the second expert meeting, contested statements were discussed until agreement was reached and the Pedi-R-MAPP could be finalised. CONCLUSION: The Pedi-R-MAPP nutrition awareness tool was developed using a modified Delphi consensus. This tool aims to support the technological transformation fast-tracked by the COVID-19 pandemic by providing a structured approach to completing a remote nutrition focused assessment, as well as identifying the frequency of follow up along with those children who may require in-person assessment.

Micronutrient status during paediatric critical illness: A scoping review.

BACKGROUND: No evidence based recommendations for micronutrient requirements during paediatric critical illness are available, other than those arising from recommended nutrient intakes (RNI) for healthy children and expert opinion. OBJECTIVES: The objective of this review is to examine the available evidence from micronutrient status in critically ill children considering studies which describe 1) micronutrient levels, 2) associations between micronutrient levels and clinical outcome, and 3) impact on clinical outcome with micronutrient supplementation during PICU admission. DESIGN: Scoping review. ELIGIBILITY CRITERIA: Any study which used a qualitative and quantitative design considering causes and consequences of micronutrient levels or micronutrient supplementation during paediatric critical illness. SOURCES OF EVIDENCE: NICE Healthcare Databases Advanced Search website (https://hdas.nice.org.uk/) was used as a tool for multiple searches, with a content analysis and charting of data extracted. RESULTS: 711 records were identified, 35 were included in the review. Studies evaluated serum micronutrient status was determined on admission day in majority of patients. A content analysis identified (n = 49) initial codes, (n = 14) sub-categories and (n = 5) overarching themes during critical illness, which were identified as: i) low levels of micronutrients, ii) causes of aberrant micronutrient levels, iii) associations between micronutrients levels and outcome, iv) supplementation of micronutrients. CONCLUSION: During critical illness, micronutrients should be provided in sufficient amounts to meet reference nutrient intakes for age. Although, there is insufficient data to recommend routine supplementations of micronutrients at higher doses during critical illness, the 'absence of evidence should not imply evidence of absence', and well designed prospective studies are urgently needed to elucidate paediatric micronutrient requirements during critical illness. The absence of reliable biomarkers make it challenging to determine whether low serum levels are reflective of a true deficiency or as a result redistribution, particularly during the acute phase of critical illness. As more children continue to survive a PICU admission, particularly those with complex diseases micronutrient supplementation research should also be inclusive of the recovery phase following critical illness.

A systematic review of the applications of artificial intelligence and machine learning in autoimmune diseases.

Autoimmune diseases are chronic, multifactorial conditions. Through machine learning (ML), a branch of the wider field of artificial intelligence, it is possible to extract patterns within patient data, and exploit these patterns to predict patient outcomes for improved clinical management. Here, we surveyed the use of ML methods to address clinical problems in autoimmune disease. A systematic review was conducted using MEDLINE, embase and computers and applied sciences complete databases. Relevant papers included "machine learning" or "artificial intelligence" and the autoimmune diseases search term(s) in their title, abstract or key words. Exclusion criteria: studies not written in English, no real human patient data included, publication prior to 2001, studies that were not peer reviewed, non-autoimmune disease comorbidity research and review papers. 169 (of 702) studies met the criteria for inclusion. Support vector machines and random forests were the most popular ML methods used. ML models using data on multiple sclerosis, rheumatoid arthritis and inflammatory bowel disease were most common. A small proportion of studies (7.7% or 13/169) combined different data types in the modelling process. Cross-validation, combined with a separate testing set for more robust model evaluation occurred in 8.3% of papers (14/169). The field may benefit from adopting a best practice of validation, cross-validation and independent testing of ML models. Many models achieved good predictive results in simple scenarios (e.g. classification of cases and controls). Progression to more complex predictive models may be achievable in future through integration of multiple data types.

GenePy - a score for estimating gene pathogenicity in individuals using next-generation sequencing data.

