Transcript Seminar

Course: Research in Biomedicine and Health III
Seminar 2: Asking a question

EBM steps
◦ Step 1: Formulating questions that can be
answered
◦ Step 2: Finding best evidence
◦ Step 3: Quick critical assessment of the
evidence
◦ Step 4: Applying evidence
◦ Step 5: Assessing effectiveness and
efficiency of the process
How to ask a clinical question?
“PICOT” model:
Patient Who is treated?
(For which patient, population or problem you need information?)
Intervention/indicator
What is given as a treatment?
(What is the intervention you are studying?)
Comparison/control
What is the comparison?
(Which is the alternative treatment?)
Outcome
What outcome is of interest to us?
(What is the effect of the intervention?)
Type of study
What is the study design?
(Which is the optimal study design ?)
Question type
Ideal study design
Therapy
RCT
Prevention
RCT > Cohort> Case-control
Diagnosis
Prospective, blind, randomized trial of
comparison with golden standard
Prognosis
Cohort> Case-control> Case series/ Case
report
Cost and efficiency
Economic analysis
Phenomenon
Qualitative study
Be precise!
P
Patient,
problem
I
C
O
intervention Comparison Outcome
(not always)
Kew words for
database search
T
Type of
study
Aim of steps 1. i 2. : find “pearls”
Examples of clinical questions
• What is causing migraine?
• How often a 40 year-old women needs to go to
mammography?
• What is the efficacy and safety of early and long-term
use of clopidogrel and aspirin compared to aspirin
alone in patients with acute coronary syndrome
without ST-segment increase?
• In patients with mechanical ventilation longer than 48
hours, is oral decontamination with chlorchexidine
(CHX) or CHX + colistine useful in decreasing the risk of
pneumonia related to ventilation?
Examples of clinical questions
How to treat a disease or condition?
Intervention
What is causing the problem?
Ethiology and risk factors
How common is the problem?
Prevalence
Has this patient a problem?
Diagnosis
Who will have a problem?
Prognosis and prediction
What is the problem?
Phenomenon

Interventions cover a wide spectrum of activities from medications and other clinical therapeutics to
change in life habits (exercise, smoking cessation)
and social activities (education programmes).

Interventions may include patient care for a single
patient but also public health activities.
28 year-old man suffers from furunculosis in the last 8 months. These episodes are
treated by drainage and 2 different atibiotics, but the disease relapses. The
patient asks if the relapses can be prevented.
Patient with recurring furunculosis
Prophylactic antibiotics
No treatment
Decrease in relapse of furunculosis
Q: In patients with recurring furunculosis, can prophylactic antibiotic
therapy in comparison to no treatment decrease the incidence of
recurring episodes?

Reading a medical journal, you find an article about imiquimode cream for the
treament of basal cell carcinoma. The notion of a cream as cancer therapy
sounds surprising, so you wonder about its efficacy, especially in regard to longterm cure rate.
Patient with basal cell carcinoma
Imiquimode cream
Active placebo – standard/gold therapy
Increase in long-term cure rate with the
cream
Q: In patients with basal cell carcinoma, can im iquimode cream , in
comparison to standad surgical intervention, increase the long-term
cure rate in these patients?



Questions about ethiology and risk factors
are questions about what causes a disease or
health condition.
Harmful outcomes of activities or exposures
Common in public health:
◦ Can intake of certain food increase risk of
cardiovascular disease?
◦ Can exposure to infrared waves increase risk of
cancer?
Jane is expecting her first baby in two months. She read about the benefits and
harms of injections of vitamin K to newborns. She is scared by the reports of
increased risk of childhood leukemia in babies administered vit. K injections.
She asks you if this is true and what the risk is for her baby.
Newborn
Vitamin K injection
No vitamin K injection
Childhood leukemia
Q: In newborns, does the administration of vit. K injection increase the risk
of childhood leukemia?


Questions about how many people have a
diseases or health condition.
Examples:
◦ What is the prevalence of hearing problems in
newborns?
◦ What is the prevalence of Alzheimer’s disease in the
population older than 70?
Maria is a 6-month baby coming to her regular check up. She was born
prematurely, at the gestational age of 35 months. You want to warn the
parents about the odds that Maria may have hearing problems.
Infant
Premature birth
Term birth
Deafness
Q: In preterm infants, in comparison to term babies, what is the prevalence
of deafness?

Questions about:
◦ How accurate are diagnostic tests in different patient groups.
◦ How accurate are diagnostic tests in comparison to other
tests.
Clara is pregnant for the second time. She had her first baby when she was 30 and had
amniocentesis to exclude Down’s syndrome. The test was negative, but she was not
satisfied because she got results in the 18th pregnancy week. She is now 35 and asks if
there is a test with a quicker answer. Your hospital has biochemical test of serum markers
and ultrasound of nuchal folds in the first trimester, but you want the check if the test is
sensitive and specific enough in comparison to amniocentesis.
Pregnant women
Serum markers biochemistry +
US of nuchal folds
Amniocentesis
Correct diagnosis (specificity and
sensitivity) of Down’s syndrome
Q: In pregnant women, is nuchal fold US + serum markers test equaly
correct as amniocentesis in diagnosing Down’s syndrome?

What are the odds for an outcome for a population
with certain characteristics (risk factors)?
◦ What are the odds that a man who feels atypical chest pain
develops heart failure?
◦ What is predicted mortality and morbidity for a person who is
diagnosed with colon cancer?
Clonic seizures in childhood are common and disturbing for parents, but the
decision to start prophylactic therapy after the first attack is difficult. To
help parents make a decision, you have to explain them the risk of new
attacks after the first clonic seizure of undetermined cause.
Children
A clonic attack of undetermined cause
No attack
Future attacks
Q: In children with a single clonic attack of undetermined origin, in
comparison to children without attacks, what is the risk of future clonic
seizures?
• For each type of clinical question there are study
designs that provide best evidence for the question.
• These are always study designs that have a
maximum decrease in bias .
Question type
Ideal study design
Therapy
RCT
Prevention
RCT > Cohort> Case-control
Diagnosis
Prospective, blind, randomized trial of
comparison with golden standard
Prognosis
Cohort> Case-control> Case series/ Case
report
Cost and efficiency
Economic analysis
Phenomenon
Qualitative study
Design
Observational
No control
Experimental
Randomized
control
Cohort
Casecontrol
Non-randomized
control
Increasing validity of evidence for
conclusions on cause and consequence
Open
Single
blind
Double
blind
Examples of disbalance between surogate and relevant outcomes:
Disease or condition
Surogate outcome
Relevant outcome
Lidocaine for arrythmia
after AMI
Decreases arrythmia
Increases mortality
Vitamin E for heart
disease
Decreases the level of
free radicals
Do not decrease
mortality
Clofibrate for
hyperlipidemia
Decreases blood lipid
levels
Do not decrease
mortality
Sodium floride for bone
fracture prevention
Increases bone densitiy
Do not decrease fracture
rate
Hormone replacement
therapy
Decreases LDL,
increasese HDL
Do not decrease
mortality, can increase
incidence of
cardiovascular events in
women over 60