Quality Improvement, Clinical Audit and Research in Healthcare
Understanding the system: process mapping, human factors and causes
Practical system diagnosis using process mapping, human factors, cause analysis, risk thinking and identification of leverage points before intervention.
Learning objectives
• Map the actual process and distinguish work-as-imagined from work-as-done. • Identify hand-offs, queues, rework, failure points and hidden dependencies. • Use cause-and-effect thinking without mistaking correlation or opinion for cause. • Apply human-factors principles to redesign systems rather than blaming individuals. • Prioritise leverage points for testing.
Learning objectives
- Map the actual process and distinguish work-as-imagined from work-as-done.
- Identify hand-offs, queues, rework, failure points and hidden dependencies.
- Use cause-and-effect thinking without mistaking correlation or opinion for cause.
- Apply human-factors principles to redesign systems rather than blaming individuals.
- Prioritise leverage points for testing.
Map before redesigning
Teams often know the policy but not the real workflow. Process mapping makes the sequence visible: start/end points, tasks, decisions, hand-offs, waiting, loops and exceptions. Observe the process at different times and with different staff. A map drawn only in a meeting can reproduce work-as-imagined rather than work-as-done.
Simple mapping method
Choose one defined pathway. Walk through a recent real case or observe several cases. Write each step on a separate note. Place steps in time order, then add decision points, waits, hand-offs and rework. Ask where information is copied, where responsibility becomes unclear, where queues form, and where staff create workarounds. Validate the map with people who do the work.
Human factors
Human factors examines how people interact with tasks, tools, environment, teams and organisational conditions. Errors are more likely when systems rely on memory, ambiguous displays, interruptions, poor hand-offs, excessive workload or conflicting goals. Redesign should make the safer action easier and more reliable rather than assuming education will overcome every system weakness.
Cause-and-effect analysis
Fishbone diagrams can organise possible contributors under headings such as people, process, equipment, environment, information and organisation. The diagram is a hypothesis map, not proof. Use data, observation and testing to determine which factors actually influence the problem. Five Whys can help explore a causal chain but should not be used to force a single “root cause” in a complex system.
Failure modes and risk
Ask what can fail at each critical step, how likely it is, how detectable it is before harm and what the consequence may be. Formal FMEA methods may be appropriate for high-risk redesign, but even a simple prospective failure-mode discussion can identify safeguards and balancing measures before testing.
Worked example: delayed imaging
A service believes radiology turnaround is slow. Mapping shows the major delay occurs before the request reaches radiology: requests wait for senior authorisation during a daily bottleneck. The original proposal to add radiology staff would not address the dominant delay. The team tests an agreed criterion-based authorisation process during defined hours and measures time from clinical decision to request, request to scan, and total pathway time.
Choose leverage points
Prioritise causes that are supported by evidence, materially affect the outcome, are feasible to influence and do not create disproportionate risk elsewhere. Avoid changing everything at once; otherwise the team cannot learn which change mattered. Small tests can distinguish plausible leverage points from attractive ideas.
Process-mapping worksheet
Define start and end. Observe at least three real journeys if feasible. Record each step and who owns it. Mark waits, hand-offs, duplication, workarounds, failure points, variation and patient/staff burden. Identify the top three uncertainties to investigate before selecting changes.
