Where does congestion collide?

In transport economics, congestion is the mutual interference that arises as traffic approaches the capacity of a road: each additional vehicle slows all the others, so the private cost a driver perceives (their own travel time) falls short of the social cost they impose (the delay added to everyone else). Walters's 1961 The theory and measurement of private and social cost of highway congestion formalised this gap and the congestion toll that would close it. The referent is a flow-versus-capacity externality on a transport network, measured in delay and speed, and it is a different thing from a data link's backlog or a patient's fluid overload.[1]

Unit

Extra travel time (and its cost) each vehicle imposes on others as flow nears road capacity

Recovery means

Congestion eases as demand falls below capacity, or when pricing or metering holds inflow at the capacity rate

Measured by

Gauged by the speed-flow relationship, delay per vehicle, and the divergence between marginal private and marginal social cost

In packet-switched networks, congestion is the collapse in useful throughput that occurs when the data offered to a link exceeds what it can carry: queues at routers fill, delays climb, and packets are dropped, which without control triggers retransmissions that make things worse. Jacobson's 1988 Congestion avoidance and control introduced the algorithms (slow start, congestion-window back-off) that let senders infer congestion from loss and throttle their rate. The referent is a load-versus-capacity breakdown inside a communication network, measured in queue length, loss, and throughput, sharing only the word with road jams or clinical congestion.[2]

Unit

Queueing delay, packet loss, and lost throughput when offered load exceeds link capacity

Recovery means

Throughput recovers as senders reduce their transmission rate (window back-off) until offered load fits the available capacity

Measured by

Tracked by round-trip time, queue occupancy, packet-loss rate, and achieved versus offered throughput

In clinical medicine, congestion is the accumulation of excess fluid or blood in a tissue or vascular bed because it cannot drain or be cleared normally, for example the pulmonary and systemic venous congestion of heart failure, where rising filling pressures back fluid up into the lungs and peripheral tissues. Gheorghiade and colleagues' 2010 scientific statement Assessing and grading congestion in acute heart failure sets out how clinicians detect and grade it. The referent is a state of fluid overload and raised venous pressure in the body, measured by signs, imaging, and pressures, unrelated to a congested motorway or an overloaded network beyond the shared image of something that cannot flow away.[3]

Unit

The degree of fluid or blood accumulation and raised filling pressure in a tissue or vascular bed

Recovery means

Congestion resolves as decongestive therapy (for example diuresis) lowers filling pressures and clears the excess fluid

Measured by

Assessed by clinical signs, chest imaging, natriuretic peptides, and invasive or estimated filling pressures