Lean Six Sigma in healthcare

A patient pathway is a queueing system with clinical work attached, and almost all of the elapsed time is queue. Referral to treatment measured honestly shows hours of clinical contact spread across weeks, and the waits concentrate at a small number of transitions: the referral being triaged, the diagnostic being reported, the decision being made in a meeting that happens weekly.

Bed occupancy is where the arithmetic bites hardest. Run a ward near capacity and delay does not rise gently, it accelerates, which is why a hospital at ninety nine percent occupancy behaves catastrophically worse than one at eighty five. The same mathematics governs theatres, scanners and clinics. Capacity that looks wasteful on a spreadsheet is what makes flow possible.

The third is discharge, which is rarely a hospital process at all. A bed occupied by someone medically fit to leave is usually waiting on a social care assessment, a package, transport or a pharmacy round, and each of those belongs to a different organisation. It is the clearest example in any sector of a constraint sitting outside the walls of the operation that suffers from it.

Define picks one pathway and one cohort, because "outpatients are slow" cannot be finished. Measure uses what the systems already hold, and the first job is nearly always reconciling three sources that disagree about when a patient arrived.

Analyse should be done with the clinicians rather than presented to them. The method has a poor reputation in parts of the NHS precisely because it has been arrived with, as a programme, by people who had not walked the pathway. Sitting in the department for a day earns more than a month of data alone.

Improve in healthcare is usually scheduling and sequencing rather than speed: pooling waiting lists rather than holding one per consultant, booking diagnostics before the clinic rather than after, moving a weekly decision meeting to twice weekly. Control has to survive winter, which is the honest test, so the chart belongs to the operational team and is reviewed inside existing governance rather than as an extra meeting.

Clinical judgement is variation to protect. Two patients with the same diagnosis are not the same case, and a pathway that removes the discretion to treat them differently is not an improvement, it is a hazard. The rule is to standardise the administrative wrapper ruthlessly and the clinical decision never.

Targets are the second danger, and healthcare has the most instructive examples in any sector. A four hour target met by admitting people who did not need admitting is a number improving while care gets worse. Every measure introduced needs a counter measure beside it, and somebody senior willing to look at both.

The third is fatigue. Staff in most trusts have been surveyed, engaged and transformed repeatedly, and arriving with a branded methodology invites justified cynicism. Doing one small thing that visibly works, in one department, buys more credibility than any launch event.