BACKGROUND: Next-generation sequencing is revolutionising diagnosis and treatment of rare diseases, however its application to understanding common disease aetiology is limited. Rare disease applications binarily attribute genetic change(s) at a single locus to a specific phenotype. In common diseases, where multiple genetic variants within and across genes contribute to disease, binary modelling cannot capture the burden of pathogenicity harboured by an individual across a given gene/pathway. We present GenePy, a novel gene-level scoring system for integration and analysis of next-generation sequencing data on a per-individual basis that transforms NGS data interpretation from variant-level to gene-level. This simple and flexible scoring system is intuitive and amenable to integration for machine learning, network and topological approaches, facilitating the investigation of complex phenotypes. RESULTS: Whole-exome sequencing data from 508 individuals were used to generate GenePy scores. For each variant a score is calculated incorporating: i) population allele frequency estimates; ii) individual zygosity, determined through standard variant calling pipelines and; iii) any user defined deleteriousness metric to inform on functional impact. GenePy then combines scores generated for all variants observed into a single gene score for each individual. We generated a matrix of ~ 14,000 GenePy scores for all individuals for each of sixteen popular deleteriousness metrics. All per-gene scores are corrected for gene length. The majority of genes generate GenePy scores < 0.01 although individuals harbouring multiple rare highly deleterious mutations can accumulate extremely high GenePy scores. In the absence of a comparator metric, we examine GenePy performance in discriminating genes known to be associated with three common, complex diseases. A Mann-Whitney U test conducted on GenePy scores for this positive control gene in cases versus controls demonstrates markedly more significant results (p = 1.37 × 10- 4) compared to the most commonly applied association tool that combines common and rare variation (p = 0.003). CONCLUSIONS: Per-gene per-individual GenePy scores are intuitive when assessing genetic variation in individual patients or comparing scores between groups. GenePy outperforms the currently accepted best practice tools for combining common and rare variation. GenePy scores are suitable for downstream data integration with transcriptomic and proteomic data that also report at the gene level.

Patient, parent and professional perception of the use of maintenance enteral nutrition in Paediatric Crohn's Disease.

AIM: Maintenance enteral nutrition (MEN) is routinely used in Paediatric Crohn's Disease (CD) to prolong remission although there is limited evidence for efficacy and a lack of formal guidelines. This study surveyed patients', parents' and professional experience with MEN. METHODS: Two questionnaires were developed to survey the experience of MEN; (i) Patients/Parents (children >10 years of age aimed to complete independently) and (ii) Dietitians. Questionnaires were sent to families prescribed MEN after exclusive enteral nutrition (EEN) between 2015-17 (n = 77) and dietitians working in paediatric regional centres in UK (n = 23). RESULTS: Response rate to the questionnaires was 53% patients, 62% parents and 83% dietitians. Patients/parents reported medical/dietetic advice to be the primary factor affecting compliance, 30% patients reported side effects. Fifty-six per cent of patients/58% parents stated a preference for dietary advice rather than MEN. Dietetic responses indicated 79% used MEN after EEN as standard procedure and 79% did not have exit criteria for MEN. Sixty-eight per cent perceived the taste was the primary factor affecting patient compliance. CONCLUSION: Patients' perception of the usefulness of MEN differs to professionals. This study highlights the extensive practice of MEN after EEN in clinical remission, which may not be nutritionally indicated. Patient preference is for dietary advice rather than MEN.

Nutritional support in paediatric Crohn's disease: outcome at 12 months.

