ORIGINAL ARTICLE
APACHE Score, Severity Index of Paraquat Poisoning, and Serum Lactic Acid Concentration in the Prognosis of Paraquat Poisoning of Chinese Patients Shuyun Xu, PhD, MD,* Hai Hu, PhD, MD,* Zhen Jiang, PhD, MD,* Shiyuan Tang, PhD, MD,* Yuangao Zhou, PhD, MD,† Jie Sheng, MD,* Jinggang Chen, MD,* and Yu Cao, PhD, MD* Abstract: Many prognostic indictors have been studied to evaluate the prognosis of paraquat poisoning. However, the optimal indicator remains unclear. To determine the value of the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, the Severity Index of Paraquat Poisoning (SIPP), and serum lactate levels in the prognosis of paraquat poisoning, we performed a prospective study that enrolled 143 paraquat patients. Data were collected from patients (161) at West China Hospital in Chengdu, China, including details about the patients' general conditions, laboratory examinations, and treatment. Receiver operating characteristic curves for predicting inpatient mortality based on APACHE II score, SIPP, and lactate levels were generated. To analyze the best cutoff values for lactate levels, APACHE II scores, and SIPP in predicting the prognosis of paraquat poisoning, the initial parameters on admission and 7-day survival curves of patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II score greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L at the time of arrival at West China Hospital were compared using the 1-way analysis of variance and the log-rank test. The APACHE II score (5.45 [3.67] vs 11.29 [4.31]), SIPP (2.78 [1.89] vs 7.63 [2.46] h · mg/L), and lactate level (2.78 [1.89] vs 7.63 [2.46] mmol/L) were significantly lower in survivors (77) after oral ingestion of paraquat, compared with nonsurvivors (66). The APACHE II score, SIPP, and lactate level had different areas under the curve (0.847, 0.789, and 0.916, respectively) and accuracy (0.64, 0.84, and 0.89, respectively). Respiratory rate, serum creatinine level, PaCO2, and mortality rate at 7 days after admission in patients with lactate levels greater than or equal to 2.95 mmol/L were markedly different compared with those of other patients (P < 0.05). The predictive value of lactic acid was superior to that of APACHE II score and SIPP for acute oral paraquat poisoning. Key Words: paraquat poisoning, receiver operating characteristic, lactate levels (Pediatr Emer Care 2015;31: 117–121)
P
araquat is an organic heterocyclic contact defoliant and herbicide, which has gained worldwide popularity.1–3 The widespread use of paraquat in the villages of developing countries such as China has led to an increase in the number of paraquat poisoning cases.2 It was reported that the number of poisoning cases reaches an estimated 877,000 every year (World Health Organization, 2003), and many health officials consider paraquat to be a public health threat because of its high toxicity.4,5 The ingestion of 40 mg of paraquat ion/kg body weight is associated with a From the Departments of *Emergency, and †Laboratory Medicine, West China Hospital, Sichuan University, Chengdu, Sichuan Province, People's Republic of China. Reprints: Yu Cao, PhD, MD, Department of Emergency, West China Hospital, Sichuan University, No 37 Guoxuexiang Road/Wuhou District, Chengdu, Sichuan Province, 610041, People's Republic of China (e‐mail: [email protected]). Disclosure: The authors declare no conflict of interest. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0749-5161
mortality rate of up to 100%.1 The clinical manifestations of paraquat poisoning can vary depending on a number of factors, including the dosage and route of entry,2,6 which can also affect the prognosis. There have been reports of survival from mild poisonings with no intervention, but highly toxic doses (35 mg/kg) are almost always fatal despite aggressive interventions, such as hemodialysis and renal replacement therapy.7 Hence, the identification of accurate prognostic indicators is very necessary. Appropriate prognostic indicators could aid physicians in evaluating the severity of poisoning and in predicting overall outcomes. Currently, most primary health care institutional guidelines lack favorable prognostic indicators for paraquat poisoning. Therefore, most physicians predict the outcome of the condition based on the clinical signs and symptoms.8 However, the onset of symptoms after acute paraquat poisoning requires some time, and by then, the poisoning can be irreversible or fatal. Thus, a simple and easy to obtain prognostic indicator for acute paraquat poisoning is crucially important. Many prognostic indictors have been studied to evaluate the prognosis of paraquat poisoning, but the optimal indicators remain unclear. The Acute Physiology and Chronic Health Evaluation II (APACHE II) score is a scoring system based on 12 parameters of acute physiology and chronic health.9 This scoring system has been commonly used for the assessment of the severity and prognosis of diseases,10 especially in intensive care units.11 The APACHE II has also been used to predict the severity and prognosis of poisoning, but Chinese physicians have rarely used it for acute paraquat poisoning cases. The Severity Index of Paraquat Poisoning (SIPP) is also a potentially valuable prognostic indicator, which is calculated by multiplying the time since the ingestion (hour) of paraquat by the paraquat concentration in the plasma (milligram per liter).12,13 Studies have suggested that patients with a SIPP of less than 10 h · mg/L have a good chance of survival, whereas those with a SIPP between 10 and 50 h · mg/L often die because of interstitial pulmonary fibrosis secondary to paraquat poisoning, and those with values greater than 50 h · mg/L die rapidly owing to circulatory collapse.14,15 Although promising, the plasma paraquat level is not widely available in China. Other studies have shown that acidosis is an important prognostic indicator of acute paraquat poisoning,16,17 particularly the concentration of serum lactic acid, which is the primary cause of acidosis. To determine the value of APACHE II scores, SIPP, and serum lactate concentrations in the prognosis of paraquat poisoning, a prospective study, enrolling 143 patients with paraquat poisoning, was performed.
