Improving gastrointestinal scoring systems for predicting short-term mortality in critically ill patients

Abstract

Shen et al’s retrospective study aims to compare the utility of two separate scoring systems for predicting mortality attributable to gastrointestinal (GI) injury in critically ill patients [the GI Dysfunction Score (GIDS) and the Acute Gastrointestinal Injury (AGI) grade]. The authors note that this study is the first proposal that suggests an equivalence between the ability of both scores to predict mortality at 28 days from intensive care unit (ICU) admission. Shen et al retrospectively analysed an ICU cohort of patients utilising two physicians administering both the AGI grade and GIDS score, using electronic healthcare records and ICU flowsheets. Where these physicians disagreed about the scores, the final decision as to the scores was made by an associate chief physician, or chief physician. We note that the primary reason for the development of GIDS was to create a clear score for GI dysfunction, with minimal subjectivity or inter-operator variability. The subjectivity inherent to the older AGI grading system is what ultimately led to the development of GIDS in 2021. By ensuring consensus between physicians administering the AGI, Shen et al have controlled for one of this grading systems biggest issues. We have concerns, however, that this does not represent the real-world challenges associated with applying the AGI compared to the newer GIDS, and wonder if this arbitration process had not been instituted, would the two scoring systems remain equivalent in terms of predicted mortality?

Key Words: Gastrointestinal injury; Critical care; Patient mortality prediction; Gastrointestinal Dysfunction Score; Acute Gastrointestinal Injury grade; Intensive care unit scoring systems

Core Tip: In order to ensure the feasible application of scoring systems in critically ill patients, they must be simplistic and pragmatic to ensure transferability from an academic setting to clinical implementation. We propose such a modification of the current system forthwith.

TO THE EDITOR

We were delighted to read this high-quality article by Shen et al[1] published in the World Journal of Gastroenterology. This study, retrospective in nature, looked at applying two separate scoring systems for predicting mortality attributable to gastrointestinal (GI) dysfunction in critically ill patients - namely the Acute Gastrointestinal Injury (AGI) grade and the GI Dysfunction Score (GIDS). The study involved all eligible patients admitted to an intensive care unit over a period of 3 months, yielding a total of 178 patients included for analysis.

Two attending physicians assessed each patients’ AGI grade and GIDS for the 1^st week of admission, utilising electronic records and critical care flowsheets. When these assessors disagreed in their scores for a patient, they were required to justify their scores in the presence of an associate chief physician/chief physician. This chief physician made the final decision on a patients AGI grade or GIDS. Twenty-eight-day mortality was then ascertained, and the ability of AGI grade and GIDS for predicting this mortality was assessed and ultimately found to be comparable.

A brief synopsis of the different scoring systems involved

As noted by Shen et al[1] , the ability for critical care physicians to stratify disease severity and prognosticate short term mortality is a key clinical concern. To this end, multiple scoring systems have developed over the years to help with these issues, perhaps the most well-known and widely utilized being the APACHE II[2,3] and SOFA scores[4] , although a myriad of others do also exist.

GI dysfunction is highly prevalent in critically ill patients and is increasingly being recognised as a predictor of poor outcomes[5]. To this end, in 2012, an ESICM working group published a statement on GI dysfunction with associated terminology. This publication also gave us the AGI grade[6]. The AGI grade ranged from 1 (risk of developing GI dysfunction or failure) up to 4 (GI failure with severe impact on distant organ function).

While useful, the AGI has been criticized for its ‘subjective assessment of patient’s condition’ by none other than some of the original authors[7], who also noted that its ‘observer dependency is high’. In addition, they remark that the AGI ‘focuses on feeding intolerance – a poorly defined entity that is dependent on local feeding practices’[8].

To this end, in 2021 the GIDS was developed in an effort to provide a ‘readily available bedside score with minimised subjectivity and maximised reproducibility’[7]. This paper involved some of the same individuals involved in the original ESICM working group that produced the original AGI grading system. This GIDS system ranged from 0 (no risk of GI dysfunction) to 4 (life threatening GI dysfunction), and involved objective measures varying from the presence or lack of bowel sounds up to GI bleeding resulting in haemorrhagic shock.

Preliminary research has shown some marked discrepancies between the values obtained for a given patient when utilising AGI grade vs GIDS – with one study showing that more than 20% of patients categorised as GIDS I-II being classified as AGI III-IV[9].

In summary, GIDS was developed as a more objective approach to quantifying GI dysfunction in the critically ill, as the older AGI score was felt to be too subjective and could be influenced by other organ dysfunctions[7].

A word on the attribution of these scores within the study

Upon reviewing Shen et al’s paper[1], we do have some observations about the manner of application of these scores to patients that may perhaps have an impact on the ultimate findings of the study. Firstly, it seems that Shen et al[1] utilised two attending physicians to assign both AGI grades and GIDS to patients included in this study. Given that one score (GIDS) has been developed to be more objective and reproducible than the more subjective AGI, we cannot help but wonder if a researcher’s assessment of the AGI could be influenced by the more objective score, they had obtained from their utilisation of the GIDS? Secondly, by comparing scores obtained for both the AGI grade and the GIDS between two attending physicians, there is potentially a danger that the inherent subjectivity of the AGI grade in particular is controlled for. This is especially true given that where such a discrepancy occurs, both physicians were required to justify their scores, with the ultimate decision for final scores resting in the hands of a third, chief or associate chief physician.

Thus, in the context of this study, the AGI grade has been subject to a high level of scrutiny by no less than three experienced physicians in circumstances where the initial assessors disagreed with each other’s findings. We feel that this level of rigor would likely be difficult to replicate in real-world situations outside of academic purposes, and that most likely the GIDS, with its less subjective approach to grading, would be easier to implement.

Recommendations and conclusion

In order to control for this issue, we advocate that the following measures could potentially be implemented: (1) Assessment of the AGI grade be carried out by one independent physician, and the assessment of the GIDS be carried out by another. In this manner, an individual assessor would be blinded to the scores obtained from the other scoring system, which may potentially have influenced their own scores if they are tasked with performing both systems in tandem; (2) This would then also obviate the need for having a third party arbitrator in the form of the associate chief or chief physician. This would have the benefit of preserving the potential subjectivity of these scoring systems within real-world limitations; and (3) If time and/or resources allowed, having two independent physicians score the AGI grade and another two independent physicians score the GIDS would permit delineation of just how subjective the AGI grade is relative to the GIDS. If the creators of these scoring systems are correct, one might expect that the scores from independent physicians assessing the GIDS to be more consistent, while the scores from independent physicians assessing the AGI grade may vary more widely owing to its inherent subjectivity.