Using administrative health data to describe colorectal and lung cancer care in New South Wales, Australia: a validation study

摘要: Background: Monitoring treatment patterns is crucial to improving cancer patient care. Our aim was determine the accuracy of linked routinely collected administrative health data for monitoring colorectal and lung care in New South Wales (NSW), Australia. Methods: Colorectal cases diagnosed NSW between 2000 2002 were identified from Central Cancer Registry (CCR) their hospital discharge records Admitted Patient Data Collection (APDC). These then two relevant population-based surveys. The main outcome measures sensitivity specificity CCR APDC disease staging, investigative procedures, curative surgery, chemotherapy, radiotherapy, selected comorbidities. Results: 2917 1580 analysed. Unknown stage more common (18%) than survey (2%). Colonoscopies captured reasonably accurately compared with surveys (82% 79% respectively; 91% sensitivity, 53% specificity) but all other or diagnostic procedures under-enumerated. Ninety-one percent had potentially surgery recorded 95% (96% 92% specificity), similar (16% 17%; 99% specificity). Chemotherapy (~40% sensitivity) radiotherapy (sensitivity≤30%) vastly under-enumerated data. only comorbidity that diabetes. Conclusions: Linked provided accurate information on surgical treatment, colonoscopies comorbidities such as Other comorbidities, chemotherapy not well enumerated sources will be required comprehensively monitor primary management patients.