B4.045 - Study Population and Data Collection Flashcards
You are preparing a survey to give to parents who have children enrolled in the local childcare program. While writing the survey, you develop the following question:
Please indicate how much you agree or disagree with the following statement about the childcare program: I feel welcomed by staff and other youth at the center.
What is the primary flaw with this question wording?
This is a double-barreled question (combination of two questions). A better way to ask would be: Please indicate how much you agree or disagree with the following statements about the childcare program: I feel welcomed by staff at the center; I feel welcomed by other youth at the center. When a question is double-barreled, two items are being addressed within a single question. When a participant is asked a double-barreled question, the resulting data are challenging to interpret.
what is a leading question
A leading question suggests to the respondent that the researcher expects or desires a certain answer. The respondent should not be able to discern what type of answer the researcher wants to hear.
You have learned that your hospital ran a breast cancer study in the 1950s, and both medical records and tissue samples were preserved. You are interested in doing new research looking at the genetic markers in these tumors, to define further the mutations that lead to oncogenesis. Since the patients have passed away, a colleague suggests you contact the patients’ families to ask for permission to do the study. What would make you unsure whether this would be the right thing to do?
Confidentiality.
The first order issue is that the patients can’t consent for themselves since they have passed. Second, if you can identify surviving spouses or more likely children or grandchildren from that time, their loved one may not have told them they had cancer. Being dead does not mean you lose the right to medical confidentiality.
Gastric cancer rates have declined over the last several decades, mainly as a consequence of the decreasing prevalence of Helicobacter pylori, decreases in tobacco consumption, improvement of food preservation (e.g., refrigeration), and a diet richer in vegetables and fruit. Nevertheless, gastric cancer remains the fifth most common cancer and the third leading cause of cancer mortality worldwide. Experimental and human studies suggest that allium vegetables and their associated organosulfur compounds (OSC) can have a chemopreventive effect on gastric cancer. To further investigate this association, researchers want to conduct a pilot study. They have limited funds to conduct this study. What study design would be best suited to the study question and the authors’ needs?
Case-control studies are generally less expensive than cohort studies and can be conducted faster. In addition, gastric cancer is a relatively rare outcome, despite being the fifth most common cancer globally. Thus, a backwards-directional study, like a case-control study, makes the most sense.
what is a case control study
In a case-control study design, researchers start with the outcome of interest and look backwards in time for the exposure of interest.
what is a cohort study
In a cohort study, a study with forward directionality, researchers first ask about exposure than follow people forward in time to look for the development of the outcome. If the outcome is rare, a very large study population would have to be used. Cohort studies are well-suited to a rare exposure;
what are case study studies useful for
case-control studies work well for rare outcomes.
The incidence of head louse infestation is high in many countries. This may be explained, in part at least, by the evolution in head lice with lower susceptibility (resistance) to older pediculicides. Two new types of head lice products have found wide acceptance in many countries: essential oil based products and products designed to “suffocate” head lice. Researchers want to compare the efficacy of three topical pediculicides: a pediculicide containing melaleuca oil (tea tree oil) and lavender oil (TTO/LO); a head lice “suffocation” product; and a product containing pyrethrins and piperonyl butoxide (P/PB). What study design is best suited to address this research question?
Randomized controlled trial.
The researchers are interested in conducting a head-to-head experiment. Because the products are already on the market, basic safety has been established. This should minimize any potential risk to study participants who are randomized to one of the treatment groups.
Studies have shown that prophylactic human papillomavirus (HPV) vaccination programs have resulted in significant reductions in vaccine-type infections. However, concerns about the effects of vaccination have slowed uptake in some countries. These include fears that it may lead to premature sexual debut. Some parents believe human papillomavirus (HPV) vaccination increases the chance of risky sexual behaviors among adolescents. Several studies have investigated this association. You are interesting in evaluating the overall body of evidence. What type of study design would you use for this process?
A systematic review is a study design that includes a structured plan for considering scientific literature around a focused study question. In this case, you could conduct a systematic review of articles looking at the association between HPV vaccination and sexual risk behaviors.
In spite of diabetes screening guidelines, nearly 30% of adults with type 2 diabetes and 90% of adults with prediabetes in the U.S. remain undiagnosed, and only half of those eligible for guideline-indicated screening complete diabetes testing. Various screening guidelines recommend targeting diabetes screening to high-risk individuals when they present for clinical care, but each guideline defines risk differently. This impacts clinicians’ ability to identify individuals in need of screening. Non-diagnostic random glucose elevations are more strongly associated with undiagnosed diabetes than traditional diabetes risk factors and may provide an early warning sign of glycemic dysregulation. What type of study design could be used to evaluate the performance of random glucose testing to identify individuals in need of further diabetes testing and compare its performance to current screening guidelines?
Cross-sectional study.
Test performance (sensitivity, specificity, positive/negative predictive value) is typically assessed using a cross-sectional study. In this case, researchers use cross-sectional data (data are collected at one time-point) from adults without diagnosed pre-diabetes or diabetes. Using a blood specimen, a random glucose test is performed. An A1C test is also run as the gold standard for diagnosing diabetes. Results from the random glucose test are compared to other risk factors collected through a survey to determine how well the random glucose test predicted a diabetes diagnosis.
You are designing a study to investigate a treatment of bulimia nervosa. You decide to include patients if they agree to suspend any other treatment they might be receiving and remain on a stable dose of any psychotropic medication they might be receiving. What is the biggest threat to the validity of your study?
Selection bias.
Requiring patients with bulimia nervosa to suspend all other treatments and remain on a stable dose of any psychotropic medication may help control for threats to internal validity (i.e., the ability to accurately characterize the treatment-response relationship), but potentially at the expense of external validity. Selection bias creates a disconnect between the study sample and the patient population by generating a sample that is not representative of the larger population to which we wish to apply results.
You are designing a case-control study to investigate the relationship between dietary intake of docosahexaenoic acid (DHA) during pregnancy and the risk of preterm birth. You plan to enroll subjects at the infant’s six-week check-up. Both cases and controls will be asked about their intake of DHA-rich foods and supplements in the first trimester of their pregnancy. What protection against bias should be incorporated into the study design?
Standardized questionnaires.
The risk of recall bias—a type of information bias that results when study participants incorrectly report actual exposures—is high for this study due to the prolonged period between exposure (first trimester of pregnancy) and data collection (~6 weeks after birth). Recall bias can be differential (differs between cases and controls) or non-differential (cases and controls have same likelihood of recall bias). Standardizing data collection tools will minimize the risk of differential recall.
