2025/12/26 by Christopher Oddy · 1 voice
Health Professions · Medicine · #Healthcare Operations and Scheduling Optimization #Sepsis Diagnosis and Treatment #Trauma and Emergency Care Studies
paper · pdf · doi:10.1111/anae.70113
openalex created_date 2025/12/26 · openalex publication_date 2025/12/26 · openalex updated_date 2026/03/11
In a previous study, we showed that surgical enhanced care facilities in the UK serve a population with different characteristics to critical care [1]. On average, patients referred to enhanced care: undergo less complex surgery; have fewer comorbidities; experience a shorter duration of hospital stay; and are referred earlier, reflecting the emphasis placed on planned care within these facilities. Although the enhanced and critical care populations show clear distinctions when regarded as a whole, the delivery of enhanced care in the UK is diverse, creating heterogeneity that prevents population averages from representing individual units. In this planned secondary analysis of our dataset, we aimed to describe how the population characteristics vary between individual enhanced care units and their respective critical care facilities. Our dataset represents a nationwide (n = 110) evaluation of enhanced and critical care services in the UK. From these data, sites with an enhanced care facility with > 10 referrals during the 3-month study period were selected for analysis. Population characteristics were displayed graphically using bidirectional stacked bar charts. Values from enhanced and critical care facilities within each institution were plotted adjacent to each other to permit direct comparison. Institutions with multiple sites or facilities of the same type were aggregated. Operative complexity was categorised according to AXA Health Specialist Procedure Codes [2] and comorbidity status for each patient was calculated using the Charlson comorbidity index [3]. Within our dataset, 28 sites for which full data were available had an enhanced care unit with > 10 referrals during the study period. At these sites, 2902 (71%) patients were referred to enhanced care and 1210 (29%) to critical care (Fig. 1). At most sites, the proportion of patients falling into the highest categories for each parameter was greater within critical care. This trend is most noticeable for the Charlson comorbidity index and duration of hospital stay. However, there was also marked heterogeneity among enhanced care facilities within our sample. The population characteristics of individual units were often more similar to those present within critical care at the same hospital than enhanced care units elsewhere. This was identifiable most clearly for surgical complexity and timing of referral. Our sample included enhanced care services caring for a patient cohort with comparable clinical acuity to critical care. The two sites caring for the populations with the highest average Charlson comorbidity index admitted all their high-risk surgical cases to their enhanced care facility. This was also the case for the sites caring for populations with the lowest average Charlson comorbidity index. These four sites differed in that the two sites caring for the most complex patient populations had critical care units on-site, whereas the lower acuity units did not. We have shown that the population characteristics of individual enhanced care facilities in the UK differ significantly between institutions. The observed differences between the enhanced and critical care populations appear to arise from small but consistent differences in case selection within institutions, with higher-risk patients streamed more frequently towards critical care. Although the clinical acuity within enhanced care was lower at most sites, none of the institutions in our sample showed binary stratification of the highest-risk patients to critical care, with several enhanced care facilities caring for a population with similar characteristics to critical care. National guidelines on the development of enhanced care services in the UK suggest tailoring these facilities to serve a population with intermediate care needs [4]. Naturally, this raises the question of how to define ‘intermediate’. There are numerous factors that influence postoperative care needs, where interactions are complex when determining risk. It is unclear, for example, whether a patient who is fit and well but undergoing complex surgery should be considered intermediate risk. Or, conversely, whether this classification should apply to an individual living with significant comorbidities undergoing minor surgery. Current guidance encourages local determination of these thresholds [4]. Our analysis has shown that, in current practice, this determination process does not separate the enhanced and critical care populations reliably based on superficial categorisations like surgical complexity. This implies that a more nuanced approach is applied, perhaps guided by risk stratification tools. The absence of an evidence base on which to determine admission thresholds within enhanced care risks limiting the value added by these services by under- or overutilisation of resources. Future work should seek to establish which patient populations can be supported safely within enhanced care and how service design is matched to their needs, with consideration of the potential economic benefits of providing tailored care within environments that are often less resource intensive than critical care [5]. This study was supported by a grant from the MPS Foundation, led by Adrian Jackson, to which the authors would like to extend their thanks. Volunteers from the charity ICUsteps supported this project in its development phase. The authors would also like to thank Christina Jones, Simon Newton-Smith, Barbara Buck and Jane Glineur for their involvement in the project. We ask that all contributors (online Supporting Appendix S1) are recognised as part of any future work where this dataset is utilised. No competing interests declared. Appendix S1. REPACC study investigators. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.