(2) Measurement Properties 1 — Hierarchy of Evidence & Study Designs

Hi

⭐ What is the Hierarchy of Evidence?

The hierarchy of evidence is basically a ranking of study types from stronger evidence at the top to weaker evidence at the bottom.

⭐ 1. Q: What is the hierarchy of evidence from highest to lowest?

From highest → lowest:

1. Systematic reviews and meta-analyses of randomized controlled trials

2. Randomized controlled trials (RCTs)

3. Cohort studies

4. Case-control studies

5. Cross-sectional studies and surveys

6. Case reports and case studies

7. Mechanistic studies

8. Editorials and expert opinion

⭐ 2. Q: What does a study’s position in the hierarchy of evidence tell you?

Studies higher in the hierarchy generally provide a higher level of evidence and are less prone to bias due to their study design.

🔎 What this means:

Going up the pyramid:

• Quality/level of evidence increases.

• Risk of bias generally decreases.

Going down the pyramid:

• Evidence is generally weaker.

• Risk of bias generally increases.

A study design that is inappropriate for the question may make its conclusions invalid

⭐ 3. Q: How does a systematic review differ from a narrative review?

A systematic review uses a standardized, comprehensive, and reproducible literature search to minimize bias, whereas a narrative review selects a less systematically identified subset of the literature.

🔎 In list form:

Systematic review:

• Comprehensive search

• Identifies, appraises, and synthesizes relevant studies

• Reports its sources and search strategy

• Search can be reviewed and reproduced

Narrative review:

• Does not use an explicitly systematic search

• May depend on study availability or author selection

• Is therefore more vulnerable to bias

Narrative reviews can still be useful, but the professor emphasizes that they are not as robust as systematic reviews.

4. Q: What is a meta-analysis?

A meta-analysis is a statistical analysis that combines the results of multiple scientific studies addressing the same question.

🔎 What this means:

Instead of treating each study separately, the results are statistically combined to produce a pooled estimate.

💡 Think:

Systematic review = systematically find and evaluate the studies.
Meta-analysis = statistically combine results from multiple studies.

The lecture places systematic reviews and meta-analyses of RCTs at the top of the evidence pyramid

⭐ 5. Q: How do descriptive and analytical studies differ?

Descriptive studies describe animals or events without a comparison group, whereas analytical studies compare groups to investigate associations or effects.

🔎 What this means:

If no controls/comparison group are included, the study is descriptive.

Examples:

Descriptive

• Case report

• Case series

• Survey

Analytical

• Observational studies such as cohort, case-control, and cross-sectional studies

• Experimental studies such as randomized controlled trials

⭐ 6. Q: How do randomized controlled trials (RCTs) and observational studies differ?

In an RCT, investigators assign the intervention at random, whereas in an observational study the exposure occurs naturally or already exists and is not assigned by the investigators.

RCT:

• Investigator assigns treatment.

• Assignment is random.

• Includes a control group.

Observational study:

• Investigator observes rather than assigns the exposure.

• Exposure may be a naturally occurring condition, behavior, previous treatment, physical characteristic, etc.

Random assignment helps reduce systematic differences between groups and therefore reduces the potential for bias.

⭐ 7. Q: What sequence of questions can be used to identify a study design?

First determine whether groups were compared, then whether investigators assigned the exposure, and if not, determine whether subjects were selected by exposure, outcome, or neither.

Step 1: Was there a comparison/control group?

• No → descriptive

• Yes → analytical

Step 2: Did the investigators assign the treatment/exposure?

• Yes → clinical trial

• Assigned randomly → RCT

• No → observational

Step 3: If observational, how were animals selected?

• By exposure → cohort

• By outcome → case-control

• By neither, with exposure and outcome assessed together → cross-sectional

Measurement properties 1. - Hie…

⭐ 8. Q: What should you look at to identify a study design?

Identify the study design by asking how the animals were selected and whether the exposure or treatment was assigned by the investigators.