METHODS Study Population Patients with acute oral paraquat poisoning were consecutively selected from those patients who were referred to the emergency department of Sichuan University (Chengdu, China) between January
Pediatric Emergency Care • Volume 31, Number 2, February 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.pec-online.com
117
Xu et al
Pediatric Emergency Care • Volume 31, Number 2, February 2015
2011 and January 2012 (n = 143), according to the following exclusion criteria: transcutaneous or intravenous exposure and undetectable levels of paraquat in the blood. In addition, individuals were also ineligible if they had concomitant diseases, including cancer, cardiac disease, pulmonary disease, renal disease, and liver disease or if they refused to participate in the study (see the flow chart). Signed informed consent was obtained from the patients, family members, or legal guardians.
12 hours until serum paraquat levels decreased to less than 200 ppm. In addition, these patients also received daily pulse therapy via intravenous infusion of 1 g of methylprednisolone, 400 mg of cyclophosphamide, the antioxidants vitamins C (2 g/d) and E (300 mg/d), and cysteine (1.2 g/d). Supplemental oxygen therapy was administered if the oxygen partial pressure was less than 50 mm Hg. Hemodialysis was performed if the patients had severe metabolic acidosis, hyperkalemia, or heart failure.
Data Collection
Statistical Analysis
All of the data were collected by experienced emergency physicians, including details of the patients' general conditions, laboratory examinations, and treatments. On admission, each patient's age, sex and other demographic data, past medical history, amount of paraquat ingested (including other toxic substances ingested), and poisoning time were documented. On admission, vital signs were recorded, and blood samples were collected for the following tests: routine blood work, liver and renal function, other biochemical blood examinations, and arterial blood gas examinations. To ensure interrater reliability, a 20% sample of the data was screened by a second reviewer. The Glasgow Coma Scale score and the APACHE II score were calculated, and the serum lactate level and plasma paraquat concentration were determined. All of the specimens were collected 1 hour after admission. A complete blood cell count and biochemistry were performed, and lactic acid levels were measured using an autoanalyzer (Roche modular P800, Mannheim, Germany). Plasma paraquat concentrations were measured using a high-performance liquid chromatographic system (Shimadzu, Japan). The elapsed time (hours) from the ingestion of paraquat to arrival and the serum paraquat levels were recorded to calculate SIPP (mean elapsed time [hours] serum paraquat level in the emergency department [milligram per liter]). Deaths that occurred in the hospital were recorded, and those patients discharged from the hospital were followed up by telephone or letter. The 60-day mortality rate was used to determine the efficacy of individual prognostic indicators, and the results were summarized by a toxicology expert.
