Predictive model for failure of autologous platelet-rich plasma injection in MRI-confirmed partial-thickness supraspinatus tendon tears– Dataset

This dataset was generated from a prospective, single-center cohort study conducted at Hai Phong International Hospital (Vietnam) between January 2024 and March 2025. Adult patients (≥18 years) with MRI-confirmed partial-thickness supraspinatus tendon tears were consecutively recruited during routine clinical practice and treated with standardized ultrasound-guided leukocyte-poor platelet-rich plasma (PRP) injection. Baseline demographic, clinical, laboratory, and imaging data were collected using standardized case report forms, and patients were followed for six months to assess pain (VAS), function (QuickDASH), and need for additional interventions. Treatment failure was defined as <30% improvement in VAS or QuickDASH scores, or requirement for further procedures. The dataset includes anonymized patient-level variables covering demographics, comorbidities, metabolic parameters, MRI features, prior treatments, adherence after injection, and clinical outcomes. It was specifically designed to support the development and internal validation of a multivariable prognostic model for PRP treatment failure, including logistic regression modeling, calibration, and decision curve analysis