🔎 What this means:

Do not identify the design just from the topic of the question.

Instead ask:

• Did the researchers assign the treatment/exposure?
  → If yes, think clinical trial/RCT.

• Were animals selected based on exposure?
  → Think cohort.

• Were animals selected based on outcome/disease status?
  → Think case-control.

• Were exposure and outcome measured at the same time?
  → Think cross-sectional.

💡 The study design comes from what the researchers actually did, not just what they wanted to find out.

⭐ 9. Q: What are an exposure and an outcome in a study?

An exposure is a factor hypothesized to affect or be associated with an outcome, while the outcome is the disease, condition, performance measure, or other result of interest.

Exposure = the possible cause/influencing factor
Outcome = the result you are watching for

🔎 What this means:

An exposure can include:

• A treatment

• A disease

• A previous event or condition

• A physical characteristic such as teat conformation

• A risk or preventive factor

An outcome can include:

• Disease

• Productivity

• Performance

• Survival/lifespan

Study animals can therefore fall into four combinations:

• Exposure + outcome

• Exposure + no outcome

• No exposure + outcome

• No exposure + no outcome

⭐ 10. Q: What do longitudinal, prospective, and retrospective mean in observational studies?

Longitudinal studies involve a time relationship between exposure and outcome; prospective studies observe outcomes that have not yet occurred, while retrospective studies examine outcomes that have already occurred.

🔎 What this means:

Longitudinal

• Exposure occurs before outcome.

• Subjects/events are considered across time.

Prospective

• Outcome has not happened yet when the study begins.

• Study moves forward toward the outcome.

Retrospective

• Outcome has already happened.

• Investigators use historical information.

The notes emphasize that cohort and case-control studies can technically be either prospective or retrospective, although you mainly need to recognize prospective cohort and retrospective case-control designs.

⭐ 11. Q: How do cohort, case-control, and cross-sectional studies differ?

Cohort studies select by exposure and assess later outcomes; case-control studies select by outcome and look back for exposures; cross-sectional studies assess exposure and outcome at the same time.

Cohort: compare exposed vs unexposed animals and their risk of developing an outcome.

Case-control: compare cases vs controls and ask whether cases were more likely to have been exposed.

Cross-sectional: take essentially a snapshot in time and look for an association between exposure and outcome.

12. Q: What questions should be asked when appraising whether a paper is useful for a clinical question?

Determine whether the study design is appropriate, what level of evidence it provides, whether its quality is adequate, and whether it is relevant to the clinical question and patient/population.

🔎 What this means:

Do not just read the conclusion and decide whether you agree.

Ask:

1. Is the study design appropriate?

2. What level of evidence does it provide?

3. Is the paper’s quality sufficient?

4. Is it relevant to my question, population, or patient?

🧠 13. Q: When constructing a question in PICO format, what does C refer to?

A. Chief complaint
B. Comparator
C. Contrast
D. Calisthenics

B: Comparator

🧠 14. Q: The I, or Intervention, in PICO may include which of the following?

A. Treatment
B. Prognostic factor
C. Exposure
D. All of the above

The correct answer is D: All of the above

The professor uses I broadly. It may represent a treatment, prognostic factor, or exposure/risk factor.

🧠 15. Q: Which PICO question best addresses whether plasma-derived colostrum affects long-term heifer performance compared with maternally derived colostrum?

A. In [calves], does [giving plasma-derived colostrum], [compared with maternally derived colostrum at birth], [negatively affect heifers’ performance long term]?

B. In [cattle], is [giving plasma-derived colostrum better than maternally derived colostrum]?

C. In [calves], does [giving plasma-derived colostrum at birth] [negatively affect heifers’ performance long term]?

D. Does [giving plasma-derived colostrum compared with maternally derived colostrum at birth] [negatively affect the animals’ performance long term]?

The correct answer is A because it contains the population, intervention, comparator, and outcome specified in the farmer’s question.

