Pubh7600 Assessment 1: Epidemiology

*It is assumed that once participants contract HIV they remain HIV positive, and that over the follow-up period some of these cases detected at baseline were lost to follow-up.

• What is the study design used in this study? [1 mark]
• Cross-sectional study
• What was the prevalence of HIV in (i) males and (ii) females aged 15-30 in 2013? [3 marks]
• The solution is obtained by calculating using the number of males and females who tested positive at baseline.

For males= 2.658%

For females = 6.02%

• What was the prevalence of HIV in (i) males and (ii) females after 3 years of follow-up? [3 marks]
• The solution is obtained by calculating using the number of males and females with HIV after three years follow up.

For males= 2.185%

For females = 5.605%

• What was the cumulative incidence of HIV among (i) males and (ii) females over the 3-year study period? [4 marks]

Cumulative incidences refer to all new rates of HIV infection that occurred over the three years divided by the population at risk. In this study, the population at risk will be regarded as the number of individuals who tested negative at baseline (The Open Learn University, 2019).

For males, population at risk = 1693-45=1648

For females, the population at risk =1909-115=1794

Cumulative incidence =

For males = 1.82%

For females = %

• What was the incidence rate of HIV among (i) males and (ii) females for the 3 years? [4 marks]

Incidence rate is calculated by dividing the incidence number by the total population under study then multiplying it by 100 to make it a percentage.

In this case, the total population reduced through death and loss in flow up. Therefore, the total male population became

1693-147= 1546

For females, the total population became

1909-123=1786

For males = 1.95%

For females = 2.575%

• Is cumulative incidence or incidence rate a better reflection of the incidence of HIV in this study population? Please justify your [2 marks]

Cumulative rate is a better reflection of HIV incidence in the study population because it incorporates changes in the total population while cumulative incidences only consider the population at risk, which might be difficult to ascertain.

• Describe the findings on HIV from this study. How would you interpret these findings? (Limit of half a page No more than half a page of text will be marked) [2 marks]

The study findings indicate that HIV prevalence is high in females as compared to males. For instance, only 45 cases were reported in males as compared to 115 in females at baseline. The same occurrences are replicated in the incidence during the follow-up period. However, more losses were reported in males during the follow-up period was high as compared to females.

• Since the completion of this study, the drug Truvada, also known as pre-exposure prophylaxis (PrEP), has been introduced in other countries to reduce the acquisition of HIV in those who regularly take the drug. You are asked to brief the South African health minister on the likely effects on the (i) short-term and long-term prevalence and (ii) incidence of HIV of a government-funded pilot program that would make Truvada (PrEP) freely available to those aged 15-30 years in South Africa. What do you tell her? (Limit of half a page. Please note: no more than half a page of text will be marked) [3 marks]

Introducing Truvada on the population will have insignificant effects on HIV prevalence in the short term. Ideally, very few people will use the drug due to lack of information. However, in the long, the prevalence of HIV will reduce significantly due to the introduction of Truvada as many people will use it and lead to low incidences of the condition. Similarly, a government-funded project making Truvada freely accessible will significantly reduce the incident rate of HIV, as many people will use it in circumstances that may be potential sources of infection.

QUESTION 2:

Question 2 relates to the following table of state-level data male suicide, stratified by quintile of socioeconomic status (SES). (Only the lowest and highest SES groups are shown). Note:   5 years of death data are presented. Population data comes from a single year (2014). It is conventional to present mortality rates as annual average rates.

• Calculate the crude suicide mortality rate (per 100,000) for (i) the lowest SES quintile and (ii) the highest SES quintile for those aged ≥15 years. [4 marks]
  Crude suicide mortality rate==

(i) The lowest SES quintile = 137.778

(ii) The highest SES quintile = 93.564

• Calculate the crude mortality rate ratio (expressed as a %) for suicide in the lowest SES quintile relative to the highest SES quintile [2 marks]

Crude mortality rate ratio in the lowest SES quintile= 137.778

Crude mortality rate ratio in the lowest SES quintile= 93.564

% ratio= 0= 14.72%

• How would you interpret the crude mortality rate ratio you calculated in Question 2b? [2 marks]

The crude mortality rate in the lowest SES quintile is 14.72 % times more as compared with the crude mortality rate in the highest SES quintile.

• How many deaths from suicide would you expect in the highest SES quintile 25-34-year-olds if they had the same suicide mortality rate as 25-34-year-old lowest SES quintile? [1 mark]

Suicide mortality rate as 25-34-year-old lowest SES quintile

=0000=171.998

Number of deaths in the highest SES quintile 25-34 year with the same rate

=  220

Number of deaths expected =220.478

=221 deaths

• Calculate the direct age-standardized suicide mortality rate (per 100,000) for the lowest SES quintile using the highest SES quintile as the standard [4 marks]

Calculating age-specific mortality rates

= 148.854

= 171.998

= 145.786

= 127.122

= 101.318

= 98.558

= 136.615

The proportion of age-specific population per age group using the high SES quintile

For 21.11%

For 219.39%

For 19.72%

For 21.59%

For 19.75%

For 18.44%

For 21.00%

Direct age-standardized suicide mortality rate=

= (21.11%*148.854)+( 171.998*19.39%)+(145.786*19.72%)+(127.122*21.59%)+(101.318*19.75%)+(98.558*18.44%)+(136.615*21.00%)=

= 685.07 suicides per 100,000 standard population.

