Study Design And Evidence Types
Help Questions
USMLE Step 1 › Study Design And Evidence Types
A cross-sectional study surveys 2,000 high school students (ages 14–18) during a single semester to test whether energy drink use is associated with anxiety symptoms. Students complete anonymous questionnaires on weekly energy drink servings, sleep duration, and a validated anxiety scale; school nurses measure resting heart rate once. Results: 29% report at least 3 energy drinks/week; elevated anxiety scores are present in 34% overall, 45% among high users vs 30% among low/nonusers. High users also report shorter sleep and more social media use. Investigators state they cannot determine directionality.
Which bias is most likely to affect the results of this study?
Observer bias because anxiety scale is laboratory-based
Temporal bias from unclear exposure-outcome sequence
Selection bias because random sampling eliminates it
Cross-contamination due to treatment switching over time
Recall bias because cases remember exposures better
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a cross-sectional study, which is appropriate for associating energy drink use with anxiety at one time point. The correct answer identifies temporal bias from unclear exposure-outcome sequence as the main issue. A common misconception, as seen in choice B, is misidentifying the study design due to assuming recall bias dominates in concurrent assessments. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
Investigators conduct a case-control study to evaluate whether prior antibiotic exposure is associated with Clostridioides difficile infection (CDI). Cases are 220 hospitalized adults (median age 67) with positive toxin assay and diarrhea. Controls are 220 hospitalized adults on the same wards without diarrhea, matched by age and length of stay. Antibiotic exposure in the prior 90 days is abstracted from the electronic medication record. Results: 154/220 (70%) cases vs 110/220 (50%) controls received fluoroquinolones. The team notes that some controls later develop diarrhea after discharge, which is not captured.
What is the most appropriate next step in this study design?
Randomize patients to antibiotics to confirm causality
Replace records with participant recall of antibiotics
Exclude matched controls to increase generalizability
Improve control definition to avoid outcome misclassification
Measure incidence of CDI prospectively in all patients
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a case-control study, which is appropriate for associating antibiotic exposure with CDI. The correct answer identifies improving control definition to avoid outcome misclassification as the next step. A common misconception, as seen in choice B, is misidentifying the study design due to suggesting randomization in an observational context. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A randomized controlled trial enrolls 300 adults aged 30–65 with newly diagnosed hypertension. Hypothesis: home blood pressure telemonitoring plus pharmacist titration improves BP control at 6 months versus usual care. Randomization is 1:1; participants cannot be blinded due to the device, but outcome measurement uses automated office BP by staff unaware of assignment. Results: controlled BP (<130/80) occurs in 62% intervention vs 48% usual care. The intervention group has more frequent contact with clinicians and reports improved medication adherence.
Which of the following best describes the study design used in this scenario?
Randomized controlled trial comparing two management strategies
Cross-sectional study measuring BP at one visit
Prospective cohort study observing telemonitoring uptake
Case-control study comparing controlled versus uncontrolled BP
Ecologic study comparing clinics by device availability
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a randomized controlled trial, which is appropriate for comparing telemonitoring to usual care for hypertension control. The correct answer identifies this as comparing two management strategies through randomization. A common misconception, as seen in choice D, is misidentifying the study design due to confusing it with observational cohorts. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A prospective cohort study in 5,600 adults aged 45–75 without colorectal cancer tests whether processed meat intake increases cancer incidence over 10 years. Baseline diet is assessed by food frequency questionnaire; exposure categories are low, moderate, and high processed meat intake. Incident colorectal cancer is confirmed by pathology reports. Results: cancer occurs in 1.8% high intake vs 1.2% low intake. High intake participants also have lower fiber intake, higher alcohol use, and less screening colonoscopy at baseline. Investigators collect colonoscopy history but some records are missing.
Which bias is most likely to affect the results of this study?
