Article 1

Critical appraisal refers to the systematic process that identifies strengths and weaknesses of an article for evaluating its validity and usefulness of the research findings (12). This report will critically analyse two articles, using the CASP tool for appraisal. Questions in the CASP tool will facilitate determination of the relevance and trustworthiness of the aforementioned articles.

The issue addressed in the study was quite relevant to the context of the research. Diabetes is a chronic health disorder prevalent globally with an incidence of 1 in 3 adults (5). According to research studies, a dramatic increase has been observed in the number of diabetic patients in Thailand from 2009 to 2014 (17). Furthermore, there are several evidences that suggest correlation between exposure to pesticides and incidence of type 2 and gestational diabetes (8). Moreover, the sudden increase in the application of pesticides in Thailand establishes selection of the issue as a correct procedure (19). 

An appropriate method was used to address the research question that can be validated by the fact that the outcomes were beneficial for the target population. The research method applied to address the question suggested that an overexposure to pesticides is responsible for increasing the susceptibility to diabetes among farmers. Consistency with previous findings suggests that the method was correct (3).

Lack of adequate information on association between diabetes and pesticide in Thailand might create negative impacts on both the case and the control by providing inaccurate results during analysis of the responses. Apart from mentioning that the cases were diagnosed with diabetes, they were not defined in a precise way. The time frame of pesticide exposure and number of cases were not sufficient. Reliable system for selecting the cases was based on diagnosis of diabetes. There was no specialty of the sample and no power calculation. However, the sample represented a defined population of diabetic rice farmers in Thailand. However, the selection of participants for the study was based on prevalence of diabetes and pesticide exposure, collected from hospital data and structured questionnaire respectively,and a reliable system was used (randomisation). Thus, the cases not recruited in an acceptable way.

Although the controls represented rice farmers without diabetes, which represented the population, their specialty was that they belonged to same gender and age that increased likelihood of pesticide exposure, due to similar environmental factors. They were matched but more in number than the sample. The authors failed to provide information on the response rate. Thus, the controls were not selected in an acceptable way.

Structured questionnaires, in addition to similar measurement in the sample and control were used. Temporal relation was correct as pesticide exposure preceded diabetes. However, there was recall bias due to collection of information based on questionnaire that might affect the responses owing to differences in recall ability of the participants. Moreover, the researchers did not blind the participants to the study, which might affect their responses and give misleading results (6). Construct validity was also not measured. Thus, the measures did not reflect true to what was claimed.

Article 2

The confounding factors that were taken into account include gender, BMI, cigarette smoking, alcohol consumption, occupation, family history and age.

The researchers had used logistic regression to analyse the adjusted odds ratio showing association between exposure to pesticide and its outcome (diabetes). Thus, use of regression analysis help in estimating the correlation between the independent variable and outcome, while holding other variables as constant (10).

Statistically significant association between rodenticide exposure and diabetes prevalence was established by the findings (OR = 1.35; 95%CI 1.04-1.76). Presence of odds ratio > 1 for the use of fungicides (OR = 2.08; 95%CI 1.03–4.20), organophosphate (OR = 2.22; 95%CI 1.17–4.19), carbamate (OR = 1.50; 95%CI 1.02–2.19), and organoclorine (OR = 1.40; 95%CI 1.01–1.95) establishes strong correlation of exposure to outcome. Thus, adjustment of the results using regression has made a big difference. 

Low p-values for BMI, occupation, smoking, alcohol consumption, and family history of diabetes signifies strong evidence against for the association between pesticide exposure and diabetes incidence (16). Presence of 95% CI for all findings suggests that sampling the same population numerous times will result in similar results and sound statistical findings.

Although high odds ratio suggest positive association between exposure and the outcome, it is essential to remove selection bias and recall bias and use a larger sample size for confirming and believing the results.

The findings can be applied in local population based on the fact that about 2 million tones of pesticides are used every year, worldwide (15). Owing to the high exposure to pesticides by people and the high prevalence of diabetes patients, worldwide, the results can prove effective in the local population.

Results show consistency with previous findings that correlated type 2 and gestational diabetes with pesticide exposure.

The issue addressed by the authors was relevant owing to the high mortality rates due to acute myocardial infarction (AMI) in Taiwan. Moreover, presence of previously conducted cohort studies and systematic review that correlated risks of myocardial infarction with chronic infection of hepatitis C also establishes the fact that the authors addressed a clearly focused issue (4) (13).

