trills Identified no difference m smoking cessation,
the proportion ol participants teducmgsnokmg
tonsunpkon, mean reduction in cigarettessmoked
per day, or harms, between e-ogaret1es and
traditional nicobne replacement therapy.Most
studies were judged to have a high risk of bias,
resulting in the overall quality of evidence aslow.
More research is necessary prior to estabbshng
recommendationsrelated to e-cigarettes assmoking
cessation tools.
Source:BasicConcepts in Prevention. Surveillance, arid Health Promotion.AFMC Primer on Population Health. Iittp:/Vphprimei.armc.ca/Pait1-
ThcoryThinkingAboutHealtli Chapter4BasicConceptslnPreventionSurveillanceAndHealthPromotion/Thestagesofprevention
Screening (Secondary Prevention)
•
"
screening is a strategy used in a population to identify the possible presence of an as-yet-undiagnosed
disease in individuals without signs or symptoms"
• screening vs. case finding:screening tests are not diagnostic tests
the primary purpose of screening tests is to detect early disease or risk factors for disease in
large numbers of apparently healthy individuals.The purpose of a diagnostic test is to establish
the presence (or absence) of disease as a basis for treatment decisions in symptomatic or screen
positive individuals (confirmatory test). Both screening and case finding seek to risk stratify for
further investigation
to minimize biases and harms, and maximize benefits,screening is best done at the population
level, not the individual clinical level, as part of a screening program (e.g. provincial breast cancer
screening program vs.screening by primary care/family physicians)
• types of screening
universal screening: screening all members of a population for a disease (e.g. phenylketonuria
(PKU ) and hypothyroidism in all newborns)
• selective screening: screening of targeted subgroups of the population at risk for a disease (e.g.
mammography in women >50 yr)
• multiphasic screening: the use of many measurements and investigations to look for many
disease entities (e.g. periodic health exam)
• types of bias in screening
• lead-time bias: overestimation of survival time ‘from diagnosis' when the estimate is made from
the time of screening, instead of the later time when the disease would have been diagnosed
without screening
length-time bias: overestimation of the survival time due to screening at one time point
including more stable cases than aggressive cases of disease, which may have shortersurvival
times
r -i
+
PHI1 Public Health and Preventive Medicine Toronto Notes 2023
Overt
Disease
Occult
Disease Onset of Disease Death from Disease
^^
4—-
Lesd Time—
A Snapshot of the Opioid Crisis in
Canada
Canada is experiencing a crisis of opioidrelated overdose and death.Between
January 2016 and September 2019.
there were more than14700 deaths in
Canada related to opioids.There were
also19490 hospitalizations and17000
emergency services.Individuals 25-34
y/o are at the greatestrisk of overdose
death (1in 6 deaths),but rates have
increased for all adult ages.Deaths are
most commonly unintentional.Heroin,
fentanyl. and hydromorphone are most
commonly invoked.The highest rates of
opioid-related overdose and death are
found in British Columbia.An estimated
300 per million British Columbians died
in relation to opioid use in 2017.More
died from opioids than homicide,motor
vehicle accidents,and suicide combined.
In 2017,deaths from opioids in Ontario
were
-1250,while deaths from motor
vehicle accidents vrere "
450.Fentanyl
or a fentanyl analogue were involved in
more than 70% of cases,increased from
55%in 2016.
Sources:JAddict Med.Measuring the Burden olQpwidreloted Mortality inOntario. Canada. Latest Trendsin
Opioid-Related Deathsin Ontario:1991to 2015.Toronto:
Ontario Orug Policy ResearchKetvnxk.Health Canada.
March 2018.Opioid-relatedharms inCanada.Health
Canada.March 2020
Screen Detected Clinically Detected
Figure 2. Lead-time bias
Table 4. Ideal Criteria for Screening Tests
Disease Test Health Care System
Causessignificant suffering and/or death
Natural history must be understood
Must have an asymptomatic stage that can be Acceptable to providers and the population
Continuously utilized
High sensitivity
Safe,rapid, easy, relatively inexpensive
Adequate capacity for reporting, lollow-up,
and treatment of positive screens
Cost effective
Sustainable program
Clear policy guidelines on who to treat
detected by a test
Early detection and intervention must result in
improved outcomes
Adapted from:Shah CP.Public Health andPreventive Medicine in Canada,5th ed.Toronto:Elsevier,2003
Health Promotion Strategies
Table 5. Disease Prevention vs. Health Promotion Approach
Disease Prevention Health Promotion
Health - absence of disease
Medical model (passive role)
Aimed mainly at high -risk groups in the population
One-shotstrategy aimed at a specific pathology
Health ~ positive and multidimensional concept
Participatory model of health
Aimed at the population in its total environment
Diverse and complementary strategies aimed at a network at issues/
determinants
Facilitating and enabling approaches by incentives offered to the
population
Focused on a person's health status and environment
Led by non-professional organizations, civic groups,local, municipal,
regional, and national governments
Directive and persuasive strategies enforced in target groups
Focused mostly on individuals
Led by professional groupsfrom health disciplines
See landmark Public Health and Preventive Medune
trialstable for more information on the Swedish TwoCounty trial, which detailsthe long-term effect ol
mammugraphic screening on breast canter mortality.
Source: Shah CP. Public Health and Preventive Medicine in Canada,5th ed. Toronto:Elsevier.2003
Healthy Public Policy
• purpose: to create a supportive environment to enable people to lead healthy lives, thereby making
healthy choices easier for citizens
•governments and non-governmental agencies need to consider the cost and acceptability of proposed
public health interventions (e.g.. more invasive or costly measures should be justified by the extent of
beneficial impacts on people'
slives)
•the Nuffield Intervention Ladder provides one way of ranking the level of intrusion and hence a need
for proportionate benefit of health promotion interventions at a population level
•methods
• fiscal: imposing additional costs (e.g. taxes on tobacco and alcohol)
• legislative:implementing legal deterrents (e.g.smoking bans, legal alcohol drinking age)
• social: improving health beyond providing universally funded health care (e.g. providing
affordable housing)
Source:International Conference on Health Promotion, Adelaide,South Australia (1998)
Transtheoretical Model Stages ol Change lor
Dietary and PhyskaI Exercise Modification ia
Mright Loss Management for Overweight and
Obese Adults
Cochrane DB Syst Dev 20I 4 CD008066
Purpose: loexplore the efficacy of dietary
and physical activity inlerventmns based n the
transtheoretical model otchange forsustained
weight loss alter one yr in overweight ov obese
adults.
