Question 1

A graduate student conducting research on the association between vitamin D₃ supplementation and cancer is reviewing the abstract of a recently published journal article. Which of the following conclusions can be drawn from the results of the study? A)BMI modifies the effect of vitamin D₃ on metastatic cancer B)Only BMI ≥30 kg/m² is significantly associated with an increase in metastatic cancers C)The effect of vitamin D₃ on cancer mortality is significant only for certain individuals D)Vitamin D₃ is significantly associated with a reduction in cancer mortality regardless of BMI level

Accepted Answer

Explanation: 11ec2b69_c0e9_c90f_8f63_e9a9c971172e_MD0005_00 This study evaluated the association between vitamin D₃ supplementation and advanced (metastatic or fatal) cancer among adults without a diagnosis of cancer at baseline. The initial results (ie, unstratified analysis) show that vitamin D₃ is associated with a significant reduction in metastatic or fatal cancers compared with placebo (HR: 0.83, 95% CI: 0.69-0.99). A subsequent stratified analysis was conducted based on participants' BMI to determine whether BMI level (categorized as <25 kg/m², 25-29 kg/m², or ≥30 kg/m²) helps explain these associations. The stratified analysis shows that some of these associations were inconsistent across BMI level. In individuals with BMI <25 kg/m², vitamin D₃ was significantly associated with cancer mortality because the CIs excluded the null value of HR = 1 (95% CI: 0.39-0.86). However, vitamin D₃ was not significantly associated with metastatic cancers because the CI included the null value (95% CI: 0.38-1.05). In individuals with BMI 25-29 kg/m², vitamin D₃ was not significantly associated with metastatic cancer or cancer mortality because all CIs included the null value (95% CI: 0.52-1.23 for metastatic cancer; 95% CI: 0.64-1.23 for cancer mortality). Among individuals with BMI ≥30 kg/m², vitamin D₃ was not significantly associated with metastatic cancer or cancer mortality because all CIs included the null value (95% CI: 0.42-1.89 for metastatic cancer; 95% CI: 0.75-1.76 for cancer mortality) (Choice B). Therefore, the effect of vitamin D₃ supplementation on cancer mortality is modified by BMI level-that is, BMI level is an effect modifier (Choice D). Effect modification (interaction) occurs when an extraneous variable called an effect modifier (eg, BMI) changes the effect that an exposure (eg, vitamin D supplementation) has on an outcome (eg, cancer mortality) in certain groups of individuals but not others. A stratified analysis based on the modifier can help detect effect modification because it typically shows different effects in each stratum. (Choice A) There is no significant association between vitamin D₃ supplementation and metastatic cancer at any BMI level because all CIs include the null value of HR = 1. Therefore, BMI does not modify the effect of vitamin D₃ supplementation on metastatic cancer. Educational objective: Effect modification occurs when the magnitude or direction of the effect of the independent variable on the dependent variable (outcome) varies depending on the level of a third variable (effect modifier). Separate (stratified) analyses should be conducted for each level of the effect modifier.

Question 2

A systematic review examines the effect of drug X, a commonly used medication, on all-cause mortality in patients with a chronic pulmonary problem.  A systematic search identified 12 retrospective cohort studies.  The pooled odds ratio of mortality with drug X use was 0.72 (95% confidence interval: 0.61-0.79).  The authors also reported the following funnel plot of the 12 studies, centered around the log of the pooled odds ratio:  The graphical findings suggest the presence of which of the following?
A)Lead-time bias
B)Placebo effect
C)Publication bias
D)Small samples in drug X treatment arms

Accepted Answer

Explanation: A funnel plot is helpful in assessing publication bias. Each study's treatment effect (on the x-axis) is plotted against a measure of that study's size or precision, usually using the standard error of the treatment effect (on the y-axis). The triangle is centered on a summary estimate of the treatment effect (eg, pooled estimate of odds ratio, usually in log scale) with its sides corresponding to standard errors. If there is no bias, any scatter between the study results should be due to sampling variation, and 95% of studies should lie within the triangle centered on the summary estimate and extending 1.96 standard errors on either side. Because a larger sample size is associated with increased precision, larger studies (more powerful) will be at the top and have a narrow spread whereas small studies will be scattered widely at the base of the triangle. Funnel plots should be symmetric in the absence of study heterogeneity and publication bias. In this figure, however, there are no studies on the right side of the funnel. This asymmetry suggests publication bias (ie, studies showing null/increased effect of drug X on mortality are noticeably absent as they are less likely to have been published). Other possible explanations for asymmetry include heterogeneity, methodological anomalies, artifact, or chance. (Choice A) Lead-time bias occurs when several disease interventions are compared and one of the interventions diagnoses the disease earlier than the others without an effect on the outcome (eg, survival). This would make it appear that the intervention prolonged survival whereas in actuality it only diagnosed the disease sooner. This funnel plot does not help to assess lead-time bias. (Choice B) The placebo effect is the phenomenon whereby a patient's symptoms are alleviated by an otherwise ineffective treatment, most likely as a result of the fact that the individual expects that the treatment will work. This funnel plot does not help to assess placebo effect. (Choice D) A small sample size (low number of patients) in the drug X treatment arms would lead to more studies at the base of the triangle because such studies would have lower power and larger standard errors. Educational objective: Asymmetry in a funnel plot suggests publication bias. References: Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. (http://www.ncbi.nlm.nih.gov/pubmed/21784880) Bias in meta-analysis detected by a simple, graphical test. (http://www.ncbi.nlm.nih.gov/pubmed/9310563)

Question 3

Investigators want to explore a possible association between the use of nonsteroidal antiinflammatory drugs (NSAIDs) and the development of Crohn disease.  They plan to administer a standardized questionnaire on NSAID use over the past year to patients with newly diagnosed Crohn disease and to age- and sex-matched healthy volunteers.  The study design is most concerning for which of the following?
A)Lack of external validity
B)Lack of statistical power
C)Length time bias
D)Misclassification of exposure
E)Misclassification of outcome

Accepted Answer

Explanation: Recall bias results from the inaccurate recall of past exposure by participants in a study. It applies mostly to case-control designs and leads to misclassification of exposure. Individuals with a disease are more likely to search their memories to try to identify whether an exposure has contributed to their condition. This study is assessing a potential association between nonsteroidal antiinflammatory drug (NSAID) use and Crohn disease. Patients with newly diagnosed Crohn disease may be more likely to recall NSAID use over the past year as they complete the questionnaire. (Choice A) External validity (or generalizability) is defined as the applicability of a study's results beyond the group that was initially assessed. It answers the question, "How generalizable are the results of a study to other populations?" External validity is not a concern as this study's results may be applicable to other individuals with Crohn disease. (Choice B) There is no information regarding sample size, which would help determine statistical power. However, recall bias would still be the primary concern even with an adequate sample size. (Choice C) Subjects with a rapidly progressive form of disease are less likely to be detected by screening compared to those with slowly progressive disease. This length time bias tends to overstate the beneficial effects of screening on length of survival and mortality. However, this study is less concerned with screening as the diagnosis of Crohn disease has already been made. (Choice E) The outcome (Crohn disease development) is known in this example, so misclassification of outcome is unlikely. Educational objective: Recall bias results from the inaccurate recall of past exposure by participants in a study. It applies mostly to case-control designs and leads to misclassification of exposure. References: Bias and causal associations in observational research. (http://www.ncbi.nlm.nih.gov/pubmed/11812579)

