|Year : 2022 | Volume
| Issue : 2 | Page : 116-121
Clinical epidemiology and predictors of outcome in chlorpyrifos poisoning in farming and allied agricultural workers in East Godavari, Andhra Pradesh
Anand Acharya1, Abinash Panda2
1 Department of Pharmacology, Konaseema Institute of Medical Sciences and Research Foundation, Amalapuram, Andhra Pradesh, India
2 Department of Pharmacology, M.K.C.G. Medical College, Berhampur, Odisha, India
|Date of Submission||22-Jan-2021|
|Date of Decision||05-Jan-2022|
|Date of Acceptance||05-Feb-2022|
|Date of Web Publication||4-Jul-2022|
Dr. Abinash Panda
Associate Professor, Department of Pharmacology, M.K.C.G. Medical College, Berhampur, Odisha - 760 004
Source of Support: None, Conflict of Interest: None
Background: Chlorpyrifos is the ninth most consumed pesticide for agricultural purposes in India. It is a chlorinated organophosphate that also causes poisoning in non-target species such as humans. There is a scarcity of published literature on chlorpyrifos intoxication in the Indian scenario. Objective: The objective of the study was to find out the clinical epidemiology and factors influencing the outcome of chlorpyrifos poisoning in the farming and allied agricultural workers. Methodology: The patient record-based cross-sectional study was carried out on 40 confirmed cases of chlorpyrifos poisoning in a tertiary care teaching hospital in East Godavari district, a predominantly agrarian region in the state of Andhra Pradesh. Clinic-demographic profile, aspects of poisoning, outcome in terms of recovery, death, ventilator support were obtained. Mortality or requirement of ventilator support was classified as a bad outcome. Carlsons' Co-morbidity Index, Poison Severity Score, and Glasgow Coma Scale score at the time of admission were assessed objectively. The strength of association between the predictor variables and the outcome was represented using the odds ratio. Statistical analysis was performed using Graph Pad Prism 8.0 trial version. Results: Chlorpyrifos poisoning constituted about 10% of pesticide poisoning and 25% of such cases had a poor outcome. The mean age was 44.73 ± 10.76 years, the majority being farmers. Younger age (<40 years), early reporting to hospital, Carlsons' Co-morbidity Index of less than 8, Glasgow Coma Scale score of more than 12, and Poison severity score of less than 2 were significantly associated with favorable outcome (P < 0.005). During admission, 60% of cases had a plasma pseudocholinesterase level of less than 1000 U/L. Lower plasma pseudocholinesterase was associated with a poor outcome (OR 9.62 CI 4.05-22.84). Conclusions: Farmers and allied agricultural workers are susceptible to chlorpyrifos intoxication. Along with clinical and laboratory parameters, Carlsons' Co-morbidity Index, Glasgow Coma Scale score, and Poison severity score at the time of admission are seen as reliable predictors of outcome.
Keywords: Carlsons' co-morbidity index, Chlorpyrifos poisoning, Glasgow coma scale score, plasma pseudocholinesterase, poison severity score, predictors of outcome
|How to cite this article:|
Acharya A, Panda A. Clinical epidemiology and predictors of outcome in chlorpyrifos poisoning in farming and allied agricultural workers in East Godavari, Andhra Pradesh. Indian J Occup Environ Med 2022;26:116-21
|How to cite this URL:|
Acharya A, Panda A. Clinical epidemiology and predictors of outcome in chlorpyrifos poisoning in farming and allied agricultural workers in East Godavari, Andhra Pradesh. Indian J Occup Environ Med [serial online] 2022 [cited 2022 Aug 14];26:116-21. Available from: https://www.ijoem.com/text.asp?2022/26/2/116/349853
| Introduction|| |
Chlorpyrifos is a chlorinated organophosphate compound having IUPAC name O, O-diethyl-O-(3, 5, 6-trichloro-2-pyridyl) -phosphorothioate. It is extensively used in developing countries, in the agricultural sector for effective crop protection because of its potential toxicity against a broad spectrum of insect species that does not affect the seed activity, it is cost effective and leaves little residue on food grains. But, along with target species, chlorpyrifos also causes poisoning in non-target species such as humans. It is commonly associated with intentional intoxication in many Asian countries., According to W.H.O toxicity classification, it is a moderately hazardous pesticide. In an Agricultural Health Study done in the United States involving 55,071 pesticide applicators, out of 1851 