Arsenic exposure from drinking water,

comprise a series of epidemiological calculations and short answer questions in addition to questions related to an epidemiological article.

Epidemiology

The questions in Section 1 relate to the article:
Argos M et al. Arsenic exposure from drinking water, and all-cause and chronic-disease mortalities in Bangladesh (HEALS): a prospective cohort study. The Lancet 2010; 376:252- 258.

Q1.    What was the aim of this study?    (1 mark)

Q2. Identify the criteria for selection into the study and discuss the  advantages  and disadvantages of these criteria.    (5 marks)

Q3.    a) Were the authors successful at minimising loss-to-follow-up?  Provide data to  support your answer.    (3 marks)

1. b) Who (which groups) would be most likely to drop out?    (2 marks)

Q4.    What was the primary outcome of the study and how was it ascertained?    (2 marks)

Q5. Comment on the advantages and disadvantages of the method of ascertainment of the outcome?    (4 marks)

Q6.    a) What were the exposure variables and how were they defined?    (3 marks)

1. b) Comment on the strengths and weaknesses of the exposure variables.    (5 marks)

Q7.  Is there evidence that BMI is a confounder of the relationship between arsenic  exposure and mortality?    (2 marks)

Q8. On page 255, 3rd paragraph, the authors report the finding that “a one-quartile increase in arsenic concentration in well water was associated with a 15% increase in all-cause mortality (95% CI 1.05-1.26)”. Explain in your own words the interpretation of this 95% CI.    (2 marks)

Q9. On page 255, 5th paragraph, the authors report  the  finding that “multivariate-adjusted Hazard Ratio (HR) for comparison of high baseline exposure to low baseline exposure was 1.46 (95% CI 1.14-1.86) for deaths occurring after follow up 1”. Explain in your own words the interpretation of the HR.    (3 marks)

Q10. Did the authors observe a change in risk of death associated with changes in arsenic concentration in urine over time?     Refer to or provide data to support your answer.
(2 marks)

Q11. What is the interpretation of the attributable proportion based upon well water for chronic-disease mortality of 24%?    (2 marks)

Q12. Imagine you are designing a RCT to evaluate the impact on urine arsenic  of  an intervention to reduce arsenic exposure in well water. The intervention is the one–off addition of a chemical to remove arsenic in the water. The follow up will be 6 months.

1. a)    Who would be your study population?    (2 marks)
2. b)    Draw a diagram to illustrate the study design.    (3 marks)
3. c)    Identify and explain one ethical issue of this study.    (3 marks)

Q13. An outbreak of gastroenteritis that appears to be related to the consumption of fast food has recently occurred in Mandurah. A case-control study was undertaken and the following results found.

Food item    Cases (n= 19)    Controls (n=17)
Crumbed chicken    14    11
Any chicken    16    16
Egg rolls    14    3
Fried rice    14    9

1. a)    Which food(s) do you suspect to be the cause of this outbreak? Give reasons for your answer.    (5 marks)
2. b)    Was a case-control study the best study design to use here? Explain your answer.
   (4 marks)

Q14.        In a matched cohort study of oral contraceptive (OC) use and breast cancer, where the exposed: non-exposed ratio was 1:1, the matched-pair findings were as follows.

No history of OC use
Breast cancer    No breast cancer

Previous OC use    Breast cancer    12    102
No breast cancer
32
87

1. a)    What is the relative risk of breast cancer for those with a history of OC use? Show all working.    (5 marks)
2. b)    The 95% confidence interval for the RR is 1.1- 26.5. What does this mean in terms of possible random error in the study?    (2 marks)

Q15.    A new screening test for HIV in blood products was administered to 700 people with clinically proven HIV and to 900 people without HIV. The screening test was positive for 670 of the proven HIV cases and 150 of the people without HIV.

1. a)    Calculate the sensitivity and specificity of the test.    (2 marks)
2. b)    The old screening test for HIV has a sensitivity of 80% and its specificity is 97%. Discuss if you would recommend the old or the new screening test for community based screening of blood donations?    (3 marks)

Q16. A number of employees at a factory in WA have recently been diagnosed with cancer and the union has demanded an investigation into whether exposure to the factory working environment is placing their members at increased risk of cancer. You are asked to determine whether an increased risk exists and if so provide an estimate of the degree that factory workers are at risk compared with the community.
A total of 16 cases of the cancer have been diagnosed in factory workers over the last 10 years. You collect the following data from those who have had a cancer diagnosis while working at the factory over this time frame.

AGE GROUP in years    PERSON YEARS AT RISK    Cases
15-19    22.2    0
20-24    325.8    0
25-29    440.1    1
30-34    505.7    2
35-39    425.1    3
40-44    322.8    4
45-49    243.8    3
50-54    211.7    2
55-59    129.6    0
60-64    79.8    1
65-69    5.0    0
70-74    3.0    0
TOTAL    2714.4    16

Recent research has reported that the incident rate of the cancer in the WA population is 3330 per 100,000 person-years, with an age distribution in the community as follows:

AGE GROUP in years    Age-specific IR in Western Australia (per 100,000PY)
15-19    0.283
20-24    1.646
25-29    11.770
30-34    44.110
35-39    103.474
40-44    196.288
45-49    335.884
50-54    413.418
55-59    503.016
60-64    569.476
65-69    565.352
70-74    585.785

Using the above information above, determine the age-standardised rate ratio (show all working) for cancer in the factory workers compared with the community and describe (in lay terms) what this result means.    (5 marks)

                                                            SAMPLE SOLUTION

The study aimed at prospectively assessing whether chronic and recent changes in arsenic exposure can lead to all-cause and chronic disease mortalities within the population of Bangladesh.

#2

Convenient sampling and longitudinal studies were employed as selection criteria. Convenient sampling was employed in selecting trained physician who interviewed and clinically assessed participants. This method of sampling enables researchers to develop theories quickly (Lamb, 2013). Results obtained from studies employing this form of sampling may be biased due to issues of personal prejudice on the part of researchers. Data may also be misinterpreted, particularly when researchers use the information from the research to prove untrue facts. Drawing complete conclusions from the results of study that is based on convenient sampling can be difficult.

Longitudinal study was employed in selecting participants for the study. Longitudinal studies help in the determination of patterns since they involve using and gathering data from long periods (Lamb, 2013). They also permit the learning of cause and effect associations. Besides, data gathered over extended periods allows for accuracy and precision of results. As such, longitudinal studies are high in terms of validity. On the contrary, these studies take long periods and require large samples of study.

#3

1. a) Researchers do not provide information on their success in minimizing loss-to-follow ups. However, the researchers conducted a biennial follow up of participants to minimize loss-to-follow up.
2. b) The most probable group to be dropped out are individuals who are unwilling to take part in the study.

#4

The study’s primary outcome was that chronic arsenic exposure via drinking water leads to an increase in the rate of mortality. Furthermore, researchers identified that recent changes is exposure did not have a significant impact on the mortality rate. This outcome was ascertained using the cox proportional hazard model.

#5

Lamb (2013) states that the key weakness associated with the use of cox proportional hazard model is that this framework does not estimate the baseline hazard. As such, the researchers cannot manage to make any comment about the hazard or any aspect relating to it such as elasticity or effects. Its principal strength lies in the fact that the researcher cannot make error during the specification of baseline hazard.