Question 4

A study investigates the prevalence of ischemia in 512 patients with chest pain undergoing stress testing.  Only patients with negative cardiac enzymes and nondiagnostic electrocardiogram changes are included.  Myocardial perfusion imaging for ischemia and coronary artery calcification (CAC) score are obtained for each participant.  The following results are reported: Which of the following best describes the study design?
A)Case-control study
B)Case series
C)Cross-sectional study
D)Prospective cohort study
E)Randomized clinical trial

Accepted Answer

Explanation: 11ec2b69_c0ea_8c61_8f63_dbb54bc5ccde_MD0005_00 This study compares the prevalence (at a specific point in time) of ischemia within different categories of coronary artery calcification (CAC) scores. This snapshot corresponds to a cross-sectional design. Cross-sectional studies can be descriptive or analytic (can infer but not prove causation). Subjects are assessed for the occurrence of risk factor exposure and for the presence of disease; some will not have been exposed or have the disease. (Choice A) A case-control study categorizes patients with and without a specific disease (eg, ischemic heart disease) and compares the risk factor frequency (eg, CAC score) between the 2 groups. (Choice B) A case series is a group of case reports regarding individual patients who had similar clinical manifestations or received similar treatment. Each individual case is typically described in detail. Case reports are sometimes used in rare diseases or novel treatments. (Choice D) A prospective cohort study would follow individuals who do not have the disease/outcome of interest (eg, ischemic heart disease) and who may or may not have a certain risk factor (eg, elevated CAC scores) over time; the study would then assess disease development (ie, incidence of ischemic heart disease). (Choice E) A randomized clinical trial would follow individuals who have been randomized to either a treatment arm (or arms) or a control arm and compare the effect of the intervention. Educational objective: In cross-sectional studies, individuals are assessed at a specific point in time to determine whether or not they have a certain risk factor and a certain disease of interest. References: Observational research methods. Research design II: cohort, cross sectional, and case-control studies. (http://www.ncbi.nlm.nih.gov/pubmed/12533370)

Question 5

Gastric adenocarcinoma is typically diagnosed using upper endoscopy and managed with radical gastrectomy and, in some cases, chemoradiation; nevertheless, long-term survival is poor.  Researchers in a region with a relatively high incidence of gastric adenocarcinoma are evaluating a novel screening test for early detection of the malignancy in high-risk patients.  The novel test is based on the measurement of a recently discovered serologic marker of gastric adenocarcinoma, MGAc.  Based on results of a large study sample of representative high-risk patients screened with MGAc and diagnosed with gastric adenocarcinoma, the mean survival of patients since time of diagnosis is 8.2 months.  A review of records shows that, compared to these patients, similar high-risk patients who were diagnosed using the traditional endoscopic modality had a statistically significant lower mean survival of 7.1 months, with no statistically significant differences in the rates of radical gastrectomy or in the average number or intensity of chemoradiation cycles.  Which of the following best explains these results?
A)Lead-time bias
B)Length-time bias
C)Low sensitivity
D)Recall bias
E)Selection bias

Accepted Answer

Explanation: 11ec2b69_c0e5_0e1e_8f63_bd0bffa64d68_MD0005_00 An ideal screening test detects the disease early enough to allow for an intervention that improves morbidity or mortality. However, bias could lead to incorrect conclusions about the test. Lead-time bias, a common bias seen with screening tests, occurs when a test diagnoses a disease earlier than another test. As a result, the time from diagnosis until death ("apparent survival period") appears prolonged, even though there is no actual improvement in survival. Compared to traditional endoscopic evaluation, use of the novel serologic test for gastric adenocarcinoma screening was associated with a statistically significant apparent increase in survival (since time of diagnosis) by ~1 month (8.2 - 7.1 months); however, the comparable gastrectomy rates and chemoradiotherapy courses suggest that the serologic marker had no impact on management or actual (as opposed to apparent) survival. In other words, the new test gives the appearance that patients' survival is increased, but only because it is able to diagnose the disease earlier; the intervention does not change how long the patients live. (Choice B) Length-time bias results when subjects with a rapidly progressive form of disease are less likely to be detected by screening compared to those with slowly progressive disease. Patients with slowly progressive disease tend to be asymptomatic for a longer period, increasing the likelihood that they will be diagnosed based on screening rather than clinical symptoms. This may give the impression that screening improves disease survival, whereas in actuality it only identifies more benign cases. There is no indication that the patients identified by the serologic marker have more benign forms of cancer given their management. Length-time bias (Choice C) Sensitivity, the ability of a test to correctly identify patients with a disease, is important for a screening test; a negative result on a sensitive test helps to rule out a disease. The comparable gastrectomy rates and chemoradiotherapy courses suggest that the sensitivity of the serologic test is likely high, as it identified patients with gastric adenocarcinoma as well as endoscopy did. (Choice D) Recall bias can distort the results of a case-control study because the investigators must rely on the accounts of exposure from the participants in the study. (Choice E) Selection bias often refers to systematic differences between groups in terms of treatment response or prognosis. The high-risk patients studied were part of a large and representative study sample, and they had gastrectomy rates and chemoradiotherapy courses comparable to those of patients diagnosed endoscopically. This makes selection bias less likely. Educational objective: Lead-time bias is a common bias seen with screening tests. It occurs when a test diagnoses a disease earlier than another test does. As a result, the time from diagnosis until death appears prolonged even though there is no actual improvement in survival.

Question 6

A large longitudinal study examined the association between obesity (defined as BMI >30 kg/m²) and risk of ischemic stroke in a cohort of patients. After assessing participant baseline characteristics, researchers followed 2,300 patients for a mean of 2.7 years and determined the incidence of major cardiovascular events. Within the entire cohort, the crude relative risk (RR) of ischemic stroke was 2.4 in obese patients as compared to nonobese patients, with a 95% confidence interval (CI) of 1.8-2.6. When the cohort patients were divided into those with diabetes mellitus and those without, among patients with diabetes, the RR of ischemic stroke was 1.1 in obese as compared to nonobese patients (CI, 0.9-1.6) whereas among patients without diabetes, the RR was 1.0 (CI, 0.7-1.5). Which of the following is the best interpretation of this study's findings? A)Diabetes confounds the association between stroke and obesity. B)Obesity increases the risk of stroke in patients with diabetes only. C)Obesity independently increases the risk of stroke in the study population. D)Obesity is confounding the association between stroke and diabetes. E)The same relative rate of stroke is seen in obese and nonobese patients.

