Exploring limits for data registration in the context of PROCARE, a quality improvement project on rectal cancer.

BACKGROUND: A high burden of registration in the context of quality improvement projects may result in registration fatigue.

METHODS: Time required for data collection and registration was measured. Quality of care indicators (QCI) were scored and factors for adjusted benchmarking were identified. The PROCARE data set was compared with 5 other European data sets.

RESULTS: Time required for data collection varied per domain while time for registration was more uniform. On average, per item 33 seconds were needed for collection and registration. The number of data to be registered per patient was 48-276, depending on the stage of the disease, resulting in a minimum of 25 minutes and a maximum of 2 hours 4 minutes per patient, follow-up not included. Focusing on 43 clinically relevant QCIs would result in a 50% reduction, using aggregate scores for performance audit in a 71% reduction. The PROCARE data set was larger than comparable European data sets. Linkage of the PROCARE database with administrative databases provided confident data on the patients' survival status, but did not appear to be a practical option for other QCIs.

CONCLUSIONS: Limiting the aim to performance audit could significantly reduce the burden of registration. In the context of a quality improvement project, the PROCARE Steering Group concluded that detailed clinical data from all centres are still required, which can be reconsidered in the future. Maintenance of a specific database remains of crucial value. Data collection and registration cannot be based on benevolence but should be compensated for.