Predictive validity and reliability of the Poisoning Severity Score in a pediatric poisoned population: a multicenter retrospective study

doi:10.21203/rs.3.rs-4209125/v1

Download PDF

Research Article

Predictive validity and reliability of the Poisoning Severity Score in a pediatric poisoned population: a multicenter retrospective study

https://doi.org/10.21203/rs.3.rs-4209125/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background: The Poisoning Severity Score (PSS) aims to assess the severity of poisoned patients based on their clinical presentation. In the absence of a validated severity score for assessing pediatric intoxications in clinical toxicology, we aimed to evaluate the predictive validity and reliability of the PSS in measuring poisoning severity and toxicity progression in poisoned children.

Methods: We conducted a multicenter retrospective cohort study using health records data from children aged 0 to 17 years poisoned by a carbo-adsorbable substance and managed in the emergency department at three academic centers in Québec, Canada. Logistic regression was used to estimate the accuracy of the initial PSS and the 12-hour delta PSS to predict hospital admission, intensive care unit admission, a hospital stay ≥ 12 hours, and the need for follow-up after discharge. Its performances were compared to that of the Pediatric Logistic Organ Dysfunction Score (PELODS) using the DeLong test for paired data. Inter-rater reliability of the PSS was measured using weighted Kappa coefficients.

Results: Of the 469 subjects included in the study, an initial PSS of 0 was observed for 97% of non-admitted subjects versus 58% of admitted subjects (p<0.05). The AUC-ROC of the PSS ranged from 0.626 (95% CI 0.581-0.670) to 0.690 (95% CI 0.649-0.506) for hospital admission. The AUC-ROC of the PSS were < 0.75 for all outcomes in the unadjusted models but significantly superior or equivalent to those of the PELODS (p≤0.05). Inter-rater reliability was good for the initial PSS (Kappa 82%) and moderate for the 12-hour delta PSS (Kappa 77%).

Conclusion: The Poisoning Severity Score appears to be a valid and reliable score to measure the severity of poisoning and the toxicity progression in poisoned children presenting to the Emergency Department within 12 hours of ingestion of a carbo-adsorbable substance, even if its discrimination remains poor.

Poisoning Severity Score

child poisoning

predictive validity

inter-rater reliability

Pediatric Logistic Organ Dysfunction Score

retrospective study

In Canada, poisoning is the third-leading cause of unintentional injury hospitalization in children [1,2]. Although deaths are uncommon, data from the United States indicate that approximately one million poison exposures occur annually among children under 12 years of age [3]. Medication is one of the most common agents responsible for child poisoning, and several drugs are known to be carbo-adsorbable (i.e., adhere to activated charcoal, a gastrointestinal decontaminant) [1-5].

For both clinical practice and research in clinical toxicology, the use of a reliable, valid, and well-calibrated severity indicator is essential to measure the severity of poisoning and the risk of toxicity progression [6-8]. However, a well-accepted method for assessing the severity of pediatric acute intoxications is still lacking [7]. The Pediatric Logistic Organ Dysfunction Score (PELODS) is the clinical reference score to assess the severity of poisoned children with acute organ failure risk [9-12]. However, this score was designed to describe the severity of organ dysfunction, and other scoring systems could be more appropriate for poisoned children, especially for less severe cases.

The Poisoning Severity Score (PSS) was developed in the 1990s to help comparison between poisoning exposures and documented poison cases [8,13]. It has been validated for telephone assessment of the severity of intoxicated subjects [13,14]. Since then, the PSS has been used worldwide in both retrospective and prospective studies [8]. Provided that it is valid and reliable, this score could be an excellent indicator of the severity of poisoning in medical practice and clinical toxicology research in children [8]. However, it has never been validated in this population [7].

We aimed to assess the predictive validity and reliability of the PSS to measure the severity of intoxication and the progression of toxicity in children poisoned by a carbo-adsorbable substance. We hypothesized that the PSS was better than the PELODS to predict poisoning severity and toxicity progression. Specific objectives were to (1) assess the predictive validity of the PSS to measure the severity of poisoning and the progression of toxicity in acutely poisoned children; (2) compare the discrimination and calibration capacities of the PSS to that of the PELODS; and (3) assess inter-observer agreement in scoring the PSS at arrival in the Emergency Department (ED) and during hospitalization.