Accepted Answer

Explanation: Confounding bias results when an exposure-disease relationship is obscured by the effect of an extraneous factor associated with both the exposure and the disease. In this case, diabetes (confounder) is associated with both obesity (exposure) and stroke (disease). When a crude analysis is performed (without taking diabetes into account), there appears to be a significantly increased risk of stroke in obese individuals as compared to nonobese individuals in the cohort, as reflected by a crude relative risk (RR) of 2.4 with a 95% confidence interval (CI) that does not include the null value (1.0 for RRs). However, when the analysis is stratified by diabetes status, there is no statistically significant association as both 95% CIs cross 1.0, as expected in the presence of a confounding variable. Therefore, the crude RR >1 was due to confounding by diabetes. If there is a confounding variable and the analysis is stratified by this variable, there typically will be no significant difference in risk between the groups because the confounding effects have been removed. Confounding threatens the internal validity of a study. The RRs reported are for stroke among obese compared to nonobese individuals, not for stroke among patients with diabetes compared to those without diabetes. These results do not imply that diabetes is not associated with ischemic stroke. Rather, they reflect that: Among cohort patients with diabetes, no difference was seen between obese and nonobese patients in terms of risk of stroke Among cohort patients without diabetes, no difference was seen between obese and nonobese patients in terms of risk of stroke The corresponding RRs were ~1.0. Therefore, obesity does not independently increase the risk of stroke (Choice C). There may be a difference in the overall risk of stroke between the cohort patients with diabetes and those without diabetes; however, the RRs reported specifically compare obese to nonobese patients within those 2 groups. (Choice B) In patients with diabetes, the RR of stroke in obese compared to nonobese individuals is 1.1, but the 95% CI includes 1.0. This means that the RR is not statistically significant. (Choice D) The stratification in this analysis is based on diabetes, not obesity. Diabetes is the confounding variable. (Choice E) Relative rates (as opposed to relative risks) involve incidence rates at a given point in time and are based on person-time calculations that are not given in this question. Educational objective: Confounding bias results when an exposure-disease relationship is obscured by the effect of an extraneous factor that is associated with both the exposure and the disease. If there is a confounding variable and the analysis is stratified by this variable, there will be no significant difference in risk between the groups because the confounding effects have been removed. References: Causation, bias and confounding: a hitchhiker's guide to the epidemiological galaxy Part 2. Principles of causality in epidemiological research: confounding, effect modification and strength of association. (http://www.ncbi.nlm.nih.gov/pubmed/18577320)

Question 7

A 33-year-old woman comes to the clinic for a routine health maintenance examination.  She has no lower abdominal pain, vaginal bleeding, or vaginal discharge.  Her past medical history is remarkable for 1 episode of uncomplicated urinary tract infection treated with a short course of antibiotics.  She is currently not taking any medications.  She has had 5 sexual partners in her life and has used barrier contraception the majority of the time.  Her latest sexual relationship began 3 months ago.  Her last cervical cancer screening examination was a Pap smear performed at the age of 23 that showed no abnormalities.  She does not use tobacco, alcohol, or illicit drugs.  During her visit, she requests screening for cervical cancer and appears extremely concerned as her older sister was recently diagnosed with it. During the study, women with abnormal Pap test results or a positive HPV test underwent colposcopy and biopsy, as did a random sample of women with negative test results.  The purpose of performing a colposcopy and a biopsy on women with negative test results is to reduce which type of bias?
A)Contamination bias
B)Observer bias
C)Selection bias
D)Susceptibility bias
E)Verification bias

Accepted Answer

Explanation: Ideally, gold standard testing would be performed in all participants; however, this may not be feasible due to ethical, time, and cost concerns, particularly if the gold standard method is invasive and associated with an increased risk of complications. For example, in a study of a new screening test for a certain liver condition where liver biopsy is the gold standard, it may not be feasible to perform a biopsy on all patients, particularly those who have a negative result on the new screening test. Such a situation can cause verification bias, or workup bias, a type of measurement bias that occurs when a study uses gold standard testing selectively in order to confirm a positive (or negative) result of preliminary testing; this can result in overestimates (or underestimates) of sensitivity (or specificity). A possible method to reduce verification bias is to perform gold standard testing in a random sample of participants with negative results, as seen in this study on cervical cancer. The results from this analysis can be used to extrapolate the likely number of cases that would have been found if all participants with negative screening tests had been fully investigated with gold standard testing. (Choice A) Contamination bias occurs when the control group unintentionally receives the treatment or the intervention, thereby reducing the difference in outcomes between the control and treatment group. (Choice B) Observer bias occurs when an observer responsible for recording results is influenced by prior knowledge about participants or study details. Blinded studies (as in this case) usually avoid this bias by preventing observers from knowing which treatment or intervention the participants are receiving; this leads to a more objective measurement of outcomes. (Choice C) Selection bias results from the manner in which study participants are selected or lost to follow-up. Randomization in a clinical trial reduces selection bias. (Choice D) Susceptibility bias is a type of selection bias where experimental and control groups differ from a prognostic standpoint, possibly due to unforeseen confounding variables. Patients in this trial were randomly selected for a specific screening method, minimizing this type of bias. Educational objective: Verification bias occurs when a study uses gold standard testing selectively in order to confirm a positive (or negative) result of preliminary testing. This can result in overestimates (or underestimates) of sensitivity (or specificity). References: Correcting for verification bias in studies of a diagnostic test's accuracy. (http://www.ncbi.nlm.nih.gov/pubmed/9871951)

Question 8

Researchers are conducting a clinical trial to assess the effect of erythropoietin on the survival of patients with end-stage renal disease receiving dialysis.  Statistical analysis of the data shows that, after adjustment for age, sex, and duration of disease, patients who did not receive erythropoietin were at increased risk of mortality compared with those who did; the calculated relative risk (RR) of survival and corresponding p-value for patients treated with erythropoietin were RR = 2.4 with p = 0.005.  Among patients with diabetes, erythropoietin use was associated with higher survival rates, with an RR of 6.2 and p < 0.001.  Among patients without diabetes, however, the corresponding RR was 0.94 with p = 0.56.  The researchers are preparing a manuscript for publication of their findings.  Which of the following recommendations is most appropriate for the manuscript?
A)Diabetes status of patients could be ignored.
B)Only the p-value for patients with diabetes should be reported.
C)Overall relative risk (RR; RR = 2.4) is the best measure of outcome to report.
D)Overall RR (RR = 2.4) should be adjusted for diabetes status.
E)Separate RRs for patients with and without diabetes should be reported.

