explaining a scientific paper

Question 1

overview of the station

Answer 1

1 - standard WIPER intro
2 - establish reason for consult. current PT understanding.
3 - pt concerns and questions
4 - let the patient lead you with their questions. they will ask you to explain lots of stuff and that is the bulk of the station.
5- essentially explaining all the jargon. explaining the methodology and results of the study, its problems and whether you can draw anything significant from it.
6 - reassure. check uunderstanding, summarise key points. check if they have any last questions. thank and finish.

Question 2

4 study types

Answer 2

1 - cross sectional - snapshot in time. often questionaires. looks for associations. lots of confounding factors and no temporal causlity.
2 - case control study - 2 groups, 1 with disease, one without and look for differences. confounding factors. will use OR misleadingly. no temporality.

3 - cohort study - 1 group with specific variable measured. follow to see who gets the disease. look for possible cause of disease. case pick up is often low which may warp OR.

4 - RCT / non-randomised interventional study. - RCT is gold standard. beware of poor interventional studies.

Question 3

what is ARR

Answer 3

absolute risk reduction

Question 4

what is RRR

Answer 4

relative risk reduction. often overstates the effect e.g. if the ARR is really low like 1% then you might only reduce the PTs risk to 0.5% but the RRR is 50% here which seems dramatic but its not really as your risk is already so small.

Question 5

what is NNT

Answer 5

number needed to treat. the number you need to treat to get one less case of disease. = 1/ARR

so if the ARR is 1% then the NNT is 100

Question 6

define sensitivity

Answer 6

proportion of people with the disease that if tested will get a positive result. if 100 people have PE then 95 get a positive test result.

Question 7

define specificity

Answer 7

the proportion of people who don’t have the disease that if tested get a negative result. 100 people without PE and 16 of them get a negative test. that’s really bad

Question 8

define PPV

Answer 8

the proportion of people with a positive test result who ACTUALLY have the disease. iff 100 people get a positive test, 21 will have the disease. poor

Question 9

define NPV

Answer 9

the proportion of people with a negative test result who actually do not have the disease. if you take 100 people with a negative result, 93 do not have the disease. so it is good for ruling out the disease here.

Question 10

some common design flaws with studies

Answer 10

the use of composite end points. whether the population used is representative. the method of follow up and how many are lost in follow up. drawing dodgy conclusions from the data. any lack of allocation concealment letting the researchers unconsciously or consciously influence who is assigned to which group.

sometimes then endpoint will be a slightly strange choice e.g. statins in heart disease. the end point might be deaths which might be reduced but we might be interested in risk of heart attacks which might not be reduced but we don’t know that as they have chosen a different end point. always scruitinise the end point.

Question 11

what does NNH stand for

Answer 11

number needed to harm. if the number needed to harm is smaller than the number needed to treat then it would be a bad idea to treat unless the harm incurred is minor. is far more people would get side effects compared to the number benefiting from the Rx.

Question 12

overview of the epidemiology station

Answer 12

1 - intro

2 - epidemiology of the study in question - either looking at RRR,ARR, NNT NNH and end points, or looking at sensitivity, specificity, PPV and NPV

3 - demographics - does the study in question apply to the demographic of the patient in front of me - ‘how does the study relate to me?’ cover age, sex, race and patient risk factors if talking about the use of an intervention

or 3 - councilling - if talking about a screening test then “do you think i should have this test?” - the risks of being investigated and worrying but not actually having the disease plus the unpleasantness of the test if it applies.
- what are my alternatives? think about managing risk factors, what they PT can do to look for warning signs, explain any relevant screening programs in current use of relevance. reassure that the can see their GP if they are ever worried.

4 - close - good rapport. ensure all their ICE are covered.

Question 13

whats a Ecological study

Answer 13

observational.

An ecological study is an observational study defined by the level at which data are analysed, namely at the population or group level, rather than individual level. Ecological studies are often used to measure prevalence and incidence of disease, particularly when disease is rare.

adv - Quick (based on routinely collected data)
Large differences in disease rates and exposures
Clues about causation for further study

disadv- Many other things vary
ECOLOGICAL FALLACY – taking data from a group and inferring causation for an individual

Question 14

key points to remember

Answer 14

• Very(likely(to(be(a(arLcle/query(about:(
– Either(a(TEST(or(
– a(TREATMENT(

• (A(staLon(about(a(TEST(tends(to(require(knowledge(on(sensiHvity/
specificity/PPV/NPV%

• A(staLon(about(a(TREATMENT(tends(to(require(knowledge(on(risk%
reducHon%

•All(staLons(will(require(you(to(idenLfy(if(the(study(is(relevant#and(of(
good(quality(or(not.(
• All(staLons(will(require(some(reassurance/(lifestyle#advice/(ICE!%

1. Know(your(definiHons%
2. IdenLfy(the(relevance(of(the(study(to(the(paLent(
   siÇng(in(front(of(you(
3. Most(scenarios(will(require(some(form(of(lifestyle%
   advice/counselling%%
4. Listen(to(the(quesHons(the(paLent(asks(and(answer(
   them(–(that’s(where(the(marks(are.(
5. Know(a(lifle(about(the(naLonal(screening(programs(
6. Know(about(the(different(types(of(studies((RCT,(caseN
   control(etc.)((