This study was approved by the Research Ethics Board of the CHU of Québec – Université Laval.

Study design, setting, and participants

We conducted a retrospective cohort study using data we had already collected in three academic centers in the province of Québec, Canada. The research program for whom data were primary collected aimed at evaluating the risks and benefits associated with the use of a single dose of activated charcoal in patients presenting to the ED in the first twelve hours after the ingestion of a potentially toxic dose of a carbo-adsorbable substance (both adults and children were involved). We applied four exclusion criteria for the initial data collection: (1) ingestion > 12 hours at arrival in the ED or unknown time of ingestion, (2) route of exposure other than ingestion, (3) the ingested substance was not carbo-adsorbable or was a corrosive agent [15], (4) a method of gastrointestinal decontamination other than oral activated charcoal was used (e.g., gastric lavage or whole bowel irrigation). For this present retrospective multicenter cohort study, we used data from the pediatric cases of this poisoned cohort. We included all children aged 0 to 17 years who presented to the ED between January 2013 and December 2016 with acute intoxication following ingestion of a potentially toxic dose of a carbo-adsorbable substance. No additional exclusion criterion was applied.

Data collection

We traced patient records in the archives of the health care institutions using ICD-10 codes related to intoxications with cross-referencing to data archived at the local poison control center. Data were extracted from electronic or paper medical records using a standardized extraction grid by trained abstractors.

Poisoning severity scores

The PSS was measured using clinical data of the poisoned cases to define severity according to five categories: 0 for asymptomatic, levels 1 to 3, respectively, for minor, moderate, and severe intoxication, and 4 for death [13]. Biological and hemodynamic parameters of the patients were used to calculate the PELODS, between 0 and 71 points. Three levels of severity were then applied for the PELODS according to literature: 1 for low severity (<10 points), 2 for moderate severity (10-19 points), and 3 for high severity (≥20 points) [9-12]. We measured both scores every 12 hours during the first 72 hours of hospitalization or until discharge, depending on the first occurring event. Data abstractors were asked to assess a minimal score of PSS and PELODS for each time according to data collected in the medical record, along with a maximal PSS and PELODS under the hypothesis of missing information in the charts. We assumed that the minimal scores were the most representative of the severity of poisoned children and used them in the main analyses under the assumption that clinicians reported significant events and symptoms in the charts. Two different measures were used for our study: the initial PSS (PSS(t0)) and the 12-hour delta PSS (∆PSS = PSS(t0) - PSS(t12)). For comparison purposes, we measured the same parameters with the PELODS. Two assessors independently reviewed all case records from one hospital facility (14% of the study population) to assess the reliability of the PSS measures.

Outcomes

Our primary outcome was hospital admission after visiting the ED. Our secondary outcomes were the need for vasopressors, need for extra renal replacement therapy, use of mechanical ventilation, admission to the intensive care unit (ICU), hospital length of stay (LOS) ≥ 12 hours (ED stay included), discharge status, and death.

Statistical analyses

Given the low rate of deaths and severely affected subjects, we aggregated levels 2 and 3 of the PSS to obtain a 3-category PSS(t0) in our analyses. We categorized the ∆PSS and the ∆PELODS as decreasing, stabilizing, or increasing scores between t0 and t12.

To address our first objective of assessing the predictive validity of the PSS, we evaluated its discrimination and internal calibration to predict our outcomes of interest. First, we performed a Receiving Operating Curve (ROC) analysis with the PSS(t0) or the ∆PSS. We used the Area Under the Receiving Operating Curve (AUC) and its 95% confidence interval (95% CI) to assess the internal discriminatory ability of each score [16,17]. We interpreted the AUCs according to the thresholds preferred in clinical epidemiology: poor discrimination if <0.70, acceptable between 0.70 and 0.79, excellent between 0.80 and 0.89, and outstanding if ≥ 0.90 [18]. Then, the plot of predicted versus observed values for each model allowed a graphical analysis of the internal calibration of each score. Spiegelhalter's goodness-of-fit test for sparse data was also performed [16,17,19,20].