Accepted Answer

Explanation: Effect modification results when an external variable (effect modifier) has a positive or negative impact on the observed effect of a risk factor (exposure) on disease status (outcome). For instance, aspirin use is associated with Reye syndrome in children but not adults; therefore, age (effect modifier) modifies the effect of aspirin (exposure) on Reye syndrome development (outcome). Stratification based on the modifier (eg, analyzing the data by age) can help detect effect modification as it typically shows different effects in each stratum (eg, children, adults); as a result, separate measures of outcome should be reported for each stratum. This study analyzed the association between erythropoietin (exposure) and survival (outcome) among patients with end-stage renal disease on dialysis. Based on the p-values, diabetes appeared to modify the effect of erythropoietin on survival: Among patients with diabetes, erythropoietin was significantly associated with higher survival rates, as the relative risk (RR) was 6.2 with a statistically significant p-value (p < 0.05). In contrast, among patients without diabetes, there was no association between erythropoietin and survival, as the p-value (p = 0.56) was nonstatistically significant (p > 0.05). Therefore, the manuscript should report separate RRs (with their corresponding p-values) for patients with and without diabetes. Effect modification is not due to flaws in the design or analysis phases of the study and should not be controlled or adjusted (Choice D). The RRs and their corresponding p-values should be reported for each stratum of the effect modifier (ie, with and without diabetes) (Choice B). Ignoring the effect modification and reporting only the overall RR would provide a false estimate of the true association between exposure and outcome (Choices A and C). Stratified analysis can help distinguish between effect modification and confounding. With confounding, there is usually no significant difference between the strata. An example of confounding is shown in the exhibit: In a primary school, bigger shoe size may appear to be associated with higher intelligence quotient on crude analysis, but with stratification into younger (eg, grade 1) and older (eg, grade 6) children, the association is no longer statistically significant. In this example, age is a confounder, a variable that is associated with both exposure (shoe size) and outcome (intelligence quotient) and that muddles the exposure-outcome relationship. exhibit Educational objective: Effect modification results when an external variable has a positive or negative impact on the observed effect of a risk factor on disease status. Stratification can help detect effect modification. Separate measures of outcome should be reported for each level of an effect modifier.

Question 9

Women with inflammatory bowel disease (IBD) appear to be at greater risk for human papillomavirus (HPV) infection and HPV-related malignancies compared to healthy counterparts.  A new study evaluated knowledge about HPV infection among women with IBD.  Potential participants were identified from among patients seeking care at local clinics, and all women age ≥15 with an established diagnosis of IBD from 2009 to 2010 were invited to participate in the study.  A questionnaire was used to collect information on demographics, lifestyle, sexual behavior, and HPV-related knowledge; simultaneously, the association between HPV-related knowledge and the other factors was analyzed.  Which of the following best describes the study design?
A)Case series
B)Case-control
C)Cross-sectional
D)Prospective cohort
E)Retrospective cohort

Accepted Answer

Explanation: 11ec2b69_c0ea_da82_8f63_7d4a372f8bd5_MD0005_00 Research studies are broadly classified as having an observational design or an experimental design depending on the control the researchers have over the independent variables. In observational studies (eg, case series, case-control, cohort, cross-sectional), the researchers have no control over the independent variables (eg, exposure to risk factors, treatments). In experimental studies (eg, randomized controlled trial), the researchers control and randomly assign the independent variables (eg, treatments). A cross-sectional study is an observational study used to estimate prevalence of disease. It may also be used to examine associations between risk factors and outcomes (eg, disease) in a well-defined population at one particular time. This type of study typically takes a snapshot of a population and measures frequency of risk factors and outcomes simultaneously. In this study, participants (ie, women with inflammatory bowel disease) were classified according to potential risk factors (eg, demographics, lifestyle, sexual behavior) and outcome status (ie, knowledge about human papillomavirus infection) at the same time (ie, snapshot). The association between risk factor and outcome was then evaluated. However, a major limitation of cross-sectional studies like this one is that the temporal relationship between risk factor and outcome is not always clear. (Choice A) In a case series, a small group of patients with a similar diagnosis or treatment is described at a point in time or followed over a certain period. A case series study is observational but has no comparison group, so it cannot establish associations between risk factors and outcome (eg, disease). (Choices B, D, and E) In a case-control study, 2 groups of individuals are initially identified as cases or controls according to an outcome (eg, disease of interest) and then assessed on their past exposure to risk factors. In a cohort study, 2 groups of individuals are initially identified as exposed or nonexposed according to their status with respect to a specific risk factor and then followed over time to assess development of the outcome. In the cross-sectional study, individuals were assessed for exposure to risk factors and outcome at the same time (ie, snapshot). Educational objective: A cross-sectional study is commonly used to estimate the prevalence of disease, but it may also be used to examine associations between potential risk factors and outcomes (eg, disease). Exposure to risk factors and outcome status may be measured simultaneously at a particular point in time (ie, snapshot), but the temporal relationship between the two is not always clear.

Question 10

The drug advertisement suggests that BoXeva should be added to metformin for the treatment of patients with type 2 diabetes mellitus. A 56-year-old woman comes to the office for follow-up. She has a history of type 2 diabetes mellitus and hypertension. The patient has been on metformin therapy for the past 6 months. Despite adhering to a diabetic diet and increasing her level of physical activity, she has been unable to lose weight. At the current visit, her blood pressure is 145/90 mm Hg, and heart rate is 72/min. Her weight is 82 kg (180 lb). HbA_1c is 7.6%, and serum creatinine is 0.9 mg/dL. The section of the drug advertisement shown does not mention side effects. BoXeva inhibits sodium-glucose cotransporter 2 (SGLT2), which is expressed in the proximal renal tubules. Based on this mechanism of action, which of the following side effects would be expected for the SGLT2 inhibitor class of medications? A)Acute inflammatory pancreatitis B)Acute lactic acidosis C)Angioedema D)Flatulence and diarrhea E)Vulvovaginal candidiasis