deaths, 588 occurred due to chlorpyrifos applicators; citing a possible association between chlorpyrifos use and external causes of death. Chlorpyrifos is commonly used throughout India in large amounts. The official data of the Directorate of Plant Protection, Quarantine and Storage Govt. of India has listed chlorpyrifos as the ninth most consumed insecticide in the country. According to the National Poison Information Centre, India, OP compounds are the third most common agent implicated in suicidal poisonings after household agents and drug. Early detection, proper medical management, and transportation to experienced medical centers might assist in improving the outcomes of patients with severe acute organophosphorus (OP) poisoning. The toxicity of OP compounds varies between different types, and the differences between OP pesticides in human self-poisoning have been documented. In our region, chlorpyrifos is used as a pesticide to be sprayed in paddy fields, vegetables, cotton, and banana. Due to ignorance about the handling and toxicity of these chemicals, people get exposed to them knowingly or unknowingly. Published literature has traditionally considered acute OP poisoning as a homogenous entity., Thus, there is a scarcity of literature on chlorpyrifos intoxication, in particular about the clinical epidemiology, presentation, and outcome in the Indian scenario. The chlorine group present on chlorpyrifos increases the compound's lipid solubility and half-life in the body, resulting in a more gradual and persistent lowering of cholinesterase (ChE) levels compared to other OP pesticides. This unique toxico-kinetic property of chlorpyrifos influences the morbidity, mortality, and response to pralidoxime treatment. Apart from the classical signs and symptoms of acute cholinergic crisis common to all OP poisoning, acute coronary syndrome, acute renal failure, delayed myelopathy, and polyneuropathy are also seen in chlorpyrifos poisoning. With this background, the present study has been designed with a primary objective to find out the association between the clinical epidemiology, clinical profile, laboratory parameters, and outcome of chlorpyrifos poisoning.
| Methodology|| |
Study design and setting
The patient record based cross sectional study was carried out in the Department of Pharmacology in collaboration with the Departments of General Medicine and Emergency Medicine in Konaseema Institute of Medical Sciences & Research Foundation, Amalapuram – a tertiary care teaching hospital having a catchment area from throughout the East Godavari district, in the state of Andhra Pradesh, India during the period from January 2015 to March 2020. The region is unique, with 28.93% of the population having agriculture as their predominant occupation, the lands being irrigated by river Godavari, having a robust network of canals, producing crops like paddy, banana, and cotton by farmers of all class like poor/rich, contract/landowning class.
The study was approved by the Institutional Ethics Committee of Konaseema Institute of Medical Science, Amalapuram (Approval No. IEC/PR/2020/002 Dated 08-01-2020). As the study involved the retrospective evaluation of existing data obtained from the patient case sheet, waiver of informed consent of the study participants was obtained from the Institutional Ethics Committee. To ensure confidentiality, identifying information of the study participants was delinked from the main clinical data set and was accessible to researchers only.
During the five years of the study period, 40 cases of chlorpyrifos poisoning were identified from the existing medical records and included as per the inclusion and exclusion criteria. Chlorpyrifos poisoning constituted 10% of all the cases of pesticide poisoning admitted to the hospital during the study period, so 40 cases were found to be adequate post factum to obtain the desired confidence level of 95% at an absolute precision of 10% for data analysis. Patients of either gender, irrespective of age and socio-demographic profile were included. Poisoning due to multiple pesticide compounds was excluded from the study.
A confirmed diagnosis of OP poisoning based on clinical features, history of exposure supported by the empty container brought by patients' attendants indicating the ingestion of chlorpyrifos type of OP compound and a low serum pseudo cholinesterase at the time of admission was defined as a case of chlorpyrifos poisoning for the present study.