Confounding by screening behavior and lifestyle factors
Cross-contamination from controls receiving chemotherapy exposure
Recall bias because cancer diagnosis changes past diet
Observer bias because pathology reports are unblinded
Temporal bias because meat intake follows cancer onset
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a prospective cohort study, which is appropriate for linking processed meat intake to colorectal cancer incidence. The correct answer identifies confounding by screening behavior and lifestyle factors as the bias. A common misconception, as seen in choice B, is misidentifying the study design due to assuming recall bias in baseline assessments. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A cross-sectional study measures vitamin D levels and reports musculoskeletal pain in 700 outpatient adults aged 18–80 during winter months. Hypothesis: low vitamin D is associated with current chronic pain. Vitamin D is measured from a single blood draw; pain is assessed with a standardized questionnaire. Results: 42% have vitamin D deficiency; chronic pain is reported by 36% overall, 44% among deficient vs 30% among nondeficient. Many participants with pain report reduced outdoor activity and higher BMI. Investigators caution about confounding and directionality.
Which bias is most likely to affect the results of this study?
Observer bias because vitamin D assays are subjective
Cross-contamination from randomized vitamin D supplementation
Temporal bias because exposure-outcome timing is unclear
Selection bias because cross-sectional studies randomize subjects
Recall bias because cases overreport vitamin D labs
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a cross-sectional study, which is appropriate for associating vitamin D levels with musculoskeletal pain concurrently. The correct answer identifies temporal bias because exposure-outcome timing is unclear as the bias. A common misconception, as seen in choice B, is misidentifying the study design due to assuming recall bias in lab-based measurements. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A multicenter randomized controlled trial enrolls 640 adults aged 45–79 with type 2 diabetes (mean A1c 8.4%, 52% women) and no prior myocardial infarction. Investigators test the hypothesis that adding once-daily “gliflozin-X” to metformin reduces major adverse cardiovascular events (MACE) over 3 years versus placebo. Participants are randomized 1:1 using a computer-generated sequence stratified by site; identical capsules are dispensed, and patients, clinicians, and outcome adjudicators remain blinded. Follow-up visits occur every 3 months; 8% are lost to follow-up, similar between groups. Results: MACE occurs in 44/320 (13.8%) in gliflozin-X vs 62/320 (19.4%) in placebo; hospitalization for heart failure occurs in 18/320 (5.6%) vs 31/320 (9.7%). Adverse events include genital infections in 27/320 (8.4%) vs 9/320 (2.8%). The investigators conclude the drug lowers MACE risk with acceptable safety in this population.
Which of the following best describes the study design used in this scenario?
Ecologic study comparing group-level diabetes outcomes
Randomized, double-blind, placebo-controlled clinical trial
Cross-sectional prevalence survey at one time point
Prospective cohort study with exposure-based follow-up
Case-control study using retrospective exposure recall
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a randomized, double-blind, placebo-controlled clinical trial, which is appropriate for testing the efficacy and safety of a new drug intervention in patients with type 2 diabetes. The correct answer identifies this as the precise study design, emphasizing randomization, blinding, and placebo use to reduce bias. A common misconception, as seen in choice A, is misidentifying the study design due to confusing it with observational cohort studies that lack intervention assignment. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A university hospital follows a prospective cohort of 2,400 adults aged 50–75 without dementia at baseline (60% women; 30% with hypertension). Investigators test the hypothesis that chronic benzodiazepine use increases incidence of Alzheimer disease over 8 years. Exposure is defined from pharmacy refill records (≥90 days supplied per year). Annual cognitive screening and neurologist adjudication determine incident Alzheimer disease. Results: Alzheimer disease develops in 96/800 (12.0%) exposed vs 144/1,600 (9.0%) unexposed. Exposed participants have higher baseline anxiety and poorer sleep, and they attend follow-up visits less often. Investigators adjust for age, education, depression, and comorbidities, but they acknowledge persistent differences between groups.
Which bias is most likely to affect the results of this study?