The cohort was recruited in an appropriate way owing to the fact that of the 186112 patients diagnosed with AMI, 4666 patients were identified with HCV infection. Moreover, the researchers displayed. Thus, the cohort represented the defined population.

The authors used appropriate methods to classify the subjects into particular groups. One-to-one matching among all participants regarding some factor such as, hypertension, age, sex, diabetes mellitus, peripheral vascular disease and heart failure demonstrates correct methodology in minimizing bias. Moreover, the NHIRD data on AMI had been validated by previous studies (2).

The subjects and the assessors were not blinded to the study, which might contribute to bias in the outcomes. Although the measures reflected the expected outcomes, lack of adequate information on minimizing bias makes it difficult to tell if all the outcomes were accurately measured.

Although the database did not include information on potential confounding factors, the researchers used a propensity score matching procedure for controlling the essential confounding variables such as, age, sex, hypertension, previous stroke, peripheral vascular disease, and dyslipidemia, which might affect the outcomes among AMI patients. Thus, all important confounding factors were identified.

The authors had accurately analysed the results using a Cox proportional hazard regression method. This helped in adjusting the confounding factors that were taken into account (14). Furthermore, use of a propensity score matching method also helped in minimizing the factors or variables. Thus, all confounding factors were taken into account in the research design.

The follow-up of the subjects was complete owing to the fact that all patients were followed for a time period of 12 years, until the outcomes of the disease was accurately observed. The fact that the authors did not mention loss of patients from the groups, it can be considered that the study began and ended with similar number of patients.

The fact that the study evaluated impacts of Hepatitis C infection on the mortality rate of patients with AMI in Taiwan for 12 years, establishes on the follow-up being long enough. Conducting the study for a period of 12 years helps to develop a valid scenario of the extent of the outcome being studied.

The results showed that mortality rate after 12 years was significantly higher among AMI patients with cirrhosis and HCV infection compared to HCV infected subjects without cirrhosis (P<0.0001) and controls (P<0.0001). Furthermore, the hazard ratio (HR) was significantly large among patients with HCV infection (HR 1.12; 95% CI 1.06 to 1.18). Presence of HR>1 indicates that the risk of mortality among AMI patients with HCV infection is greater (18).

The results were also precise due to presence of 95% CI in the role of HCV infection on the long term mortality of all patients with AMI (HR 1.12; 95% CI 1.06-1.18). Presence of 95% CI indicates that similar results will be obtained on performing the research for innumerable times on the same population.

Although utilization of data from the NHIRD database that contains more information for more than 23,000,000 patients and reliability of the AMI data by previous studies suggest that the results are true, it cannot be believed completely (2). This occurs due to the fact that the retrospective design of the study, did not provide information on atherosclerosis burden, and used a database which did not include information on the major confounding factors such as, family history, body weight, height, glucose levels, lipid and viral load, and actual reason for death. Apart from consistency with previous studies that established correlation between HCV and coronary atherosclerosis, no other principles of the Bradford Hill criteria.

The study can be applied to the local population due to the fact that millions of cases of AMI are observed globally. Furthermore, high incidence and a higher prevalence of hepatitis C throughout the world accounts for the fact a similar cohort study can be conducted in the population to estimate effects of the infection on morality due to AMI (20).

The results of the study fit with previous findings in that antiplatelets and statins are less used among HCV group than control group. Similarity is also observed in previous reports on the link between carotid atherosclerosis and HCV seropositivity and AMI mortality among liver cirrhosis patients (7) (1).

Results of the study can have positive implications in that it can create awareness among healthcare professionals for considering the impacts of Hepatitis C infection while administering therapeutic treatments to AMI patients. However, more robust evidences are required to establish the findings.