Study tool and data collection
The data obtained from the patient case sheet was captured in a pre-designed case record form. Demographic profile, clinical presentation, and baseline laboratory parameters obtained at the time of admission were recorded. All the aspects of the poisoning like circumstances, amount, mode of ingestion of poison were entered in the case record form. Time lag in getting the patient to the hospital, treatment protocol adopted, the outcome in terms of recovery or death, requirement of ventilator support, and hospital stay were also recorded. Any patient with mortality or requiring ventilatory support during treatment was classified as having a bad outcome. Patients recovering without any sequel, and not requiring ventilatory support were classified as those with good outcomes. Carlsons' Co-morbidity Index (CCI) that predicts the 10-year survival in patients with multiple co-morbidities was calculated for each patient based on their records and using online https://www.mdcalc.com/charlson-comorbidity-index-cci software. The severity grade at the time of admission was assessed using the Poison Severity Score (PSS). The Glasgow Coma Scale (GCS) score was used to objectively assess the extent of impaired consciousness in patients at the time of admission. Data extraction from the patient case record was done by the investigators. This included the patient demographic profile, clinical presentation, and baseline laboratory parameters at the time of admission, circumstances, amount, mode of ingestion of the poison, the time lag for hospitalization, outcome, the requirement of ventilator support, and duration of hospital stay.
Statistical analysis plan
Descriptive statistics were used to present the clinic-demographic profile. Chi-square test was used to find out the association between the prognostic factors and the outcome. The strength of association between the predictor variables and the outcome was represented using the odds ratio. The data were analyzed using Graph Pad Prism 8.0 trial version. A P value ≤0.05 was considered to be statistically significant.
| Results|| |
The record based retrospective study was carried out on 40 hospitalized patients with chlorpyriphos poisoning. The outcome was stratified into good and poor outcomes based on the mortality or requirement of ventilatory support during treatment. During the study period, 416 cases of poisoning were reported to the hospital out of which 254 cases were due to OP compounds. Forty cases were attributed to chlorpyrifos compounds. Ten (25%) study participants had poor outcomes. The clinico-demographic characteristics of the study participants with chlorpyrifos intoxication revealed that the mean age of patients was 44.73 ± 10.76 years; the range being 22 to 66 years. Twenty-eight patients were less than 40 years of age, out of which 25 (89%) had a good outcome, thus younger age (<40 years) was associated with a statistically significant (P ≤ 0.05) better outcome as compared to older patients where about one-third of the patients had a poor prognosis. Thirty five (87%) patients were males of which 24 (68%) had a good outcome. The outcome was better in females. But the association between gender and outcome was not statistically significant. Majority of the study participants were farmers and daily wage earners who received steady work and wages from the employer. The mean time interval for the arrival of the patient in the hospital was 3.43 ± 1.39 hrs; the inter-quartile range was 4 hrs. Early reporting to the hospital was associated with significantly favorable outcomes (95%). Few patients had co-morbidities like diabetes mellitus (10%) and hypertension (25%). But, these were associated with poor outcomes of 50% and 60%, respectively, but this association was not statistically significant. Presence of habits like smoking and alcoholism were associated with poor prognosis, though statistically insignificant.
All intoxications were by oral ingestion. The average amount was 29.38 ± 17.40 ml. Patients reporting an ingested amount of fewer than 50 ml had a significantly favorable outcome. There was a significant association (P = 0.0006) between the amount of chlorpyriphos ingested and the outcome. Mean duration of hospitalization was 9.07 ± 3.96 days. Twenty two (55%) of patients were discharged in less than 8 days. Duration of hospitalization did not influence the outcome significantly [Table 1]. At the time of admission, 35 patients (86%) in this study had Carlsons' Co-morbidity Index (CCI) of less than 8 and this was significantly associated with a good prognosis (OR 3.5, CI 1.97-6.91). Thirty three patients (83%) had a Glasgow Coma Scale (GCS) score of more than 12. Patients with a GCS score of more than 12 had a good outcome (OR 37.92, CI 17.02-84.45). About 83% of the patients had Poison Severity Score (PSS) of less than 2. This was significantly associated with favorable outcomes (P < 0.005) [Table 2]. Most common presentation after intoxication was an acute cholinergic crisis (50%). Acute cholinergic crisis, respiratory failure, acute renal failure, and occurrence of seizure were associated with poor outcomes (P ≤ 0.05) [Table 3]. It was observed that, at the time of admission, 60% of the patients had a low plasma pseudocholinesterase level <1000 U/L. The laboratory reference range of plasma pseudocholinesterase used in the present study was 3930-10800 U/L and 4620-11500 U/L for females and males respectively and was estimated at the time of admission in all the patients using the DGKC method (LiquiChek). A lower plasma pseudocholinesterase was associated with a poor outcome (OR 9.62 CI 4.05-22.84). Patients with normal blood urea nitrogen, serum creatinine, potassium, sodium, WBC count, acid base balance, and arterial blood gas analysis had significantly better outcomes (P ≤ 0.05). Abnormal laboratory values were a significant risk factor for poor prognosis after chlorpyrifos intoxication [Table 4].