Temporal bias because dementia precedes exposure definition
Cross-contamination because controls receive benzodiazepines
Selection bias from differential loss to follow-up
Recall bias because exposure is self-reported
Observer bias because outcomes are laboratory values
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a prospective cohort study, which is appropriate for measuring incident Alzheimer disease in relation to benzodiazepine exposure over time. The correct answer identifies selection bias from differential loss to follow-up, as exposed participants attend less often. A common misconception, as seen in choice C, is misidentifying the study design due to assuming recall bias in a prospective setup with records. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A randomized controlled trial enrolls 210 adults (ages 19–55) with moderate persistent asthma and frequent nighttime symptoms despite inhaled corticosteroids. Investigators test the hypothesis that adding a new leukotriene antagonist reduces exacerbations over 24 weeks. Participants are randomized 1:1 using sealed, opaque envelopes prepared by an independent pharmacist; tablets are identical, and both patients and clinicians are blinded. Exacerbations are defined as need for systemic steroids or urgent visit. Results: 22/105 (21%) in treatment vs 35/105 (33%) in placebo have an exacerbation. However, 18% of participants correctly guess their assignment due to mild headache occurring more often in the active group (26% vs 10%).
What is the primary advantage of the study design described?
It estimates disease prevalence in the population
It is fastest for studying rare outcomes
It proves causation without any potential bias
It reduces confounding through random assignment
It avoids loss to follow-up by design
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a randomized controlled trial, which is appropriate for evaluating the efficacy of a new leukotriene antagonist in reducing asthma exacerbations. The correct answer identifies the study design's advantage, such as reducing confounding through random assignment. A common misconception, as seen in choice C, is misidentifying the study design due to overestimating its efficiency for rare outcomes. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A dermatology group performs a case-control study to test whether indoor tanning is associated with melanoma. Cases are 260 adults aged 25–60 with newly diagnosed cutaneous melanoma from a cancer registry. Controls are 260 adults without melanoma recruited from orthopedic clinics, frequency-matched on age and sex. Exposure is assessed by interview about lifetime tanning bed sessions and sunburn history. Results: 48% of cases report indoor tanning vs 30% of controls. Investigators worry orthopedic clinic controls may have different outdoor activity patterns than the source population.
Which bias is most likely to affect the results of this study?
Confusion of terms between incidence and prevalence
Temporal bias because tanning occurs after melanoma
Observer bias because pathology is not standardized
Selection bias from nonrepresentative control recruitment
Cross-contamination from controls receiving tanning exposure
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a case-control study, which is appropriate for associating indoor tanning with melanoma. The correct answer identifies selection bias from nonrepresentative control recruitment as the main issue. A common misconception, as seen in choice B, is misidentifying the study design due to assuming temporal bias in retrospective exposure. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.
A prospective cohort study enrolls 3,300 adults aged 35–70 without prior stroke from community clinics. Hypothesis: atrial fibrillation (AF) detected on wearable devices increases risk of ischemic stroke over 4 years. Exposure is AF episodes ≥30 seconds confirmed by patch monitor; outcomes are strokes confirmed by imaging and neurologist review. Results: stroke occurs in 52/600 (8.7%) with AF vs 96/2,700 (3.6%) without AF. Participants with AF are older and have more hypertension and diabetes. Investigators adjust for these factors but note some participants start anticoagulation after AF detection.
Which bias is most likely to affect the results of this study?
Observer bias because stroke diagnosis is blinded
Temporal bias because stroke occurs before AF detection
Recall bias because AF is self-reported from memory
Cross-contamination from anticoagulation affecting stroke risk
Sample size error because large cohorts lack power
Explanation
This question tests understanding of study design and evidence types in biostatistics and epidemiology. Study design involves choosing the appropriate methodology to answer a research question while minimizing bias. In this vignette, the described study design is a prospective cohort study, which is appropriate for linking AF to stroke incidence over time. The correct answer identifies cross-contamination from anticoagulation affecting stroke risk as the bias. A common misconception, as seen in choice E, is misidentifying the study design due to confusing temporal bias in a prospective setup. Teaching strategies include emphasizing study design features and common biases, using real-world examples to illustrate abstract concepts, and practicing vignettes to identify design characteristics.