References

1. Abougergi MS, Karagozian R, Grace ND, Saltzman JR, Qamar AA. ST elevation myocardial infarction mortality among patients with liver cirrhosis: a nationwide analysis across a decade. Journal of clinical gastroenterology. 2015 Oct 1;49(9):778-83.
2. Cheng CL, Lee CH, Chen PS, Li YH, Lin SJ, Yang YH. Validation of acute myocardial infarction cases in the national health insurance research database in taiwan. Journal of epidemiology. 2014 Nov 5;24(6):500-7.
3. Evangelou E, Ntritsos G, Chondrogiorgi M, Kavvoura FK, Hernández AF, Ntzani EE, Tzoulaki I. Exposure to pesticides and diabetes: a systematic review and meta-analysis. Environment international. 2016 May 1;91:60-8.
4. Forde KA, Haynes K, Troxel AB, Trooskin S, Osterman MT, Kimmel SE, Lewis JD, Re VL. Risk of myocardial infarction associated with chronic hepatitis C virus infection: a population?based cohort study. Journal of viral hepatitis. 2012 Apr 1;19(4):271-7.
5. Guariguata L, Whiting DR, Hambleton I, Beagley J, Linnenkamp U, Shaw JE. Global estimates of diabetes prevalence for 2013 and projections for 2035. Diabetes research and clinical practice. 2014 Feb 1;103(2):137-49.
6. Hróbjartsson A, Thomsen AS, Emanuelsson F, Tendal B, Hilden J, Boutron I, Ravaud P, Brorson S. Observer bias in randomised clinical trials with binary outcomes: systematic review of trials with both blinded and non-blinded outcome assessors. Bmj. 2012 Feb 27;344:e1119.
7. Ishizaka N, Ishizaka Y, Yamkado M. Atherosclerosis as a possible extrahepatic manifestation of chronic hepatitis C virus infection. Clinical Medicine Insights: Cardiology. 2014 Jan;8:CMC-S17069.
8. Jaacks LM, Staimez LR. Association of persistent organic pollutants and non-persistent pesticides with diabetes and diabetes-related health outcomes in Asia: A systematic review. Environment international. 2015 Mar 1;76:57-70.
9. Juntarawijit C, Juntarawijit Y. Association between diabetes and pesticides: a case-control study among Thai farmers. Environmental Health and Preventive Medicine. 2018 January; 23:3.
10. Knol MJ, Le Cessie S, Algra A, Vandenbroucke JP, Groenwold RH. Overestimation of risk ratios by odds ratios in trials and cohort studies: alternatives to logistic regression. Canadian Medical Association Journal. 2012 May 15;184(8):895-9.
11. Kuo SH, Hung WT, Tang PL, Huang WC, Yang JS, Lin HC, Mar GY, Chang HT, Liu CP. Impact of hepatitis C virus infection on long-term mortality after acute myocardial infarction: a nationwide population-based, propensity-matched cohort study in Taiwan. BMJ open. 2018 January; 8:e017412.
12. Morse JM. Critical analysis of strategies for determining rigor in qualitative inquiry. Qualitative health research. 2015 Sep;25(9):1212-22.
13. Pothineni NV, Rochlani Y, Vallurupalli S, Kovelamudi S, Ahmed Z, Hakeem A, Mehta JL. Comparison of angiographic burden of coronary artery disease in patients with versus without hepatitis C infection. American Journal of Cardiology. 2015 Oct 1;116(7):1041-4.
14. Sedgwick P. Cox proportional hazards regression. BMJ: British Medical Journal (Online). 2013 Aug 9;347.
15. Shorette K. Outcomes of global environmentalism: longitudinal and cross-national trends in chemical fertilizer and pesticide use. Social Forces. 2012 Jul 25;91(1):299-325.
16. Sullivan GM, Feinn R. Using effect size—or why the P value is not enough. Journal of graduate medical education. 2012 Sep;4(3):279-82.
17. Thai NCD network. Situation of NCDs (2nd report):“Kick off to the Goals.” Nontaburi: The International Health Policy Program, Thailand (IHPP); 2016.
18. Uno H, Claggett B, Tian L, Inoue E, Gallo P, Miyata T, Schrag D, Takeuchi M, Uyama Y, Zhao L, Skali H. Moving beyond the hazard ratio in quantifying the between-group difference in survival analysis. Journal of clinical Oncology. 2014 Jun 30;32(22):2380-5.
19. Wanwimolruk S, Kanchanamayoon O, Phopin K, Prachayasittikul V. Food safety in Thailand 2: Pesticide residues found in Chinese kale (Brassica oleracea), a commonly consumed vegetable in Asian countries. Science of the Total Environment. 2015 Nov 1;532:447-55.
20. Zidan A, Scheuerlein H, Schüle S, Settmacher U, Rauchfuss F. Epidemiological pattern of hepatitis B and hepatitis C as etiological agents for hepatocellular carcinoma in iran and worldwide. Hepatitis monthly. 2012 Oct;12(10 HCC).