Nurses Face an Interruption Every Six Minutes While Giving Medication, Landmark Chinese Hospital Study Reveals

In the crowded corridors of a 2,600-bed academic teaching hospital in Northwest China, a nurse preparing a patient’s medication is interrupted roughly every six minutes. A new time-and-motion study published in BMC Nursing has now captured, second by second, exactly what happens next — and the findings paint a vivid picture of the split-second judgment calls that shape medication safety on hospital wards. Over 204 hours of direct observation, researchers from Lanzhou University, the Chinese Academy of Medical Sciences and Peking Union Medical College logged a staggering 2,091 interruptions experienced by 37 registered nurses during medication tasks, averaging 10.25 interruptions per hour with a mean duration of 41.9 seconds each.

The sheer density of interruptions recorded in the study — one roughly every six minutes of medication-related work — places nursing among the most interruption-prone occupations ever studied. In high-reliability industries such as aviation and nuclear power, interruption rates of this magnitude would be considered unacceptable for safety-critical tasks, and procedures are deliberately designed to protect them. Medication administration, by contrast, has historically been treated as a task that nurses must juggle alongside a constant stream of ward communication, and the Chinese data make visible just how costly that cultural assumption may be.

The temporal clustering of interruptions is one of the most operationally revealing findings. With 84 percent of all interruptions occurring between 7:30 and 12:00, the morning medication round emerges as the single most contested window of the nursing day. This concentration is not accidental: morning hours combine the largest medication administration workload of the day, physician rounds, shift handovers, diagnostic scheduling, and family visits, all competing for the same narrow slice of time. The result is a predictable daily bottleneck in which the tasks most dependent on sustained attention — calculating doses, checking medications against patient records, preparing infusions — coincide with peak ambient demand for nurses’ attention.

The source distribution of interruptions carries particular weight for hospital management. Because 63.9 percent of interruptions originated from healthcare providers themselves — staff nurses, nurses interrupting their own task flow, and other health professionals — the problem is largely internal to the clinical team rather than driven by patients or visitors. This reframes the issue from one of individual vigilance to one of system design. Communication accounted for 62 percent of interruptions during medication tasks, yet only 38 percent of those communication interruptions were professionally necessary. In other words, a substantial share of the traffic flowing toward nurses during medication work was avoidable, suggesting that the timing and channel of routine clinical communication, rather than its content, is the primary lever for improvement.

The intent data add nuance to this picture. Reminding and questioning, each accounting for roughly a third of interruptions, are largely benign in motivation but potentially hazardous in effect. A colleague reminding a nurse about an upcoming procedure or asking a quick question about a patient is behaving reasonably by ordinary workplace standards; the danger arises because the recipient is mid-task in a safety-critical sequence. This asymmetry — socially appropriate behavior that is individually harmless but collectively risky — is precisely the kind of problem that cannot be solved by asking individuals to be more careful, and instead requires structural solutions such as protected medication zones, visual signals indicating a nurse is preparing medications, or asynchronous messaging systems that queue non-urgent queries until the medication round is complete.

The response data are arguably the study’s most novel contribution. Nurses responded immediately to 85.7 percent of interruptions, with only 14.3 percent blocked — split almost evenly between postponement (7.2 percent) and outright rejection (7.1 percent). Immediate response typically meant task-switching or multitasking, both of which carry documented cognitive costs. Research on attention and human performance has repeatedly shown that resuming an interrupted task involves a period of reduced accuracy and increased error risk, and that the probability of omitting a step in a procedural task rises sharply when the interruption occurs mid-sequence. In medication administration, where a single omitted verification step can translate into a wrong dose or wrong patient, the near-universal pattern of immediate response is a genuine safety concern.

Yet the finding that nurses almost never block interruptions should be interpreted with care rather than read as a failure of individual judgment. Nurses work within strong professional norms of responsiveness and teamwork; declining a colleague’s approach can feel socially costly, and nurses often cannot know whether an interruption is urgent without hearing it out. The study’s observation that response strategies varied significantly with the primary task being performed, the source and intent of the interruption, and the interruption type suggests that nurses are in fact engaged in a rapid, context-sensitive triage process — weighing what they are doing against who is asking and why. The management implication drawn by the authors, that nurses should be empowered to exercise clinical judgment and defer non-critical interruptions, therefore amounts to giving institutional permission and protection for a triage decision many nurses are already attempting to make implicitly.