Accepted Answer

Explanation: The sodium-glucose cotransporter 2 (SGLT2) promotes glucose reabsorption at the level of the proximal renal tubule. SGLT2 inhibitors can therefore reduce hyperglycemia via glucosuria. A complication frequently seen in clinical trials is vulvovaginal candidiasis (in up to 15% of women), possibly in the setting of higher urinary glucose levels. Other effects include polyuria and increased urinary tract infections. (Choices A and C) Dipeptidyl-peptidase 4 (DPP4) inhibitors should not be used in patients with a history of pancreatitis and should be discontinued in those who develop pancreatitis while on therapy. Some DPP4 inhibitors have also been associated with hypersensitivity responses, including angioedema. (Choice B) Metformin has been associated with lactic acidosis, particularly in acute renal failure. (Choice D) Flatulence and diarrhea are gastrointestinal side effects commonly seen with initiation of biguanide (eg, metformin) or α-glucosidase inhibitor (eg, acarbose) therapy. Educational objective: Sodium-glucose cotransporter 2 inhibitors reduce glucose levels by promoting renal glucose excretion and are associated with vulvovaginal candidiasis, urinary tract infections, and polyuria. References: Evaluation of vulvovaginal symptoms and Candida colonization in women with type 2 diabetes mellitus treated with canagliflozin, a sodium glucose co-transporter 2 inhibitor. (http://www.ncbi.nlm.nih.gov/pubmed/22632452) Canagliflozin: Improving diabetes by making urine sweet. (http://www.ncbi.nlm.nih.gov/pubmed/24186885)

Question 11

The proportion of colonoscopies performed by a physician with at least 1 detected adenoma is called the adenoma detection rate.  A study examined the association between the adenoma detection rate and the risk of colorectal cancer 6 months to 10 years after colonoscopy.  In the graph below, physicians were sorted into quintiles based on their adenoma detection rate, with quintile 1 being the lowest rate of detection and quintile 5 being the highest.  There were no significant differences in rates of complications.  The hazard ratios (HRs), shown along with their corresponding 95% confidence intervals (CIs), were adjusted for other variables including patient age, sex, comorbidities, and indication for colonoscopy.  Which of the following statements best describes the study results?
A)Adenoma detection rate could be a useful quality measure of a physician's colonoscopy performance
B)Adenoma detection rate is not associated with the incidence of colorectal cancer
C)Patients with no adenomas can be safely followed for 10 years without repeat colonoscopy
D)Physicians in quintile 3 prevent significantly more colorectal cancers than those in quintile 1
E)The risk of colorectal cancer appears to increase with higher adenoma burden

Accepted Answer

Explanation: The hazard ratio (HR) is defined as the likelihood of an event occurring in a treatment group relative to the control group. The null value for HR is 1.0. A HR <1.0 indicates an event is less likely to occur in a treatment group than the control group. In this study, the cases of colorectal cancer are listed per quintile; and the HR for incident colorectal cancer for each quintile, with quintile 1 used as a control group (given that HR = 1.0), is plotted on the graph and listed along with its corresponding 95% confidence interval. The HRs appear to decrease with higher quintiles (ie, increasing adenoma detection rates), with the decrease becoming statistically significant in quintiles 4 and 5, where the confidence intervals (0.52-0.9 and 0.32-0.69) do not include the null value. These results, which were adjusted for various variables including patient comorbidities and indications for colonoscopy, suggest that physicians with higher adenoma detection rates are more likely to prevent colorectal cancer in their patients during the next 6 months to 10 years of follow-up. Therefore, adenoma detection rates could reflect useful quality measures for colonoscopy performance. HR is similar to relative risk (RR), except that RR is usually calculated at the end of a study (or other defined endpoint) to convey the risk of an event occurring within that time frame. In contrast, HRs are a measure of the instantaneous risk of an event occurring, usually during a subset of the total study period. (Choice B) The graph shows that increasing adenoma detection rate is associated with decreased risk of developing colorectal cancer. This trend reaches statistical significance in quintiles 4 and 5. Note that, although lack of overlap in the confidence intervals of similar parameters means that there is a statistically significant difference between these parameters, the converse is not necessarily true: If confidence intervals of similar parameters overlap, there may or may not be a statistically significant difference between these parameters. (Choices C and E) No information is provided in the question regarding the relationship between the number of adenomas detected on colonoscopy and the future incidence of colorectal cancer. (Choice D) The 95% confidence interval of the HR in quintile 3 includes the null value of 1.0. Therefore, there is no statistically significant difference between quintile 3 and quintile 1 (despite the downward trend in colorectal cancer incidence). Educational objective: The hazard ratio (HR) is defined as the likelihood of an event occurring in a treatment group relative to the control group. HRs <1.0 indicate an event is less likely to occur in a treatment group than the control group. HRs >1.0 indicate an event is more likely to occur in a treatment group than the control group.

Question 12

A 45-year-old man is diagnosed with dilated nonischemic cardiomyopathy.  Late gadolinium enhancement cardiac magnetic resonance (CMR) imaging shows a left ventricular ejection fraction (LVEF) of 20% and evidence of extensive midwall myocardial scarring.  A recent study determined prognosis in patients with dilated cardiomyopathy using CMR-derived LVEF and evidence of midwall myocardial scarring (fibrosis).  Two models were assessed: one model was based on LVEF alone, and the other incorporated the presence or absence of fibrosis.  The following 5-year prediction curves were reported:  Based on the study results, what is the approximate predicted 5-year all-cause mortality for this patient?
A)10%
B)20%
C)30%
D)40%
E)50%

Accepted Answer

Explanation: 11ec2b69_c0eb_9dd4_8f63_bff8a7e3bdf5_MD0005_00 In these prediction curves, the 5-year risks of mortality or cardiac transplantation were calculated based on the left ventricular ejection fraction (LVEF) derived from cardiac magnetic resonance imaging and on the presence or absence of fibrosis (myocardial scarring). The dotted purple curve was constructed based on information about LVEF alone; the red and dotted blue curves additionally incorporated information about the presence or absence of fibrosis. The top chart provides the risk of all-cause mortality and the bottom chart provides the risk of cardiovascular mortality or transplantation. The top chart is the appropriate choice for all-cause mortality calculations. Given the presence of fibrosis (red curve) and a LVEF of 20% (x-axis), the 5-year all-cause mortality risk (y-axis) was 40%. A patient with LVEF of 20% and no fibrosis had an all-cause mortality risk of about 20% (dotted blue line in top chart) (Choice B). In the absence of information regarding presence or absence of fibrosis, the all-cause mortality risk of a patient with LVEF 20% was about 30% (dotted purple curve in top chart) (Choice C). Thus, the additional knowledge regarding the presence or absence of fibrosis changed mortality estimates. This may be useful in risk classification of patients with heart failure. (Choice A) The approximate 5-year risk of all-cause mortality for a patient with LVEF of 30% (rather than 20%) and absence (rather than presence) of fibrosis was about 10% (dotted blue curve in top chart). (Choice E) The approximate 5-year risk of cardiovascular mortality or transplantation (rather than all-cause mortality) for a patient with a LVEF of 20% and history of fibrosis was about 50%. Educational objective: Mortality prediction curves are useful for providing mortality risk while accounting for a number of variables. To interpret these curves correctly, each variable must be accounted for to arrive at the accurate predicted mortality.