The collected data are presented as mean (SD) for continuous variables and as frequencies for categorical variables. A P value of less than 0.05 was considered to be statistically significant, with all of the statistical analyses performed using SPSS software, version 18.0. Differences in covariates between survivors and nonsurvivors were tested by Student t test for continuous variables and by the χ2 test or Fisher exact probability method for categorical variables. To investigate which prognostic indicator (APACHE II, SIPP, and serum lactic acid level) was the most valuable predictor of prognosis for acute paraquat poisoning, receiver operating characteristic (ROC) curves for predicting inpatient mortality were generated. The areas under the ROC curve (AUCs), sensitivities, specificities, cutoff values, accuracies, and indicators of the predictive value of the tests were calculated using GraphPad Prism software, version 4.0. To analyze the best cutoff values for APACHE II score, SIPP, and lactate level to predict the prognosis of paraquat poisoning, 1-way analysis of variance was used to assess the differences in initial parameters at the time of West China Hospital (WCH) arrival among paraquat poisoning patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II scores greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L. In addition, the 7-day survival curves of patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II scores greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L were compared.
RESULTS Treatment As summarized in Table 1, all of the patients presenting within 6 hours of paraquat ingestion received gastric lavage, followed by 200 mL of 20% mannitol via gastric tube. The patients who presented within 12 hours of ingestion received hemoperfusion every
Patient Characteristics During the study period, 161 patients fulfilled the criteria for acute oral poisoning. Eighteen of these patients were excluded because of missing data or loss to follow-up. The remaining 143
TABLE 1. General Characteristics of 143 Oral Paraquat Poisoning Patients at the Time of WCH Arrival
No. patients* Age, y† Sex* Time between exposure and WCH arrival, h† Estimated amount of paraquat dichloride (24.5% concentrate) exposure, mL† Comorbid disease* Psychiatric condition* Underlying disease* Concomitant alcohol ingestion* Serum lactates, mmol/L† APACHE II score† SIPP, h · mg/L†
All Patients
Survivors
Nonsurvivors
P
143 30.6 (12.7) 54/89 9.9 (5.1) 65.6 (14.1) 47 25 7 1 5.29 (2.54) 8.65 (4.63) 81.07 (67.42)
77 28.0 (12.9) 28/49 11.2 (4.3) 50.1 (21.2) 27 16 5 1 2.78 (1.89) 5.45 (3.67) 15.39 (10.6)
66 33.6 (11.9) 26/40 7.4 (6.9) 76.9 (18.2) 20 9 2 0 7.63 (2.46) 11.29 (4.31) 147.85 (78.56)
0.190 0.666 0.106 0.227 0.546 0.262 0.35 1.000
APACHE Score, Severity Index of Paraquat Poisoning, and Serum Lactic Acid Concentration in the Prognosis of Paraquat Poisoning of Chinese Patients Shuyun Xu, PhD, MD,* Hai Hu, PhD, MD,* Zhen Jiang, PhD, MD,* Shiyuan Tang, PhD, MD,* Yuangao Zhou, PhD, MD,† Jie Sheng, MD,* Jinggang Chen, MD,* and Yu Cao, PhD, MD* Abstract: Many prognostic indictors have been studied to evaluate the prognosis of paraquat poisoning. However, the optimal indicator remains unclear. To determine the value of the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, the Severity Index of Paraquat Poisoning (SIPP), and serum lactate levels in the prognosis of paraquat poisoning, we performed a prospective study that enrolled 143 paraquat patients. Data were collected from patients (161) at West China Hospital in Chengdu, China, including details about the patients' general conditions, laboratory examinations, and treatment. Receiver operating characteristic curves for predicting inpatient mortality based on APACHE II score, SIPP, and lactate levels were generated. To analyze the best cutoff values for lactate levels, APACHE II scores, and SIPP in predicting the prognosis of paraquat poisoning, the initial parameters on admission and 7-day survival curves of patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II score greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L at the time of arrival at West China Hospital were compared using the 1-way analysis of variance and the log-rank test. The APACHE II score (5.45 [3.67] vs 11.29 [4.31]), SIPP (2.78 [1.89] vs 7.63 [2.46] h · mg/L), and lactate level (2.78 [1.89] vs 7.63 [2.46] mmol/L) were significantly lower in survivors (77) after oral ingestion of paraquat, compared with nonsurvivors (66). The APACHE II score, SIPP, and lactate level had different areas under the curve (0.847, 0.789, and 0.916, respectively) and accuracy (0.64, 0.84, and 0.89, respectively). Respiratory rate, serum creatinine level, PaCO2, and mortality rate at 7 days after admission in patients with lactate levels greater than or equal to 2.95 mmol/L were markedly different compared with those of other patients (P < 0.05). The predictive value of lactic acid was superior to that of APACHE II score and SIPP for acute oral paraquat poisoning. Key Words: paraquat poisoning, receiver operating characteristic, lactate levels (Pediatr Emer Care 2015;31: 117–121)
P
