CPH EXAM STUDYGUIDE 2023
n n n
Attributablen riskn -n ANSWER
Raten of n diseaseninnexposednindividualsnthatn cannbenatt
ributedn ton then exposure.n Orn thenproportionnof nallncasesnthatn cann ben attributedn ton an par
ticularn exposure.
Adjustedn raten -n ANSWER
Effectsn of n differencesn inncompositionnofn popsnbeingn comparedn
haven been n minimizedn byn statistical n methods.
ex:n regression n analysisn andn strandardization
-often n usedn on n ratesn orn relativen risks
Ecological n Fallacyn -n ANSWER
Biasn thatn mayn occurn becausenann associationnobservednbe
tween n variablesnorn annaggregaten level ndoesnnotnrepresentnthenassociationnthatnexistsn atnanni
ndividual n level
Confidencen Interval n -n ANSWER
95%n confidentn thatn thentruenvaluen ofn an variablen isn
containedn within n then interval.
-usedn ton accountn forn samplingn variability
-
itn isn an pointn estimaten +_n marginnof nerror,n wherenthenpointn estimaten isn thenbestn estimaten of n te
h n unknownn parametern andnthen marginnofn errorn isn thenproductn ofn thenconfidencen level n andn t
hen standardn error.
if n an 95%n CInfornthendifferencesn innmeann doesn notnincluden0n (thennulln value)n thenntheren isnevei
dencen of n an statisticallyn significantn differencen atn sigma=0.05
Clinical n Trial n Phasesn -n ANSWER 1.n Safetyn andn Pharmacologicn profiles
2. pilotn efficacyn studies
3. extensiven clinical n trials
4. aftern then FDAnapproves,n lookn atnspecificneffectsn tonestablishn incidencen ofn adversen r
eactions,n etc.n longtermn usen effects.
interpretation n of n studiesn -n ANSWER
temporality:n causen precedesneffectn
,Specificity:n importantn in n assessingn then possibilityn of n biases.
Consistency:n several n studiesn showingn similarn results.n homogeneityn statistically.
,Confoundersn -n ANSWER -non-
causal n association n between n exposuren andn outcomen asn an resultn of n an thirdn variable.
-distortion n of n effectn byn othern factors
-mustn ben relatedn ton exposuren AND n outcome
-notn an n intermediaten variablen on n causal n pathway
Controllingn forn confoundersn -n ANSWER
beforen datan collection:n randomncollection,n
individual n matching,n frequencyn matching
Aftern datan collection:n directn adjustment,n indirectn adjustment,n mantel-
haenszel,n regression n techniques
Qualityn Assurancen vs.n Qualityn Control n -n ANSWER
QA:n ensuren qualitynbeforendatan
collection
QC:n monitorn andn maintain n qualityn duringn study
reliabilityn vs.n validityn -n ANSWER
R:nprecision,nreproducibilityn
V:n accuracy,n absencen of n bias
systematicn errorn -n ANSWER
(lackn of n validity)n ifn there'sn andifferencen betweennwhatn isnact
uallyn beingn estimatednandn whatn isn intendednton benmeasured.n Increasingnsamplensizen doe
sn'tn help.
Randomn Errorn -n ANSWER (lackn of n precision)n occurs,n butn increasingn samplen sizen helps.
RCTn studiesn -n ANSWER
Testsn efficacyn orn effectivenessnof nhealthcarenservices.n randomn
allocation n ofn participantsn ton differentn treatments.n Includesnblinding,n placebo.n goldn standard
n forn evidence.
Communityn Intervention/clustern RCTn -n ANSWER community-widen basisn orn groupwide
Case-Crossovern RCTn design n -n ANSWER -casesn serven asn theirn own n control
, -exposuren hasn transientn effect
n n n
Attributablen riskn -n ANSWER
Raten of n diseaseninnexposednindividualsnthatn cannbenatt
ributedn ton then exposure.n Orn thenproportionnof nallncasesnthatn cann ben attributedn ton an par
ticularn exposure.
Adjustedn raten -n ANSWER
Effectsn of n differencesn inncompositionnofn popsnbeingn comparedn
haven been n minimizedn byn statistical n methods.
ex:n regression n analysisn andn strandardization
-often n usedn on n ratesn orn relativen risks
Ecological n Fallacyn -n ANSWER
Biasn thatn mayn occurn becausenann associationnobservednbe
tween n variablesnorn annaggregaten level ndoesnnotnrepresentnthenassociationnthatnexistsn atnanni
ndividual n level
Confidencen Interval n -n ANSWER
95%n confidentn thatn thentruenvaluen ofn an variablen isn
containedn within n then interval.
-usedn ton accountn forn samplingn variability
-
itn isn an pointn estimaten +_n marginnof nerror,n wherenthenpointn estimaten isn thenbestn estimaten of n te
h n unknownn parametern andnthen marginnofn errorn isn thenproductn ofn thenconfidencen level n andn t
hen standardn error.
if n an 95%n CInfornthendifferencesn innmeann doesn notnincluden0n (thennulln value)n thenntheren isnevei
dencen of n an statisticallyn significantn differencen atn sigma=0.05
Clinical n Trial n Phasesn -n ANSWER 1.n Safetyn andn Pharmacologicn profiles
2. pilotn efficacyn studies
3. extensiven clinical n trials
4. aftern then FDAnapproves,n lookn atnspecificneffectsn tonestablishn incidencen ofn adversen r
eactions,n etc.n longtermn usen effects.
interpretation n of n studiesn -n ANSWER
temporality:n causen precedesneffectn
,Specificity:n importantn in n assessingn then possibilityn of n biases.
Consistency:n several n studiesn showingn similarn results.n homogeneityn statistically.
,Confoundersn -n ANSWER -non-
causal n association n between n exposuren andn outcomen asn an resultn of n an thirdn variable.
-distortion n of n effectn byn othern factors
-mustn ben relatedn ton exposuren AND n outcome
-notn an n intermediaten variablen on n causal n pathway
Controllingn forn confoundersn -n ANSWER
beforen datan collection:n randomncollection,n
individual n matching,n frequencyn matching
Aftern datan collection:n directn adjustment,n indirectn adjustment,n mantel-
haenszel,n regression n techniques
Qualityn Assurancen vs.n Qualityn Control n -n ANSWER
QA:n ensuren qualitynbeforendatan
collection
QC:n monitorn andn maintain n qualityn duringn study
reliabilityn vs.n validityn -n ANSWER
R:nprecision,nreproducibilityn
V:n accuracy,n absencen of n bias
systematicn errorn -n ANSWER
(lackn of n validity)n ifn there'sn andifferencen betweennwhatn isnact
uallyn beingn estimatednandn whatn isn intendednton benmeasured.n Increasingnsamplensizen doe
sn'tn help.
Randomn Errorn -n ANSWER (lackn of n precision)n occurs,n butn increasingn samplen sizen helps.
RCTn studiesn -n ANSWER
Testsn efficacyn orn effectivenessnof nhealthcarenservices.n randomn
allocation n ofn participantsn ton differentn treatments.n Includesnblinding,n placebo.n goldn standard
n forn evidence.
Communityn Intervention/clustern RCTn -n ANSWER community-widen basisn orn groupwide
Case-Crossovern RCTn design n -n ANSWER -casesn serven asn theirn own n control
, -exposuren hasn transientn effect
