| Course | NU 602 Epidemiology, Research, and Theory |
|---|---|
| Unit | Unit 4 |
| Paper type | Study design justification |
| Length | About 830 words, 6 pages |
| Format | APA 7 student paper |
| School | Herzing University |
| Program | MSN |
| Updated | October 2026 |
Free sample paper for NU 602 Unit 4
Fluoroquinolones and Aortic Aneurysm in Older Adults: Justifying an Active-Comparator, New-User Cohort Design
Student Name
Department of Nursing, Herzing University
NU 602: Epidemiology, Research, and Theory
Instructor Name
February 11, 2026
Fluoroquinolones and Aortic Aneurysm in Older Adults: Justifying an Active-Comparator, New-User Cohort Design
The Question
Among adults aged 65 and older in the United States who fill a prescription for an oral antibiotic for a respiratory or urinary infection, does a fluoroquinolone, compared with amoxicillin-clavulanate, increase the risk of aortic aneurysm or aortic dissection within 60 days?
The question comes from practice. In primary care, fluoroquinolones are still prescribed to older adults for urinary and respiratory infections, and in 2018 the Food and Drug Administration warned about an increased risk of aortic aneurysm and dissection. Published studies from Taiwan and Sweden reported higher risk with fluoroquinolones (Lee et al., 2015; Pasternak et al., 2018), but older adults in the United States differ in prescribing patterns, comorbidities, and the comparators available. The question is written so that a reader could tell whether a study had answered it: it names the population, the exposure, the comparison, the outcome, and the time window.
What the Question Requires
Four features of the question constrain the design. The outcome is rare: aortic dissection and aneurysm events occur in a small fraction of older adults each year, and far fewer within any 60-day window, so a study must include very large numbers to observe enough events. The exposure is common and brief: millions of antibiotic courses are filled each year, each lasting about a week. The question is about harm, not benefit. And the obvious confounder is the reason for prescribing: patients given fluoroquinolones may be sicker, or have different infections, than patients given other antibiotics, and infection itself may affect vascular risk.
Designs Considered
Each candidate design was judged against those four features.
| Design | Strength for this question | Decisive problem |
|---|---|---|
| Randomized controlled trial | Would remove confounding by indication | Unethical to randomize patients to a drug to test whether it causes a serious harm; would need hundreds of thousands of patients for a rare outcome |
| Prospective cohort with primary data collection | Could measure smoking and blood pressure directly | Would need an enormous cohort followed for each antibiotic course; cost and time prohibitive |
| Cross-sectional survey | Quick and inexpensive | Cannot establish that the antibiotic came before the aneurysm; prevalence of aneurysm reflects years of disease, not a 60-day risk |
| Case-control study | Efficient for rare outcomes | Recalling antibiotic use months later is unreliable; with complete pharmacy records, the efficiency advantage is smaller |
| Case-crossover study | Each case is its own control, removing fixed confounders | Cannot control for the infection that prompted the antibiotic, which changes within the person over time |
| Active-comparator, new-user cohort in claims data | Very large numbers; prescriptions recorded when filled; comparator with the same indications | Lacks clinical detail such as smoking and aneurysm size; outcome defined by diagnosis codes |
Why the Active-Comparator, New-User Cohort
The chosen design uses Medicare claims to identify older adults who fill a first prescription for either a fluoroquinolone or amoxicillin-clavulanate after a period of at least a year with no prescription for either, and follows them for 60 days for a hospital diagnosis of aortic aneurysm or dissection. Three features of the design answer the question's requirements.
First, size. Medicare claims include millions of antibiotic prescriptions to older adults each year, which is what a rare outcome requires. Prescriptions are recorded when filled, so exposure does not depend on memory, and the timing of exposure relative to the outcome is known.
Second, the active comparator. Comparing fluoroquinolone users with people who received no antibiotic would mix the effect of the drug with the effect of having an infection. Comparing them with users of another antibiotic prescribed for the same infections makes the two groups more alike in the reason for treatment (Lund et al., 2015). Amoxicillin-clavulanate is used for similar respiratory and urinary infections and has no known association with aortic disease. Pasternak and colleagues (2018) used amoxicillin as the comparator in their Swedish cohort for the same reason.
Third, new users. Restricting the study to people starting a course after a year without either drug avoids comparing long-standing users, who have already survived early risks, with new starters, and makes follow-up begin at the same point for everyone.
Why Not a Nested Case-Control Study
A case-control study nested within the same claims data is the strongest alternative and is the design Lee and colleagues (2015) used. When the outcome is rare and the cost of measuring each participant is high, sampling controls is efficient. In claims data, however, exposure is already recorded for everyone at almost no additional cost, so the efficiency gain is small. A cohort also gives absolute risks, the number of additional events per 100,000 courses, which matter for a prescribing decision in a way that an odds ratio alone does not. A nested case-control design would become the better choice if confirming each outcome required reviewing medical records, since only cases and sampled controls would need review.
What the Design Cannot Do
Claims data do not record smoking, blood pressure control, or known aneurysm size, all of which affect risk. If these differ between fluoroquinolone and amoxicillin-clavulanate users, residual confounding will remain. The analysis will adjust for recorded proxies, such as diagnoses of hypertension and chronic lung disease, and use propensity scores to balance measured characteristics, but it cannot balance what is not measured. The outcome, defined by hospital diagnosis codes, will include some misclassified events, so a sample of cases will be validated against records. And some aneurysms are found incidentally on imaging done for the infection itself; restricting the outcome to dissection or emergency repair would reduce that bias.
Conclusion
The question asks about a rare harm from a common, brief exposure, with strong confounding by the reason for prescribing. A trial is not ethical, primary data collection is not feasible, and cross-sectional data cannot establish timing. An active-comparator, new-user cohort in Medicare claims provides the size, timing, and comparability the question needs, at the cost of clinical detail, which the analysis will address as far as the data allow and report as a limitation where it cannot.
