Development and internal validation of a model for postoperative morbidity in adults undergoing major elective colorectal surgery: the peri-operative quality improvement programme (PQIP) colorectal risk model

Critical care unit bed availability and postoperative outcomes: a multinational cohort study

BACKGROUND

Critical care beds are a limited resource, yet research indicates that recommendations for postoperative critical care admission based on patient-level risk stratification are not followed. It is unclear how prioritisation decisions are made in real-world settings and the effect of this prioritisation on outcomes.

METHODS

This was a prespecified analysis of an observational cohort study of adult patients undergoing inpatient surgery, conducted in 274 hospitals across the UK and Australasia during 2017. The primary outcome was postoperative morbidity at day 7. Logistic regression models were used to evaluate the relationship between critical care admission and patient and health system factors. The causal effect of critical care admission on outcome was estimated using variation in critical care occupancy as a natural experiment in an instrumental variable analysis.

RESULTS

A total of 19,491 patients from 248 hospitals were eligible for analysis, of whom 2107 were directly admitted to critical care postoperatively. Postoperative morbidity occurred in 2829/19,491 (15%) patients. Increasing surgical risk was associated with critical care admission, as was increased availability of critical care beds (odds ratio (95%CI) 1.04 (1.01-1.06), p = 0.002) per available bed; however, the probability of admission varied significantly between hospitals (median odds ratio 3.05). There was no evidence of a difference in postoperative morbidity with critical care admission (odds ratio (95%CI) 0.91 (0.57-1.45), p = 0.710).

DISCUSSION

Postoperative critical care admission is variable and related to bed availability. Statistical methods that adjust for unobserved confounding lowered the estimates of harm previously reported to have been associated with postoperative critical care admission. Our findings provide a rationale for a clinical trial which would evaluate any potential benefits for postoperative critical care admission for patients in whom there is no absolute indication for admission.

Risk factors for complications after emergency surgery for paediatric appendicitis: a national prospective observational cohort study

Appendicectomy is a common procedure in children with a low risk of mortality, however, complication rates and risk factors are largely unknown. This study aimed to characterise the incidence and epidemiology of postoperative complications in children undergoing appendicectomy in the UK. This multicentre prospective observational cohort study, which included children aged 1-16 y who underwent surgery for suspected appendicitis, was conducted between November 2019 and January 2022. The primary outcome was 30-day postoperative morbidity. Data collected included: patient characteristics; comorbidities; and physiological status. Multivariable regression analysis was used to identify independent risk factors for poor outcomes. Data from 2799 children recruited from 80 hospitals were analysed, of which 185 (7%) developed postoperative complications. Children from black and 'other' minority ethnic groups were at significantly higher risk of poor outcomes: OR (95%CI) 4.13 (1.87-9.08), p < 0.001 and 2.08 (1.12-3.87), p = 0.021, respectively. This finding was independent of socio-economic status and type of appendicitis found on histology. Other risk factors for complications included: ASA physical status ≥ 3 (OR (95%CI) 4.05 (1.70-9.67), p = 0.002); raised C-reactive protein (OR 95%CI 1.01 (1.00-1.01), p < 0.001); pyrexia (OR (95%CI) 1.77(1.20-2.63), p = 0.004); and peri-operative oxygen supplementation (OR (95%CI) 4.20 (1.44-12.24), p = 0.009). In the UK NHS, which is a universally accessible healthcare system, ethnicity, but not socio-economic status, was associated with an increased risk of postoperative complications in children having surgery for acute appendicitis. Further evaluations and interventions are required to address this health inequality in keeping with NHS and international priorities.

Systematic development and validation of a predictive model for major postoperative complications in the Peri-operative Quality Improvement Project (PQIP) dataset

Complications are common following major surgery and are associated with increased use of healthcare resources, disability and mortality. Continued reliance on mortality estimates risks harming patients and health systems, but existing tools for predicting complications are unwieldy and inaccurate. We aimed to systematically construct an accurate pre-operative model for predicting major postoperative complications; compare its performance against existing tools; and identify sources of inaccuracy in predictive models more generally. Complete patient records from the UK Peri-operative Quality Improvement Programme dataset were analysed. Major complications were defined as Clavien-Dindo grade ≥ 2 for novel models. In a 75% train:25% test split cohort, we developed a pipeline of increasingly complex models, prioritising pre-operative predictors using Least Absolute Shrinkage and Selection Operators (LASSO). We defined the best model in the training cohort by the lowest Akaike's information criterion, balancing accuracy and simplicity. Of the 24,983 included cases, 6389 (25.6%) patients developed major complications. Potentially modifiable risk factors (pain, reduced mobility and smoking) were retained. The best-performing model was highly complex, specifying individual hospital complication rates and 11 patient covariates. This novel model showed substantially superior performance over generic and specific prediction models and scores. We have developed a novel complications model with good internal accuracy, re-prioritised predictor variables and identified hospital-level variation as an important, but overlooked, source of inaccuracy in existing tools. The complexity of the best-performing model does, however, highlight the need for a step-change in clinical risk prediction to automate the delivery of informative risk estimates in clinical systems.