Methods: RCTs comparing the use ol weight loss
or physical activity intervention grounded m the
transtheoretical model otchange to usual care for
weight lossin adults whowere overweight or obese
were eligible lor inclusion.Interventions had to he
carried out by healthcare professionals ov trained lay
people.Weight loss or change m 8MI was required as
an outcome measure.
Results: llneestudesincluding a total ol
29)1 participants were included in thrsreview.
Interventions grounded kr Ihismodeldd have positive
effects on physical activity and dietary habitsthat
included increased exercise duration and lieguency,
reduced lat intake, and increased fru t a - d vegetao e
consumption,theevidence for sustained weight loss
at oneyr was inconclusive (mean difference in favour
ol the transtheoretical model was between 2.1 kg and
0.2 kg at 24 mo).
Behaviour Change
•behaviour is a result of three factors
1. predisposing factors: knowledge, attitude, beliefs, values, intentions
2. enabling factors:skills,supports
3. reinforcing factors: health care professionals and the social context of family and community
•health education serves to:increase knowledge and skills and promote healthy behaviours
Health Belief Model (1975)
•a psychological model that explains and predicts individual short- and long-term health behaviours
based on one'
s beliefs and attitudes
•based on the assumption that one will adopt a beneficial health behaviour if the following three beliefs
are present:
• the negative health outcome is avoidable
expects that the health outcome can be prevented if the recommended health behaviour is
adopted
• the individual can he successful in adopting the health behaviour
•six concepts:
• four concepts influencing one'
s
"readiness to act"
- perceived susceptibility, perceived severity,
perceived benefits, perceived barriers
cues to action:stimuli that can trigger health action
• self-efficacy: confidence in one’s ability to take a health action
c
+
PH12 Public Health and Preventive Medicine Toronto Notes 2023
Stages of Change Model
• provides a framework in which the Health Belief Model is applied to facilitating behaviour change (e.g.
quitting smoking) Principles of Standardization
• When comparing a health measure
(e.g. mortality) between two
populations (or the same population
at different time points) that
differ in characteristics known to
influence that outcome (e.g.age),
standardization is used to control for
the effect of thatfactor
• Standardization is either direct or
indirect
• Indirect standardization is expressed
asstandardized outcome ratio.For
example.Standardized Mortality
Ratio (SMR) is calculated using
age specific ratesfor a reference
population, as well as age structure
and total casesfor a sample/
known population,(e.g.an SMR
of 100 signifies that deaths are at
the expected level, a SMR of 110
indicates a death rate10% higher
than expected)
• Direct standardization is expressed
as a rate (i.e. using age specific
rates in a known/sample population
against a standard population)
1.Precontemplation:me individual is not seriously considering change p
(for various reasons) and is not interested in any kind of intervention
,2.Contemplation:the individual begins to seriously consider making
|be change within the foreseeable future (often defined assix
^
Lithsl
'»
\
Relapse:
\i possible
*
at any
'
* /
stage
iration:the individual begins experimenting, making
tanqes. he or she resolves to make a serious attempt
in tb^
fere ( usually defined as 30 days)
sml f-.
'
r
4. Actio*
the chani
individual is actively involved in making
king different techniques
v
> /
5. Maintena
success's
*
to the previi
fee individual must learn to
^
felMemptations to return
behaviour pattern
Figure 3.Stages of change model
Source:Prochaska JO.DiClenerte CC.and NorcrossJC.In search of how people change.Applicationsto Addictive Behaviours.Am Psychol
1992:47:1102-1114
Risk Reduction Strategies
•risk reduction:lower the risk to health without eliminating it (e.g. avoiding sun to lower risk of skin
cancer)
•harm reduction: a set of strategies aimed to reduce the negative consequences of drug use and other
risky behaviours (e.g. needle exchange programs)
Source:Shah,CP.Concepts,Determinants,and Promotion of Health.Public Health and Preventive Medicine in
Canada,5e.Toronto:Elsevier,2003
Community Needs Assessment
•a community needs assessmentstudies a community’
s health gaps and pairs identification of that
community’
s existing resources and strengths to find solutions to addressthose gaps.This assessment
strongly valuesinterviewing community members to gather their concerns and proposed solutions.
Steps include:
1. define the community and understand its history and demographic characteristicsto
formulate context forsubsequent data collection
2. understand what mattersto com munity stakeholders(e.g.interviews,surveys,focus groups)
3. use evidence (e.g.mortality rate,feasibility), prioritize each concern
4. identify barriers that may prevent a concern from being addressed and propose solutions
using community-based resources
Measurements of Health and Disease in a
Population
MEASURES OF DISEASE OCCURRENCE
Rates, Ratios, and Proportions
• a rate measuresthe frequency of an event in a defined population over a specific period of time (e.g.
