Amyloidosis is a serious health condition that occurs when a protein called amyloid folds incorrectly and accumulates in different parts of the body. Currently treatment options are limited and mainly focus on symptom management, whereas an early diagnosis of the disease is highly important for effective intervention and improved patient care. Amyloidosis are diagnosed at both primary and secondary care settings, but different clinical codings are usually applied, resulting in discrepant diagnoses and many patients not captured early in their disease pathway. We aim to study how amyloidosis and its sub-types are coded in clinical practice between primary and secondary care and compare characteristics of each group of patients. We will also explore ways to more clearly define/predict amyloidosis sub-types, as well as developing new tools for diagnosis. Overall, the hope is that this study will help understand common coding practice and issues in clinical settings, which can lead to a higher case detection rate through novel techniques, early intervention opportunities and ultimately better care for amyloidosis patients.
Amyloidosis, characterized by the accumulation of misfolded amyloid proteins, encompasses AL (Primary), AA (Secondary), and transthyretin (ATTR) amyloidosis. ATTR, a clinically diverse and fatal disease, arises from the aggregation of transthyretin (TTR) amyloid fibrils in various organs and tissues. The study seeks to elucidate the landscape of amyloidosis clinical coding in both primary and secondary care in the UK.
This observational study will encompass patients aged over 18 years at diagnosis, using a defined list of diagnosis codes (ICD-10, Read/SNOMET-CT, and product codes) in CPRD Aurum & GOLD data, linked with Hospital Episode Statistics admission data (HES APC), Office for National Statistics (ONS) Death Registration data, and patient-level and practice-level index of multiple deprivation (IMD) data.
The primary objectives involve delineating the agreement and disagreement between existing clinical coding frameworks, as well as comparing patient characteristics across datasets. A questionnaire aimed at healthcare professionals (HCPs) will be used to gather insights on specific diagnosis codes selected for amyloidosis, enabling an assessment of the perceived efficacy of coding practices and facilitating recommendations for amyloidosis coding in primary care. The results from the questionnaire will be linked back to patients' CPRD records to evaluate the concordance in coding practices.
Additional goals include contrasting patient characteristics between CPRD and National Amyloidosis Centre (NAC) datasets, and employing machine-learning techniques to derive amyloidosis subtypes. Concordance and discordance between primary and secondary care diagnoses will be calculated, with descriptive statistics presented for patient demographics and clinical characteristics, as needed. Results from the GP questionnaire will be summarized as proportions. For algorithm performance, a machine learning approach will be adopted, and the derived clusters will be summarized using descriptive statistics. The area under the receiver operating curve (AUROC) will be leveraged to predict case definitions at varying sensitivities and specificities.
The following will be measured at amyloidosis diagnosis:
• Concordance and discordance between primary care and secondary care clinical coding
• Descriptive summary of responses from the GP questionnaire
• Patient characteristics similarities and differences between CPRD and NAC datasets
• New potential sub-groups within the amyloidosis patient population
Jil Billy Mamza - Chief Investigator - AstraZeneca Ltd - UK Headquarters
Jil Billy Mamza - Corresponding Applicant - AstraZeneca Ltd - UK Headquarters
Eleanor Axson - Collaborator - CPRD
He Gao - Collaborator - AstraZeneca Ltd - UK Headquarters
Holly Bennett - Collaborator - CPRD
Kathleen White - Collaborator - CPRD
Lisa Anderson - Collaborator - St George's University Hospitals NHS Foundation Trust
Mike Lonergan - Collaborator - CPRD
Rachael Williams - Collaborator - CPRD
Ruiqi Zhang - Collaborator - AstraZeneca Ltd - UK Headquarters
Jil Billy Mamza - Collaborator - AstraZeneca Ltd - UK Headquarters
Justin Chan - Collaborator - CPRD
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation;Practice Level Index of Multiple Deprivation