To address our second objective, we compared the predictive validity of the PSS to that of the PELODS, first by repeating the ROC analysis using the PELODS(t0) and the ∆PELODS instead of the PSS, and then independently of the other potential predictors of the outcomes studied. For this second step, we added to our logistic models seven potential predictors previously identified using a directed acyclic graph constructed from the literature and validated by experts: sex, age group, presence of comorbidities, ingestion of ≥ 2 toxic substances, poisoning mechanism, the delay between the intoxication and arrival at the ED, and the ED physician's main specialty. We compared the ROC curves of the PSS and the PELODS in all models, with and without the other predictors, using the DeLong test for paired data [21], and we graphically assessed the calibration goodness-of-fit [19]. We performed the Hosmer and Lemeshow test on adjusted models [19,22].

We measured the inter-rater reliability for our third objective by calculating the Cicchetti-Allison weighted Kappa coefficients for the PSS at each measurement time [23]. The level of agreement was considered moderate between 60 and 79%, high between 80 and 90%, and almost perfect if > 90% [24].

Subgroup and sensitivity analyses

We performed a subgroup analysis for children aged 0 to 5 years to investigate the internal validity of the PSS in these children who are particularly vulnerable to accidental poisonings [2,4]. We performed sensitivity ROC analyses by excluding subjects transferred from another hospital, subjects who left the ED against medical advice, and subjects treated with antidotes or activated charcoal. We also performed a sensitivity analysis by using the mean PSS and PELODS values at ED arrival and after 12 hours in each model, instead of the minimum values. These mean values consisted of an average between the minimum and the maximum score for each time. We equally performed a complete case analysis for comparison between the two poisoning scores.

Sample size

To be able to detect a difference ≥ 1.5% between the AUCs of PSS and PELODS using the DeLong test for paired data, we calculated that we needed to include from 442 to 454 children assuming that respectively 25% to 20% of the children presented the clinical outcome of interest (type I error of 5% and power of 80%).

Treatment of missing data

Data on the PSS and PELODS scores at 12 hours were missing for 280 unadmitted participants (59.7%) discharged before the 12-hour assessment and one admitted child (0.002%). Considering that the 12-hour missing data could be directly explained by baseline information and discharge status, the mechanism of Missingness At Random (MAR) was plausible [25]. We simulated these missing data using multiple imputations [25], with the Multiple Imputations by Chained Equations (MICE) method [25,26]. We entered all independent variables and the potential baseline predictors of the missing values in the imputation models and simulated 60 imputations for each missing value [25]. We performed all statistical analyses using SAS software v9.4 (SAS Institute Inc., Cary, NC, USA). A type I error of 5% was considered.

Patient selection

Between January 2013 and December 2016, 488 children presented to the ED of the three academic centers for poisoning with a carbo-adsorbable substance. Of these, 19 were excluded (3.9%) for a final study sample of 469 children (Figure 1).

Poisoning severity assessment and study outcomes

On arrival at the ED, 457 patients (97,4%) were asymptomatic or had minor symptomatology according to the PSS, and 455 subjects (97.0%) were rated with a low-severity PELODS (Table 1). Almost 96% of the admitted patients experienced a decreased or stable toxicity progression, whereas 59 unadmitted patients (17.9%) had a slight increase of their PSS from 0 to 1 (n = 57) or from 0 to 2 (n=2) (Table 1, Figure 2). Among the 469 poisoned children, 109 were admitted to the ICU (23.3%), 189 were hospitalized for 12 hours or more (40.3%), and 83 required a follow-up after discharge or a psychiatric facility hospitalization (17.7%). Only two of the admitted patients needed vasopressors (0.4%), eight subjects required mechanical ventilation (1.7%), but none required extrarenal epuration (Table 1). No deaths occurred in our cohort (Table 1).