|Table 1: Clinico-demographic characteristics of the study participants and outcome|
Click here to view
| Discussion|| |
There is a paucity of literature that explores the spectrum, clinical course, and outcome of chlorpyrifos intoxication, the majority being discrete case reports, though there is an abundance of published literature on OP poisoning as a class. OP pesticide poisoning is common in the developing world. The highest incidence is seen in India. Suicidal and non-suicidal organophosphate poisoning is a major problem in rural areas of India, with a rapidly increasing incidence rate. In the present study, males were predisposed more to poison. Some other studies have also reported a similar trend. In a study done by Banday et al. in a south Indian hospital, they have observed that the male to female ratio to be 3.2:1. Studies by Safdar et al., and Hussain et al. have also seen a similar trend. This may be due to the higher proportion of agrarian workers being males and their accessibility of the organophosphates. The mean age of patients was 44.73 ± 10.76 years, the range being 22 to 66 years. Other studies have reported a similar age predisposition where the majority of the cases (80%) were young people, predominantly males from the age group 13-40 years. However, Hayden et al., showed an age range from 13-47 years with a mean age of 23 years. The predisposition of people at a younger age group to intoxication may be due to high ambitions and vulnerability to emotional issues. Reduced plasma pseudocholinesterase at the time of admissions was seen in 24 (60%) patients and was associated with a poor outcome. Other studies have made a similar observation. Banday et al. have found that mortality was highest in cases when plasma pseudocholinesterase was less than 700 U/L.Goswamy et al. have concluded that apart from clinical indicators, low serum cholinesterase levels were of greatest predictive value for ventilation in OP poisoning. However, Nouira et al. did not find any statistically significant difference in mean serum cholinesterase levels in those mechanically ventilated and those who do not need ventilator support. There was a significant relationship between abnormal laboratory values and poor prognosis after chlorpyrifos intoxication. Similar were the findings in another Asian study done in Sri Lanka. Pseudocholinesterase enzyme which is responsible for breaking down acetylcholine is inhibited by chlorpyrifos. Consequently, with the persistence of acetylcholine, there is over-stimulation of the post-synaptic nerves, muscles, and exocrine glands leading to adverse outcomes. Thus, plasma pseudocholinesterase levels are useful for assessing the severity and prognosis of acute exposures. Reduced plasma pseudocholinesterase levels indicate severe poisoning. Most common presentation after intoxication was emesis followed by tachycardia, salivation, and acute cholinergic crisis. Other studies have observed that the most frequent signs noted were miosis, increased salivation, anxiety and restlessness, bronchospasm, and incontinence in that order.,, In our study, 25% of the cases had a poor outcome based on the mortality or ventilatory support requirement during treatment. Yamashita et al. have reported a poor outcome in the range of 4-30%.The proportion of poor outcomes was 5.5% in a study by Malik et al. The reason for variation in mortality rates may be due to factors like time taken for arrival in the hospital, not receiving any treatment at a peripheral health centre before referral to the higher hospital, lack of awareness regarding the seriousness and mortality associated with OP poisoning, and non-availability of intensive care unit facilities. Early reporting to the hospital within 4 hr of poisoning was associated with a significantly favorable outcome. In the study done in south India, it became evident that higher mortality was seen in the patients when there was a delay between consumption of the organophosphate and initiation of treatment. Similar was the finding of the study done by Suleman et al. Ingestion of quantity less than 50 ml had a favorable outcome. In a study done in south India, it was observed that the mortality was higher in those cases that had consumed a large amount of OP substances (50-100 ml) and were highest in cases who consumed >100 ml.