number of opioid overdosesin Canada in one year)
• a ratio compares the magnitude of one quantity to another (e.g. ratio of women to men with lupus)
• a proportion is a ratio where the numerator is a part of the denominator (e.g.proportion of deliveries
complicated by placental abruption)
Incidence Rate
• number of new cases in a population over a specific period of time
Prevalence
• total number of cases in a population over a defined period of time
• two forms of prevalence
point prevalence: assessed at one point in time
• period prevalence:assessed over a period of time, therefore including new cases and excluding
cases that terminate (cure or death)
• a function of the incidence rate and disease duration from onset to termination
• favours the inclusion of chronic over acute cases and may underestimate disease burden if those with
short disease duration are missed
• prevalence estimates are useful for measuring disease burden and therefore help in the planning of
facilities and services
+
PH13 Public Health and Preventive Medicine Toronto Notes 2023
Age-Standardized Rate
• adjustment of the crude rate of a health-related event using a “standard” population
• standard population is one with a known number of persons in each age and sex group
• standardization prevents bias that can occur when crude ratesfrom two dissimilar populations are
compared (e.g. crude death rates over a number of decades are not comparable as the population age
distribution has changed with time)
• this allows for the calculation of a Standardized Mortality Ratio (SMR), where SMR (observed
number of dcaths)/(expected number of deaths)
MEASURES OF MORTALITY
Life Expectancy
• the expected number of years to be lived by a newborn based on age-specific mortality rates at a
selected time
Crude Death Rate
• mortality from all causes of death per 1000 in the population
Infant Mortality Rate (IMR)
• number of reported deaths among children <1 yr of age during a given time period divided by the
number of reported live births during the same time period and expressed as per 1000 live births per
year
Maternal Mortality Rate (MMR)
• “number of deaths of women during pregnancy and due to puerperal causes|...|per 1000 live births
in the same year"
MEASURES OF DISEASE BURDEN
Potential Years of Life Lost (PYLL)
• calculated for a population using the difference between the actual age at death and a standard/
expected age at death
• increased weighting of mortality at a younger age
Disability Adjusted Life Year (DALY)
• number of years lost due to premature mortality + number of years lost due to disability, where 0 = a
year of perfect health and 1 = death
• both premature death and time spent with disability accounted for; these disabilities can be physical
or mental
• used to assess burden of diseasesin a population
Top 10 Causes of DALYs in Canada,
2019
1. Neoplasms
2. Cardiovascular diseases
3. Musculoskeletal disorders
4. Neurological disorders
5. Mental disorders
6. Other non-communicable diseases
7. Unintentional injuries
8. Chronic respiratory diseases
9. Diabetes and kidney diseases
10.Substance use disorders
Source:QcCi.lBurden ot Oiseese CompareIVic Hub
[MnnelJ. Seattle|WA|:tlrnelsitf ot WeUangton.
hnbtule1«Health Metliuand tahaliou(lHMtl;2021
Idled 2021Mar 20|. Available boot MtpWMduib.
hrelltidala.oijj/glMj
-compare/
Quality Adjusted Life Year (QALY)
• years of life weighted by quality (utility is a proxy for quality), ranging from 0 (= death) to 1 (= perfect
health). Weights are assigned based on large studies that assessed the effect of various conditions on
quality of life (e.g. blindness = 0.3)
• it is possible to have “states worse than death” (e.g. QALY <0 for extremely serious conditions)
• usually used as an economic measure to assess the value for money of medical interventions
For additional rate calculations see Steps to Control an Outbreak, PH24
Consult the Public Health Agency of Canada for examples and latest statistics
Government of Canada:Chaplet 3: The chief public health officer'
s report on the state of public health in Canada 2008•our health [lnlernct|.Our
population,our health,and the distributions of our health;[updated 2008 Jun 6|. Available from: https://www.canada.ca/enrpublic- heallh/corporate/
publicalions/chref public heallli offker reporls state public heallli canada /report on slate public health Canada 2008ichapler-3b.html
Sources:Shah.Cf Health indicators and data sources. Public Health and Preventive Medicine in Canada.Se. Toronto: Elsevier. 2003
The Association ol Faculties of Medicine of Canada Public Health Educators' Hettvork. AFMC primer on population health(Internet). Methods:measuring
health:[cited 2006 Mar 25].Available from https://phprimer.afmc.ca/en/
Epidemiology
Population
• a defined collection of Individuals/regions/lnstitutions/ctc. (e.g. individuals defined by geographic
region,sex, age)
Sample
• a selection of individuals from a population
• types
LJ
SPIN:use a Specific test to rule IN a
hypothesis.Note that specific tests have
very few false positives.If you get a
positive test, it is likely a true positive
SNOUT: use a SENsitive test to rule OUT
a hypothesis. Note that sensitive tests
have very few false negatives.If you get
a negative test, it is likely a true negative
random: all members are equally likely to be selected
systematic:an algorithm is used to select a subset
stratified: population is divided into subgroups that are each sampled
cluster:grouped in space/time to reduce costs
convenience: non-random inclusion, for populations that are difficult to reach (e.g. people with
precariousliving conditions)
+
PHH Public Health and Preventive Medicine Toronto Notes 2023
Sample Size
• increasing the sample size increasesthe statistical precision of the observed estimate,resulting in
more narrow confidence intervals
• increasing the sample size decreases the probability of type I and type 11 errors
• increasing sample size does not alter the risk of bias/confounding
Bias
• systematic error leading to an incorrect estimate of the true association between exposure and
outcome
• can occur atseveral points in study execution (e.g. collection, analysis, interpretation, publication, or
review of data)
• selection bias: a systematic error in the recruitment or retention of study participants
• Berkson’s bias occurs in a case-control study using hospitalized controls, as they may not be a
representative sample of the population due to the complexity that led to their hospital admission
• non-response bias occurs when participants differ from non-participants in a study, in that those
who volunteer may be healthier
• loss to follow-up bias occurs when dropout rates differ between study groups and patients who
dropped out are different from those who did not
information bias: the way in which information is collected about study participants is inadequate
• recall bias occurs when individuals with disease may be more likely to incorrectly recall/ believe they
were exposed to a possible risk factor than those who are free of disease
• interviewer bias occurs when interviewers are unblinded to outcome status and this knowledge
biases their behaviour
• observer bias occurs when knowledge of exposure status (e.g. race, gender) biases the observer
towards a diagnosis; this occurs more commonly with subjective diagnoses like those found in
psychiatry
Confounder
• a variable that is related to both the exposure and outcome but is not a mediator in the exposureoutcome relationship
• distorts the estimated effect of an exposure if not accounted for in the study design/analysis (e.g.