• Is age an important factor in understanding differences in suicide between low and high SES groups? Using the age-specific data provided, make a specific comment about the differences between the crude suicide rates and the direct age-standardized suicide rate. [4 marks]

Age is an essential factor in understanding differences in suicide between low and high SES groups because financial and economic inequality varies across age groups (Statistics Canada, 2017). Moreover, there are ages where economic responsibilities increase thereby putting pressure on individuals.

Direct age-standardized suicide rate offers a more detailed and effective analysis of deaths as compared to the crude suicide rates. This is because the crude suicide rates only consider a single population within the same period whereas the direct age-standardized suicide rate compares deaths at different times, periods, and populations.

QUESTION 3:

For the following abstracts, please indicate (i) what the study design is and (ii) provide a justification for how you determined the study design

• National level data on dietary fish consumption from 15 countries was obtained from the Food and Agricultural organization. National level osteoarthritis prevalence data was obtained from national health surveillance records from the same countries and age standardized. The correlation between fish consumption and osteoarthritis was calculated to examine whether fish consumption is associated with risk of osteoarthritis. [2 marks]
• Meta-analysis- the study design combines data from different sources and research studies for statistical analysis.  
• Queensland Health conducted a telephone survey of randomly selected Queenslanders to examine the prevalence of lifestyle risk factors (g. physical activity, diet) and a number of chronic diseases (e.g. heart disease, cancer, arthritis). They also calculated the strength of associations between each lifestyle factors and each chronic disease. [2 marks]
• A randomized controlled trial. The design focused on clinical factors in which participants were selected by chance (LaMorte, 2018). Moreover, it was controlled to ensure that all respondents were Queenslanders.

QUESTION 4:

The following table shows data from an epidemiological study:

• What is the likely study design? [1 mark]

Cohort study

 (b) Calculate the relative risk of disease associated with having the exposure [1.5 marks]

Relative risk of disease

=

=23.3

 (c) Calculate the rate difference (per 100 person years) [1.5 marks]

Rate difference = Incidence rate of exposed- Incidence rate unexposed

= - =

==

=40 per excess 100 person years.

(d) How would you interpret this rate difference? [2 marks]

Persons who were exposed to the disease had more 40 infections per 100 person-years as compared to those who were not exposed to the disease. It also means that additional cases of infection occurred in those who were exposed to the disease as compared to those who were not exposed.
(e) How many cases of disease (per 100 person-years) in the total study population can be attributed to the exposure [2 marks?]

==

==

=

70 cases of the disease in the population can be attributed to exposure in 100 person-years.

QUESTION 5:

A study was conducted to investigate the effect of recent cannabis use on the risk of being injured in a road traffic accident.

Researchers recruited 488 people (aged 18+) who had been involved in a road traffic accident and presented to an Emergency Department at any of the major hospitals in Brisbane.

At the same time, researchers also recruited 488 people (aged 18+) who had not been injured in a road traffic accident. These participants were also recruited from the same Brisbane hospital Emergency Departments as those in injured in traffic accidents, but were being treated for causes other than road traffic accidents.

65 people tested positive for cannabis. 42 of these people were involved in road traffic accidents.

• What is the study design [1 mark]
• Case-control Study

• In your own words, what are the key advantages of this study design? [1 mark]

The case-control design allows researchers to choose participants that are only relevant to the study (Sargeant, Kelton, & Oconnor, 2014). Moreover, identification of the specific target group for the study makes the design cost-effective and time conscious.

• Construct a 2x2 table conveying the information above and calculate an appropriate relative measure of the strength of the association between cannabis use and being injured in a road traffic accident [3 marks]

Calculating the relative probability of being injured in an accident after cannabis use

Relative probability measure =

=

=   0.646

Therefore, the relative probability of being involved in a road accident after marijuana use is =   0.65

• How would you interpret your findings in Question 5c? [1 mark]

It indicates that using marijuana increases the risk of being involved in an accident by 0.65, which translates to 65%.

• Calculate the population attributable fraction of road traffic accidents that is associated with cannabis use [2 marks]

Population attributable fraction = *100

= *100

= *100

=14.58%

• How would you interpret the finding in Q5e? [2 marks]

Seven out of 48 individual involved in road accidents tested positive for cannabis use. Similarly, it means that 14.8% of road accident cases are attributable to cannabis use.

• This study was critiqued by other researchers for not selecting an appropriate comparison group (e. those not injured in road traffic accidents). Do you agree with this comment? Please justify your answer by describing characteristics of the comparison group and the study design principles of comparison group selection.
  [3 marks]

Yes, I agree with the critique stating that the comparison group was not appropriate because it does not, in any way, link incidences of road accidents with cannabis use. Instead, it focuses on admissions into a hospital’s emergency section.

In the study, the risk is being involved in a road accident, and the risk factor is cannabis use. Therefore, the comparison group ought to have experienced or been exposed to the risk either itself or the risk factor (Troyanskaya et al., 2018). Consequently, it would have helped in developing a functional relationship between the risk factor and the risk.

No. of participants

recruited

Prevalent cases at baseline

Prevalent cases remaining at follow-up*

Person-years of follow-up

Loss to follow-up in those HIV negative at baseline

Deaths in those HIV negative at baseline

Incident HIV cases at follow-up