How Management Made Medicine: The Evolution of ‘Quality Improvement’ from Industrial Production to Medical AI

• In Event: 50220 - Section on Organizations, Occupations, and Work Roundtables
  In Refereed Roundtable (60 minute): Table 19: Medical Work

Abstract

For decades, U.S. healthcare professionals have worshipped at the altar of clinical trials, the “gold standard” of academic medicine. Yet in the contemporary healthcare field, many activities that resemble scientific work are classified as quality improvement (QI), a framework for improving the efficiency and efficacy of healthcare delivery that derives from mid-twentieth century management science. While QI is ubiquitous in academic medicine, it differs from clinical research in that QI activities are largely exempt from formal ethical oversight. Furthermore, the origins and evolution of this paradigm are poorly understood. This paper traces the evolution of QI from its beginnings in statistical quality control methods to its current use as a framework for governing cutting-edge healthcare work, including management strategies that attempt to reduce burnout among hospital workers through greater reliance on AI. Drawing on archival and ethnographic materials, I ask: how has a management science paradigm initially devised to optimize industrial production been adapted for use in modern medicine, and what are the consequences for healthcare workers? First, the paper explores how QI originated with statistical techniques for measuring quality that were pioneered at Bell Labs and later applied to Japanese industrial production in the 1950s. By the 1980s, these techniques had been imported back to the U.S. by companies including Ford, Xerox, and AT&T. It was at that point that the healthcare industry took notice of the revolution in quality management, and experts working with the Harvard Community Health Plan commenced a project that brought together clinicians, healthcare executives, and quality control experts in service of standardizing a framework for quality improvement. Today, many of the attempts to automate hospital work are classified as QI work rather than clinical research, which suits the needs of AI model developers but can pose risks to patients and clinicians.

Author

• Zachary Webster Griffen, New York University