Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. BJS 2015; 102: 57-66.

Published: 10th November 2014

Authors: S. Huddart, C. J. Peden, M. Swart, B. McCormick, M. Dickinson, M. A. Mohammed et al.

Background

Emergency laparotomies in the UK, USA and Denmark are known to have a high risk of death, with accompanying evidence of suboptimal care. The emergency laparotomy pathway quality improvement care (ELPQuiC) bundle is an evidence‐based care bundle for patients undergoing emergency laparotomy, consisting of: initial assessment with early warning scores, early antibiotics, interval between decision and operation less than 6 h, goal‐directed fluid therapy and postoperative intensive care.

Method

The ELPQuiC bundle was implemented in four hospitals, using locally identified strategies to assess the impact on risk‐adjusted mortality. Comparison of case mix‐adjusted 30‐day mortality rates before and after care‐bundle implementation was made using risk‐adjusted cumulative sum (CUSUM) plots and a logistic regression model.

Results

Risk‐adjusted CUSUM plots showed an increase in the numbers of lives saved per 100 patients treated in all hospitals, from 6·47 in the baseline interval (299 patients included) to 12·44 after implementation (427 patients included) (P < 0·001). The overall case mix‐adjusted risk of death decreased from 15·6 to 9·6 per cent (risk ratio 0·614, 95 per cent c.i. 0·451 to 0·836; P = 0·002). There was an increase in the uptake of the ELPQuiC processes but no significant difference in the patient case‐mix profile as determined by the mean Portsmouth Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity risk (0·197 and 0·223 before and after implementation respectively; P = 0·395).

Conclusion

Use of the ELPQuiC bundle was associated with a significant reduction in the risk of death following emergency laparotomy.

Full text

Your comments

4 Comments

Kirsty McFarlane

3 years ago

Sir,

We were very pleased to see a trial of a quality improvement programme published in a surgical journal by Huddart and colleagues (1). Increasingly we believe that quality improvement methodology will become the way to improve outcomes in surgery rather than relying solely on randomized controlled trials and the patchy uptake of their outcomes. However, we have concerns about a potentially confounding factor that directly affects the conclusion of the paper.

Before the intervention 24.7% of the patients operated on where classified as ASA grade IV or V but this number dropped in the group after the intervention to 18.5%. This can be contrasted to the change in the P-POSSUM pooled score which increase from 0.197 to 0.223. These two results are conflicting and there needs to be an explanation for this finding. Was the P-POSSUM score recorded in the same way before and after the intervention? Some of the factors that go into the P-POSSUM score are subjective and if different groups (i.e. clerical and medical) were calculating the score then this could introduce bias (2,3). This would be a significant finding, as the crude mortality in the study did not statistically decrease and it was only when the results were adjusted for P-POSSUM that statistical significance was reached. So if the P-POSSUM score is biased then the whole conclusion of this paper could be wrong. If the ASA grading in this study is a more accurate reflection of the health status it would suggest that any significant reduction in mortality was due to healthier patients being operated on rather than better care being delivered.

R. Bethune
A. Mayor

South West Academic Health Science Network
Unit 204, Innovation Centre
University of Exeter
Exeter, EX4 4RN
UK
rob.bethune@nhs.net

References:
1. Huddart S, Peden CJ, Swart M, McCormick B, Dickinson M, Mohammed MA, et al. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg 2015; 102(1): 57-66.
2. Igari K, Ochiai T, Yamazaki S. POSSUM and P-POSSUM for risk assessment in general surgery in the elderly. Hepatogastroenterology 2013; 60(126): 1320-1327.
3. Protopapa KL, Simpson JC, Smith NC, Moonesinghe SR. Development and validation of the Surgical Outcome Risk Tool (SORT). Br J Surg 2014; 101(13): 1774-1783.

Kirsty McFarlane

3 years ago

We read with interest the recent paper by Huddart et al. (1). This study addresses an important area for improving outcome after emergency surgery. However, we seek clarification on a few points.

There were no data regarding patient case mix, pathology encountered, timing of surgery or operation type. Significant differences in these variables before and after implementation of the ELPQuiC bundle may have contributed to the observed improvement in short-term outcome. It is unclear by what mechanism increased consultant input influenced overall care – was this by better peri-operative management, or selection of appropriate surgical candidates? It would be interesting to know if more high risk patients were treated non-operatively after the implementation of the bundle thereby improving the overall characteristics of the surgical group. We acknowledge the use of ASA and P-POSSUM to assess risk in these patients. However, P-POSSUM score over predicts mortality in patients (2) and ASA is a poor marker for overall patient frailty.

