Quality Improvement, Clinical Audit and Research in Healthcare
Finding the right problem: context, lived experience, equity and priorities
How to turn concerns into defensible problem statements using data, lived experience, context, equity and organisational priorities.
Learning objectives
• Turn a concern or anecdote into a measurable problem statement without embedding a preferred solution. • Use baseline data, observation and lived experience to understand who is affected and how. • Identify inequity, exclusion and variation hidden by overall averages. • Prioritise improvement problems using impact, feasibility, risk and strategic relevance. • Avoid starting with a tool or intervention before understanding the problem.
Learning objectives
- Turn a concern or anecdote into a measurable problem statement without embedding a preferred solution.
- Use baseline data, observation and lived experience to understand who is affected and how.
- Identify inequity, exclusion and variation hidden by overall averages.
- Prioritise improvement problems using impact, feasibility, risk and strategic relevance.
- Avoid starting with a tool or intervention before understanding the problem.
From concern to problem statement
A good problem statement describes the gap, population, setting and consequence. “Staff need a checklist” is a proposed solution. “Only 54% of high-risk patients receive documented falls reassessment within 24 hours of transfer, with variation between wards” is a problem statement. Keeping the problem separate from the solution leaves room to discover the actual causes.
Triangulate evidence
No single data source gives the whole picture. Combine routine data, case review, observation, staff experience and patient/service-user feedback. Quantitative data shows scale and variation; qualitative information explains how and why the process behaves as it does. Where data conflict, treat the disagreement as information rather than choosing the source that confirms the team’s belief.
Context is part of the intervention
A change that succeeds on one ward may fail elsewhere because staffing, digital systems, leadership, workload, patient mix or physical layout differ. Record contextual conditions before and during the project. Context helps explain why a result occurred and whether a change can spread safely.
Equity by design
Overall averages can hide unequal access or outcomes. Ask who is missing from the data, who finds the pathway difficult to use, whether reasonable adjustments are available, and whether the proposed change shifts burden to people with fewer resources. Where lawful and proportionate, stratify data by relevant characteristics or service segments. Do not collect sensitive data without a clear purpose and governance basis.
Priority setting
A useful priority assessment considers harm, frequency, inequity, patient importance, staff burden, strategic relevance, evidence of unwarranted variation, feasibility and the degree to which the team can influence the problem. High visibility is not the same as high value. A small but frequent process failure may create more cumulative harm than a rare dramatic event.
Worked example: missed appointments
An outpatient team sees a 14% “DNA” rate and proposes reminder texts. Before acting, they examine language needs, appointment lead time, transport, digital exclusion, clinic type and deprivation. The rate is highest in two groups with different barriers. A single reminder intervention would not address the whole problem. The project becomes a set of targeted tests: clearer multilingual information for one group and scheduling redesign for another, with balancing measures for staff workload and appointment capacity.
Patient and staff interviews as diagnostic tools
Short structured conversations can reveal barriers not visible in electronic records. Keep questions focused on the process rather than asking participants to design the solution for you. Summarise themes transparently, preserve confidentiality and avoid implying that a small convenience sample represents everyone.
Problem-framing template
Current state: what is happening now? Population: who experiences it? Setting: where/when? Evidence: what supports the claim? Consequence: why does it matter? Variation/equity: who does better or worse? Scope: what is outside the project? Unknowns: what do we still need to understand before choosing changes?