Experience emerged as a meaningful differentiator. Nurses with more than ten years of experience handled interruptions differently from their less experienced colleagues, a pattern consistent with the development of tacit expertise in task prioritization. Veteran nurses may be better able to judge which interruptions can safely wait, or more practiced at resuming a disrupted task without losing their place in the sequence. Less experienced nurses, still consolidating the procedural steps of medication administration itself, may have less cognitive capacity available to manage concurrent demands, making them both more vulnerable to disruption and less equipped to deflect it. Any intervention to reduce interruption harm should therefore consider experience level, perhaps pairing junior staff with structured protocols that reduce reliance on in-the-moment judgment during high-risk medication tasks.

The finding that self-interruptions carried a higher rate of positive effect than interruptions from any other source, with an odds ratio of 2.84, deserves attention. Self-interruptions — a nurse pausing their own medication task to attend to something else — are self-initiated and therefore typically aligned with the nurse’s own assessment of priority. That these interruptions were more often judged to have positive effects suggests that the problem is not interruption per se, but interruption imposed from outside without knowledge of the interrupted person’s current cognitive state. This supports the case for communication systems that let the interrupter convey urgency asynchronously, allowing the nurse to decide when to engage rather than being forced to switch tasks at the interrupter’s chosen moment.

Methodologically, the time-and-motion approach gives the study a credibility that self-report designs cannot match. Retrospective surveys of interruption frequency are notoriously unreliable, because interruptions that end badly are remembered while routine ones are not, and because the sheer volume of small disruptions defies accurate recall. Direct structured observation, sustained over 204 hours across the full sequence of medication tasks — preparation, discussion, administration, documentation, health education, and materials disposal — captures the texture of real workflow, including the brief, seemingly trivial interruptions that accumulate into substantial cognitive load. The finding that medication preparation was the most frequently interrupted phase, at 32.3 percent of all events, is particularly important, since preparation involves the dose calculations and checks where errors are most consequential.

The study’s setting also matters for interpretation. A 2,600-bed academic teaching hospital in Northwest China is an environment of exceptional scale and acuity, and interruption rates there may not generalize to smaller community hospitals or to health systems with different staffing ratios, medication delivery technologies, or ward layouts. Teaching hospitals additionally carry a heavy supervisory and educational load, with trainees frequently approaching senior staff with questions — a dynamic that may inflate the questioning category of interruptions. Readers should therefore treat the specific figures as a detailed portrait of one demanding environment rather than a universal benchmark, while recognizing that the underlying dynamics — clustered morning workload, internally generated communication pressure, and near-universal immediate response — are likely present in attenuated form in many other settings.

For safety science more broadly, the study contributes to a growing recognition that medication error prevention cannot rely solely on individual accuracy at the point of administration. Interruption management sits at the intersection of human factors engineering, communication design, and workplace culture. The authors’ proposals — protective zones for medication preparation and asynchronous communication systems — reflect established human-factors principles: separate safety-critical work from ambient traffic, and convert synchronous demands into queued ones that the worker can address at a safe moment. The evidence that nurses already attempt to triage interruptions contextually suggests such interventions would be working with, rather than against, the grain of existing clinical judgment.

Finally, the study highlights a measurement gap in patient safety monitoring. Most medication safety metrics track errors and adverse events after the fact, while the micro-level dynamics of interruption and response — the 41.9-second exchanges, the task switches, the deferred questions — remain invisible to routine surveillance. Time-and-motion evidence of this kind offers a way to observe the upstream conditions under which errors become more likely, opening the possibility of leading indicators for medication safety: interruption rates per medication hour, the proportion of interruptions during preparation, and the balance between immediate and deferred responses. As hospitals seek to reduce preventable harm, understanding not just whether nurses are interrupted but how they respond, and under what conditions that response is safe, may prove as consequential as any single technological fix.

Subject of Research: Nurses' response strategies to interruptions during medication administration observed in a time-and-motion study

Article Title: How do nurses respond to interruptions during medication administration? A time-and-motion study in a Chinese hospital

Article References: Wang, Q., Yang, Y., Ding, X., Ma, S., Yang, Q., Luo, Z., Zhu, M., Chen, H., Zhang, H., & Li, Z. (2026). How do nurses respond to interruptions during medication administration? A time-and-motion study in a Chinese hospital. BMC Nursing. https://doi.org/10.1186/s12912-026-05513-y

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05513-y

Keywords: nurses, medication administration, interruptions, time-and-motion study, patient safety, multitasking, nursing workflow, hospital, China, medication safety, task-switching, healthcare communication