Signal Detection of Potentially Drug-Induced Liver Injury in Children Using Electronic Health Records

Use of Electronic Health Record Data for Drug Safety Signal Identification: A Scoping Review

INTRODUCTION

Pharmacovigilance programs protect patient health and safety by identifying adverse event signals through postmarketing surveillance of claims data and spontaneous reports. Electronic health records (EHRs) provide new opportunities to address limitations of traditional approaches and promote discovery-oriented pharmacovigilance.

METHODS

To evaluate the current state of EHR-based medication safety signal identification, we conducted a scoping literature review of studies aimed at identifying safety signals from routinely collected patient-level EHR data. We extracted information on study design, EHR data elements utilized, analytic methods employed, drugs and outcomes evaluated, and key statistical and data analysis choices.

RESULTS

We identified 81 eligible studies. Disproportionality methods were the predominant analytic approach, followed by data mining and regression. Variability in study design makes direct comparisons difficult. Studies varied widely in terms of data, confounding adjustment, and statistical considerations.

CONCLUSION

Despite broad interest in utilizing EHRs for safety signal identification, current efforts fail to leverage the full breadth and depth of available data or to rigorously control for confounding. The development of best practices and application of common data models would promote the expansion of EHR-based pharmacovigilance.

Causal Evaluation of Post-Marketing Drugs for Drug-induced Liver Injury from Electronic Health Records

Drug-induced liver injury (DILI) is one of the most common and serious adverse drug reactions that can lead to acute liver failure and death. Detection of DILI and causal estimation of drug-hepatotoxicity association are of great importance for patient safety. This paper proposes a framework for causal estimation of post-marketing drugs for DILI from real-world electronic health record (EHR) data. Randomized clinical trials were replicated at scale by automatically generating different user and non-user cohorts for each potential drug, and average treatment effects (ATEs) of drugs were estimated using targeted maximum likelihood estimation. Ten years of real-world EHRs were used to validate the framework. Of all 1199 single-ingredient drugs analyzed, 7 novel and 7 known drug-hepatotoxicity associations were found to be causal.

Risk evaluation of carbapenem-induced liver injury based on machine learning analysis

INTRODUCTION

Information regarding carbapenem-induced liver injury is limited, and the rate of liver injury caused by meropenem (MEPM) and doripenem (DRPM) remains unknown. Decision tree (DT) analysis, a machine learning method, has a flowchart-like model where users can easily predict the risk of liver injury. Thus, we aimed to compare the rate of liver injury between MEPM and DRPM and construct a flowchart that can be used to predict carbapenem-induced liver injury.

METHODS

We investigated patients treated with MEPM (n = 310) or DRPM (n = 320) and confirmed liver injury as the primary outcome. We used a chi-square automatic interaction detection algorithm to construct DT models. The dependent variable was set as liver injury from a carbapenem (MEPM or DRPM), and factors including alanine aminotransferase (ALT), albumin-bilirubin (ALBI) score, and concomitant use of acetaminophen were used as explanatory variables.

RESULTS

The rates of liver injury were 22.9% (71/310) and 17.5% (56/320) in the MEPM and DRPM groups, respectively; no significant differences in the rate were observed (95% confidence interval: 0.710-1.017). Although the DT model of MEPM could not be constructed, DT analysis showed that the incidence of introducing DRPM in patients with ALT >22 IU/L and ALBI scores > -1.87 might be high-risk.

CONCLUSIONS

The risk of developing liver injury did not differ significantly between the MEPM and DRPM groups. Since ALT and ALBI score are evaluated in clinical settings, this DT model is convenient and potentially useful for medical staff in assessing liver injury before DRPM administration.

Drug-induced kidney injury in Chinese critically ill pediatric patients

Background: Drug-induced acute kidney injury (DIKI) is a common adverse drug reaction event but is less known in pediatric patients. The study explored the DIKI in Chinese pediatric patients using the Pediatric Intensive Care database (PIC).

Method: We screened pediatric patients with acute kidney injury (AKI) using the KDIGO criteria from the PIC and then assessed the relationship between their drugs and DIKI using the Naranjo scale. For the fifteen frequently used DIKI-suspected drugs, we divided patients into drug-exposed and non-exposed groups, using the outcome of whether DIKI was presented or not. Propensity score matching (PSM) was used to control for the effects of four confounders, age, gender, length of hospital stay, and major diagnosis. Unconditional logistic regression was used to identify statistically significant differences between the two groups.