Question 13

A three-arm, randomized crossover trial is conducted to determine whether the profile of circulating micro RNAs (miRNAs) is altered after acute resistance training with blood flow restriction (BFR), as compared with unrestricted low- and high-volume training.  Eighteen healthy volunteers (mean age 25) are enrolled.  The arms are single bouts of leg flexion/extension resistance training at: 70% of the individual single-repetition maximum (1RM)30% of the 1RM30% of the 1RM with BFR (artificially applied by a cuff at 300 mm Hg)
During each training intervention, load-associated outcomes (fatigue, heart rate, and exhaustion) are monitored.  Acute effects (circulating miRNAs and lactate) are determined using pre- and postintervention measurements.  Preliminary results show that lower lactate concentrations tend to correspond with lower miR-143-3p expression, and that the relationship is very strong and statistically significant.  Based on this information, which of the following statements is most correct about the correlation coefficient r for this relationship?
A)r is negative and probably closer to −1 than to 0
B)r is negative and probably closer to 0 than to −1
C)r is positive and probably closer to 0 than to 1
D)r is positive and probably closer to 1 than to 0

Accepted Answer

Explanation: 11ec2b69_c0ea_3e40_8f63_9becc5aa5d10_MD0005_00 The correlation coefficient r describes the direction (negative or positive) and the strength (values closer to −1 or 1 indicate stronger relationships) of the linear relationship between 2 quantitative variables: r < 0 indicates that as one variable increases, the other variable decreases. This means that lower values of one variable tend to relate to higher values of the other variable, and vice versa. r > 0 indicates that both variables increase (or decrease) together. This means that lower values of one variable tend to relate to lower values of the other variable, and vice versa. In this study, results reveal that lower lactate concentrations tend to correspond with lower miR-143-3p expression. This means that both variables increase (or decrease) together, so the correlation coefficient must be positive (Choices A and B). Study results also show that the relationship is very strong and statistically significant. Because the relationship is described as very strong, the correlation coefficient must be closer to 1 than to 0 (Choice C). Educational objective: The correlation coefficient r describes the direction (negative or positive) and the strength (values closer to −1 or 1 indicate stronger relationships) of the linear relationship between 2 quantitative variables. It does not necessarily imply causality.

Question 14

A 5-day-old girl is brought to the office by her adoptive parents for her first newborn visit.  The girl has been doing well.  The parents are concerned about her risk of having hemophilia A.  The birth father and his brother (the paternal uncle) have hemophilia A.  The birth mother does not, and there is no history of bleeding in her family, including in male relatives.  The girl is well-appearing and has no abnormalities on examination.  Which of the following is the most appropriate response?
A)The girl has a 25% chance of having hemophilia and a 75% risk of being a carrier
B)The girl has a 50% chance of having hemophilia and a 50% risk of being a carrier
C)The girl has a 75% chance of having hemophilia and a 25% risk of being a carrier
D)The girl has hemophilia
E)The girl will be a carrier

Accepted Answer

Explanation: 11ec2b69_c0ec_6126_8f63_91b95fe67485_MD0005_00 Hemophilia A is an X-linked recessive disorder that occurs almost exclusively in males. Females are usually asymptomatic carriers of the disease, although mild bleeding is possible in carriers. A female may acquire hemophilia in the rare situation in which her father has hemophilia and her mother is a carrier who transmits the abnormal allele to her. This patient's birth mother has no family history of bleeding, including in male relatives; therefore, it is highly unlikely that the birth mother is a carrier. A father who has hemophilia will transmit the abnormal allele to all his daughters and, consequently, they will be carriers of the disease. Sons will be unaffected, as they receive only a single, normal allele on the X chromosome inherited from their mother. If the mother is a carrier, her daughters have a 50% chance of being carriers, and her sons have a 50% chance of having hemophilia. Educational objective: Hemophilia A is an X-linked recessive disorder that occurs almost exclusively in males. Daughters born to an affected father and a normal mother will be carriers. Sons born to a carrier mother have a 50% chance of having hemophilia. References: Hemophilia: a practical approach to genetic testing. (http://www.ncbi.nlm.nih.gov/pubmed/16295028)

Question 15

A 62-year-old man with a history of dyslipidemia and type 2 diabetes mellitus comes to the emergency department due to severe substernal chest pain and diaphoresis that began 2 hours ago.  His ECG shows ST- segment elevation in leads V1 to V4.  He is taken to the cardiac catheterization lab and a complete occlusion of the mid-left anterior descending artery (LAD) is diagnosed.  A drug-eluting stent is placed in the LAD with complete restoration of flow, and the patient is started on aspirin and clopidogrel.  On the third day of hospitalization, the option of adding rivaroxaban is discussed with the patient.  He asks about the benefits and risks of the added therapy.  Based on the research abstract, The statistical methods section of the full research paper indicates that the analysis was performed using an "intention-to-treat" approach.  The primary purpose of such an approach is most likely which of the following?
A)Decrease the placebo effect
B)Eliminate observer bias
C)Enhance external validity
D)Preserve randomization
E)Reduce type II errors

Accepted Answer

Explanation: Important features of randomized clinical trials are randomization (eg, to eliminate bias in treatment assignments), blinding, and intention-to-treat (ITT) analysis. The basic premise of ITT analysis is that participants in trials should be analyzed in the groups to which they were randomized, regardless of whether they received or adhered to the allocated intervention and regardless of whether they withdrew from treatment. The ITT principle holds that randomization is of paramount importance and that deviation from the original randomized groups can contaminate the treatment comparison. ITT analyses are performed to avoid the effects of crossover and dropout, which may break randomization and affect the outcome. For example, if the sickest patients drop out at a higher rate, even an ineffective treatment may appear beneficial if analysis is performed on only those who finished the treatment. (Choice A) The placebo effect is the phenomenon whereby a patient's symptoms are alleviated by an otherwise ineffective treatment (placebo), most likely due to the individual expecting the treatment to work. The ITT principle does not affect the placebo effect. (Choice B) Observer bias occurs when the observer may be influenced by prior knowledge or details of the study that can affect the results. Blinded studies, such as this one, usually avoid this bias by preventing the observers from knowing which treatment or intervention the participants are receiving, thereby measuring outcomes more objectively. (Choice C) The generalizability or external validity of a study reflects the applicability of the obtained results beyond the cohort that was studied. External validity is dependent on the adequacy of the sample. If the sample is representative of the desired population, then the results will generalize. Although ITT analysis can improve the external validity by including the entire sample originally selected for the trial (regardless of participation or compliance with the intervention), the primary purpose of ITT is to preserve randomization rather than to enhance external validity. (Choice E) Type II error represents failure to reject a null hypothesis when it is false. Type II errors are dependent on the power of a study, which is in turn dependent on sample size. As the sample size and power increase, the likelihood of a type II error decreases. Educational objective: The basic premise of the intention-to-treat principle is that participants in trials should be analyzed in the groups to which they were randomized, regardless of whether they received or adhered to the allocated intervention and regardless of whether they withdrew from treatment. References: Intention-to-treat analysis in clinical trials: principles and practical importance. (http://www.ncbi.nlm.nih.gov/pubmed/12522981)

