Monitoring quality control: can we get better data?

Quality control is important in many fields, especially industrial production. Major research has been developed with regard to industrial quality control to ensure reliable and consistent products. We adapt and develop methodology in quality control to monitor data collection in epidemiologic studies. There are no procedures currently used by epidemiologists to evaluate quality control during the actual process of data collection; methods are implemented only after the data have been collected. We focus on procedures that can be used during data collection: instrument calibration and population sampling. For the first, we propose methods utilizing Shewhart control charts and Westgard stopping rules. For evaluating population sampling, we present methods utilizing regression analysis. We provide a motivating example to highlight the utility of these methods. The proposed methodology may help investigators to identify data quality problems that can be corrected while data are still being collected, and also to identify biases in data collection that might be adjusted later.