araquat is an organic heterocyclic contact defoliant and herbicide, which has gained worldwide popularity.1–3 The widespread use of paraquat in the villages of developing countries such as China has led to an increase in the number of paraquat poisoning cases.2 It was reported that the number of poisoning cases reaches an estimated 877,000 every year (World Health Organization, 2003), and many health officials consider paraquat to be a public health threat because of its high toxicity.4,5 The ingestion of 40 mg of paraquat ion/kg body weight is associated with a From the Departments of *Emergency, and †Laboratory Medicine, West China Hospital, Sichuan University, Chengdu, Sichuan Province, People's Republic of China. Reprints: Yu Cao, PhD, MD, Department of Emergency, West China Hospital, Sichuan University, No 37 Guoxuexiang Road/Wuhou District, Chengdu, Sichuan Province, 610041, People's Republic of China (e‐mail: [email protected]). Disclosure: The authors declare no conflict of interest. Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0749-5161
mortality rate of up to 100%.1 The clinical manifestations of paraquat poisoning can vary depending on a number of factors, including the dosage and route of entry,2,6 which can also affect the prognosis. There have been reports of survival from mild poisonings with no intervention, but highly toxic doses (35 mg/kg) are almost always fatal despite aggressive interventions, such as hemodialysis and renal replacement therapy.7 Hence, the identification of accurate prognostic indicators is very necessary. Appropriate prognostic indicators could aid physicians in evaluating the severity of poisoning and in predicting overall outcomes. Currently, most primary health care institutional guidelines lack favorable prognostic indicators for paraquat poisoning. Therefore, most physicians predict the outcome of the condition based on the clinical signs and symptoms.8 However, the onset of symptoms after acute paraquat poisoning requires some time, and by then, the poisoning can be irreversible or fatal. Thus, a simple and easy to obtain prognostic indicator for acute paraquat poisoning is crucially important. Many prognostic indictors have been studied to evaluate the prognosis of paraquat poisoning, but the optimal indicators remain unclear. The Acute Physiology and Chronic Health Evaluation II (APACHE II) score is a scoring system based on 12 parameters of acute physiology and chronic health.9 This scoring system has been commonly used for the assessment of the severity and prognosis of diseases,10 especially in intensive care units.11 The APACHE II has also been used to predict the severity and prognosis of poisoning, but Chinese physicians have rarely used it for acute paraquat poisoning cases. The Severity Index of Paraquat Poisoning (SIPP) is also a potentially valuable prognostic indicator, which is calculated by multiplying the time since the ingestion (hour) of paraquat by the paraquat concentration in the plasma (milligram per liter).12,13 Studies have suggested that patients with a SIPP of less than 10 h · mg/L have a good chance of survival, whereas those with a SIPP between 10 and 50 h · mg/L often die because of interstitial pulmonary fibrosis secondary to paraquat poisoning, and those with values greater than 50 h · mg/L die rapidly owing to circulatory collapse.14,15 Although promising, the plasma paraquat level is not widely available in China. Other studies have shown that acidosis is an important prognostic indicator of acute paraquat poisoning,16,17 particularly the concentration of serum lactic acid, which is the primary cause of acidosis. To determine the value of APACHE II scores, SIPP, and serum lactate concentrations in the prognosis of paraquat poisoning, a prospective study, enrolling 143 patients with paraquat poisoning, was performed.
METHODS Study Population Patients with acute oral paraquat poisoning were consecutively selected from those patients who were referred to the emergency department of Sichuan University (Chengdu, China) between January
Pediatric Emergency Care • Volume 31, Number 2, February 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.pec-online.com
117
Xu et al
Pediatric Emergency Care • Volume 31, Number 2, February 2015
2011 and January 2012 (n = 143), according to the following exclusion criteria: transcutaneous or intravenous exposure and undetectable levels of paraquat in the blood. In addition, individuals were also ineligible if they had concomitant diseases, including cancer, cardiac disease, pulmonary disease, renal disease, and liver disease or if they refused to participate in the study (see the flow chart). Signed informed consent was obtained from the patients, family members, or legal guardians.
12 hours until serum paraquat levels decreased to less than 200 ppm. In addition, these patients also received daily pulse therapy via intravenous infusion of 1 g of methylprednisolone, 400 mg of cyclophosphamide, the antioxidants vitamins C (2 g/d) and E (300 mg/d), and cysteine (1.2 g/d). Supplemental oxygen therapy was administered if the oxygen partial pressure was less than 50 mm Hg. Hemodialysis was performed if the patients had severe metabolic acidosis, hyperkalemia, or heart failure.