Question 15

arr v rrr

Answer 15

• RRR(tends(to(make(the(treatment(‘look(befer’(
• Especially(if(the(risk(in(a(group(is(low(to(start(
with.(
• E.g.(Risk(of(MI(in(lowNrisk(2.6%.(Taking(a(staLn(
may(reduce(the(risk(by(50%((RRR),(but(this(
would(only(be(a(1.3%(absolute(risk(reducLon(
which(doesn’t(look(very(impressive.((

Question 16

What(is(the(current(naLonal(breast(cancer(

screening(programme?(

Answer 16

• 47-73(yo,(every(3(years,(get(a(mammogram(
• If(any(concerns(–(triple(assessment:(
– ExaminaLon(
– Imaging(
– Biopsy(/(FNA(

Question 17

what is the hidden question in the station about a new test /screening programme

Answer 17

• In(all(of(these(staLons(remember(to(ICE(the(
paLent,(and(broaden(out(to(health(promoLon:(
– Are(you(parLcularly(concerned(about(having(
breast(cancer?(

– Screen(for(risk(factors:(
• Age(
• Personal(/(family(Hx(
• OCP(/(HRT(

– Promote(selfNexaminaLon(and(early(presentaLon(
to(GP(if(noLce(any(lumps.(

Question 18

A(Note(on(Screening(Programmes(

Answer 18

• There(are(3(UK(cancer(screening(programmes((BBC):(
– Breast(cancer(
– Bowel(cancer(
– Cervical(cancer(
• There(is(no(naLonal(screening(programme(for(
Prostate(cancer(in(the(UK.(
• Instead(there(is(an(“informed(choice(programme”(
which(informs(paLents(on(the(pros(and(cons(of(a(PSA(
test.(This(can(be(deliver(via(GPs(or(informaLon(
leaflets.(

Question 19

What(is(the(difference(between(relaLve(risk(reducLon(and(

absolute(risk(reducLon?(

Answer 19

Both(absolute(risk(reducLon((ARR)(and(relaLve(risk(reducLon(
(RRR)(describe(the(change(in(probability(of(an(event(occurring(
when(on(treatment(vs(risk(of(an(event(occurring(when(not(on(
treatment,(but(they(measure(the(change(in(the(risk(
differently.((

• ARR(describes(the(actual(change(in(risk.(In(this(study(being(on(
a(simavasLn(reduces(the(risk(of(a(CVD((e.g.(a(heart(afack)(by(
11%.(((it(dropped(from(a(22%(probability(when(off(treatment(
to(only(11%(when(on(treatment).(

• Whereas(RRR(describes(the(proporLonal(change(in(risk.(The(
results(in(this(study(show(that(paLents(on(simvastaLn(are(
50%(less(likely(to(have(a(CVD((e.g.(A(heart(afack),(or(in(other(
words(about(half(as(likely.((half(off(22%(is(11%).(

Question 20

2. (What(does(95%(confidence(interval(mean?(

Answer 20

• A(confidence(interval(indicates(how(sure(we(can(be(
that(the(true%value%in%a%populaHon%(in(this(case(the(
true(reducLon(in(risk)(lies(with(in(certain(range(of(
values.(

• This(study(has(used(a(95%(confidence(interval,(so(we(
can(be(sure(that(95%(of(the(Lme(the(true(relaLve(risk(
reducLon(in(a(populaLon(will(be(between(41%(and(
59%.((

• The(smaller(the(range(of(the(confidence(interval(the(
more(precise(the(value(obtained(in(the(study(is.((

Question 21

Does(this(trial(mean(taking(simvastaLn(will(stop(me(from(having(a(
heart(afack?(

Answer 21

• This(trial(shows(that(for(paLents(with(high(cholesterol,(ages(
between(65(and(80,(with(a(10(year(CVD(of(20%(taking(simvastaLn(
reduces(their(risk(of(having(a(CVD(event((e.g.(A(heart(afack(or(
stroke)(by(half.(It(will(not(stop(them(having(a(heart(afack,(but(it(will(
reduce(the(probability(of(them(having(one.(

• Remember%–%is%the%trial%valid?(
• You%need%to%then%establish%whether%this%trial%is%relevant%to%the%
paHent.%%
%
• Explain(to(paLent(this(would(involve(working(out(their(10(year(CVD(
risk(which(is(a(calculaLon(that(takes(into(account(a(paLents(age,(
gender,(cholesterol(levels,(blood(pressure(and(smoking(status.(

Question 22

what is the hidden question in the station about a new drug or procedure

Answer 22

• In(all(of(these(staLons(remember(to(ICE(the(
paLent,(and(broaden(out(to(health(promoLon:(

– Do(you(have(any(concerns(about(taking(a(staLn(

– Health(promoLon:(
• Good(diabetes(control(if(diabeLc(
• Good(diet(
• Exercise(
• Stop(smoking(
• Lose(weight(