Question 16

A cohort study of 4,000 patients examines the role of vitamin D supplements on the incidence of colon cancer.  Relative risk (RR) calculations with their corresponding 95% confidence intervals (CI) are reported for subjects taking vitamin D versus controls after a 5-year follow-up period: A separate double-blind randomized clinical trial assigns 3,900 subjects to vitamin D supplementation or placebo.  The following results are reported for subjects taking vitamin D versus placebo after a 5-year follow-up period:
Which of the following best describes the results of both studies?
A)Cancer prevention using vitamin D lacks biological plausibility
B)Lead-time bias best explains the discrepancy in study results
C)Residual confounding best explains the discrepancy in study results
D)Results of both studies support using vitamin D for colon cancer prevention
E)The clinical trial was underpowered to detect any difference in outcomes

Accepted Answer

Explanation: In the cohort study, use of vitamin D supplements was associated with statistically significant lower relative risks (RR) of colon cancer and colon cancer mortality (when adjusted for age and gender) as the RRs were both <1 with confidence intervals (CIs) that do no include the null value (1.0 in the case of RRs). However, with the randomized trial, there was no association between vitamin D supplement use and colon cancer or colon cancer mortality as the CIs included the null value. Confoundingoccurs when the exposure-disease relationship is obscured by the effect of an extraneous factor that is associated with both (in this case, exposure is vitamin D, disease is colon cancer). Randomization helps to remove the effects of both known and unknown confounders. However, for the cohort study, the analysis accounted only for the confounding effects of age and gender, but additional confounding factors (residual confounders) may not have been taken into account, leading to residual confounding. Such factors could include intake of specific nutritients (eg, red meat consumption may be associated with colon cancer and may be less likely among patients who are health-conscious and take supplements, including vitamin D), alcohol consumption (eg, there may be an association between alcohol and colon cancer, and patients who consume large amounts of alcohol may be less likely to use vitamin supplements), or factors that have not been identified. (Choice A) The presence or absence of an association between vitamin D supplementation and colon cancer development is not related to biological plausibility (invoked in trying to prove causation). (Choice B) Lead-time bias is the overestimation of survival due to earlier disease diagnosis. These studies are not evaluating survival. (Choice D) For the randomized study, the confidence intervals for the hazard ratios cross 1.0, so the results are not statistically significant. (Choice E) Sample size is typically determined using an estimate of the expected effect of an intervention along with acceptable error margins. In this trial, the large sample size (3,900 subjects) and relatively high precision (narrow confidence intervals) suggest adequate power. A nonsignificant result (confidence intervals that cross the null) does not imply lack of power. Educational objective: Unknown residual confounders may impact the statistical analysis of study results. Randomization removes the effects of known and unknown confounders.

Question 17

Several clinical trials studied the effect of a new antiviral agent on symptomatic relief of genital herpes simplex virus infection.  All trials were based on physician report of symptom improvement; valacyclovir was used as a comparator.  The pooled analysis of studies that blinded only patients favored valacyclovir with an odds ratio for valacyclovir of 1.65 (95% confidence interval, 1.53-1.75).  The pooled analysis of studies that blinded both patients and treating physicians favored the new agent with an odds ratio for valacyclovir of 0.64 (95% confidence interval, 0.38-0.79).  Which of the following best explains the difference in the reported odds ratios?
A)Effect modification
B)Lead-time bias
C)Observer bias
D)Publication bias
E)Randomization failure

Accepted Answer

Explanation: There is a difference in odds ratio between the studies that only blinded patients and those that blinded both patients and physicians. This is most likely due to observer bias, which occurs when observers misclassify data due to individual differences in interpretation or preconceived expectations regarding the treatment. Observer bias is particularly likely when the studied outcome is qualitative (eg, symptomatic relief of herpes simplex virus infection in this study). In this example, unblinded physicians had knowledge about which drug their patients were taking. It is possible that some of these physicians had a preconceived preference regarding valacyclovir and reported more symptom improvement in the patients taking this drug. (Choice A) Effect modification occurs when an external variable differentially modifies (positively or negatively) the effect of a risk factor on the disease of interest in different groups. (Choice B) Lead-time bias occurs when 2 disease interventions are compared and one of the interventions diagnoses the disease earlier than the other. This leads to the erroneous presumption that there is an effect on the outcome (eg, the intervention prolongs survival outcome). In actuality, the intervention only diagnoses the disease sooner and has no effect on survival. (Choice D) Publication bias occurs when trials with significant positive results are published but trials with negative/null results are not. (Choice E) Randomization failure is not likely to occur in all of the studied trials. Observer bias is a more likely explanation for the observed difference. Educational objective: Observer bias generally occurs in the absence of blinding, when observers misclassify data due to individual differences in interpretation or preconceived expectations regarding the treatment. Observer bias is particularly likely when the studied outcome is qualitative. References: Observer bias in randomised clinical trials with binary outcomes: systematic review of trials with both blinded and non-blinded outcome assessors (http://www.ncbi.nlm.nih.gov/pubmed/22371859)

Question 18

A 71-year-old man comes to the office due to slowly progressive shortness of breath for the past year. He enjoys working in his garden, but shortness of breath has forced him to reduce the time he spends there. He also has a mild morning cough productive of whitish sputum. He estimates that his physical activity level is <30 min/day. The patient has a history of hypertension. He is a retired middle school teacher. He smoked 2 packs of cigarettes daily for 20 years and quit 15 years ago. Blood pressure is 127/85 mm Hg, pulse is 78/min, respirations are 14/min, and pulse oximetry is 93% on room air. BMI is 24 kg/m². Occasional end-expiratory wheezes are present on lung examination, and the heart examination demonstrates a regular rate and rhythm. Chest x-ray shows hyperinflation, and pulmonary function test results are consistent with an obstructive defect. Based on the results of the study, which of the following is most likely to occur if the patient increases his physical activity level to >60 min/day? A)A statistically significant decrease in anxiety and depression based on HADS scores. B)A statistically significant decrease in anxiety based on HADS-A score. C)A statistically significant decrease in dyspnea based on MMRC scale score. D)A statistically significant increase in 6MWD. E)A statistically significant increase in physical functioning based on SF-36 score.

