Derivation and internal validation of a clinical prediction score to predict major effect or death in acute metamfetamine toxicity

Abstract

Objective

Metamfetamine use can cause serious complications or death. We aimed to derive and internally validate a clinical prediction score to predict major effect or death in acute metamfetamine toxicity.

Methods

We performed secondary analysis of 1,225 consecutive cases reported from all local public emergency departments to the Hong Kong Poison Information Centre between 1 January 2010 and 31 December 2019. We split the entire dataset chronologically into derivation (first 70% of cases) and validation (the remaining 30% of cases) cohorts. Univariate analysis was conducted, followed by multivariable logistic regression in the derivation cohort to identify independent predictors of major effect or death. We developed a clinical prediction score based on the regression coefficients of the independent predictors in the regression model and compared its discriminatory performance with five existing early warning scores in the validation cohort.

Results

The MASCOT (Male, Age, Shock, Consciousness, Oxygen, Tachycardia) score was derived based on the six independent predictors: male gender (1 point), age (≥35 years, 1 point), shock (mean arterial pressure <65 mmHg, 3 points), consciousness (Glasgow Coma Scale <13, 2 points), need for supplemental oxygen (1 point), and tachycardia (pulse rate >120 beats/min, 1 point). The score ranges from 0–9, with a higher score indicating higher risk. The area under the receiver operating characteristic curve of the MASCOT score was 0.87 (95% CI 0.81–0.93) in the derivation cohort and 0.91 (95% CI 0.81–1.00) in the validation cohort, with a discriminatory performance comparable with existing scores.

Conclusions

The MASCOT score enables quick risk stratification in acute metamfetamine toxicity. Further external validation is warranted before wider adoption.

Keywords:

• Metamfetamine
• poisoning
• clinical prediction rules
• emergency department
• poison centres

Acknowledgements

The authors would like to thank the Ms Annie Chan and Ms Agatha Ho for their help in data collection and the staff in Hong Kong Poison Information Centre for their help in data retrieval.

Author contributions

RPKL, CKC, and MLT conceived the study, designed the study, and obtained research funding. CKC, MLT, MSHT, and THR supervised the conduct of the trial and data collection. RPKL undertook recruitment of patients and managed the data, including quality control. RPKL, EHYL, ZD, and THR provided statistical advice on study design and analyzed the data. RPKL drafted the manuscript, and all authors contributed substantially to its revision. RPKL takes responsibility for the paper as a whole.

Meetings

This work has not been presented in any meetings

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was supported by a grant from the Beat Drugs Fund Association from the Narcotics Division of the Security Bureau of the Government of the Hong Kong Special Administrative Region (BDF190053). The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.