Cases with the Glasgow Coma Scale score of more than 12 had a good outcome in the present study. In a study by Acikalin et al. low Glasgow Coma Scale score was a poor prognostic factor for organophosphate intoxication. Sam et al. had also indicated a significant negative correlation between the Glasgow Coma Scale score and mortality rate. Poison Severity Score (PSS) is used for the severity grading of acute poisoning regardless of the type and number of agents involved. Severity Grades include 0 (none) being no symptoms or signs related to poisoning, 1 (minor) mild, transient and spontaneously resolving symptoms, 2 (moderate) pronounced or prolonged symptoms, 3 (severe) severe or life-threatening symptoms, 4 (fatal) death. In this study, a score of less than two at the time of admission was significantly associated with a favorable outcome. A study performed in Nepal by Shrestha et al. reported that PSS could recognize high-risk patients, and help the health care facility to provide prompt treatment to the poisoned patients. The higher the PSS, the greater was the probability of mortality.Carlsons' Co-morbidity Index (CCI) predicts the 10 year mortality by weighing 17 categories of medical co-morbidities. A higher score indicates greater co-morbidity and risk of death., Earlier studies have used CCI in acute settings like ICU and burn units. It was in our study that at the time of admission 35 (86%) of cases in this study had a CCI score of less than 8 and this was associated with a good prognosis. This may be due to the lower prevalence of multiple co-morbidities in agrarian workers who mostly perform physical labor. The major limitation of the study is its small sample size. The cross-sectional nature could establish only the odds of association between the predictor and outcome variables. A follow-up study could have identified the definite risk factors for a bad prognosis in chlorpyrifos poisoning. The data was collected from the medical records of only one tertiary care hospital and there is a likelihood that all cases of chlorpyrifos poisoning may not have come to the current hospital, hence the findings could not be generalized to the entire agrarian community of East Godavari district.
| Conclusion|| |
Chlorpyriphos poisoning was associated with a poor outcome in 25% of cases. Farmers and allied agricultural workers are susceptible to intoxication. Apart from clinical and laboratory parameters, CCI, GCS score, and Poison severity score at the time of admission are the predictors of outcome. They can be used for early identification and timely referral of high-risk patients at the first point of care in chlorpyriphos poisoning.
Declaration of patient consent
The authors certify that, since the study involved the retrospective evaluation of existing data obtained from the patient case sheet, waiver of informed consent of the study participants was obtained from the Institutional Ethics Committee. To ensure confidentiality, identifying information of the study participants was delinked from the main clinical data set and due efforts has been made to conceal their identity.
The authors would like to thank the medical record section of Konaseema Institute of Medical Sciences and Research Foundation, Amalapuram for their support.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Rathod AL, Garg RK. Chlorpyrifos poisoning and its implications in human fatal cases: A forensic perspective with reference to Indian scenario. J Forensic Leg Med 2017;47:29-34.
Liu SH, Lin JL, Shen HL, Chang CC, Huang WH, Weng CH, et al.
Acute large-dose exposure to organophosphates in patients with and without diabetes mellitus: Analysis of mortality rate and new-onset diabetes mellitus. Environ Health 2014;13:11.
World Health Organization (2010). The WHO Recommended classification of pesticides by hazard and guidelines to classification 2009 (Report). World Health Organization.
[Last accessed on 2020 Jan 09].
Lee WJ, Alavanja MC, Hoppin JA, Rusiecki JA, Kamel F, Blair A, et al.
Mortality among pesticide applicators exposed to chlorpyrifos in the Agricultural Health Study. Environ Health Perspect 2007;115:528–34.
Srivastava A, Peshin SS, Kaleekal T, Gupta SK. An epidemiological study of poisoning cases reported to the National Poisons Information Centre, All India Institute of Medical Sciences, New Delhi. Hum Exp Toxicol 2005;24:279-85.
Dong N, Liu J, Wang Z, Gao N, Pang L, Xing J. Development of a practical prediction scoring system for severe acute organophosphate poisoning. J Appl Toxicol 2020;40:889-96.
Eddleston M, Eyer P, Worek F, Mohamed F, Senarathna L, von Meyer Buckley L. Differences between organophosphorus insecticides in human self-poisoning: A prospective cohort study. Lancet 2005;366:1452–9.
Liu HF, Ku CH, Chang SS, Chang CM, Wang IK, Yang HY, et al.
Outcome of patients with chlorpyrifos intoxication. Hum Exp Toxicol 2020;39:1291-300.
Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47:1245-51.