late maternal age could be a confounder in an investigation of birth order >4 and risk of developing
Trisomy 21)
• randomization, stratification, matching, and regression modelling can help minimize confounding
effects
The Association of Faculties ol Medicine ol Canada Public Health Educators' Network. AFMC primer on population health [Internet],
Assessing evidence and information. Available from https://phprimer.afmc.ca/en/part-ii/chapter-5l
Figure 5. Understanding sensitivity
and specificity
Source:Loony TW. Understanding sensitivity and
specificity with the right side ol the brain.8MJ
2003:327:716-719
O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
O O O O O O O O O O
Interpreting Test Results O — well person
•
- person with disease
TP -True positive TN - True negative FP -False positive FN -False negative Figure 5a. Hypothetical population
Disousc
O O O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
Prosont Nogative
O O
Tost Result Posilivo TP FP O O
Negative FN TN O O
Sensitivity O O =TP/ITP-FN)
Specificity -TN/ITN+FPI O O
O O
Likelihood Ratio (LH|
• Likelihood that a given test result would be expected in a patient with disease compared with the likelihood
that the same result would be expected in a patient without disease
• LR.indicates how much tho probability of diseaso increases if Iho tost is positivo
• LR indicates how much tho probability of disoasa decreasesif the tost is nogotivo Dork grey positive test resrit
Light grey negative test result
LR. Sensitivity ITP/TP.FNII
i spQcmcuy IFP/ITN+ FPI1
Sensitivity |FN/(TP+ FN )|
specificity (TTWmTFFTT
LR
Figure 5b. Results of diagnostic test
on hypothetical population
Positive Predictive Value (PPV)
• Proportion of people with a positive test who have the disease
PPV = TP
TP + FP
Negative Predictive Value ( NPV)
• Proportion of pooplo with a nogativo tost who oro froo of disoaso
NPV TN
Figure 5c. Sensitivity of test
(e.g. 24/30 = 80% sensitive)
TN A FN
0 239 LR+ -3.85 Advanced Neoplasia 0.938
r -t O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
O O O O O O O O
o o
Present Negative L J
LR- JJLSE
0.938 -0.81 o o
Test Result Positive 68 147 O O
Negative 216 2234 PPV 68 O O = - 31.6% (68 + 1471 O O
Total 284 2381
O O
Sonsitivrty 68/284 - 23 9% NPV 2234 O O +
Spacrficity • 2234/2381 . 93.8%
Figure 4. Interpreting test results:practical example using FOBT testing in advanced colon cancer
Source*:Collins J.lieberman D. Durbin T. et al.Accuracy ol screening for fecal occult blood on a single stool sample obtained by digital rectal
examination:a comparison with recommended sampling practice.Ann Intern Med 2005:142:81-85
91.2%
(2234 216)
Figure 5d. Specificity of test
(e.g. 56/70 80% specific)
PH15 Public Health and Preventive Medicine Toronto Notes 2023
Sensitivity
• proportion of people with disease who have a positive test $
• Sensitivity and specificity are
characteristics of the test
• LR depends on the test
characteristics, not the prevalence
• PPV and NPV depend on the
prevalence of the disease in the
population
Specificity
• proportion of people without disease who have a negative test
Pre-Test Probability
• probability that a particular patient has a given disease before test/assessment results are known
Post-Test Probability
• a revision of the probability of disease after a patient has been interviewed/examined/tested
• post
-test odds = pre-test odds x l.K i , or, pre-test odds x LR-; for a positive test or negative test,
respectively (recall odds » probability/!1- probability))
• the post-test probability from clinical examination is the basis of consideration when ordering
diagnostic tests or imaging studies
• after each iteration, the resultant post-test probability becomes the pre-test probability when
considering new investigations
PRE • TEST
PROBABILITY
POST - TEST
PROBABILITY
0 001 -i 1- 0 939
. U 998
- 0 997
. U 999
- 0 993
- 0 39
0 002
-
0003 -
0 005 -
0 007 -
001 -
LIKELIHOOD
RATIO
Effectiveness of Interventions 0 98
- 0 97
- 0.95
- 0 93
0 02 -
003 1000
500 - -
200 - -
100 - -
Effectiveness, Efficacy, Efficiency
• three measurements indicating the relative value (beneficial effects vs. harmful effects) of an
intervention
efficacy:the extent to which a specific intervention produces a beneficial result under ideal
conditions(e.g. RCT)
ideal conditions include adherence, close monitoring, access to health resources,etc.
effectiveness: measures the benefit of an intervention under usual conditions of clinical care
considers both the efficacy of an intervention and its actual impact on the real world,
taking into account access to the intervention, whether it is offered to those who can benefit
from it,its proper administration, acceptance of intervention, and degree of adherence to
intervention
efficiency:a measure of economy of an intervention with known effectiveness
considers the optimal use of resources (e.g. money, time, personnel, equipment)
Disease (e.g. lung cancer)
0 05 -
0 07
-
0.1 - 0.9
S 02 ' - 0.8
\V]
PSA 0.3 - 0.7 - ?E
*
V i \
- -
0.4 - 0.6
05 - 0.5
b 5 * - PSA
°
v\
J
o.i
v -
x
0.05 ,
- x
06 - 0.4
0.7
- - 0.3
0.8 - - 0.2
0.02 - 0.01 ‘ '
- -\ \
.. % PSA\A
+ \
3 M
0.9 - 0.1
- 0.07
- 0.05
0.93 -
0.95 -
0.005
0.003 - - \
0.97 0.001 - 0.03
- 0.02 \
Present Absent Total 0.98
- \
V
099 - 0 01
- 0 007
- 0 005
Present A B A + B
PSA
=0.5 ”
^
0 993 -
0 995 -
Exposure
(eg. smoking) Absent C D C + D
0 997 -
0 998 -
- 0 003
- 0 002 Total A + C B + D A + B + C + D
J
993 J
Figure 7. Fagan's likelihood ratio
nomogram:practical example using
PSA levels to calculate post-test
probability of prostate cancer
Modified from source:Holmstrom
B,Johansson M,Bergh A,et
al.Prostate specific antigen
for early detection of prostate
cancer:longitudinal study. BMJ
2009;339:b3537
LOOOIJ Case-Control Study
odds ratio (OR)
'= *
-jj— =>
Cohort Study
—— - incidence rate of health outcome in exposed —£— - incidence rate of health outcome in non-exposed
ArB C + D
AxO
B
"
xC
-
A
- +
_
L
_
AiB C + D
'Ratio ol tho odds in favour of Itio hoalth outcome amongIlia oxposodto tho oddsIn favour among the unoxposod
"Ratio of tho risk of a hoalth outcomo among oxposodto tho risk among tho unoxposnd
'"Rato of hoalth outcomo inoxposod individuals that contra attributed to tho exposure
Figure 6. Measures of effect by study type
Number Needed to Treat (NNT)
• number of patients who need to be treated to achieve one additional favourable outcome
• only one of many factors that should be taken into account in clinical or health system decision
making (e.g. must take into account cost, ease, feasibility of intervention)
a condition with death as a potential outcome can have a higher NNT (and be acceptable), as
compared to an intervention to prevent an outcome with low morbidity, in which a low NNT
would be necessary
A C relative risk -
(RRT
attributable risk.
