Skip to Main Content


Every day, medical directors and administrators of all varieties worry about throughput, census, and patient flow. This concern may show up in many guises, across all care settings:

  • “There are no new patient appointment slots in primary care for the next six weeks.”

  • “We have no urgent care slots in cardiology left open this week.”

  • “The emergency department is backing up, and our average wait time is 6 hours.”

  • “Patients are waiting in the operating room because the recovery room is full. We can’t get patients out of the recovery room because the hospital floor is full.”

  • “The intensive care units (ICUs) are full, and cardiac surgery is backing up. A patient just got intubated on the general medicine floor. There are 10 ICU patients ready to go to the floor, but the floor is full so we can’t move them. How did it get like this?”

External Resource image

The Institute for Healthcare Improvement’s white paper Achieving Hospital-wide Patient Flow is available at

They also have in-person coursework available, with information at

There are many approaches to improving patient flow, most of which do not require sophisticated analytics. A number of effective techniques are reviewed in Chapter 9, which focuses in particular on Lean production/Toyota Production System (TPS) approaches to improvement. In addition, the Institute of Healthcare Improvement (IHI), the National Health Service, and others have both practical advice and courses on improving patient flow.

External Resource image

NHS England’s Good Practice Guide: Focus on Improving Patient Flow is available at

However, sometimes understanding patient flow in a detailed and more quantitative way can be remarkably helpful. For this purpose, we turn to a different set of modeling tools than we’ve discussed so far. These come out of the field known as operations research. These kinds of tools are particularly well suited to characterizing the dynamic aspects of patient flow. In spite of their relative lack of uptake, their potential value has been recognized for more than a decade. In 2005, the Institute of Medicine and the National Academy of Sciences collaborated to develop a joint report, Building a Better Delivery System: A New Engineering/Health Care Partnership.1 This document highlighted the usefulness of operations research—a topic that would come up again in subsequent National Academy reports focusing on developing and engineering a learning health system. Operations research is a field that developed over the past century. Its tools—including one of its oldest branches, queuing theory—help provide us a mathematical foundation with which we can understand, quantify, and predict patient flow across the healthcare system.


It is easy to mistake patient flow, census, and throughput as simply bureaucratic problems—things that help determine revenue and that make administrators happy, but nothing else. Patient flow certainly influences these things (it ...

Pop-up div Successfully Displayed

This div only appears when the trigger link is hovered over. Otherwise it is hidden from view.