Predictive validity of the PSS

Discrimination of the PSS was poor (<0.70) for all outcomes, both for the PSS(t0) and the ∆PSS (Tables 2-3). In terms of calibration, the PSS(t0) tended to underestimate the probability of hospital admission (Figure 3). In contrast, it slightly overestimated the risk of a LOS ≥ 12 hours (Appendix. Figure 2). The ∆PSS showed an excellent calibration for the risk of hospital admission (Figure 3) and acceptable calibration for ICU admission and LOS (Appendix, Figures 1-2). Both initial and delta scores were not well calibrated when predicting the probability of follow-up after discharge (Appendix, Figure 3). Nevertheless, all goodness-of-fit tests were non-significant, and the 95% confidence interval included the reference axis in all graphics.

Comparison to the PELODS

In the primary ROC analysis, the PSS was significantly superior to the PELODS to predict hospital admission, ICU admission, and LOS (Tables 2-3). For post-discharge follow-up, PSS(t0) was significantly superior to PELODS(t0) (Table 2), whereas the 12-hour delta scores were equivalent (p>0.05) (Table 3). When adjusting for other potential predictors, the discrimination and calibration of both PSS and PELODS increased to >0.90, considered to be excellent (Tables 2-3, Figure 2). However, the PSS remained significantly superior to the PELODS for predicting hospital admission (Tables 2-3). The superiority or equivalence of the PSS to the PELODS remained unchanged in all sensitivity analyses performed, including in the subgroup of 261 poisoned children aged 0 to 5 years (Appendix, Table 1) when using the mean values of the poisoning scores (Appendix, Tables 5a-5b), and after exclusion of the different subgroups of patients mentioned above (Appendix, Tables 2-4). However, a type II error due to smaller sample sizes cannot be excluded for sensitivity analysis, including the complete case analysis (Appendix, Tables 6a-6b).

Inter-rater reliability

The inter-rater agreement for the initial PSS was high (>0.80), and it was moderate (77%) at the 12-hour measurement (Table 4).

In our population of 469 children acutely intoxicated by a carbo-adsorbable substance, the PSS showed good internal calibration but poor discrimination to predict hospital admission, ICU hospitalization, and a longer stay in the hospital. However, the PSS performed significantly better than the PELODS. In addition, inter-rater agreement of the PSS score in the first 12 hours of hospitalization showed good reliability.

In 1998, the Birmingham Center of the United Kingdom National Poisons Information Services evaluated the predictive value of the PSS for the first time in a prospective study [14]. The initial PSS was reported as accurate to assess clinical severity and the risk of subsequent deterioration [14]. However, no ROC curve analysis was performed [14]. In the last decade, a few studies have compared the PSS to other severity scores, such as the Glasgow Coma Scale [27,28] and the APACHE II [28-30]. Four studies included organophosphates [27-29,31] or carbamate poisonings [29], which are uncommon exposures in North America [2,3]. They observed that the PSS performed well compared to other severity scores. Still, suboptimal analytical methods were used [27-29,31], except for one retrospective study of 396 patients poisoned by organophosphates that observed poor discrimination of the PSS with an AUC for mortality of 0.67 [30]. More recently, the PSS was compared to the Sequential Organ Failure Assessment (SOFA) score in adults patients poisoned by pesticides [32], carbon monoxide [33], or non-specified toxic agents [34]. The PSS was an accurate prognostic tool for various clinical outcomes, but the SOFA score was preferred in most cases for its simplicity and objectivity [32,34]. However, none of these studies focused on child poisoning or carbo-adsorbable poisonings, thus limiting comparability with our results.