(AR)*
"
A iB C + D
Equations to Assess Effectiveness
CER - control group event rate
EER - experimental group event rate
ARR • absolute risk reduction
RR • relative risk
NNT x
number needed to treat
RR -
EERCER
ARR = CER -EER
NNT x VARR
r1
L J
Number Needed to Harm (NNH)
• number of patients who, if they received the experimental treatment, would lead to one additional
patient being harmed, compared with patients who received the control treatment
Adherence (formerly compliance)
• degree to which a patient’s behaviour and lifestyle concords with the recommendations of healthcare
providers (e.g. the extent to which a patient takes medications as directed)
+
NNT
Consult http://www.thennt.com for quick
summaries of evidence-based medicine
(includes NNT. LR,and risk assessments)
PI116 Public Health and Preventive Medicine Toronto Notes 2023
Coverage
• extent to which the services rendered cover the potential need for these services in a community
Sources:Shah.CP.Health Indicators and data sources.Public Health and Preventive Medicine in Canada,5e.Toronto:Elsevier.2003
The Association olFaculties of Medicine olCanada Public HealthEducators' Network.Assessing Evidence and Information.AFMC Primer onPopulation
Health
Beware
Do not be swayed by a large RR or
odds ratio,as it may appear to be large
if event rate is small to begin with.In
these cases AR is more important (e.g.
a drug which lowers an event which
occurs in 0.1% of a population to 0.05%
can boast a RR of 50%,and yet the AR
is only 0.05%.which is not nearly as
impressive)
Types of Study Design
Qualitative vs. Quantitative
Table 6. Qualitative vs.Quantitative Study Designs
Qualitative Quantitative Formulating a Research Question
Often used to generate hypotheses (Why? What does it mean?)
"Bottom-up" approach
Observation •pattem tentative hypothesis • theory
Sampling approach to obtain representative coverage of ideas,
concepts,or experiences
Narrative:rich, contextual,and detailed information from a small
number of participants
Often tests hypotheses (What? How much/many?)
“Top-down" approach
Theory »hypothesis
*
observation -»confirmation
Sampling approach to obtain representative coverage of people in the
population
Numeric:frequency, severity,and associationsfrom a large number of
participants
PICO
PopulatioruPatient characteristics
Intervention/exposure of interest
Comparison group or control group
Outcome that you are trying to prevent
or achieve
Source: Adapted trom httptrphprimer.almc.cn
Souice: The Association ot Faculties of Medicine of Canada Public HealthEducators' Network. Assessing evidence andInformation. AFMC Primer
on Population Health
Quantitative Research Methods
Were exposures assigned by the investigator?
I
£ i
(
~
Yes No
Experimental Study (
I
ObservationalStudy
T
Random allocation to groups? Testing a hypothesis?
I I
T T
[ Yes Yes No [ No
\ f 2 '
Randomized Analytical Study
Controlled
Non- Descriptive
Randomized Study
Trial Designs
Sampling based on
1
Exposure Outcome Neither
1 l .
Cross-Sectional
Study
Cohort Case-Control
Study Study
Figure 8.Quantitative study designs
Source:adapted from The Association of Faculties of Medicine of CanadaPubic HealthEducators’Network.AFMC Primer onPopulationHealth
[Internet].About the primer onpopulation health.Available from https://phprimer.afmc.ca/en/
Observational Study Designs
•observational studiesinvolve neither the manipulation of the exposure of interest nor randomization
of the study participants
•there are two main subtypes of observational studies: descriptive and analytic studies
Descriptive Studies
•describe the events and rates of disease with respect to person, place, and time; estimates disease
frequency and time trends
•includes case reports, on one person or event, or a case series, which assesses exposures and outcomes
•can be used to generate an etiologic hypothesis and for policy'planning
An ecological fallacy is an erroneous
conclusion made when extrapolating
population level data to txplaln
phenomena occurring inindividuals.An
example of an ecological fallacy would
be concluding that red wine drinking
leads to lower risk of death (tom CVS
disease based on an ecological study
showing that countries with a higher rate
of red wine consumption have a lower
rate of death from CVS causes
r i
L J
+
PH17 Public Health and PreventiveMedicine Toronto Notes 2023
Analytic Studies
• observational studies used to test a specific hypothesis
• includes ecological studies,cohort studies, case-control studies, and cross-sectional studies
Table 7. Observational Study Designs
Type of Study Ecological Cross-Sectional Case-Control Cohort
Definition Units of analysis are
populations or groups
of people,rather than
individuals
Use individual data on Samples a group of people
exposures and outcomes who already have a
gathered allhe same time particular outcome (cases)
and compares them to
a similar sample group
without that outcome
(controls)
Two or more samples
of individuals with and
without the outcome|s)
of interest (i.e.cases and
controls)
Subjects are sampled and,
as a group,classified on (he
basis of presence or absence
of exposure to a particular
risk factor
Subjects Aggregated groups (e.g. Sample of a population
cities)
One or more cohorts
Cohort:group of people with
common characteristics
(e.g.year of birth,region of
residence)
Divided into measured
exposed vs.unexposed
groups
Collect information on factors
from all persons at the
beginning of the study
Subjects are foilowedfor
a specific period of time to
determine development of
disease in each exposure
group
Prospective:measuring from
the exposure at present lo the
future outcomes
Retrospective:measuring
forward in time from
exposures in the past to later
outcomes
Use statistical models to
test associations between
exposures and disease or
other measured outcomes
Provides estimates of
incidence,relative risk,
attributable risk
Shows an association
between risk factoi(s) and
outcome(s)
Stronger evidence for
causation
Can consider a variety of
exposures and outcomes
Descriptions of the average Collect information
exposure or risk of disease from each person at one
particular time
Can use regression models Tabulate the numbers in
lo lest associations
between area-level
Select sample olcases of
a specific disease during a
specific timeframe
Representative of
groups (e.g.by presence or spectrum of clinical
absence of disease/factor disease
Methods
lor a population
predictors and aggregate of interest)
outcomes
Select control(s)
Make tables and compare Represent the general
population
To minimice risk of bias.