Persson H, Sjöberg G, Haines J, Pronczuk de Garbino J. Poisoning severity score: Grading of acute poisoning. J Toxicol Clin Toxicol 1998;36:205-13.
Yanagisawa N, Morita H, Nakajima T. Sarin experiences in Japan: Acute toxicity and long-term effects. J Neurol Sci 2006;249:76–85.
Rodgers ML. OP poisoning. Am J Emerg Med 2006;22:335–44.
Gupta SK, Peshin SS, Srivastava A, Kaleekal T, Pandian TV. An epidemiological pattern of poisoning in India. Pharmacoepidemiol Drug Saf 2002;11:73–4.
Banday TH, Tathineni B, Desai MS, Naik V. Predictors of morbidity and mortality in organophosphorus poisoning: A case study in rural hospital in Karnataka, India. N Am J Med Sci 2015;7:259-65.
Safdar A, Saeed A, Muhammad NR. Organophosphorus poisoning: Emergency management in intensive care unit. Prof 2003;10:308–14.
Hussain AM, Sultan ST. Organophosphorus insecticide poisoning: Management in surgical intensive care unit. J Coll Physicians Surg Pak 2005;15:100–2.
Hayden KM, Norton MC, Darcey D, Ostbye T, Zandi PP, Breitner JC, et al.
Occupational exposure to pesticides increases the risk of incident AD: The Cache County study. Neurology 2010;74:1524–30.
Goswamy R, Chaudhuri A, Mahashur AA. Study of respiratory failure in organophosphate and carbamate poisoning. Heart Lung 1994;23:466–72.
Nouira S, Abroug F, Elatrous S, Boujdaria R, Bouchoucha S. Prognostic value of serum cholinesterase in organophosphate poisoning. Chest 1994;106:1811–4.
van der Hoek W, Konradson F. Analysis of 8000 Hospital admission for acute poisoning in rural areas of Sri Lanka. Clin Toxicol (Phila) 2006;44:225–31.
Tahir MH, Raja JI, Haq IU. Acute organ phosphorus poisoning — An experience. Pak Armed Forces Med J 2006;56:150–6.
Karki P, Ansari JA, Bhandary S, Koirala S. Cardiac and electrocardiographical manifestation of acute organophosphate poisoning. Singapore Med J 2004;45:385–9.
Yamashita M, Yamashita M, Tanaka J, Ando Y. Human mortality in organophosphate poisoning. Vet Hum Toxicol 1997;39:84–5.
Malik GM, Mubarik M, Romshoo GJ. Organphosphorus poisoning in the Kashmir Valley, 1994-1997. N Engl J Med 1998;338:1078.
Numidasa UA, Gawarammana IB, Kularatne SA, Kumarasiri PV, Goonasekera CD. Survival pattern in patients with acute organophosphate poisoning receiving intensive care. J Toxicol Clin Toxicol 2004;42:343–7.
Suleman MI, Jibran R, Rai M. The analysis of organophosphate poisoning case treated at Bhawal Victoria Hospital, Bahawalpur 2000-2003. Pak J Med Sci 2006;22:244–9.
Acikalin A, Disel NR, Matyar S, Sebe A, Kekec Z, Gokel Y, et al.
Prognostic factors determining morbidity and mortality in organophosphate poisoning. Pak J Med Sci 2017;33:534–9.
Sam KG, Kondabolu K, Pati D, Kamath A, Pradeep Kumar G, Rao PG. Poisoning severity score, APACHE II and GCS: Effective clinical indices for estimating severity and predicting outcomeof acute organophosphorus and carbamate poisoning. J Forensic Leg Med 2009;16:239–47.
Shrestha TM, Aacharya RP, Neupane RP. Predicting outcome of acute organophosphorus poisoning using poison severity score in patients presenting at a university hospital in Nepal. J Inst Med 2018;40:17-21.
Benoit MF. Acuity in a gynecologic oncology patient population: A score adjusted charlson comorbidity index and comparison with other predictive medical models in an effort to document need and support health care resource use. Clin Obstet Gynecol Reprod Med 2018;4:1-5.
Quan H, Li B, Couris CM, Fushimi K, Graham P, Hider P, et al.
Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 2011;173:676-82.
[Table 1], [Table 2], [Table 3], [Table 4]