The ACE (Albumin, CRP, and Endoscopy) Index in Acute Colitis: Simplify to a Better Prognostic Prediction

To the Editors,

We read with great interest your article entitled “The ACE (Albumin, CRP and Endoscopy) Index in Acute Colitis: A Simple Clinical Index on Admission that Predicts Outcome in Patients With Acute Ulcerative Colitis.” ¹ In fact, we believe it is an issue of great clinical relevance in the management of patients with acute ulcerative colitis (AUC). It is crucial to identify patients with a predicable steroid nonresponse (SNR) in order to anticipate the need for rescue therapies and exclude contraindications. The available scores, due to their complexity, are often difficult to use in clinical practice. Moreover, some of these scores included variables only available days after patient admission, not allowing an attempt at prediction of clinical course.

This issue has already been a topic of investigation by our work group who analyzed the accuracy of the different scores available to evaluate the severity of AUC. We performed a retrospective study including all consecutive patients admitted for severe AUC with need for intravenous steroids (IVS; n = 281); 32% (n = 91) had SNR at day 3, with need for medical rescue therapy in 95.6% of cases (88.5% infliximab and 4.6% ciclosporin; 6.9% received vedolizumab; absolute contraindication for other therapies); 7% required surgery (4 of those after nonresponse to infliximab). After calculating the different scores such as Mayo Endoscopic Score (MES), Simple Clinical Colitis Activity Index (SCCAI), Mayo Score (MS), Oxford Score (OS), Lindgren Score (LS), and Edinburgh Score (ES), we concluded that, as previously described by Bernardo S et al, the SL and SE scores were superior in predicting SNR.² After analyzing biochemical and endoscopic variables at admission, we found increased CRP (P < .001), reduced albumin (P < .001), and an increased endoscopic severity (P < .001) are significant to predict SNR. However, when we apply the “ACE Index,” the results were quite inferior, namely in terms of sensitivity (S; score 3, 35%), specificity (SE; 93%), positive predictive value (PPV; 70%), and negative predictive value (NPV; 76%; score 0, S 26%; SE 89%; PPV 82%; VPN 37%). The worst performance of ACE index in our population may be related to an unsystematic evaluation of biochemical variables in the emergency department, namely serum albumin.

As we consider the “ACE index” of extreme clinical relevance in this group of patients, in which timely decisions must be undertaken and should be easy to apply in everyday practice, we propose to prospectively evaluate the accuracy of this score in predicting the SNR in our AUC population.

Conflicts of Interest

All authors disclose no personal conflicts of interest or financial relationships relevant to this publication.

Funding

There is no funding declaration.

References