Results: A total of 238 drugs were used 1,863 times by the 81 patients with DIKI during their hospital stay. After screening the Naranjo scale to identify the top 15 suspected DIKI drugs with a high frequency of use, we found that furosemide injection (p = 0.001), midazolam injection (p = 0.001), 20% albumin prepared from human plasma injection (p = 0.004), fentanyl citrate injection (p = 0.001), compound glycyrrhizin injection (p = 0.026), vancomycin hydrochloride for intravenous (p = 0.010), and milrinone lactate injection (p = 0.009) were associated with DIKI.

Conclusion: In critically ill pediatric patients, DIKI is more likely to occur after using furosemide injection, midazolam injection, 20% albumin prepared from human plasma injection, fentanyl citrate injection, compound glycyrrhizin injection, vancomycin hydrochloride for intravenous, milrinone lactate injection.

Detection of pediatric drug-induced kidney injury signals using a hospital electronic medical record database

Background: Drug-induced kidney injury (DIKI) is one of the most common complications in clinical practice. Detection signals through post-marketing approaches are of great value in preventing DIKI in pediatric patients. This study aimed to propose a quantitative algorithm to detect DIKI signals in children using an electronic health record (EHR) database.

Methods: In this study, 12 years of medical data collected from a constructed data warehouse were analyzed, which contained 575,965 records of inpatients from 1 January 2009 to 31 December 2020. Eligible participants included inpatients aged 28 days to 18 years old. A two-stage procedure was adopted to detect DIKI signals: 1) stage 1: the suspected drugs potentially associated with DIKI were screened by calculating the crude incidence of DIKI events; and 2) stage 2: the associations between suspected drugs and DIKI were identified in the propensity score-matched retrospective cohorts. Unconditional logistic regression was used to analyze the difference in the incidence of DIKI events and to estimate the odds ratio (OR) and 95% confidence interval (CI). Potentially new signals were distinguished from already known associations concerning DIKI by manually reviewing the published literature and drug instructions.

Results: Nine suspected drugs were initially screened from a total of 652 drugs. Six drugs, including diazepam (OR = 1.61, 95%CI: 1.43–1.80), omeprazole (OR = 1.35, 95%CI: 1.17–1.54), ondansetron (OR = 1.49, 95%CI: 1.36–1.63), methotrexate (OR = 1.36, 95%CI: 1.25–1.47), creatine phosphate sodium (OR = 1.13, 95%CI: 1.05–1.22), and cytarabine (OR = 1.17, 95%CI: 1.06–1.28), were demonstrated to be associated with DIKI as positive signals. The remaining three drugs, including vitamin K1 (OR = 1.06, 95%CI: 0.89–1.27), cefamandole (OR = 1.07, 95%CI: 0.94–1.21), and ibuprofen (OR = 1.01, 95%CI: 0.94–1.09), were found not to be associated with DIKI. Of these, creatine phosphate sodium was considered to be a possible new DIKI signal as it had not been reported in both adults and children previously. Moreover, three other drugs, namely, diazepam, omeprazole, and ondansetron, were shown to be new potential signals in pediatrics.

Conclusion: A two-step quantitative procedure to actively explore DIKI signals using real-world data (RWD) was developed. Our findings highlight the potential of EHRs to complement traditional spontaneous reporting systems (SRS) for drug safety signal detection in a pediatric setting.

Signal Detection of Potential Hepatotoxic Drugs: Case-Control Study Using Both a Spontaneous Reporting System and Electronic Medical Records

Drug-induced liver injury (DILI) is a common adverse drug event. Spontaneous reporting systems such as the Japanese Adverse Event Report Database (JADER) have been used to evaluate the association between drugs and adverse drug events. However, the association of drugs with adverse drug events may be overestimated due to reporting biases. Therefore, it is important to objectively evaluate the association using liver function test values. The aim of the present study was to predict potential hepatotoxic drugs using real-world data including electronic medical records and the JADER database. A total of 70009 (2779 with DILI and 67230 without DILI) and 438515 (10235 with DILI and 428280 without DILI) Japanese adult patients were extracted from electronic medical records and the JADER database, respectively. Drugs with ≥100 DILI patients in both of the two databases were regarded as suspected drugs for DILI. We used multivariate logistic regression to evaluate the association between the suspected drugs and increased risk of DILI. Among the suspected drugs, broad-spectrum antibiotics such as meropenem, tazobactam/piperacillin and ceftriaxone were significantly associated with an increased risk of DILI, and meropenem had a greater risk of DILI in both of the two databases. Additionally, there were significant associations of mosapride and L-carbocisteine with increased risk of DILI. In addition to well-known associations between antibiotic drugs and DILI, mosapride and L-carbocisteine were found to be new potential signals of drugs causing hepatotoxicity. This study indicates potential hepatotoxic drugs that require further causality assessment.