Our study has several strengths. First, the similar results observed between the primary analyses using the minimal scores and the sensitivity analysis with mean scores favor a limited information bias. Second, results were consistent across all the clinical outcomes assessed, attesting to the robustness of our main conclusion of the superiority or equivalence of the PSS to the PELODS in our study population. The low-severity profile of carbo-adsorbable poisoning is consistent with the epidemiological knowledge of child poisoning in North America, thus limiting the presence of a selection bias [1,3]. It is also part of the explanation why the PELODS had such a poor discriminatory ability in our cohort: this score was designed to assess the risk and severity of multiple-organ failure in children admitted to the ICU [35]. Even if 23% of our cohort was admitted to the ICU, only 7.8% had a moderate to severe intoxication, thus limiting the discriminating contribution of the PELODS. Although we had to deal with missing data for the 12-hours scores, results in complete-case analyses were consistent with those observed with imputed data, so our strategy allowed us to maximize the power of statistical analyses while limiting attrition bias.

Nonetheless, we are aware of some limitations in our research. We treated both scores as categorical variables and not as continuous: first, to account for the nonparametric distribution of the scores and, second, to facilitate the clinical interpretation of both scores. However, this led to a loss in precision, along with the fact that we had to aggregate some severity levels of the PSS because of the low prevalence of high severe poisonings in our population. Also, the choice of the discharge status as a severity outcome might not have been appropriate considering the role of human and organizational factors not related to the poisoning itself in referring the patient for psychiatric care. The use of retrospective data collected for another purpose is associated with some exclusion criteria that limit the generalizability of our results. Finally, the choice of the PELODS as a comparator is criticizable, given its generalist nature. Still, it seemed the most obvious choice given that organ failure is one of the significant risks associated with acute intoxication. Also, there is no gold-standard reference for assessing the severity of poisonings in children.

Based on our results, the PSS presents some issues but has an interesting potential, and the possibilities of its use are important. Our study opens the path to developing a decision tree incorporating the PSS with other clinical predictors of poisoning severity to contribute to managing poisoned children in the ED. Moreover, we could also derive a PSS score from the Canadian poison control centers' databases to get a standard measure useful for clinical toxicology research in pediatrics. With future research focusing on prospectively validating the PSS for various intoxication profiles and healthcare settings, we will be able to assess the global validity of the PSS in child poisoning and to determine its usefulness in both practice and research in clinical toxicology.

In our multicenter retrospective study, the PSS showed a superior or similar ability as that of the PELODS in predicting hospital admission, ICU admission, and a length of stay of more than 12 hours. Our results suggest that the PSS could be an interesting tool to measure the severity of poisoning and the progression of toxicity in children acutely poisoned by a carbo-adsorbable substance. Future prospective studies are needed to confirm our findings.

AUC: Area Under the Receiving Operating Curve

ED: Emergency Department

ICU: intensive care unit

LOS: hospital length of stay

MAR: Missingness At Random

MICE: Multiple Imputations by Chained Equations

PELODS: Pediatric Logistic Organ Dysfunction Score

PSS: Poisoning Severity Score

ROC: Receiving Operating Curve

SOFA: Sequential Organ Failure Assessment

Contributors: EC, LM, AFT, and MSO were involved in the conception and design. MSO and ECh were involved in the data collection. EC was responsible of the data analysis. EC, LM, AFT, and MSO participated in the interpretation of data. EC drafted the manuscript. EC, LM, AFT, ECh, and MSO revised the manuscript and approved the version published.

-This study was approved by the Research Ethics Board of the CHU of Québec – Université Laval. Given the retrospective analysis of anonymised data, consent was not required.

-All authors consented to publication.

-The authors report no conflict of interest.

-No specific funding was provided for the study, Dr Turgeon is the chairholder of the Canada Research Chair in Critical Care Neurology and Trauma. Dr Moore and Dr St-Onge are recipients of a salary support award from the Fonds de la Recherche du Québec – Santé (FRQS).

-Data are currently stocked in a private database at the CHU de Québec – Université Laval and are not publicly available.

-This work is part of a PhD thesis. The work will be available at Université Laval only after April 13^th 2024.