Data Collection
Statistical Analysis
All of the data were collected by experienced emergency physicians, including details of the patients' general conditions, laboratory examinations, and treatments. On admission, each patient's age, sex and other demographic data, past medical history, amount of paraquat ingested (including other toxic substances ingested), and poisoning time were documented. On admission, vital signs were recorded, and blood samples were collected for the following tests: routine blood work, liver and renal function, other biochemical blood examinations, and arterial blood gas examinations. To ensure interrater reliability, a 20% sample of the data was screened by a second reviewer. The Glasgow Coma Scale score and the APACHE II score were calculated, and the serum lactate level and plasma paraquat concentration were determined. All of the specimens were collected 1 hour after admission. A complete blood cell count and biochemistry were performed, and lactic acid levels were measured using an autoanalyzer (Roche modular P800, Mannheim, Germany). Plasma paraquat concentrations were measured using a high-performance liquid chromatographic system (Shimadzu, Japan). The elapsed time (hours) from the ingestion of paraquat to arrival and the serum paraquat levels were recorded to calculate SIPP (mean elapsed time [hours] serum paraquat level in the emergency department [milligram per liter]). Deaths that occurred in the hospital were recorded, and those patients discharged from the hospital were followed up by telephone or letter. The 60-day mortality rate was used to determine the efficacy of individual prognostic indicators, and the results were summarized by a toxicology expert.
The collected data are presented as mean (SD) for continuous variables and as frequencies for categorical variables. A P value of less than 0.05 was considered to be statistically significant, with all of the statistical analyses performed using SPSS software, version 18.0. Differences in covariates between survivors and nonsurvivors were tested by Student t test for continuous variables and by the χ2 test or Fisher exact probability method for categorical variables. To investigate which prognostic indicator (APACHE II, SIPP, and serum lactic acid level) was the most valuable predictor of prognosis for acute paraquat poisoning, receiver operating characteristic (ROC) curves for predicting inpatient mortality were generated. The areas under the ROC curve (AUCs), sensitivities, specificities, cutoff values, accuracies, and indicators of the predictive value of the tests were calculated using GraphPad Prism software, version 4.0. To analyze the best cutoff values for APACHE II score, SIPP, and lactate level to predict the prognosis of paraquat poisoning, 1-way analysis of variance was used to assess the differences in initial parameters at the time of West China Hospital (WCH) arrival among paraquat poisoning patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II scores greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L. In addition, the 7-day survival curves of patients with lactate levels greater than or equal to 2.95 mmol/L, APACHE II scores greater than or equal to 15.22, and SIPP greater than or equal to 5.50 h · mg/L were compared.
RESULTS Treatment As summarized in Table 1, all of the patients presenting within 6 hours of paraquat ingestion received gastric lavage, followed by 200 mL of 20% mannitol via gastric tube. The patients who presented within 12 hours of ingestion received hemoperfusion every
Patient Characteristics During the study period, 161 patients fulfilled the criteria for acute oral poisoning. Eighteen of these patients were excluded because of missing data or loss to follow-up. The remaining 143
TABLE 1. General Characteristics of 143 Oral Paraquat Poisoning Patients at the Time of WCH Arrival
No. patients* Age, y† Sex* Time between exposure and WCH arrival, h† Estimated amount of paraquat dichloride (24.5% concentrate) exposure, mL† Comorbid disease* Psychiatric condition* Underlying disease* Concomitant alcohol ingestion* Serum lactates, mmol/L† APACHE II score† SIPP, h · mg/L†
All Patients
Survivors
Nonsurvivors
P
143 30.6 (12.7) 54/89 9.9 (5.1) 65.6 (14.1) 47 25 7 1 5.29 (2.54) 8.65 (4.63) 81.07 (67.42)
77 28.0 (12.9) 28/49 11.2 (4.3) 50.1 (21.2) 27 16 5 1 2.78 (1.89) 5.45 (3.67) 15.39 (10.6)
66 33.6 (11.9) 26/40 7.4 (6.9) 76.9 (18.2) 20 9 2 0 7.63 (2.46) 11.29 (4.31) 147.85 (78.56)
0.190 0.666 0.106 0.227 0.546 0.262 0.35 1.000
