Managing Waiting in Health Services: Strategies, Techniques and Measurement

Waiting is not a malfunction of health care; it is a structural feature of it. Demand arrives unevenly, capacity cannot flex in the short term, and clinical urgency differs between patients, so every system carries some level of waiting. The goal of waiting management is therefore not to eliminate waiting but to manage it on several fronts at once: how long it lasts, how it is experienced, how safe it is and how fairly it is shared. This chapter, written as a narrative training text from a structured Web of Science search (440 records screened to a working pool of 243), organises that field for managers, with direct implications for cancer services where delay can change outcomes.

Objective time, perceived time and expectation

The chapter’s founding distinction is between three variables. Objective waiting time is the clock or calendar interval between a reference event (referral, request, listing) and the start of care. Perceived waiting time is how long the patient feels they waited, and it departs systematically from the clock depending on information, environment, occupation, uncertainty and fairness. Expectation is the third: under the disconfirmation paradigm, satisfaction depends less on absolute length than on the gap between the expected and the experienced wait (Thompson & Yarnold, 1995). A classic emergency department study found that perceived, not objective, waiting predicted satisfaction. The managerial consequence is liberating: it is possible to improve the patient’s experience even before the wait itself can be shortened.

Waiting also has stages, each with its own logic: referral waiting, access waiting, in-facility waiting, diagnostic pathway waiting, waiting for test results, and list waiting. Waiting for a diagnosis or a test result carries a qualitatively different psychological load. Anxiety rises as news approaches (Sweeny & Andrews, 2014), and, importantly, a short wait does not automatically remove it: patients with short waits before bypass surgery also reported marked fear, partly because they had too little time to prepare.

Four families of strategy

The chapter proposes its own synthesis (explicitly not an established taxonomy), classifying interventions by the variable through which they are expected to work:

• Family A: reducing duration. Targets supply-demand balance and flow, for example direct booking, open access and rapid process redesign. A Cochrane review found these promising but with low certainty of evidence (Ballini et al., 2015). Their effect depends on the capacity constraint: without fixing the structural bottleneck, gains are usually temporary.
• Family B: perception and communication. Changes the experienced wait and expectation while objective time stays the same. Communication quality buffered the link between perceived wait and satisfaction in breast cancer patients (Lee et al., 2020); waiting-area design affects satisfaction, perceived time and behaviour. Activities that create a sense of flow help more than passive distraction: a television is not the same as a task that engages the patient.
• Family C: active waiting. Uses the waiting period for clinical preparation, education or support: nurse-led shared care on surgical lists, prehabilitation before transplantation, pre-clinic group education that reduced both waiting time and cost, digital self-help to bridge mental health waiting lists, mindfulness for coping, and social prescribing.
• Family D: prioritisation and fairness. Changes how waiting is distributed between patients without changing the total. Patients, surgeons and general practitioners weigh priority criteria differently; public disclosure of waiting times changes behaviour; and, often overlooked, the absence of explicit priority rules transfers an ethical burden to individual clinicians.

A rough map of the evidence pool reveals an imbalance. Active waiting is the fastest-growing family, roughly five times as many records after 2015 as before. Yet the weakest family in design terms is B: perception and communication interventions have the most records but no randomised trials in the pool. Because they are cheap and intuitively plausible, they are often implemented untested. The chapter recommends that every such intervention be evaluated at least with a before-and-after measurement.

Demand, capacity and variability

The operational foundation is queueing theory, presented as standard operations knowledge rather than evidence synthesis. The key relationship for managers is that waiting grows not linearly but steeply as utilisation approaches 100%; small fluctuations in demand or service time then produce disproportionate delays. “High utilisation equals high efficiency” is therefore a fallacy: waiting management needs a deliberate capacity buffer. Before choosing an intervention, the manager must diagnose why the list is growing: excess demand, lost capacity, variability, inappropriate referral, or an outdated list.

A six-dimension measurement system

The most common mistake is to monitor a single indicator, usually the mean. Waiting distributions are right-skewed, so the mean hides the patients who wait longest. The chapter proposes:

• Operational: median with 90th and 95th percentiles, share within target, stage-by-stage intervals, and abandonment (left without being seen, no-shows, withdrawal).
• Perception and experience: perceived time and its gap from objective time, expectation-reality gap, acceptability, adequacy of information.
• Psychological burden: anxiety, depression and distress screening, with an ethical precondition: do not screen without a defined response pathway.
• Clinical outcome and safe waiting thresholds: in radiotherapy, longer waits lower tumour control probability in a dose-response fashion (Waaijer et al., 2003). Every programme needs a list of time-critical indications, a maximum safe waiting threshold for each that triggers automatic acceleration, and a schedule for reassessing patients on the list.
• Equity: report waiting by insurance type, ethnicity, geography and other axes, with predefined minimum cell sizes and confidence intervals.
• Economic: institutional, service-system and societal costs, assessed separately for cost-effectiveness, budget impact and opportunity cost.

Every programme should also carry at least one balancing indicator. And one rule deserves emphasis: active waiting programmes may never be used to legitimise delaying necessary treatment. Family C sits beside Family A, not in its place.

Reading the evidence and evaluating change

An attitude survey shows acceptability, not effectiveness; a protocol shows intention, not impact. A fall in waiting after an intervention does not prove the intervention caused it: seasonality, falling demand, staffing changes or a new measurement method can do the same. In-house projects should aim at least for an interrupted time series design, accumulating several measurements before the change.

Digital tools and patient involvement

Machine-learning predictions of waiting time and status notifications can help, provided some practical rules are followed: give a range rather than a single number (“40 to 60 minutes” rather than “45 minutes”), state when the estimate was last updated, and check model calibration in subgroups. Patients should be involved from problem definition onwards, because what managers assume bothers patients often differs from what actually does.

An eight-step decision framework

• Identify the stage of waiting.
• Diagnose the cause.
• Assess clinical risk and define safe waiting thresholds.
• Test the capacity constraint.
• Start low-cost experience interventions in parallel.
• Define a balancing indicator for every intervention.
• Set up the evaluation design in advance.
• Involve patients from the start.

Source chapter: Kurutkan, M. N. (2026). Managing waiting in health services: Strategies, techniques, tools and measurement approaches. Oncology Hospital Management Training Series, supplementary chapter (original in Turkish).

References