Accepted Answer

Explanation: The odds ratio (OR) is a measure of association between exposure (eg, risk factor) and outcome (eg, disease), with OR >1 meaning that exposure is associated with higher odds of outcome, OR <1 meaning that exposure is associated with lower odds of outcome, and OR = 1 meaning that exposure has no effect on odds of outcome. A 95% confidence interval (CI) that does not contain the null value (in this case, OR = 1) is often used as a proxy for statistical significance when the p-value is not provided. In this example, univariate analyses revealed that all risk factors apart from anxiety were statistically associated with high-level (ie, ≥60 min/day) physical activity (PA) given CIs excluding the null value (for anxiety, the association was nonstatistically significant given 95% CI: 0.13, 1.05). However, univariate analysis does not take into account confounders, additional independent variables that may also be associated with exposure and outcome and that may mask the real association between exposure and outcome if not adjusted. Multiple regression is one of the techniques used during the data analysis phase of a study to adjust for confounders; it takes into account the effect from additional independent variables included in the analysis and estimates "adjusted" ORs. An adjusted OR provides information about the strength of the association between exposure and outcome once all other potential confounders that might mask the real association have been controlled for in the regression model. Based on the abstract, after adjusting for all independent variables considered, high-level PA was significantly associated only with depression based on HADS-D (OR = 0.48, 95% CI: 0.24, 0.98) and dyspnea based on the MMRC scale (OR = 0.39, 95% CI: 0.17, 0.93) (Choices A, B, D, and E). This means that, if the patient increases his PA level to ≥60 min/day, he will have a statistically significant decrease (because OR <1) in dyspnea based on his MMRC score and in depression based on his HADS-D score. Educational objective: Confounding bias affects the statistical analysis of study results. Confounders may be controlled during the data analysis phase with multiple regression. Factors statistically associated with the outcome in univariate analysis may not necessarily be significant in the multivariable analysis once all other factors have been adjusted for in the model.

Question 19

A clinical trial was conducted to compare the effects of ACE inhibitors and diuretics in elderly patients with hypertension.  The investigators believed that ACE inhibitors provide some additional survival benefits over diuretics use in elderly subjects because of their neurohumoral effects.  The study lasted 5 years and the investigators presented the following results after the completion of the study: According to the information given, which of the following is the best interpretation of the study results?
A)ACE inhibitors have no overall advantage over diuretics
B)ACE inhibitors are associated with higher risk of heart failure
C)ACE inhibitors may protect from some cardiovascular events
D)ACE inhibitors decrease the risk of mortality
E)ACE inhibitors should be used with caution in elderly patients

Accepted Answer

Explanation: The scenario emphasizes the basics of study result interpretation. A hazard ratio (HR) is a measure of effect used in survival analysis (or time-to-event analysis). It has almost the same interpretation as relative risk; therefore, the study results presented can be interpreted in terms of risk. The null value for HR is 1.00: HR = 1.00 means that there is no difference in risk between the 2 groups. HR <1.00 indicates a protective effect, whereas HR >1.00 indicates a detrimental effect. The HRs for all 3 outcomes (ie, all-cause mortality, non-fatal cardiovascular events, heart failure) indicate a protective effect of ACE inhibitors; however, only the HR for non-fatal cardiovascular events is statistically significant because the 95% confidence interval does not include the null value. (Choice A) ACE inhibitors may have an overall advantage over diuretics in this age group. (Choices B and D) According to the study results, ACE inhibitors can decrease the risk of non-fatal cardiovascular events compared to diuretics in elderly patients with hypertension; although the results suggest a lower risk of all-cause mortality and heart failure, these are not statistically significant. Educational objective: A hazard ratio (HR) is a measure of effect used in survival analysis (or time-to-event analysis). The null value for HRs is 1.00: HR = 1.00 means that there is no difference in risk between the 2 groups. HR <1.00 indicates a protective effect, whereas HR >1.00 indicates a detrimental effect.

Question 20

A 65-year-old man comes to the clinic for follow-up treatment of hypertension. He is on 2 antihypertensive agents and says he is compliant with his medication regimen. The patient has no other medical problems. Blood pressure is 134/88 mm Hg and pulse is 78/min and regular. BMI is 32 kg/m². Physical examination is otherwise unremarkable. His fasting blood glucose levels on 2 recent laboratory workups were 108 mg/dL and 111 mg/dL. He is concerned about his risk of dementia because his father suffers from Alzheimer dementia. After excluding from the analysis participants with diabetes who had an unusual clinical course such as high fluctuations in reported glucose levels, the investigators repeat their calculations of the hazard ratio of dementia. This analytical step is best referred to as which of the following? A)Blinding B)Linear regression C)Matching D)Propensity scoring E)Sensitivity analysis

Accepted Answer

Explanation: Sensitivity analysis refers to repeating primary analysis calculations after modifying certain criteria or variable ranges; the goal is to determine whether such modifications significantly affect the results initially obtained. With any study, a number of decisions are made based on certain assumptions or by consensus (eg, choosing 5-year rolling windows as opposed to 2- or 8-year rolling windows). If results of sensitivity analyses conducted with slight changes in certain variables (eg, changing rolling window period, changing cutoffs for average glucose levels by ±10%) are similar to the initial results, the investigators can be more confident of the robustness of their results. In this case, patients with diabetes who had high fluctuations in glucose levels were excluded. If, on repeat analysis excluding those patients, the p-value for the association between average glucose levels and dementia remains statistically significant among the remaining patients with diabetes, this would further confirm the investigators' conclusions as it would make it less likely that high fluctuations in glucose levels explain the different hazard ratios obtained. (Choice A) Participants can be blinded so that they are unaware of which intervention they are receiving; observers/investigators can be blinded so that they are unaware of which intervention the participants are receiving. Blinding protects against bias that could favor one study arm if participants or observers/investigators were aware of the intervention. Unlike sensitivity analysis, blinding is part of initial study design rather than analysis. (Choice B) Linear regression models the linear relationship between a dependent variable and one or more independent variables. For example, multiple linear regression could be used to quantify the effects of alcohol use, tobacco smoking, and charred food consumption (independent variables) on the incidence of gastric cancer (dependent variable). (Choices C and D) Propensity scoring typically weighs different variables (eg, severity of different comorbidities) in both the treatment and the control groups to ensure that these variables are balanced between both groups. An individual in the treatment group can be matched with an individual in the control group who has a similar propensity score. Matching can also be conducted based on similar variables (eg, age, sex) even in the absence of propensity scoring. Educational objective: Sensitivity analysis refers to repeating primary analysis calculations in a study by modifying certain criteria or variable ranges to determine whether such modifications significantly affect the results initially obtained. References: A tutorial on sensitivity analyses in clinical trials: the what, why, when and how. (http://www.ncbi.nlm.nih.gov/pubmed/23855337)

Exam 2: Foundations of Independent Practice

A graduate student conducting research on the association between vitamin D3 supplementation and cancer is reviewing the abstract of a recently published journal article. Which of the following conclusions can be drawn from the results of the study?

A graduate student conducting research on the association between vitamin D₃ supplementation and cancer is reviewing the abstract of a recently published journal article. Which of the following conclusions can be drawn from the results of the study?