In addition, we agree 30-day mortality is a clinically significant outcome measure. However, other clinically relevant outcomes such as unplanned return to theatre, quality of life (e.g. QALY score), and longer-term mortality data were not presented.

Finally, there were a few areas of potential methodological bias in the study. The small sample size of the retrospectively collected pre-bundle implementation group and the predominance of one unit contributing the most data are areas likely to have significantly influenced these results.

Stephen Magill
Jennifer McIIhenny
Alastair McKay
Janso Ouseph Padickakudi
Nathan Stephens
Andrew Jackson
Christopher Ray
Raymond Oliphant
On behalf of the West of Scotland Virtual Journal Club

Department of General Surgery
Gartnavel General Hospital
1053 Great Western Road
Glasgow, G12 0YN
UK
stephen_m1987@hotmail.com

References:
1. S. Huddart, C. J. Peden, M. Swart, B. McCormick, M. Dickinson, M. A. Mohammed, N. Quiney and the ELPQuiC Collaborator Group. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg 2015; 102: 57–66.
2. E. Leung, K McArdle and L.S. Wong. Risk-adjusted scoring systems in colorectal surgery. Int J Surg 2011; 9: 130–135.

Kirsty McFarlane

3 years ago

We would like to thank Drs Behune and Mayor for their comments. Patients undergoing emergency laparotomy vary widely in their perioperative risk and post-operative outcome (1,2). Crude mortality was less in all hospitals in our study after introducing the care bundle, however none reached statistical significance. This alone cannot provide evidence of improvement, or lack thereof, and may be a reflection of the relatively small sample sizes. Risk-adjustment is necessary to assess impact on outcomes, and minimise any impact of differences in risk between the two groups.

All data was collected according to strictly-defined parameters, by designated trained staff. PPOSSUM data was collected, and the risk of death calculated by data collectors. ASA data was that which clinical staff had recorded at the time of operation. Therefore our ASA data is likely to be subject to more inter-rater variability than PPOSSUM. This may explain the discrepancies between ASA and PPOSSUM.

PPOSSUM is a validated risk-prediction tool and is currently recommended for by the Royal College of Surgeons of England (3,4,5). PPOSSUM predicts a (continuous linear) risk of death for each patient based on well-defined physiological and surgical variables. In our study we used pre-operative physiological variables and actual operative findings to predict outcome.

The ASA score is a non-linear, categorical variable. The categories are broadly defined by the impact of concurrent illnesses on functional capacity. It does not take into account the emergence of a procedure, or any patient-specific parameters. It is arguably more subjective than PPOSSUM, and has been reported as lacking accuracy in predicting outcomes in heterogeneous populations (5).

For these reasons we actively chose PPOSSUM, over ASA, to risk adjust outcomes. We hope this reassures readers that observed improvements in risk-adjusted outcome are related to improvements in care rather than differences in risk between the two cohorts.
S Huddart
C Peden
N Quiney
On Behalf of the ELPQuiC Collaborator Group

Department of Anaesthesia
Royal Surrey County Hospital NHS Foundation Trust
Egerton Road
Guildford GU2 7XX
UK
samhuddart@nhs.net

References:
1. Huddart S, Peden CJ, Swart M, McCormick B, Dickinson M, Mohammed MA, et al. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg 2015; 102(1): 57-66.
2. Saunders DI, Murray D, Pichel AC, Varley S, Peden CJ; UK Emergency Laparotomy Network. Variations in mortality after emergency laparotomy: the first report of the UK EmergencyLaparotomy Network. Br J Anaesth 2012; 109(3): 368-375.
3. Anderson ID, Eddleston J, Grocott MPW, Lees NP, Lobo D, Loftus I et al. The Higher Risk General Surgical Patient: Towards Improved Care for a Forgotten Group. Royal College of Surgeons of England and Department of Health: London, 2011.
4. Moonesinghe SR, Mythen MG, Das P, Rowan KM, Grocott MPW. Risk stratification tools for predicting morbidity and mortality in adult patients undergoing major surgery: qualitative systematic review. Anesthesiology 2013; 119: 959-981.
5. Neary WD, Prytherch D, Foy C, Heather BP, Earnshaw JJ. Comparison of different methods of risk stratification in urgent and emergency surgery. Br J Surg 2007; 94: 1300-1305.

