Project Summary. The goal of this study is to create a sustainable and expandable Pediatric Glomerular Disease Learning Network to address fundamental limitations of current observational cohorts and registries and focus on practice pattern variation, observational outcomes and comparative effectiveness research, and ultimately pragmatic clinical trials. The objectives of this study are: to develop and validate a computable phenotype for the identification of glomerular disease in PEDSnet, to establish a collaborative of nephrologists from the 8 participating institutions committed to the creation and maintenance of a Pediatric Glomerular Disease Learning Network, and to promote the standardization of data collected in the course of clinical care.
Study Design. This study is an observational, cross-sectional analysis using previously collected data.
Engagement. Dr. Denburg’s work in mineral metabolism in glomerular disease has been supported by the NephCure Foundation. In 2014, NephCure established a Patient Powered Research Network (PPRN), the NephCure Kidney Network (NKN) Patient Registry, to include the patient perspective in driving new research. Drs. Denburg, Furth and Forrest hope to partner with NephCure on this project. Drs. Denburg and Furth are already seeking engagement with industry for this work.
Data Elements. We will examine two approaches for defining primary glomerular disease, one with an emphasis on specific diagnostic entities and one with an emphasis on inclusivity. The definition of glomerular disease for the latter will be intentionally comprehensive and broad, so as not to exclude data on patients with rare or poorly characterized clinical entities.
1) The first approach uses diagnosis codes for primary nephrotic syndrome, with the intention of capturing nephrotic syndrome due to biopsy-based diagnoses of minimal change disease (MCD), focal segmental glomerulosclerosis (FSGS), membranous nephropathy (MN), membranoproliferative glomerulonephritis/C3 glomerulopathy (MPGN), and IgA nephropathy (IgAN), as well as for the condition of nephrotic syndrome without biopsy since the majority of children with steroid-sensitive nephrotic syndrome will not undergo kidney biopsy.
2) The second approach uses a broader list of diagnosis codes to identify any glomerular diagnosis:
• based on kidney biopsy or nephrectomy including, but not limited to, the conditions above (e.g. Henoch-Schonlein (HSP) nephritis, pauci-immune glomerulonephritis, crescentic glomerulonephritis, post-infectious glomerulonephritis, anti-glomerular basement membrane disease, Alport/hereditary nephritis, diffuse mesangial sclerosis, C1q nephropathy, fibrillary nephropathy).
• based on a clinical diagnosis of glomerulonephritis (e.g. post-infectious, hereditary, presumed IgAN/HSP nephritis).
Computable Phenotype for Cases
EHR queries: We will query EHRs to identify all PEDSnet patients (at least one in-person clinical encounter since January 2009) who meet case definitions.
Case definition 1: Primary nephrotic syndrome
• ≥2 entries of any ICD-9-CM/ICD-10 diagnosis included in List A
• ≥1 CPT/ICD code for kidney biopsy AND ≥1 subsequent entry of any ICD-9-CM/ICD-10 diagnosis included in List A
Case definition 2: Broad definition of glomerular disease
• ≥2 entries of any ICD-9-CM/ICD-10 diagnosis included in List B
• ≥1 CPT/ICD code for kidney biopsy AND ≥1 subsequent entry of any ICD-9-CM/ICD-10 diagnosis included in List B
• For both case definitions, we will exclude those with diagnosis codes for systemic lupus erythematosus
Computable Phenotype for Non-Cases
Non-cases will be any PEDSnet patient with at least 3 outpatient visits to a Nephrologist, excluding the inclusion diagnoses.