Parachute. Unintentional injury trends for Canadian children. Parachute. Toronto: Ontario.2016. Available from: https://parachute.ca/wp-content/uploads/2019/06/SKW-Trend-Report.pdf
Jiang A, Belton KL, Fuselli P. Evidence Summary on the Prevention of Poisoning in Canada. Parachute. Toronto: Ontario.2020 [2020-06-07]. Available from: https://parachute.ca/wp-content/uploads/2020/11/Evidence-Summary-on-Poisoning-in-Canada-UA.pdf
Gummin DD, Mowry JB, Beuhler MC, et al. 2020 Annual Report of the American Association of Poison Control Centers’ National Poison Data System (NPDS): 38th Annual Report. Clinical Toxicology. 2021;59(12):1282-1501.
Parachute. White paper on the prevention of poisoning of children in Canada. Parachute. Toronto: Ontario.2011. Available from: https://parachute.ca/wp-content/uploads/2019/07/WhitePaper_Poisoning.pdf
Groth Hoegberg LC, Shepherd G, Wood DM, et al. Systematic review on the use of activated charcoal for gastrointestinal decontamination following acute oral overdose. Clinical Toxicology. 2021;59(12):1196-1227.
Bouch DC, Thompson JP. Severity scoring systems in the critically ill. Continuing Education in Anaesthesia Critical Care & Pain. 2008;8(5):181-185.
Schwarz ES, Kopec KT, Wiegand TJ, et al. Should We Be Using the Poisoning Severity Score? Journal of Medical Toxicology. 2017;13(2):135-145.
Cairns R, Buckley NA. The Poisoning Severity Score: If It Did Not Exist, We Would Have To Invent It. Journal of Medical Toxicology. 2017;13(2):131-134.
Leteurtre S, Duhamel A, Salleron J, et al. PELOD-2: An Update of the PEdiatric Logistic Organ Dysfunction Score. Critical Care Medicine. 2013;41(7):1761-1773.
Leteurtre S, Duhamel A, Grandbastien B, et al. Paediatric logistic organ dysfunction (PELOD) score. The Lancet. 2006;367(9514):897.
Leteurtre S, Martinot A, Duhamel A, et al. Validation of the paediatric logistic organ dysfunction (PELOD) score: prospective, observational, multicentre study. The Lancet. 2003;362(9379):192-197.
Leteurtre S, Duhamel A, Grandbastien B, et al. Daily estimation of the severity of multiple organ dysfunction syndrome in critically ill children. Canadian Medical Association Journal. 2010;182(11):1181-1187.
Persson HE, Sjöberg GK, Haines JA, et al. Poisoning Severity Score. Grading of Acute Poisoning. Journal of Toxicology: Clinical Toxicology. 1998;36(3):205-213.
Casey PB, Dexter EM, Michell J, et al. The Prospective Value of the IPCS/EC/EAPCCT Poisoning Severity Score in Cases of Poisoning. Journal of Toxicology: Clinical Toxicology. 1998;36(3):215-217.
AACT-EAPCCT. Position Paper: Single-Dose Activated Charcoal. Clinical Toxicology. 2005;43(2):61-87.
Collins GS, Reitsma JB, Altman DG, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD): the TRIPOD Statement. British Journal of Surgery. 2015;102(3):148-158.
Moons KG, Altman DG, Reitsma JB, et al. Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD): explanation and elaboration. Annals of internal medicine. 2015;162(1):W1-W73.
Vittinghoff E, Glidden DV, Shiboski SC, et al. Regression methods in biostatistics: linear, logistic, survival, and repeated measures models (Second Edition). Springer Science & Business Media; 2011. (Gail M, Krickeberg K, Samet JM, et al., editors. Statistics for Biology and Health).
Crowson CS, Atkinson EJ, Therneau TM. Assessing calibration of prognostic risk scores. Statistical Methods in Medical Research. 2013;25(4):1692-1706.
Rufibach K. Use of Brier score to assess binary predictions. Journal of Clinical Epidemiology. 2010;63(8):938-939.
DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the Areas under Two or More Correlated Receiver Operating Characteristic Curves: A Nonparametric Approach. Biometrics. 1988;44(3):837-845.
Nattino G, Lemeshow S, Phillips G, et al. Assessing the Calibration of Dichotomous Outcome Models with the Calibration Belt. The Stata Journal. 2017;17(4):1003-1014.