🧠 ⭐ 16. Q: When deciding on a treatment, which listed study design provides the highest level of evidence?

A. Expert opinion
B. Cohort study
C. Randomized controlled trial
D. Case-control study

The correct answer is C: Randomized controlled trial.

A systematic review/meta-analysis of RCTs would rank even higher, but it is not one of the choices in this question. Measurement properties 1. - Hie…

🧠 ⭐ 17. Q: Which listed study type provides the best evidence for answering a treatment question?

A. Case study
B. Textbook
C. Systematic review
D. Case-control study
E. Cross-sectional study

C: Systematic review.

🔎 What this means:

The evidence pyramid places systematic reviews/meta-analyses of RCTs above individual RCTs

🧠 ⭐ 18. Q: Cows are enrolled based on poor versus good teat conformation and followed for one lactation to see whether they develop mastitis. What is the study design?

A. RCT
B. Cohort
C. Case-control
D. Cross-sectional

The correct answer is B: Cohort study.

🔎 What this means:

The cows are selected based on the exposure:

• Poor teat conformation

• Good teat conformation

Then they are followed to see who develops the outcome, mastitis.

Exposure → outcome = cohort.

🧠 ⭐ 19. Q: Calves are randomly assigned to no DFM, 1.0 g DFM, or 0.5 g DFM. What is the study design?

A. RCT
B. Cohort
C. Case-control
D. Cross-sectional

The correct answer is A: Randomized controlled trial.

🔎 What this means:

DFM means direct-fed microbial, which the slide describes as a probiotic.

The investigators assign the treatments at random, which identifies the study as an RCT

🧠 ⭐ 21. Q: Investigators select dogs with pancreatitis and control dogs without pancreatitis, then look back in medical records for dietary and other exposures. What is the study design?

A. RCT
B. Cohort
C. Case-control
D. Cross-sectional

The correct answer is C: Case-control study.

🔎 What this means:

Animals are selected based on outcome status:

• Cases = pancreatitis

• Controls = no pancreatitis

Then investigators look backward for previous exposures.

Outcome → previous exposure = case-control.

🧠 ⭐ 22. Q: Cows are tested for BLV and their milk-production data are recorded on the same day. What is the study design?

A. RCT
B. Cohort
C. Case-control
D. Cross-sectional

The correct answer is D: Cross-sectional study.

🔎 What this means:

The exposure (BLV status) and outcome (milk production) are assessed at essentially the same point in time.

That is the defining feature of a cross-sectional design.

🧠 ⭐ 23. Q: Investigators identify rabbits with and without a rare cancer and then ask owners about the rabbits’ diets during the previous three years. What study design is this?

A. Cross-sectional study
B. Cohort study
C. Case-control study
D. Case series study

The correct answer is C: Case-control study.

🔎 What this means:

They begin with:

• Cases: rabbits with cancer

• Controls: rabbits without cancer

Then they look backward for the possible exposure, which is the previous diet.

That is classic case-control logic.

🧠 24. Q: When deciding whether to apply a paper’s findings in practice, do you only need to decide whether you agree with its conclusions?

A. No
B. Yes

The correct answer is A: No.

🔎 What this means:

You also need to appraise:

• Study design

• Level of evidence

• Quality

• Relevance to your question and patient/population

Simply agreeing with the authors’ conclusion is not sufficient appraisal. Measurement properties 1. - Hie…

🧠 ⭐ 25. Q: Which observational study designs may be used to assess the risk of an outcome?

A. Cross-sectional study
B. Case-control study
C. Cohort study
D. All of the above

The correct answer is D: All of the above.

🔎 What this means:

All three study designs can ask whether an exposure is related to an outcome, but they approach that relationship differently.

A risk-related question does not automatically determine the study design.

Cohort, case-control, and cross-sectional studies can all investigate associations between exposures and outcomes; you identify the design based on how animals were selected and when exposure/outcome were assessed.