Kirsty McFarlane

3 years ago

Dear Sir,

We would like to thank Dr Magill et al. for their interesting comments.

Our quality improvement project was designed to address the high mortality rate identified by the Emergency Laparotomy Network (ELN) and others (1–3). These publications have highlighted problems of delivery of care across the spectrum of emergency general surgery laparotomy patients. For this reason we used the inclusion criteria established by the ELN.

Perioperative risk is dictated by a combination of existing patient factors, pathophysiology and management. This group encompasses a very wide range of all three. PPOSSUM is a validated risk prediction tool that takes into account these factors, including intraoperative findings. In our project there is no significant difference in PPOSSUM between the two groups (4). PPOSSUM is acknowledged to be the best current tool for mortality prediction, despite its tendency to over predict, and is recommended by Royal College of Surgeons of England (RCSE) (5,6).

Dr Magill et al. seek to identify the mechanism by which increased involvement of senior clinicians might improve outcomes. The aim of our quality improvement project was to implement recommendations by the RCSE and test their impact on risk-adjusted mortality (5). It was not our intention to test the efficacy of individual elements. Reliably identifying patients who did not undergo surgery after consultant review (who would previously have done so) proved almost impossible, despite attempts to do so. The project was not intended to change clinical decision-making. We have not identified any evidence to suggest that increased senior involvement pre-operatively significantly changed the cohort of patients after implementation. However, one of the aims of promoting formal risk scoring prior to surgery as recommended by the RCSE is to ensure multidisciplinary awareness of the patient’s potential risk and to plan accordingly (5).

Long-term survival and quality of life measures were unfortunately not within the scope or resources of our project. These will be excellent topics of further research.

Finally, Dr Magill et al. are concerned at the sample size of the populations used and the retrospective data collection. We believed an eight-month collection period would be long enough to see if the use of our care bundle would significantly affect mortality rates. Three of the four hospitals individually achieved statistically significant improvements in risk adjusted outcomes, and unadjusted mortality, both 30-day and in-hospital, improved at all centres (4). We acknowledge in the paper that a limitation is the differing sample size and time intervals of the sample groups. We encourage others to validate our work and further develop solutions to this complex issue.

Dr S Huddart
Professor C Peden
Dr M Swart
Dr B McCormick
Dr M Dickinson
Dr N Quiney
On behalf of the ELPQuiC Collaborator Group

Department of Anaesthesia
Royal Surrey County Hospital NHS Foundation Trust
Egerton Road
Guildford GU2 7XX
UK
samhuddart@nhs.net

References:
1. Saunders DI, Murray D, Pichel a C, Varley S, Peden CJ. Variations in mortality after emergency laparotomy: the first report of the UK Emergency Laparotomy Network. Br J Anaesth 2012; 109(3): 368-375.
2. Al-Temimi MH, Griffee M, Enniss TM, Preston R, Vargo D, Overton S, et al. When Is Death Inevitable After Emergency Laparotomy? Analysis of the American College of Surgeons National Surgical Quality Improvement Program Database. J Am Coll Surg Am Coll Surg 2012; 215(4): 503-511.
3. Vester-Andersen M, Lundstrøm LH, Møller MH, Waldau T, Rosenberg J, Møller a M. Mortality and postoperative care pathways after emergency gastrointestinal surgery in 2904 patients: a population-based cohort study. Br J Anaesth 2014; 1–11.
4. Huddart S, Peden CJ, Swart M, McCormick B, Dickinson M, Mohammed M a, et al. Use of a pathway quality improvement care bundle to reduce mortality after emergency laparotomy. Br J Surg 2015; 102(1): 57-66.
5. The Royal College of Surgeons of England, Department of Health. The Higher Risk General Surgical Patient [Internet]. 2011. Available from: http://www.rcseng.ac.uk/publications/docs/higher-risk-surgical-patient/@@download/pdffile/higher_risk_surgical_patient_2011_web.pdf
6. Moonesinghe SR, Mythen MG, Das P, Rowan KM, Grocott MPW. Risk Stratification Tools for Predicting Morbidity and Qualitative Systematic Review. Anaesthesiology 2013; 119: 959-981.