Minimum Data Set for a Poisoning Registry: A Systematic Review

Prediction the prognosis of the poisoned patients undergoing hemodialysis using machine learning algorithms

BACKGROUND

Hemodialysis is a life-saving treatment used to eliminate toxins and metabolites from the body during poisoning. Despite its effectiveness, there needs to be more research on this method precisely, with most studies focusing on specific poisoning. This study aims to bridge the existing knowledge gap by developing a machine-learning prediction model for forecasting the prognosis of the poisoned patient undergoing hemodialysis.

METHODS

Using a registry database from 2016 to 2022, this study conducted a retrospective cohort study at Loghman Hakim Hospital. First, the relief feature selection algorithm was used to identify the most important variables influencing the prognosis of poisoned patients undergoing hemodialysis. Second, four machine learning algorithms, including extreme gradient boosting (XGBoost), histgradient boosting (HGB), k-nearest neighbors (KNN), and adaptive boosting (AdaBoost), were trained to construct predictive models for predicting the prognosis of poisoned patients undergoing hemodialysis. Finally, the performance of paired feature selection and machine learning (ML) algorithm were evaluated to select the best models using five evaluation metrics including accuracy, sensitivity, specificity the area under the curve (AUC), and f1-score.

RESULT

The study comprised 980 patients in total. The experimental results showed that ten variables had a significant influence on prognosis outcomes including age, intubation, acidity (PH), previous medical history, bicarbonate (HCO3), Glasgow coma scale (GCS), intensive care unit (ICU) admission, acute kidney injury, and potassium. Out of the four models evaluated, the HGB classifier stood out with superior results on the test dataset. It achieved an impressive mean classification accuracy of 94.8%, a mean specificity of 93.5 a mean sensitivity of 94%, a mean F-score of 89.2%, and a mean receiver operating characteristic (ROC) of 92%.

CONCLUSION

ML-based predictive models can predict the prognosis of poisoned patients undergoing hemodialysis with high performance. The developed ML models demonstrate valuable potential for providing frontline clinicians with data-driven, evidence-based tools to guide time-sensitive prognosis evaluations and care decisions for poisoned patients in need of hemodialysis. Further large-scale multi-center studies are warranted to validate the efficacy of these models across diverse populations.

Comparative Study of Demographic and Toxico-Clinical Factors of Patients with Acute Poisoning Admitted to General Intensive Care Unit versus Specific Intensive Care Unit for Poisoning Cases

Background

There is no dedicated specific intensive care unit (ICU) for poisoning cases due to the small number of poisoned patients in some poisoning centers and patients may hospitalized in the general ICU. In this study, we compared the outcome of hospitalization in poisoning and general ICU, in matched patients to demographical and toxico-clinical factors.

Materials and Methods

This historical cohort study was conducted from September 2020 to January 2022 in the general and poisoning ICUs of Khorshid Hospital affiliated to the University of Medical Sciences, Isfahan, Iran. Patient characteristics, clinical, and toxicological information as well as the therapeutic measures and outcome were collected from hospital medical records and analyzed.

Results

Totally, 178 (60.1% male and 39.9% female) patients met inclusion criteria. Medicines (56.2%) and opioids (25.3%) followed by pesticides (14%) were the most common substances. Suicide was the type of exposure in 78.7% of the cases. Most patients suffered from lung (19.1%) and kidney (15.2%) injuries. The mortality rate was 23.6%. The median length of hospital stay (P-value < 0.001) and duration of ventilator usage was higher (P-value < 0.001) in general ICU compared to specific ICU for poisoning cases. No significant difference with respect to demographic, toxico-clinical variables and mortality rate was found between the two groups.

Conclusion

Among poisoned patients admitted to ICU, reported mortality rate was relatively high. Patients who hospitalized in the specific ICU for poisoning cases have lower length of hospital stay and duration of mechanical ventilation compared to general ICU.

Design and Development of a Hospital-Based Coronary Artery Disease (CAD) Registry in Iran

Background

The incidence of coronary artery disease (CAD), the leading cause of mortality in most developed and developing countries, is increasing. The adoption of hospital registries can improve care delivery and facilitate the management of CAD through better planning, as well as help with outcome assessment through more effective data management.

Objectives

The present study is aimed at designing a hospital-based CAD registry for managing CAD data.

Methods

This developmental study was conducted in three phases. Initially, sources related to CAD registries were reviewed, the results of which were published in two studies. In the next phase, the prerequisites and requisites of the software were determined through a qualitative study. In this phase, the registry dataset was determined by using a questionnaire. Finally, the developed conceptual model of the software was validated. The software was then developed based on the validated conceptual model.

Results

The registry data elements were classified into 13 main categories, including identification data, medical history, and risk factors. The dataset included 171 data elements, including data related to surgical and nonsurgical procedures. The conceptual model was approved by field experts, and the software was developed accordingly.

Conclusion

The steps followed in the present study for developing the CAD registry can be used as an appropriate approach for designing similar hospital-based registries. Considering the pivotal role of the registry in the management of CAD, the routine and systemic use of the registry is suggested in all healthcare centers.

Providing a Population Based Registry Model of Drug Poisoning in Iran

Background

The prevalence of drug poisoning is on the rise in Iran due to the increased public access to drugs. A national drug poisoning registry system is a suitable tool for better management, control, and prevention of drug poisoning.

Objectives

This study aimed to propose a national drug poisoning registry model for Iran.

Methods

This was an applied research conducted in two major phases. In the first phase, all sources pertaining to drug poisoning registries were reviewed, and a national drug poisoning registry model was proposed. In the second phase, this model was validated and finalized using a researcher-made questionnaire and through a two-stage Delphi technique.

Results

The focus of national drug poisoning activities and registry management reached the 100% consensus of experts at the Drug and Poison Information Center of the Food and Drug Organization (Ministry of Health and Medical Education). Goals, data sources, registry system structure, data set, standards, data exchange, registry features, and processes of the proposed model also achieved unanimous expert consensus.

Conclusions

Given the importance of a national drug poisoning registry in gathering, storing, analyzing, and reporting the data of patients, it is essential to provide a framework for evaluating and controlling drug poisoning and for generating valuable data for decision-making. The model proposed herein can offer the information infrastructure for designing and implementing such a system.

Developing a Multifaceted Evaluation Tool for Electronic Prescribing System: A Study from a Developing Country

: Evaluation of electronic prescribing systems (EPS) can contribute to their quality assurance, and motivate users and policy-makers to implement these systems, directly influencing the health of society. An appropriate evaluation tool plays a determining role in the identification of proper EPS. The present study aimed to develop a multifaceted evaluation tool for assessing the EPS. This study was conducted in two main steps in 2018. In the first step, we conducted a literature review to find the main features and capabilities of the prosperous EPS. In the second step, a Delphi method was used for determining the final criteria for evaluating EPS. After preparing a primary questionnaire based on the first step results, 27 expert stakeholders from related fields participated in this 3-phase Delphi study. The narrative content analysis and descriptive statistics were used for data analysis. The final evaluation tool consists of 61 questions in 10 main dimensions, including practical capabilities of the process/user and patient safety, data storage and transfer, prescription control and renewal, technical functions, user interfaces, security and privacy, reporting, portability, hardware and infrastructure, and system failure/recovery. The evaluation tool developed in this study can be used for the critical appraisal of features of EPS. It is recommended that this multifaceted evaluation tool be employed to help buyers compare different systems and assist EPS software vendors in prioritizing their activities regarding the system development. By using this tool, healthcare organizations can also choose a system that improves many aspects of health care.