AIM: Paediatric Crohn's disease (CD) is associated with growth delay and poor nutritional status. Maintenance enteral nutrition (MEN) supplementation is a potential adjunct to improve growth/prolong remission. METHODS: Newly diagnosed CD patients were identified. Anthropometry, treatments and outcomes were collected for 12 months following diagnosis. Data are presented as medians. RESULTS: A total of 102 patients were identified (age = 13 years, 76% male), 58 (57%) completed exclusive enteral nutrition (EEN) as induction therapy, and 77 (75%) entered clinical remission. Following induction, 58 (57%) of all patients continued MEN and 44 (43%) consumed normal diet (ND). BMI Z-score increased (diagnosis-12 months) for EEN (-1.41 to -0.21 (p = <0.0001)) and steroid groups (-0.97 to -0.11 (p = 0.001)). BMI Z-score increased (post induction - 12 months) for MEN (-0.62 to -0.44 (p = 0.04)) but not ND (-0.33 to -0.4 (p = 0.79)). Height Z-score did not increase for any treatment group over 12 months. MEN and ND group relapse rates were similar at six months, MEN = 21/58 (36%); ND = 21/44 (48%) (p = 0.24) and 12 months, MEN = 24/58 (41%); ND = 13/44 (30%) (p = 0.22). Fewer patients treated with EEN then MEN relapsed less than six months, 14 of 43 (33%), compared to patients treated with steroids then ND 16/29 (55%) (p = 0.09). CONCLUSION: BMI Z-score increased but height Z-score remained unchanged over 12 months for the MEN group. Use of MEN was not associated with prolonged time to relapse. Prospective studies are required to examine the utility of MEN.

Classification of Paediatric Inflammatory Bowel Disease using Machine Learning.

Paediatric inflammatory bowel disease (PIBD), comprising Crohn's disease (CD), ulcerative colitis (UC) and inflammatory bowel disease unclassified (IBDU) is a complex and multifactorial condition with increasing incidence. An accurate diagnosis of PIBD is necessary for a prompt and effective treatment. This study utilises machine learning (ML) to classify disease using endoscopic and histological data for 287 children diagnosed with PIBD. Data were used to develop, train, test and validate a ML model to classify disease subtype. Unsupervised models revealed overlap of CD/UC with broad clustering but no clear subtype delineation, whereas hierarchical clustering identified four novel subgroups characterised by differing colonic involvement. Three supervised ML models were developed utilising endoscopic data only, histological only and combined endoscopic/histological data yielding classification accuracy of 71.0%, 76.9% and 82.7% respectively. The optimal combined model was tested on a statistically independent cohort of 48 PIBD patients from the same clinic, accurately classifying 83.3% of patients. This study employs mathematical modelling of endoscopic and histological data to aid diagnostic accuracy. While unsupervised modelling categorises patients into four subgroups, supervised approaches confirm the need of both endoscopic and histological evidence for an accurate diagnosis. Overall, this paper provides a blueprint for ML use with clinical data.

Outcome of nutritional screening in the acute paediatric setting.

INTRODUCTION: Multiple nutritional risk assessment tools are available, but there are limited data on their application in the acute setting. We explored the validity of two tools in a tertiary Children's Hospital's acute unit and the cohort's nutritional status using WHO definitions. METHODS: Prospective study n=300 (median 38 months; 44.6% female; 25.7% ≤12 months). Participants had standard anthropometry measured, all were screened using the Screening Tool for the Assessment of Malnutrition in Paediatrics (STAMP), the Paediatric Malnutrition Screening Tool (PMST) (modified STAMP) and 125 were additionally screened using the Paediatric Yorkhill Malnutrition Screening (PYMS) tool. RESULTS: The percentages with medium/high nutritional risk were as follows: STAMP 73.1%, PMST 79.3% and PYMS 30%. Height/weight were normally distributed with: 3.4% stunted (height-for-age z-score <-2); aged ≤ 5 years, 6.8% wasted (weight-for-height z-score (WHZ) <-2), 17.9% overweight (WHZ 1-2) and 6.2% obese (WHZ >2); aged >5 years, 5.8% thin (body mass index (BMI)-z-score (BAZ) <-2), 17.3% overweight (BAZ 1-2) and 5.8% obese (BAZ >2). The tools showed poor specificity and variable sensitivities when compared with WHO malnutrition criteria, with positive predictive values of <50%. κ-Analysis also showed poor agreement between the tools and the WHO cut-offs. CONCLUSION: These results suggest that nutritional screening tools have poor sensitivity and are difficult to interpret in the acute setting. It may be more effective to include the assessment of weight and height and nutritional intake in the context of the acute presentation as part of routine clinical assessment rather than relying on screening tools to identify those at risk.