Assessing Frailty in Gastrointestinal Cancer: Two Diseases in One?

PURPOSEOF REVIEW

This review examines the challenges of treating gastrointestinal cancer in the aging population, focusing on the importance of frailty assessment. Emphasized are the rise in gastrointestinal cancer incidence in older adults, advances in frailty assessments for patients with gastrointestinal cancer, the development of novel frailty markers, and a summary of recent trials.

RECENT FINDINGS

Increasing evidence suggests that the use of a Comprehensive Geriatric Assessment (CGA) to identify frail older adults and individualize cancer care leads to lower toxicity and improved quality of life outcomes. However, the adoption of a full CGA prior to chemotherapy initiation in older cancer patients remains low. Recently, new frailty screening tools have emerged, including assessments designed to specifically predict chemotherapy-related adverse events. Additionally, frailty biomarkers have been developed, such as blood tests like IL-6 and performance tracking through physical activity monitors. The relevance of nutrition and muscle mass is discussed. Highlights from recent trials suggest the feasibility of successfully identifying patients most at risk of serious adverse events. There have been promising developments in identifying novel frailty markers and methods to screen for frailty in the older adult population. Further prospective trials that focus on and address the needs of the geriatric population for early identification of frailty in cancer care, facilitating a more tailored treatment approach. Practicing oncologists should select a frailty assessment to implement into their routine practice and adjust treatment accordingly.

Mismatch of populations between randomised controlled trials of perioperative interventions in major abdominal surgery and current clinical practice

BACKGROUND

Demographics of patients undergoing major abdominal surgery are changing. External validity of relevant RCTs may be limited by participants not resembling patients encountered in clinical practice. We aimed to characterise differences in age, weight, BMI, and ASA grade between participants in perioperative trials in major abdominal surgery and patients in a reference real-world clinical practice sample. The secondary aim was to investigate whether time since trial publication was associated with increasing mismatch between these groups.

METHODS

MEDLINE and Embase were searched for multicentre RCTs from inception to September 2022. Studies of perioperative interventions in adults were included. Studies that limited enrolment based on age, weight, BMI, or ASA status were excluded. We compared trial cohort age, weight, BMI, and ASA distribution to those of patients undergoing major abdominal surgery at our tertiary referral hospital during September 2021 to September 2022. We used a local, single-institution reference sample to reflect the reality of clinical practice (i.e. patients treated by a clinician in their own hospital, rather than averaged nationally). Mismatch was defined using comparison of summary characteristics and ad hoc criteria based on differences relevant to predicted mortality risk after surgery.

RESULTS

One-hundred and six trials (44,499 participants) were compared to a reference cohort of 2792 clinical practice patients. Trials were published a median (IQR [range]) 13.4 (5-20 [0-35]) years ago. A total of 94.3% of trials were mismatched on at least one characteristic (age, weight, BMI, ASA). Recruitment of ASA 3 + participants in trials increased over time, and recruitment of ASA 1 participants decreased over time (Spearman's Rho 0.58 and - 0.44, respectively).

CONCLUSIONS

Patients encountered in our current local clinical practice are significantly different from those in our defined set of perioperative RCTs. Older trials recruit more low-risk than high-risk participants-trials may thus 'expire' over time. These trials may not be generalisable to current patients undergoing major abdominal surgery, and meta-analyses or guidelines incorporating these trials may therefore be similarly non-applicable. Comparison to local, rather than national cohorts, is important for meaningful on-the-ground evidence-based decision-making.

Performance and usability of pre-operative prediction models for 30-day peri-operative mortality risk: a systematic review

Estimating pre-operative mortality risk may inform clinical decision-making for peri-operative care. However, pre-operative mortality risk prediction models are rarely implemented in routine clinical practice. High predictive accuracy and clinical usability are essential for acceptance and clinical implementation. In this systematic review, we identified and appraised prediction models for 30-day postoperative mortality in non-cardiac surgical cohorts. PubMed and Embase were searched up to December 2022 for studies investigating pre-operative prediction models for 30-day mortality. We assessed predictive performance in terms of discrimination and calibration. Risk of bias was evaluated using a tool to assess the risk of bias and applicability of prediction model studies. To further inform potential adoption, we also assessed clinical usability for selected models. In all, 15 studies evaluating 10 prediction models were included. Discrimination ranged from a c-statistic of 0.82 (MySurgeryRisk) to 0.96 (extreme gradient boosting machine learning model). Calibration was reported in only six studies. Model performance was highest for the surgical outcome risk tool (SORT) and its external validations. Clinical usability was highest for the surgical risk pre-operative assessment system. The SORT and risk quantification index also scored high on clinical usability. We found unclear or high risk of bias in the development of all models. The SORT showed the best combination of predictive performance and clinical usability and has been externally validated in several heterogeneous cohorts. To improve clinical uptake, full integration of reliable models with sufficient face validity within the electronic health record is imperative.