HW1_SOHAIB_KHURAM_09052023

docx

School

Virginia Commonwealth University *

*We aren’t endorsed by this school

Course

500

Subject

Industrial Engineering

Date

Dec 6, 2023

Type

docx

Pages

5

Report

Uploaded by SohaibK7

BIOS 500 Homework 1 Name: Sohaib Khuram Although you are encouraged to work together in groups, collaboration does not mean copying a classmate’s response to a question. What you turn in must represent your understanding of the concepts learned. 1. Descriptive vs inferential. Do Problems 1.10 (p. 8) and 1.24 (p. 10) in the text- book. (a) 1. 10 is an inferential statistics example since the data is sampled from a part of the population and is used to estimate a national rate. (b) 1. 24 is also a problem of descriptive statistics since the rates (death per 100,00 population) is a way to describe the rates of death by cause and does not involve any hypothesis testing nor any generalizations are made. 2. Observational studies vs intervention studies/experiments. Do Problems 1.18 and 1.20 on p. 9. (a) 1.18 is a case of an observational study as there was no variable that was controller for by the researchers. They were merely looking at differences in rate of death amongst left vs right-handed individuals. (b) 1.20 is clearly an experimental study design as researchers randomly assign individuals to different treatment groups of aspirin vs placebo to test the 10-year incidence rate of a major cardiovascular event amongst 45+ yr old women. 3. Choose the correct answer. (a) Which is cheaper to conduct: case-control or prospective cohort? (b) Which takes longer to conduct: case-control or prospective cohort? (c) Which is longitudinal: case-control, cohort, or both? (d) Which requires an intervention: observational study, clinical trial, or both? (e) Which requires randomization: cohort, clinical trial, RCT, or all of these? (f) Which is prospective: case-control, clinical trial, or both? (g) Which requires that the outcome status is known at the onset of study: case- control, cohort, or both? (h) Which is the gold standard for demonstrating efficacy of an intervention: cohort, clinical trial, or RCT? 1
4. Two surgical procedures were compared in a certain prospective cohort study. The outcome was mortality status (i.e., ‘dead’ or ‘alive’ during the last day of observation). Baseline was set at the date of surgery. Subject follow-up time differed from subject to subject. The last follow-up visit for some patients was as short as 1 month; for some patients it was as long as 3 years. In determining which surgical procedure had a lower mortality risk, a researcher divided the number of people who died in each surgical group by the number of people in that group. Then she compared the two mortality risks by calculating the ratio of the two risks. What did this researcher fail to take into account? (a) Variable follow-up time means that the ratio of people from each surgery might be vastly different and data might be censored. So people with a 1 month follow-up are not contributing the same data as the 3 year follow-up. So the mortality risk can inaccurately account for which surgery was better/less risky. 5. Meinert (1986) lists the requirements of a sound treatment allocation scheme. Flipping a coin to determine whether a subject receives the test treatment or the control treatment violates which requirement? (a) Flipping a coin violates multiple requirements that might not seem consequential. First, while a coin flip is random, it is not balanced. We can’t ensure each group has a fair sample size. Secondly, this isn’t a reproducible method and there doesn’t seem to be any exclusionary criteria. Also, both side know which group they are in which can lead to concealment issues. 6. Using the NIH definition of a clinical trial, which of the following disqualifies the study as a clinical trial? (There may be more than one answer.) (a) A study does not randomize the assignment of treatment to patients. (b) A study does not have a comparison group. 2
(c) A study uses mice as subjects. (d) A study does not have an intervention; the study is observational. (e) A study is cross-sectional. 7. Cohort studies may be prospective or retrospective. What is the similarity be- tween prospective and retrospective cohort studies? What is the difference? (a) Both are observational studies, there are groups/cohorts based on exposure status, data on exposure and outcome status are collected to better understand the relationship between the two. Retrospective studies look back in time to those who have an outcome of interest already and try to connect to different exposures while a prospective study looks to collect data before an outcome of interest occurs. 