Nutritional perspectives of children with Crohn's disease: a single-centre cohort observation of disease activity, energy expenditure and dietary intake.

BACKGROUND/OBJECTIVES: Children with Crohn's disease often demonstrate nutritional recovery during primary therapy at diagnosis, but long-term nutritional support is sometimes necessary. Evidence to inform best nutritional practice including energy and micronutrient requirements is limited. The principal objective of this study was to determine how energy expenditure and physical activity vary with disease activity over the first year following diagnosis. SUBJECTS/METHODS: Twenty children were studied at diagnosis with Crohn's disease and were followed up over 1 year while receiving treatment according to national guidelines. The majority of children (13) were treated with exclusive enteral nutrition. At study visits, height, weight, bioelectrical impedance, resting energy expenditure by indirect calorimetry, tri-axial accelerometer and blood investigations were performed alongside clinical assessment. RESULTS: There was no significant effect of disease activity on resting energy expenditure (REE). Physical activity was greater after primary therapy (Z=3.31, P<0.01). Median wPCDAI fell from 58 at diagnosis to 7.5 after primary therapy and was 7.5 at 1 year. Weight s.d.s increased from -1.67 to -0.86 and lean index s.d.s increased from -2.93 to -1.64, although the increase was mostly in the first 2 months. Median height s.d.s was unchanged throughout this study. There was a significant association between dietary intake and weight gain (r=0.8 P<0.01) but not height gain. Persistent micronutrient deficits beyond diagnosis were seen for both iron and vitamin D. CONCLUSIONS: This study has demonstrated that REE does not change significantly through different phases of disease activity, but physical activity is low at diagnosis. Children with Crohn's disease should be screened for deficiencies of iron and vitamin D.

Practical management of functional abdominal pain in children.

Functional abdominal pain (FAP) is common in childhood, but is not often caused by disease. It is often the impact of the pain rather than the pain itself that results in referral to the clinician. In this review, we will summarise the currently available evidence and discuss the functional dimensions of the presentation, within the framework of commonly expressed parental questions. Using the Rome III criteria, we discuss how to classify the functional symptoms, investigate appropriately, provide reassurance regarding parental worries of chronic disease. We outline how to explain the functional symptoms to parents and an individualised strategy to help restore function.

Presenting phenotype of paediatric inflammatory bowel disease in Wessex, Southern England 2010-2013.

AIM: There has been at least a twofold increase in the incidence of paediatric inflammatory bowel disease (PIBD) over the last 20 years; we report the presenting features from 2010 to 2013 and compare with previous data. METHODS: All patients diagnosed with PIBD at University Hospitals Southampton from 2010 to 2013 were identified from an in-house database. Data were obtained from paper and electronic notes. Height, weight and BMI SDS are presented as median values (95% CI). RESULTS: One hundred and seventy-two patients were included (median age at diagnosis 13.5, 115 male); Crohn's disease (CD) - 107, UC - 50, inflammatory bowel disease unclassified (IBDU) - 15. The most common presenting features of CD were abdominal pain (86%), diarrhoea (78.5%) and weight loss (56.1%); 42.1% of patients had all three. In UC blood in stool (92%), diarrhoea (92%) and abdominal pain (88%) were the most common; all three in 76% of patients. CD presented with ileocolonic disease in 52.5%. UC presented with pancolitis in 64%. There was growth delay in CD: height -0.37 (-0.60 to -0.14); weight -1.09 (-1.35 to -0.83). Growth was maintained in UC: height 0.53 (0.19 to 0.87); weight 0.14 (-0.20 to 0.48). CONCLUSION: Paediatric inflammatory bowel disease phenotype remains as extensive despite increasing incidence. Although the classical phenotype is common, a reasonable proportion present with atypical features, normal growth and normal blood markers.

An unlikely cause of severe malnutrition in a 3-year-old girl with previous gastroschisis.

