An ethnographic study of improving data collection and completeness in large-scale data exercises

Abstract

Background: Large-scale data collection is an increasingly prominent and influential feature of efforts to improve healthcare delivery, yet securing the involvement of clinical centres and ensuring data comprehensiveness often proves problematic. We explore how improvements in both data submission and completion rates were achieved during a crucial period of the evolution of two large-scale data exercises. Methods:  As part of an evaluation of a quality improvement programme, we conducted an ethnographic study involving 90 interviews and 47 days of non-participant observation of two UK national clinical audits in a period before submission of data on adherence to clinical standards became mandatory. Results: Critical to the improvements in submission and completion rates in the two exercises were the efforts of clinical leaders to refigure “data work” as a professionalization strategy. Using a series of strategic manoeuvres, leaders constructed a cultural account that tied the fortunes of the healthcare professions to the submission of high-quality data, proposing that it would demonstrate responsibility, transparency, and alignment with the public interest. In so doing, clinical leadership deployed tactics that might have been seen as unwarranted managerial aggression had they been imposed by parties external to the profession. Many residual challenges were linked not to principled objection by clinicians, but to mundane problems and frustrations in obtaining, recording, and submitting data. The cultural framing of data work as a professional duty was important to resolving its status as an abject form of labour. Conclusions: Improving data quality in large-scale exercises is possible, but requires cooperation with clinical centres. Enabling professional leadership of data work may offer some significant advantages, but attention is also needed to mundane and highly consequential obstacles to participation in data collection.