groups
Estimate prevalence
Use regression models to may select more lhan one
test associations between control groupand/or match
predictors and outcomes controls to cases (e.g.age.
gender)
Assess past exposures
of interest
(e.g.EMR, questionnaire)
Association canbe
concluded between the
risk factor and the disease
(odds ratio)
Ourck.easy to do
Uses readily available data between variables
Generates hypothesis Quick and uses lewer
resources
Surveys with validated
questions allows
comparison between
studies
Determinesassociation Often used when disease
in population is rare (less
than 10% of population)
due toincreased efficiency
or when lime to develop
disease is long
Less cosily and lime
consuming
Recall bias
(see Bias, Wf14)
relationship or offer strong Confounding
evidence for causation Selection bias lor cases
and controls
Only one outcome can be
measured
Advantages
Confounding may occur due
lo individuals self-selecting
the exposure,or unknown/
unmeasured factors are
associated with the measured
exposure and outcome
Cost and duration of time
needed lo follow cohort
Selection bias
A famous cohort study is the
Framingham Heart Study,
which assessed the long-term
cardiovascular risks of diet,
exercise,and medications
such as ASA, etc.
Disadvantages Poor generalicabilily lo Does not allow lor
Individual level (not direct assessment of temporal
assessment ol causal
relationship)
Ecological fallacy: an
Incorrect Inference from
groups to individuals
Confounding
between variables
Confounding
Selection bias
Recall bias
(see Bias.W14)
A study lira!examines Ihe A famous case control
distribution of SMI by age study published by Sir
in Ontario at a particular Richard Doll demonstrated
the link between tobacco
smoking exposure and
lung cancer cases at the
individual level
Examples A study looking allhe
association between
smoking rales and lung
cancer rates in dillerenl poinl in time
countries at the population
level without individual
data on both factors
Sources:Shah,CP.Measurement and investigation. Public Health and Preventive Medicine In Camida,5e. Toronto:Elsevier, 2003.
The Association of Faculties ol Medicine at Canada Public Health Educators'Network. AFMC primer on population health (Internet). Assessing
evidence and information.Available from https://phpfirner.alinc.ca/enrpart
-iirchopter
-5/
RothmanKJ.Greenland SG.lash Tl. Modem epidemiology. 3e.Philadelphia:Wolters Kluwer. 2012.
+
PH18 Public Health and Preventive Medicine Toronto Notes 2023
Experimental Study Designs Study begins by sampling
~
subjects based on outcome |
Disease
(cases)
•not discussed here are non-randomized controlled trials(e.g. allocation by clinic or other non-random
basis- performed when randomization is not possible)
No Disease
(controls)
:
RANDOMIZED CONTROLLED TRIAL (RCT) 1
i
i
Definition I
•participants are assigned by random allocation to two or more groups, one of which is the control
group and the other group(s) receive(s) an intervention
Participants
•individuals are selected using explicit inclusion/exclusion criteria and recruitment targets are guided
by sample size calculations
Methods
•random allocation of individuals into two or more treatment groups through a centralized concealed
process
•method of assessment to reduce bias
single-blind: participant does not know group assignment (intervention or placebo)
double-blind: participant and observer both unaware of group assignment
triple-blind:participant,observer, and analyst unaware of group assignment
•control group receives standard of care or placebo if no standard of care exists
•one or more group(s) receive(s) the intervention(s) understudy
•baseline covariate(s) and outcome(s) are measured and the groups are compared
•all other conditions are kept the same between groups
Advantages
•“gold standard" ofstudies, upon which the practice of LBM isfounded
•provides the strongest evidence for effectiveness of intervention
•threats to validity are minimized with sufficient sample size and appropriate randomization
•randomization is one of few methods that can eliminate confounding (including unmeasured
confounders) and self-selection bias
•allows prospective assessment of the effects of intervention
Disadvantages
•sonic exposures are not amenable to randomization (e.g. cannot randomize participants to poverty/
wealth or to harmful exposuressuch as smoking) due to ethical or feasibility concerns
•can be difficult to randomly allocate groups (e.g. communities, neighbourhoods)
•difficult to study rare eventssince RCTs require extremely large sample sizes
•contamination, co-intervention,and loss to follow-up can all limit causal inferences
•can have poor generalizability (e.g.when trial participants are healthier than the average patient
population)
•costly
Shah. CP.Measurement andinvestigation.Public Health andPreventive Medicine inCanada. Se.Toronto:Elsevier. 2003.
the Association ol Faculties olMedicine olCanada Public Health Educators' Network. AFMC primer on population health(Internet]. Assessing evidence and
inhumation. Available from hltps:/lphpiimei.afmc.ca/cnlpait-iilchapter-S/
i
5
A A e
Exposed Unexposed Exposed Unexposed
Classify Exposure
Figure 9.Case-control study
Adapted Irom http://phptlrner.almc.ca
Study begins
Unexposed groupf
J
Exposed group
J
:
1
A A :
Disease No disease Disease No disease
^
Figure10.Cohort study
Adopted from http://phprliner.iilmc.ca
Analysis
Per-Protocol Analysis (PP)
Strategy of analysisin which only
patients who complete the entire study
are counted towardsthe results
Intention-to-Treat Analysis (ITT)
When groups are analyzed exactly as
they existed upon randomization (l.e.