A 3-year-old girl with previous gastroschisis associated with jejunal and colonic atresia presented with severe oedema, abdominal pain and diarrhoea. Clinically she was malnourished. Serum albumin and concentrations of micronutrients were low. A barium meal examination showed jejunal dilation. A stricture was suspected and the patient was taken to theatre where an 8 cm length of bowel was resected including a jejunal stricture at the point of previous atresia repair. Inside the proximal dilated jejunum was a large trichobezoar (hairball), thought to be acting as a ball valve inside the bowel. This girl made a rapid recovery after surgery. Her nutritional state improved, symptoms resolved and serum biochemistry normalised. She remains well at follow-up with normal blood results, normal albumin and no diarrhoea.

Rising incidence of paediatric inflammatory bowel disease (PIBD) in Wessex, Southern England.

BACKGROUND: There has been a significant increase in the incidence of paediatric inflammatory bowel disease (PIBD) over the last 25 years although there is no recent data from England. We aimed to analyse changes in incidence within a defined English population over the last decade and compare this to recent and historical incidence data from comparable studies. METHODS: The new diagnosis incidence of PIBD (age less than or equal to 16 years) was recorded from a prospective database for a geographically defined area within Southern England (2002-2012). Data were analysed for two separate time periods (cohort 1:2002-2006 and cohort 2:2008-2012) and compared to data from the British Paediatric Surveillance Unit (BPSU) survey in 1998/1999. Data were analysed by age, sex and disease type. RESULTS: There has been an increase in incidence of PIBD from 6.39/100,000/year during cohort 1 to 9.37/100,000/year during cohort 2 (p=0.0002). This compares with the BPSU incidence data in England (1998-1999) of 5.2/100,000/year. There was no statistically significant difference in median age of diagnosis between cohorts (p=0.46). The incidence of Crohn's disease (CD) was 3.8/100,000/year in cohort 1 rising to 5.85/100,000/year in cohort 2 (p=0.001). The incidence of ulcerative colitis (UC) was 2.01/100,000/year in cohort 1 rising to 2.62/100,000/year in cohort 2 (p=0.1458). Overall PIBD incidence is higher in males in cohort 1 (male-to-female ratio 1.35:1) and cohort 2 (male-to-female ratio 1.5:1). CONCLUSIONS: The incidence of PIBD continues to increase with a rise of almost 50% in the last decade in Southern England. The reasons for this increase remain unclear.

Surgery for children with Crohn's disease: indications, complications and outcome.

BACKGROUND: Outcome data for surgery in paediatric Crohn's disease are limited. We report 10 years' experience at a regional paediatric gastroenterology centre. METHODS: Children undergoing surgery for Crohn's disease (January 2002-December 2012) were identified from an onsite patient register. Patients were followed until transition to adult services. Data were obtained from medical records and are expressed as median (range). RESULTS: Sixty-nine children, aged 13.8 years (6.3-17.0) at diagnosis, were included. 42 were male (61%). Follow-up was 1.8 years (27 days-6 years). Surgery followed diagnosis by 9 months (0 days-7 years). 52 children (75%) received thiopurines and 5 (7%) anti-TNF (tumour necrosis factor) therapy preoperatively. 58 (84%) underwent intra-abdominal surgery (40 right hemicolectomy, 8 stoma formation, 2 subtotal colectomy, 2 small bowel resection, 6 other) and 10 (14%) underwent surgery for perianal disease. The commonest indications for intra-abdominal surgery were stricturing disease 35 (60%) and unresponsive luminal disease 13 (22%). There were 13 (22%) early, and 5 (8.6%) late, complications following intra-abdominal surgery. Nine children had disease relapse, five required further surgery. Height SD scores (SDS) did not increase between diagnosis; -0.5 (-3.4-2.1) and most recent follow up; -0.4 (-3.0-1.1). Body Mass Index (BMI) SDS increased from -1.0 (-6.3-1.5) to -0.3 (-3.3-2.0) (p<0.05). CONCLUSIONS: Surgery was associated with a 22% early complication rate and a 15% risk of relapse. 21% of patients required a second unplanned intra-abdominal procedure. Surgical intervention was associated with an increase in BMI SDS, but not in height SDS.