ESPE2024 Poster Category 1 Adrenals and HPA Axis 1 (9 abstracts)
Impact of Fludrocortisone on Blood Pressure in Young Children Under 5 with CAH in the I-CAH Registry: Autoregressive Causal Inference Analysis
Neil R. Lawrence 1 , Irina Bacila 1 , Joseph Tonge 1 , Jeremy Dawson 1 , Gary S. Collins 2 , Zi-Qiang Lang 1 , Jillian Bryce 3 , Malika Alimussina 3 , Minglu Chen 3 , Salma R. Ali 4,3 , Safwaan Adam 5 , Erica L.T. van den Akker 6 , Tânia A.S.S. Bachega 7 , Federico Baronio 8 , Niels H. Birkebæk 9 , Walter Bonfig 10,11 , Hedi Claahsen–van der Grinten 12,13 , Martine Cools 14 , Eduardo C. Costa 15 , Miguel Debono 16 , Liat de Vries 17,18 , Christa Flück 19 , Gabriella Gazdagh 20 , Ayla Güven 21 , Sabine Hannema 22 , Violeta Iotova 23 , Hetty J. van der Kamp 24 , Ruth Krone 25 , Sofia Leka-Emiri 26 , María Clemente-León 27 , Corina R. Lichiardopol 28 , Renata L. Markosyan 29 , Tatjana Milenkovic 30 , Mirela C. Miranda 7 , Uta Neumann 31 , John Newell-Price 1,32 , Şükran Poyrazoğlu 33 , Ursina Probst-Scheidegger 34 , Rodolfo Rey 35,36 , Gianni Russo 37 , Luisa De Sanctis 38,39 , Sumudu N. Seneviratne 40 , Marianna R. Stancampiano 37 , Rieko Tadokoro-Cuccaro 41 , Ajay Thankamony 41 , Ana Vieites 35 , Malgorzata Wasniewska 42 , Diego Yeste 43 , Jeremy Tomlinson 2 , S. Faisal Ahmed 3 & Nils Krone 1,44
1University of Sheffield, Sheffield, United Kingdom. 2University of Oxford, Oxford, United Kingdom. 3Royal Hospital for Children & Queen Elizabeth University Hospital, Glasgow, United Kingdom. 4University of Glasgow, Glasgow, United Kingdom. 5The Christie Hospital, Manchester, United Kingdom. 6Sophia Children’s Hospital, Rotterdam, Netherlands. 7University of Sao Paulo, Sao Paulo, Brazil. 8IRCCS AOUBO, Bologna, Italy. 9Aarhus University Hospital, Aarhus, Denmark. 10Technical University Munich, Munich, Germany. 11Klinikum Wels-Grieskirchen, Wels, Austria. 12Radboud University Medical Centre, Nijmegen, Netherlands. 13Amalia Children’s Hospital, Nijmegen, Netherlands. 14Ghent University and Ghent University Hospital, Ghent, Belgium. 15Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil. 16, Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom. 17Schneider's Children Medical Center of Israel, Petah-Tikvah, Israel. 18Tel-Aviv University, Tel-Aviv, Israel. 19Medizinische Universitätskinderklink, Bern, Switzerland. 20Southampton General Hospital, Southampton, United Kingdom. 21Baskent University Istanbul Hospital, Istanbul, Turkey. 22Sophia Children's Hospital, Rotterdam, Netherlands. 23Medical University of Varna, Varna, Bulgaria. 24University Medical Centre Utrecht, Utrecht, Netherlands. 25Birmingham Women's & Children's Hospital, Birmingham, United Kingdom. 26"P.& A. KYRIAKOU" Children's Hospital, Athens, Greece. 27University Hospital Vall d'Hebron, Barcelona, Spain. 28University of Medicine and Pharmacy Craiova, Craiova, Romania. 29Yerevan State Medical University, Yerevan, Armenia. 30Institute for Mother and Child Healthcare of Serbia “Dr Vukan Čupić”, Belgrade, Serbia. 31Charite - Universitätsmedizin, Berlin, Germany. 32Sheffield Teaching Hospitals NHS Foundation Trust, Sheffield, United Kingdom. 33Istanbul University, Istanbul, Turkey. 34Kantonsspital Winterthur, Winterthur, Switzerland. 35Hospital de Niños Ricardo Gutiérrez, Buenos Aires, Argentina. 36University of Buenos Aires, Buenos Aires, Argentina. 37Scientific Institute San Raffaele, Milan, Italy. 38Regina Margherita Children's Hospital, Torino, Italy. 39University of Torino, Torino, Italy. 40University of Colombo, Colombo, Sri Lanka. 41University of Cambridge, Cambridge, United Kingdom. 42University of Messina, Messina, Italy. 43Hospital Universitario Vall d'Hebron, Barcelona, Spain. 44Sheffield Children's Hospital NHS Foundation Trust, Sheffield, United Kingdom
Background: The International Congenital Adrenal Hyperplasia (CAH) Registry provides rich longitudinal data from CAH patients with 21 Hydroxylase deficiency. We estimated the average total causal effect of fludrocortisone on blood pressure in patients with CAH by modelling within a causal inference framework.
Methods: We constructed a directed acyclic graph (DAG) with domain experts (using Dagitty.net) to establish appropriate covariate adjustment. We used a linear mixed effects model (LMEM) to report average blood pressure and assess a change with age, and autoregressive linear regression (ARX) to quantify the causal effect of absolute fludrocortisone dose on blood pressure. Missing data was addressed using spline interpolation between longitudinal points for height and weight, last observation carried forward or next observation carried backward for dosing, and joint modelling multilevel multiple imputation for blood pressure.
Results: The DAG highlighted sex, age, height, weight, hydrocortisone dose and fludrocortisone dose as confounding variables requiring appropriate covariate adjustment. Sodium, potassium and renin were highlighted as mediating variables and not adjusted to avoid collider bias. A total of 358 patients (177 male and 181 female) with 2981 clinic visits from 22 centres under the age of 5 years were available for modelling. Complete case analysis allowed modelling of 20% of visits, increasing to 92% when appropriate imputation was employed. LMEM average BP was 103/63 not varying with age, and fludrocortisone dose largely consistent with international recommendations (median 124 micrograms, Q1-Q3 100-200). ARX multivariable modelling with a patient level random effect showed an absolute dose of 100 micrograms fludrocortisone increased systolic blood pressure by 1.3mmHg (95% confidence interval 0.6-2.1mmHg) and diastolic blood pressure by 0.6mmHg (0.1-1.1mmHg), with an interaction term showing an insignificant effect of sex (P >0.05, R 2 marg/cond = Systolic: 0.06/0.19, Diastolic: 0.04/0.13)
Discussion: Relatively low model fit explaining <20% of the variance indicates random measurement error frequently encountered when measuring blood pressure in children, leading to regression dilution bias. Whilst over-replacement with fludrocortisone is likely to drive hypertension, the small average effect size may indicate that appropriate mineralocorticoid replacement does not cause clinically significant increases in blood pressure.
Conclusion: Causal inference modelling of real world data has shown a smaller than expected average effect size of fludrocortisone on blood pressure that varies significantly between patients. This reinforces the importance of adequate and appropriately individualised mineralocorticoid replacement. Investment in controlled trials to assess fludrocortisone in CAH would help to improve dosing guidance.
Article tools
My recent searches
My recently viewed abstracts
Authors
ESPE Abstracts
© Bioscientifica 2024 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more