Cicchetti DV. Assessing Inter-Rater Reliability for Rating Scales: Resolving some Basic Issues. British Journal of Psychiatry. 1976;129(5):452-456.
McHugh ML. Interrater reliability: the kappa statistic. Biochemia medica (Zagreb). 2012;22(3):276-282.
Graham JW. Missing Data Analysis: Making It Work in the Real World. Annual Review of Psychology. 2009;60(1):549-576.
White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Statistics in Medicine. 2011;30(4):377-399.
Akdur O, Durukan P, Ozkan S, et al. Poisoning severity score, Glasgow coma scale, corrected QT interval in acute organophosphate poisoning. Human & Experimental Toxicology. 2010;29(5):419-425.
Churi S, Ramesh M, Bhakta K, et al. Prospective Assessment of Patterns, Severity and Clinical Outcome of Indian Poisoning Incidents. Chemical and Pharmaceutical Bulletin. 2012;60(7):859-864.
Sam KG, Kondabolu K, Pati D, et al. Poisoning severity score, APACHE II and GCS: Effective clinical indices for estimating severity and predicting outcome of acute organophosphorus and carbamate poisoning. Journal of Forensic and Legal Medicine. 2009;16(5):239-247.
Peter JV, Thomas L, Graham PL, et al. Performance of clinical scoring systems in acute organophosphate poisoning. Clinical Toxicology. 2013;51(9):850-854.
Churi S, Bhakta K, Madhan R. Organophosphate Poisoning: Prediction of Severity and Outcome by Glasgow Coma Scale, Poisoning Severity Score, Acute Physiology and Chronic Health Evaluation II Score, and Simplified Acute Physiology Score II. Journal of Emergency Nursing. 2012;38(5):493-495.
Yuan S, Gao Y, Ji W, et al. The evaluation of acute physiology and chronic health evaluation II score, poisoning severity score, sequential organ failure assessment score combine with lactate to assess the prognosis of the patients with acute organophosphate pesticide poisoning. Medicine. 2018;97(21):e10862-e10862.
Wang IJ, Yeom S-R, Park S-W, et al. Poison severity score and sequential organ failure assessment score: Carbon monoxide poisoning prognosis. PLOS ONE. 2019;14(3):e0212025.
Ebrahimi K, Vaisi Raigani AA, Jalali R, et al. Determining and Comparing Predictive and Intensity Value of Severity Scores - "Sequential Organ Failure Assessment Score," "Acute Physiology and Chronic Health Evaluation 4," and "Poisoning Severity Score" - in Short-Term Clinical Outcome of Patients with Poisoning in an ICU [Article]. Indian Journal of Critical Care Medicine. 2018;22(6):415-421.
Leteurtre S, Lampin ME, Grandbastien B, et al. Les scores de gravité généraux et de dysfonctions d’organes en réanimation pédiatrique : quoi de neuf en 2016 ? Médecine Intensive Réanimation. 2016;25(6):604-618.

Tables 1 to 4 are available in the Supplementary Files section.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
09 Jul, 2024
Reviews received at journal
26 Jun, 2024
Reviews received at journal
15 Jun, 2024
Reviewers agreed at journal
01 Jun, 2024
Reviewers agreed at journal
01 Jun, 2024
Reviewers invited by journal
29 May, 2024
Editor invited by journal
07 May, 2024
Editor assigned by journal
07 Apr, 2024
Submission checks completed at journal
07 Apr, 2024
First submitted to journal
02 Apr, 2024

You are reading this latest preprint version

Predictive validity and reliability of the Poisoning Severity Score in a pediatric poisoned population: a multicenter retrospective study

Status:

Version 1

Abstract

Figures

Background

Methods

Study design, setting, and participants

Data collection

Poisoning severity scores

Outcomes

Statistical analyses

Subgroup and sensitivity analyses

Sample size

Treatment of missing data

Results

Patient selection

Poisoning severity assessment and study outcomes

Predictive validity of the PSS

Comparison to the PELODS

Inter-rater reliability

Discussion

Conclusion

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1