References
Lee, C.-C., Lee, M.-T. G., Chen, Y.-S., Lee, S.-H., Chen, Y.-S., Chen, S.-C., & Chang, S.-C. (2015). Risk of aortic dissection and aortic aneurysm in patients taking oral fluoroquinolone. JAMA Internal Medicine, 175(11), 1839-1847. https://doi.org/10.1001/jamainternmed.2015.5389
Lund, J. L., Richardson, D. B., & Stürmer, T. (2015). The active comparator, new user study design in pharmacoepidemiology: Historical foundations and contemporary application. Current Epidemiology Reports, 2(4), 221-228. https://doi.org/10.1007/s40471-015-0053-5
Pasternak, B., Inghammar, M., & Svanström, H. (2018). Fluoroquinolone use and risk of aortic aneurysm and dissection: Nationwide cohort study. BMJ, 360, Article k678. https://doi.org/10.1136/bmj.k678
Reading the NU 602 Unit 4 instructions
In Unit 4 you choose and justify a research design for a specific question. Expect to state the question in measurable terms, identify the features that constrain the design, consider several designs, justify your choice and acknowledge what it cannot do. Many versions ask you to compare experimental, quasi-experimental and observational options. Check whether your section expects a design you could actually run or allows a hypothetical study with large data sets. Use methods literature to support the design choice and published studies on the same question to show what others have done. A clear question is the foundation, since the design follows from the outcome's frequency, the exposure's timing and the likely confounders.
How this NU 602 Unit 4 example is built
The paper begins with the question written precisely, naming population, exposure, comparator and outcome. The next section lists four features that any design must handle, which turns the choice into reasoning rather than preference. A table judges each candidate design, from a randomized trial to a case-control study, against those features. The justification section explains how the active comparator reduces confounding by indication and how restricting to new users avoids prevalent-user bias. A section takes the strongest alternative seriously and explains why the cohort was preferred for this question. The limits section names what claims data lack and how that could bias the result. The conclusion restates the match between question and design in a few sentences.
Reading the NU 602 Unit 4 rubric
Herzing's rubric for a design justification typically scores the research question, the analysis of design requirements, the consideration of alternatives, the justification, limitations and writing. Question points require a specific, answerable question. The requirements criterion rewards explicit features such as outcome frequency and confounding. Alternatives earn credit when at least two designs are weighed fairly. The justification row decides most of the grade: top marks need the choice explained with methods literature and tied to each feature of the question. A design picked by habit, with no reference to the outcome's rarity or the confounding it faces, rarely earns more than partial credit. Limitations are scored on specific threats. The writing criterion covers organization, precise research vocabulary and APA 7 references, including Pasternak and colleagues' Swedish cohort on the same question.
NU 602 Unit 4 help: mistakes that cost points
The usual Unit 4 problem is choosing a randomized trial by default without considering whether it is ethical or feasible for the question. Another is a question too vague to drive a design. Students often describe designs in textbook terms without linking them to the specific outcome or exposure. Confounding by indication is frequently missed in drug safety questions. Alternatives are sometimes dismissed in a sentence. Limitations listed generically, such as small sample, do not fit large database studies. Finally, cite methods sources for the design features you rely on, such as active comparators and new-user restriction, because graders want to see that the choice rests on established methods.
Get a NU 602 Unit 4 example written to your instructions
Send the Unit 4 instructions and the rubric from your NU 602 classroom, plus the question or clinical problem your section assigned. We write a custom example that sharpens the question, lists what a study would have to establish, rules out the alternatives on stated grounds and defends one design with its limits, returned in 24 to 48 hours. The first custom sample is free.
More NU 602 papers and MSN samples
- NU 602 Unit 1: Rate calculation worksheet
- NU 602 Unit 2: Prevalence and duration brief
- NU 602 Unit 3: Standardized rate comparison
- NU 602 Unit 5: Bias and confounding review
- NU 602 Unit 6: Screening test evaluation
- NU 602 Unit 7: Theoretical framework application
- NU 602 Unit 8: Study protocol outline
- NU 725 Unit 4: Systems evaluation
- NU 726 Unit 4: Conflict case analysis
- NU 730 Unit 4: Quality improvement proposal
- NU 760 Unit 4: Evidence review
NU 602 Unit 4 questions, answered
Do I have to choose a design I could really run?
Usually yes, and feasibility is part of the grade rather than an excuse. A design requiring twenty years of follow-up and a national sample is not a serious answer for a course project, and saying so out loud is better than pretending otherwise. Where your instructions ask for the ideal design regardless of resources, argue for it and name what makes it unreachable in practice.
Can an observational design ever support a causal claim?
It can support a causal argument, which is different from proving one. Strength of association, correct temporal sequence, a dose relationship, biological plausibility and consistency across settings all push in that direction, and none of them removes the possibility of an unmeasured confounder. Say what your design supports, name what is left open, and let the wording match the method.
How specific does the question have to be?
Specific enough that somebody could tell whether a finished study answered it. Name the population, the exposure or intervention, whatever it is being compared against and the outcome you will measure, including how you will measure it. Anything vaguer leaves every design defensible, and a justification where nothing could be ruled out has not justified the one you picked.
Where can I find a free NU 602 Unit 4 sample paper?
The complete NU 602 Unit 4 fluoroquinolone design paper is available here free, with the design table and the case against the nested case-control alternative. A custom justification can be built around your own research question.
What is an active-comparator new-user design?
It compares people starting the drug of interest with people starting a different drug for the same reason, counting only new users. This sample compares older adults starting a fluoroquinolone with those starting amoxicillin, which reduces confounding by the reason for prescribing.