Six sigma tools for a patient safety-oriented, quality-checklist driven radiation medicine department.

INTRODUCTION

The purpose of this work was to develop and implement six sigma practices toward the enhancement of patient safety in an electronic, quality checklist-driven, multicenter, paperless radiation medicine department.

METHODS AND MATERIALS

A quality checklist process map (QPM), stratified into consultation through treatment-completion stages was incorporated into an oncology information systems platform. A cross-functional quality management team conducted quality-function-deployment and define-measure-analyze-improve-control (DMAIC) six sigma exercises with a focus on patient safety. QPM procedures were Pareto-sorted in order of decreasing patient safety risk with failure mode and effects analysis (FMEA). Quantitative metrics for a grouped set of highest risk procedures were established. These included procedural delays, associated standard deviations and six sigma Z scores. Baseline performance of the QPM was established over the previous year of usage. Data-driven analysis led to simplification, standardization, and refinement of the QPM with standard deviation, slip-day reduction, and Z-score enhancement goals. A no-fly policy (NFP) for patient safety was introduced at the improve-control DMAIC phase, with a process map interlock imposed on treatment initiation in the event of FMEA-identified high-risk tasks being delayed or not completed. The NFP was introduced in a pilot phase with specific stopping rules and the same metrics used for performance assessments. A custom root-cause analysis database was deployed to monitor patient safety events.

RESULTS

Relative to the baseline period, average slip days and standard deviations for the risk-enhanced QPM procedures improved by over threefold factors in the NFP period. The Z scores improved by approximately 20%. A trend for proactive delays instead of reactive hard stops was observed with no adverse effects of the NFP. The number of computed potential no-fly delays per month dropped from 60 to 20 over a total of 520 cases. The fraction of computed potential no-fly cases that were delayed in NFP compliance rose from 28% to 45%. Proactive delays rose to 80% of all delayed cases. For potential no-fly cases, event reporting rose from 18% to 50%, while for actually delayed cases, event reporting rose from 65% to 100%.

CONCLUSIONS

With complex technologies, resource-compromised staff, and pressures to hasten treatment initiation, the use of the six sigma driven process interlocks may mitigate potential patient safety risks as demonstrated in this study.