Surgical Hubs Raise Physician Productivity by 14.5%
Ring-fenced elective facilities increased cost-weighted output per physician input while cutting average waits by 10 days, with larger gains from more complete separation.
Underlying Paper
High Volume Low Complexity Surgical Hubs in England: Can They Improve Physician Productivity?
Whether organisational separation of elective and emergency care improves physician productivity remains an open question. Most existing evidence relies on cross-sectional comparisons or volume-based outcomes that cannot isolate efficiency gains from input expansion or provider selection. In contrast, this paper exploits the staggered rollout of NHS England's surgical hub programme, a set of dedicated ring-fenced elective facilities introduced as part of its Elective Recovery Plan, to provide causal estimates of the effect of elective-emergency separation on physician productivity. Using this rollout to implement a heterogeneity-robust difference-in-differences design, we estimate the impact of hub adoption on physician productivity measured as cost-weighted elective output per unit of salary-weighted physician input. We find that hub adoption increases physician productivity by 14.5% relative to the no-hub counterfactual, accompanied by a 10-day (7.6%) reduction in average patient waiting times. The size of these gains depends on how completely elective care is insulated from emergency pathways. Standalone hubs situated on dedicated elective-only sites achieve relatively larger productivity gains than integrated hubs located within main acute hospital sites. Moreover, operating multiple hubs yields gains more than double the overall average, whereas a single hub shows no statistically significant effect. For health systems seeking to raise physician productivity and address elective backlogs, these findings suggest that how separation is implemented, not simply whether it is adopted, shapes the productivity gains it delivers.
Elective surgery competes with emergency admissions for theatres, beds, staff time, and scheduling capacity. England's surgical hub programme was designed to remove part of that interference by creating dedicated elective facilities, but prior evidence has struggled to distinguish an efficiency gain from added inputs, larger providers, or differences in case mix. This study uses the staggered adoption of hubs across NHS hospital trusts to estimate whether separating elective and emergency pathways changes physician productivity.
The authors define productivity as cost-weighted high-volume, low-complexity elective output divided by salary-weighted physician full-time-equivalent input, with output and labour costs expressed in 2019/20 prices. Their preferred estimate implies that hub adoption raises this ratio by 0.468: treated trusts generated about £0.47 more elective output for every £1 of physician input than their estimated no-hub counterfactual. That is a 14.5% increase from a counterfactual productivity level of 3.217 to an observed post-adoption level of 3.685.
Core Contribution
The contribution is an attempt to identify productivity rather than volume alone. A hospital can increase elective activity by hiring more clinicians or treating a different mix of patients without becoming more productive; the paper's numerator-and-denominator measure is intended to separate those mechanisms. The analysis also treats the design of separation as substantive rather than incidental: standalone hubs on dedicated elective-only sites show larger gains than hubs integrated within acute sites, and trusts operating multiple hubs show gains exceeding the programme-wide average, whereas a single hub has no statistically significant effect.
That distinction matters for policy. The evidence does not support reading a hub designation by itself as sufficient. The result is more consistent with the operational claim that protection from emergency-pathway disruption is the active ingredient.
Technical Approach
The study uses monthly hospital-level data from April 2014 through March 2024, comparing hub-adopting trusts with 61 never-treated controls. It applies a heterogeneity-robust, staggered difference-in-differences estimator, comparing each adopting trust with its imputed untreated counterfactual while including hospital fixed effects, month fixed effects, and patient case-mix controls. Standard errors are clustered by hospital.
The event-study sample includes 30 treated units through post-adoption month 10, then declines at longer horizons as late adopters leave the available treatment window. This design is preferable to a simple before-and-after comparison because adoption dates differ and COVID-19 severely disrupted elective activity. The descriptive series shows nearly identical productivity trajectories for adopting and never-treated trusts before the first hub opened in April 2020, followed by a sharp pandemic collapse in both groups. The authors additionally model region-specific effects of the first, Alpha, Delta, and Omicron COVID waves, exclude the April 2020–March 2021 disruption window in one check, and test sensitivity to the 2023–24 junior-doctor industrial action.
The full pre-treatment event study finds individually insignificant lead coefficients, which is consistent with the parallel-trends assumption used for identification.
Results and Analysis
Across six productivity specifications, the estimated average treatment effect remains positive and statistically significant with and without case-mix adjustment and with alternative cost- and salary-weighting choices. In the preferred specification, the 0.468 estimate corresponds to the 14.5% productivity gain. The paper also reports a 10-day reduction in average waiting times, or 7.6% relative to the no-hub counterfactual. Taken together, the output-per-input and waiting-time results make a stronger operational case than a volume result alone: more elective activity appears to be delivered without a proportional increase in physician input.
The timing evidence is less uniform than a single average effect suggests. Effects are modest in the early post-adoption months, become individually significant after month 12, and are directionally consistent later in the horizon. The all-period event study rises at longer horizons, but the authors correctly assign less weight to those estimates because late adopters leave the available post-treatment window and the effective sample declines after month 10.
Industrial-action checks leave the preferred estimate essentially unchanged: 0.467 when strike months are removed, 0.427 when January 2023–February 2024 is excluded, and 0.469 when pre-pandemic cancellation exposure is interacted with strike months. That stability reduces one concern about contemporaneous disruption, though it cannot turn a non-random programme rollout into an experiment. The paper's evidence is persuasive for England's participating trusts and its observed period; extending the estimate to other health systems, specialties, or hub configurations requires further evidence.
Evidence Box
moderateKey Claims
- •Elective-emergency separation increases physician productivity
- •Dedicated elective-only sites produce larger gains than integrated hubs
- •Operating multiple hubs is associated with greater productivity gains
- •Hub adoption reduces elective waiting times
Key Results
- •Preferred productivity ATT of 0.468, equal to a 14.5% increase over a 3.217 counterfactual
- •Observed post-adoption productivity of 3.685 versus 3.217 without hub adoption
- •Average waiting times reduced by 10 days, or 7.6%
- •Industrial-action checks yield ATTs of 0.467, 0.427, and 0.469 versus 0.468 at baseline
Limitations & Caveats
- •Non-random rollout leaves causal identification dependent on parallel trends and model assumptions
- •Later event-study estimates use a declining treated sample after month 10
- •Evidence is limited to English NHS trusts observed through March 2024
- •Elective-cancellation event studies are descriptive because pre-trends depart from zero and 2020/21 data are absent