using data from all patients. Including
those who did not complete the study)
Summary Study Designs
ft
META-ANALYSIS
An example of an RCT isthe SPARCL
trial,which demonstrated intense
lipid-lowering with atorvastatin
reducesthe risk of cerebrovascular
and cardiovascular events in patients
with and without carotid stenosis when
compared to placebo
Definition
• a form ofstatistical analysis that aggregates all relevantstudies addressing the same research question
in order to increase statistical precision
Participants
• all the studies identified through a systematic literature review
Methods
• selection of relevantstudiesfrom the published literature which meet quality criteria
• statistical models used to combine the results of each independentstudy
• provides a summary statistic of overall results as well as graphic representation of included studies
(forest plot)
Advantages
• attempts to overcome the problem of reduced power due to small sample sizes of individual studies
• can address questions (e.g.subgroup analyses) that the original studies were not powered to answer
Disadvantages
• studies may be heterogeneous and therefore inappropriate to combine (e.g. different patient
populations, exposure classification/measurement, outcome assessment)
An example of a meta analysis is one
that comparesthe effects of ACEIs,
calcium channel blockers, and other
antihypertensive agents on mortality
and major cardiovascular events by
compiling and analyzing data from a full
set of reported RCTs
n
LJ
+
PH19 Public Health and Preventive Medicine Toronto Notes 2023
• reliance on published studies may increase the potential conclusion of an effect as it can be difficult to
publish studies that show no significant results (publication bias)
Shah. CP. Measurement andinvestigation. Public Health andPrevenlivo Medicine inCanada. Se. Toronto:Elsevier. 2003.
the Association ol Faculties olMedicine olCanada Public HealthEducators' Network. AFMC primer on population healthllnternet]. Assessing evidence and
information. Available from https://phprinier.afmc.ca/crCpart ii/chaplet SI
Methods of Analysis
Distributions
• a distribution describes the frequency at which each value (or category) occurs in a study population
• distributions can take characteristic shapes (e.g. normal (Gaussian) or non-normal (binomial,
gamma, etc.))
• characteristics of the normal distribution
mean = median = mode
68% of observationsfall within one standard deviation of the mean
95% of observations fall within two standard deviations of the mean
• measures of central tendency
mean:sum of each observation’s data (e.g.ages) divided by total number of observations
» median: value of the 50th percentile;a better reflection of the central tendency for a skewed
distribution
• mode: most frequently observed value in a series
• measures of dispersion
• range: the largest value minus the smallest value
variance: a measure of the spread of data
• standard deviation: the average distance of data points from the mean (the positive square root of
variance)
• given the mean and standard deviation of a normal distribution curve, a description of the entire
distribution of data is obtained
Consult the Cochrane Library of
Systematic Reviews (http://www.
cochranelibrarycom) for high-quality
systematic reviews and meta-analyses
Example Calculation
Data set:17,14,17.10, 7
Mean -
(17 +14 + 17 + 10 + 7)
+ 5-13
Median (write the list in order,median is
the number in the middle)
- 7,10, 14,17,17 -
14
Mode (number repeated most often)- 17
Range - 17- 7- 10
Variance - [(17- 13)2 + (14 -13)2
(17 - 13)2 + (10 -13)2
(7 - 13)21 * 5 - 19.5
Standard Deviation -
-/variance - V19.5
-
4.42
Data Analysis
s> lean-Median-Mode
Statistical Hypotheses
• null (Ho)
the default hypothesis;often statesthere is no relationship between two variables
• alternative (HI)
the hypothesis that we are interested in;often states there is a relationship between two variables
we can find evidence against Ho but we can never‘prove’ Hi
Type I Error (a Error)
• the null hypothesis is falsely rejected (Le.concluding an intervention X is effective when it is not,or
declaring an observed difference to be real rather than by chance)
• the probability of this error is denoted by the p-value
• studies tend to be designed to minimize thistype of error since a type I error can have larger clinical
significance than a type II error
• e.g. in a study exploring a drug’
s effectiveness on lowering blood pressure, the data may indicate the
drug is effective and therefore lowers blood pressure, when in reality the drug is ineffective
Type II Error (0 Error)
• the null hypothesis is falsely accepted (i.e.stating intervention X is not effective when it is, or
declaring an observed difference/effect to have occurred by chance when it is present)
• by convention a higher level of error is often accepted for most studies
• can also be used to calculate statistical power
• e.g. in a study exploring the effectiveness of a COVID-19 vaccine, the data suggests the vaccine is
ineffective and therefore does not protect against CGVJD-19 infection, when in reality it does
Power
• probability of correctly rejecting a null hypothesis when it is, in fact, false (i.e. the probability of
finding a specified difference to be statistically significant at a given p-value)
• power increases with an increase in sample size
• power = 1 - p, and istherefore equal to the probability of a true positive result
Statistical Significance
• the probability that the statistical association found between variablesis due to random chance alone
(i.e. there is no association)
• the preset probability issetsufficiently low that one would act on the result;frequently p<0.05
• when statistical tests result in a probability less than the preset limit, the results are said to be
statistically significant (denoted by the a-value)
-
Normal Distribution
Negatively Skewed
Figure 11. Distribution curves
Type
*
I (a) Error
“There Is An Effect" where in reality
there is none
n
L J
+
PH20 Public Health and Preventive Medicine Toronto Notes 2023
Clinical Significance
• measure of clinical usefulness (e.g. 1 mmHg BP reduction may be statistically significant, but may not
be clinically significant)
• depends on factorssuch as cost, availability,patient adherence, and side effects in addition to
statistical significance
Confidence Interval (Cl)
• provides a range of values within which the true population result (e.g.the mean) lies, bounded by the
upper and lower confidence limits
• frequently reported as 95% Cl (e.g. if thisstudy were repeated 100 times, estimates would fall within
the 95% Cl 95 out of 100 times)
A wider confidence interval implies
more variance than a tighter confidence
interval given the same critical value
Good reliability
Good validity
Poor reliability £
Data Good validity <5
• there are 2 types of quantitative data
• continuous data (e.g. height in cm)
discrete data (e.g. number of patients in the 1CU)
• information collected from a sample of a population
• there are 4 overall levels of measurement for quantitative data
• categorical (e.g. blood type,marital status)
ordinal (e.g. low, medium, high)
interval (e.g.°C, time of day)
ratio (e.g.serum cholesterol, hemoglobin, age)
Validity/Accuracy (of a measurement tool)
• how closely a measurement reflects the entity it claims to measure
Reliability/Precision
how consistent multiple measurements are when the underlying subject of measurement has not
changed
• may be assessed by different observers at the same time (inter-rater reliability) or by the same
observer under different conditions(test-retest reliability)
Internal Validity
• degree to which the findings of the sample truly represent the findings in the study population
• dependent on the reliability',accuracy, and absence of other biases
External Validity (i.e. Generalizability)
• degree to which the results of the study can be generalized to other situations or populations
Good reliability
Poor validity
Poor reliability
Poor validity
Figure 12. Validity vs. reliability
What's the difference between Pearson
and Spearman correlation?
