PURPOSE: The Mucopolysaccharidosis Type I (MPS I) Registry is an international observational database that tracks the natural history and the outcomes of patients with MPS I. The Registry was a regulatory requirement following the approval of laronidase enzyme replacement therapy for MPS I in 2003. All data are collected voluntarily after informed consent from the patient or family. Data are checked through queries, monthly reviews, and electronic audits to identify missing, inconsistent, or invalid data. This analysis sought to determine overall data accuracy in the Registry through source document verification (SDV). METHODS: Two phases of SDV were performed. In each phase, Registry data were compared against source documents at sites in Europe, Latin America, and North America. Three patients were randomly selected for SDV at each of the selected sites among all patients enrolled ≥18 months and ever receiving laronidase. Key parameters central to MPS I and its treatment were examined from the baseline and the last available assessments. RESULTS: Results indicate an overall source-to-database error rate in the MPS I Registry of 2.7% (47 discrepancies out of 1715 items; 95% confidence interval [2.2%, 3.5%]) in Phase 1 and 3.7% (64 discrepancies out of 1732 items; 95% confidence interval [2.9%, 4.7%]) in Phase 2. No systematic errors were found. CONCLUSIONS: The overall error rates in both phases of SDV demonstrate acceptable data accuracy in the MPS I Registry within the data fields that were assessed.
PURPOSE: The Mucopolysaccharidosis Type I (MPS I) Registry is an international observational database that tracks the natural history and the outcomes of patients with MPS I. The Registry was a regulatory requirement following the approval of laronidase enzyme replacement therapy for MPS I in 2003. All data are collected voluntarily after informed consent from the patient or family. Data are checked through queries, monthly reviews, and electronic audits to identify missing, inconsistent, or invalid data. This analysis sought to determine overall data accuracy in the Registry through source document verification (SDV). METHODS: Two phases of SDV were performed. In each phase, Registry data were compared against source documents at sites in Europe, Latin America, and North America. Three patients were randomly selected for SDV at each of the selected sites among all patients enrolled ≥18 months and ever receiving laronidase. Key parameters central to MPS I and its treatment were examined from the baseline and the last available assessments. RESULTS: Results indicate an overall source-to-database error rate in the MPS I Registry of 2.7% (47 discrepancies out of 1715 items; 95% confidence interval [2.2%, 3.5%]) in Phase 1 and 3.7% (64 discrepancies out of 1732 items; 95% confidence interval [2.9%, 4.7%]) in Phase 2. No systematic errors were found. CONCLUSIONS: The overall error rates in both phases of SDV demonstrate acceptable data accuracy in the MPS I Registry within the data fields that were assessed.