8. Using PICOT, compare and contrast the two successful studies that looked at the use of monoclonal antibody cocktail in the treatment of patients in the early stage of SARS-CoV2 infection: Bamlanivimab+Etesevimab ( https://www.nejm.org/doi/full/10.1056/NEJMoa2102685 ) and Casirivimab+Imdevimab ( https://www.nejm.org/doi/full/10.1056/NEJMoa2035002 ) Bamlanivimab+Etesevim ab Casirivimab+Imdevima b P opulation Ambulatory patients with mild or moderate covid 19 who were at high risk for progressing to severe disease 275 18+ yr olds, symptomatic, non- hospitalized patients with covid 19 (outpatients) I nterventio n Single intravenous infusion of a neutralizing monoclonal antibody combination agent (2800 mg of bamlanivimab + 2800 mg Etesevimab) Placebo vs 2 different groups for the neutralizing monoclonal antibody serums (2.4g or 8 g of REGN-COV2) C ompariso n Both groups have a lab diagnosis of severe acute respiratory syndrome coronavirus 2 infection Both groups had to be outpatients with no more than 3 days since COVID positive result and no more than 7 days since symptom onset before randomization O utcome Overall clinical status of patients (Covid 19 related hospitalization of death from any cause by day 29) Quantitative virologic analysis, COVID 19 serum antibody testing, and measurement of two component REGN- COV2 in serum was measured to see if viral load was any different in the groups. Specifically, time-weighted average change in the viral load (in log10 copies per milliliter) from baseline (day 1) through day 7 T ime September 4 2020 to December 8 2020 was the enrollment period (3 month period) From May till September 4 th was the interim period for this study that is published
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
9. During a feverish illness, young children may sometimes have a febrile convul- sion. Febrile convulsions are very frightening to parents, but they do not mean that the child is epileptic. They sometimes recur. Wallace and Smith (1980) car- ried out a study to see whether treating children with one of two anti-epileptic drugs, phenobarbitone or valproic acid, would reduce the risk of further con- vulsions. The subjects in this trial were 121 children who had had their first convulsions during a feverish illness of 38 or more, and were consecutive ad- missions to 5 pediatric units. Children were thought to be at an increased risk of another fit if they were under 19 months, had an abnormal neurological state, had a family history of seizures, or the initial convulsion was multiple, prolonged, or on one side of the body only. Parents of these high- risk chil- dren were advised alternatively to give either phenobarbitone or valproic acid. Low-risk children were prescribed either phenobarbitone, or valproic acid, or no treatment. Allocation was sequential. The researchers do not say why they split the children into high- and low-risk groups, but a possible reason was that they thought it unethical not to offer treatment to the children they thought to be at high risk. We could analyze this study as two separate trials, one in high-risk and another in low-risk children. Things started to go wrong. Many parents could not be persuaded that their child should be given the drugs, even though they were in the high-risk group. The researchers decided to reclassify the children into three groups: receiving phenobarbitone (as allocated), receiving valproic acid (as allocated), or receiv- ing no treatment (as allocated, or because parents refused). The distinction
between high- and low-risk was abandoned. In other words, the researchers assumed that those who refused the treatment were comparable to those who accepted it, and placed the refusals in the control group. This means that the researchers could not distinguish between the effects of the treatment and the effects of differences in health behavior, etc. between groups. Refusals in trials usually have a worse prognosis than acceptors, and the control group might have more fits than the treated groups, irrespective of treatment. The analysis that the researchers performed was biased. Why? In this particular case, how does analysis by intention to treat minimize the bias? By going with allocation-based assignment as opposed to randomization or even based on placing subjects based on high/low risk, we introduce a confounding factor such as having high risk patients in the placebo group biases the number of convulsions to be higher in the non-treatment group, thus making it seem like the treatment is more effective than it actually is. There’s also the bias that the health behaviors of those who would refuse treatment, even when considered as high risk and having the most to benefit from treatment, are different from those who would not refuse treatment thus making the comparison unfair. By using Intention to Treat, researchers would randomize subjects to the groups they originally designed for and analyze them regardless of whether they followed the treatment. There might be data loss and other issues to resolve but the researchers wouldn’t have to deal with the bias introduced with the procedure they chose to go with. 10. Did you use artificial intelligence (AI) software to complete this assignment? If you did, describe how you used the tool and write the prompts you used to generate results. (a) I used AI more like a search engine to get definitions into specific questions I had. Some specific prompts I used are as follows: What are some of the differences between clinical trials and randomized clinical trials NIH definition of a clinical trial List Meinert’s requirements for good treatment allocation What is considered intention to treat analysis