A risk factor‐based predictive model for linezolid‐induced anaemia: A 7‐year retrospective study

Abstract

What is known and objective: The primary adverse reaction of linezolid is haematological toxicity, leading to thrombocytopenia and anaemia. This study aimed to investigate the risk factors of linezolid-induced anaemia (LI-AN) and establish a predictive model by multivariate logistic regression model analysis to predict LI-AN risks in Chinese adult patients.Methods: Demographic and clinical data of patients who underwent linezolid therapy for more than three days between

Electronic health record-based prediction models for in-hospital adverse drug event diagnosis or prognosis: a systematic review

“… 34 Twenty studies (80%) reported about prognostic models 31–39 , 41–50 , 54 and 5 studies (20%) about diagnostic models. 30 , 40 , 51–53 While 17 studies (68%) investigated a single ADE, 31 , 33–41 , 43–47 , 49 , 50 8 studies (32%) investigated multiple ADEs. 30 , 32 , 42 , 47 , 51–54 Only 1 study (4%) reported the involvement of key stakeholders before model development.…”

“…The included studies used various approaches to retrieve or generate ADE labels and could broadly be categorized into 3 categories: 5 category I studies using ADEs that were registered by healthcare providers (20%), 30 , 33 , 51–53 3 category II studies using adverse event signals followed by a(n) (formal) ADE causality assessment (12%), 43 , 44 , 46 and 15 category III studies using adverse event signals without a(n) (formal) ADE causality assessment (60%). 31 , 34–39 , 41 , 42 , 45 , 46 , 48–50 , 54 Studies in the first category retrieved International Classification of Diseases version 9 or 10 codes (ICD-9 or ICD-10) 55 , 56 that explicitly describe a causal relation between a drug and an adverse symptom (eg, ICD-10 code I95.1: Hypotension due to drugs ).…”

“…In 2 of these studies, formal ADE causality assessments using the Naranjo probability scale were applied. 43 , 44 Studies in the last category only retrieved adverse event signals without a subsequent (formal) ADE causality assessment. Various abnormal clinical findings were used as adverse event signals, among which: ICD-9 or ICD-10 codes that only described certain symptoms, abnormal lab values, abnormal electrocardiogram (ECG) findings, medication administrations combined with abnormal ECG findings, and medication administrations combined with mental status change assessments or skin and pain assessments.…”

“…The total number of participants and the number of participants with 1 or more ADEs (ADE cases) in each EHR database showed high variability. For example, Choudhury et al included 86 674 participants and 1314 ADE cases, 31 while Qin et al 43 included 298 participants and only 32 ADE cases. The high variability in the number of ADE cases across studies resulted in a high variability in the EPV.…”

“…For 4 studies (16%)—however—information about the former fact was implicitly reported as they excluded patients with any missing value. 35 , 36 , 39 , 43 Sixteen studies (64%) discussed the handling of missing data in the candidate predictors, 30 , 34–37 , 39–43 , 45–48 , 50 , 54 most of which used missing data indicators or mean imputation ( Supplementary Appendix S7 ).…”