Electronic calculation of quality measures in the EMR is highly dependent on data capture in structured fields: the Oncology Care Model (OCM) quality reporting experience

Background: Reporting of quality measures is ideally achieved electronically. This is dependent on presence of data elements in structured fields. We report on the variability of quality measure reporting in the OCM for quality measures OCM 8 (adjuvant chemotherapy in stage 3 colon cancer), 9 (adjuvant chemotherapy in HR- breast cancer), 10 (adjuvant trastuzumab for Her2+ breast cancer), and 11 (adjuvant hormonal therapy for HR+ breast cancer). Methods: Quality measures were calculated for 30,524 patients across 35 practices participating in the OCM. We examined the calculation of quality metrics OCM-8, 9, 10, and 11 based on presence or absence of data elements in the medical record. Results: For quality measures OCM-8 and 9 that require documentation of date of surgery, only 1.3% and 1.6%, respectively, of patients could be included in the calculation due to absence of surgical date in a structured field. For quality measures OCM 9 and 11 that require attestation of surgical staging of breast cancer, this data element was documented in 0% of patients. For OCM 9 and 10 that require stage documentation of breast cancer, this data element was documented in 63% and 75%, respectively, of patients (range 12-100%). In cases where staging criteria were not met, 58% and 44%, respectively, of patients were removed due to total absence of staging data in structured fields. Conclusions: Quality measure calculation and reporting is highly dependent on documentation of data elements in structured fields. Lack of documentation of key elements distorts calculation of the measure. Modification of physician workflow or restructuring of measure calculation to optimize use of well documented structured elements will improve generalizability of quality measure reporting. Manual abstraction of data does not represent a feasible alternative.