Doctor strikes in England coincided with shorter emergency department stays
Industrial action in the English NHS between December 2022 and July 2024 produced a counterintuitive pattern in emergency care. Across 20 major type 1 emergency departments, attendances were shorter on days affected by several strike types, especially resident doctor strikes and consultant strikes. The effect was not uniform across staff groups: when resident and consultant doctors were on strike at the same time, average emergency department time increased instead of falling.
The analysis drew on linked emergency department, inpatient, and imaging data from August 2022 through July 2024 and used a retrospective causal mediation framework. In practical terms, the work did not simply ask whether strikes changed throughput. It examined how much of the change could be traced to emergency inpatient occupancy, shifts in attendance volume, and altered efforts to meet the 4-hour target. That approach matters because emergency care is rarely governed by one lever alone. Bed availability, referral pressure, staffing mixes, and managerial response can move in opposite directions on the same day.
Shorter stays did not always mean smoother care
Resident doctor strikes were associated with a 12.6% reduction in average emergency department duration, with a 95% confidence interval of -10.4% to -14.8%. Consultant strikes were linked to a 5.6% reduction, and paramedic strikes to a 4.5% reduction. The combined resident-and-consultant doctor strike was the exception: average stays increased by 7.2%. Nurse strikes did not produce a significant change in duration, with an estimate of -1.7% and a confidence interval spanning from -1.5% to 4.7%.
Those swings point to a system that responds differently depending on which link in the care chain is disrupted. Doctor action can reduce downstream admissions and alter decision-making at the front door, while ambulance disruption may curb incoming flow. Nursing strikes, by contrast, may not alter the same decision bottlenecks to the same degree in this setting, at least not enough to shift overall duration measurably across the cohort studied.
Bed pressure accounted for part of the effect, but not all of it
Emergency inpatient occupancy explained 10.4% of the change in average ED duration during resident doctor strikes, 53.3% during consultant strikes, 5.4% during combined doctor strikes, and 14.3% during paramedic strikes. Those figures offer a concrete physiological and operational mechanism: when inpatient beds are less congested, patients can move out of the emergency department faster, and when beds are full, boarding slows flow regardless of front-door demand.
Even so, the mediation was incomplete. Only 1.5% of the change in average ED duration observed during resident doctor strikes was explained by the net effect of the modeled mediators. During combined doctor strikes, the figure was 16.8%; during paramedic strikes, 28.8%. A large fraction of the resident doctor effect therefore remained outside the measured pathways. In a real hospital, that could reflect changes in clinical threshold for admission, altered escalation habits, different senior review patterns, or frontline adaptation by teams trying to preserve flow with fewer hands available.
One striking result cuts across the rest of the analysis.
During resident doctor strikes, patients were more likely to receive treatment in the emergency department, yet less likely to be admitted.
The adjusted incident risk ratio for treatment was 1.013, with a 95% confidence interval of 1.008 to 1.018, while the adjusted odds ratio for admission was 0.963, with a 95% confidence interval of 0.946 to 0.980. That combination suggests a sharper boundary between patients who could be stabilized and discharged from the department and those who truly required inpatient care. In a strained system, fewer admissions can ease crowding, but the trade-off may be a more selective use of beds rather than a simple gain in efficiency.
Throughput pressure rose when doctor strikes hit
Efforts to meet the 4-hour target intensified during resident, consultant, and combined doctor strikes. During nurse and paramedic strikes, those efforts fell. The shift in throughput behavior explained 2.7% of the change in average ED duration during resident doctor strikes and 17.1% during consultant strikes.
This is one of the clearest signals in the analysis. When senior and resident doctors were absent, teams appear to have pushed harder to move patients through the department, likely because staffing shortages threatened visible performance metrics and operational flow. Faster disposition does not require a single mechanism; it can arise from more direct senior decision-making, narrower admission thresholds, or more rapid prioritization of patients who can safely leave the department. The data do not identify which of these dominated, but the pattern is consistent with a system reacting to scarcity by compressing decision time.
Planned care fell sharply without explaining the ED changes
Planned hospital care dropped substantially during resident, consultant, and combined doctor strikes, yet those reductions did not meaningfully account for the shorter emergency department stays. That separation is important. It suggests that lower elective or planned activity did not simply free enough inpatient capacity to explain the full ED effect, even though hospital-wide activity clearly shifted during strike periods.
The practical implication is narrower than a simple “less elective care, better ED flow” story. Emergency departments seem to have been influenced by a mix of occupancy, demand, and real-time operational changes rather than by planned care suppression alone. For hospital leaders, that means the front door cannot be understood through elective cancellations only. The structure of staffing, the timing of admissions, and the pressure to meet performance targets all shaped the observed outcomes.
Unmeasured adaptation remains a major part of the story
The most important unresolved point is the portion of the resident doctor effect that the mediation model could not explain. Because the study found a sizable reduction in ED duration with only a small fraction attributable to the measured pathways, the remaining effect likely sits in operational behavior that is harder to capture in routine data. That may include altered diagnostic intensity, different patterns of consultant oversight, modified use of observation areas, or changes in discharge practice during periods of service disruption.
For clinical services, the message is direct: industrial action did not affect all emergency departments in the same way, and it did not do so through one simple channel. Doctor strikes, especially when resident and consultant staff were involved, were associated with shorter stays in this dataset, but the mechanism was only partly explained by bed availability and throughput pressure. Nurse strikes did not show the same signal, and planned care reductions did not drive the observed emergency department changes. The study leaves hospitals with a blunt but important question about resilience: which parts of emergency flow can be compressed without harming care quality, and which apparent efficiency gains merely reflect operational strain that routine data cannot fully reveal?