Different types of correlation are used
for different levels of measurement.
Rrarson is for continuous and Normal
data,Spearman is for ordinal or
non-Normal data.There are other
forms of correlation for other levels
of measurement (e.g.tetrachoric/
polychoric)
Common Statistical Tests
Table 8. Statistical Tests
Two Sample Z-Test Analysis of
Variance (ANOVA)
Chi
-Squared Test (y2) Linear Regression Logistic Regression PearsonProduct
-Moment
Correlation (Pearson's r)
What are you trying to show?
Compare the mean Compare the mean Tests if two categorical variables Model relationship
values of an outcome values of an outcome are independent or not (e.g.
variable between two variablebetween two association between family history variable and one or
groups (eg.difference or more groups (e.g. of breast cancer and having breast more explanatory
in average BP between differencein average cancer)
men and women) BP between persons in
three towns)
Model relationship Assesses the strength of the linear
between a continuous between a categorical relationship between two variables.
variable and one or more Ranges from -1(perfect negative
explanatory variables association,increases in one variable
are associated with decreases
inanother) to1(perfect positive
association,increases in one variable
are associated with increases in the
other). A correlation of 0 indicates no
relationship
variables
What kind of variables do youmeasure?
Continuous Categorical(2 or morel
'
ordinal Continuous Categorical (outcomes Continuous
usually dichotomous)
ContinuousfOrdinalf Continuous
Categorical
Dependent
Variable
Continuous
CategoricalOrdinal(2 CategoricalOrdinal (2 or more)
or more)
Continuous/Ordinal/
Categorical
Independent
Variable
Assumptions
Dichotomous
Data follow a normal/t- Normal distribution of Expected counts rust be at least 5 Dependent variable's linearity (on logit scale) Underlying relationship is linear
error term has normal No influential values Data for both variables are normally
Model has adequate distributed
goodness-of-fit
Data are independent
distribution dependent variable's for aflcells innbyn table r m
L J Equal variances
Data are independent Data are independent
Data are independent distribution
linear relationship
between variables
Homoscedasticity
No influentialvalues
Data are independent
error term
Data are independent
+
PH21 Public Health and Preventive Medicine Toronto Notes 2023
Causation
Criteria for Causation (Bradford Hill Criteria)
1. strength of association: the frequency with which the factor is found in the disease and the
frequency with which it occurs in the absence of disease
2. consistency: is the same relationship seen with different populations orstudy design?
3. specificity: is the association particular to your intervention and measured outcome?
4. temporal relationship: did the exposure occur before the onset of the disease?
5. biological gradient:finding a dose-response relationship between the exposure-outcome
6. biological plausibility: does the association/causation make biological sense?
7. coherence: can the relationship be explained/accounted for based on what we know about
science,logic, etc.?
8. experimental evidence:does experimental evidence support the association (e.g. is there
improvement?)
9. analogy:do other established associations provide a model for thistype of the relationship?
Note:Not all criteria must be fulfilled to establish scientific causation, and the modern practice of EBM
emphasizes‘experimental evidence’ assuperior to other criteria for experimental causation review.
However,many causation questions in health cannot be answered with expe
Source:Bradford Hill A.the environment and disease:association or causation.(hoc It Sue Med 1965;58(5):295-3
rimental methods
0 0
Assessing Evidence
• critical appraisal is the process ofsystematically examining research evidence to assess validity,
results, and relevance before using it to inform a decision
FILTERED
INFORMATION
(Evidence Syntheses)
itically-Appraised Individual \
Articles (Article Synopses)
\
Randomized Controlled Trials
(RCTs)
UNFILTERED Cohort Studies INFORMATION
Case-Controlled Studies
Case Series / Reports
Background Information / Expert Opinion
Figure13. Pyramid of pre-appraised evidence
: Copyright 2006.Trustees ul Dartmouth CollegeS Yale University.All rightsreserved. Produced by Glover J..IHO D.,Odalo K..and Wang L
A. Arc the results of the study valid?
•see below for classifications of evidence that has already been assessed
B. What arc the results?
•what was the impact of the treatment effect?
•how precise was the estimate of treatment effect?
•what were the confidence intervals and power of the study?
C. Will the results help me in caring for my patients?
•are the results clinically significant?
•can l apply the results to my patient population?
•were all clinically important outcomes considered?
•are the likely treatment benefits worth the potential harm and costs?
n
i j
+
PI122 Public Health ami Preventive Medicine Toronto Notes 2023
Levels of Evidence: Classifications Cited in Guidelines/Consensus Statements
Level I evidence: based on RCTs (or meta-analysis of RCTs) big enough to have low risk of incorporating FP or FN results
Level II evidence: based on RCTs too small to provide Level ! evidence:may show positive trends that are non-significant or have a high
risk of FN results
Level III evidence: based on non-randomired. controlled or cohort studies:case series: case-controlled:or cross-sectional studies
Level IV evidence: based on opinion of respected authorities or expert committees, as published consensus conferences/guidelines
Level V evidence: opinions of the individuals who have written/reviewed the guidelines (i.e. Level IV evidence!, based on expeuence/-
knowledge of literature/peer discussion
Notes:These 5 levels of evidence are not direct evaluations of evidence quality or credibility: they refle
RCTstend to be most credible (with <1111.level III evidence gains credibility when multiple studies from different locations and/or time periods
report consistent findings.Level IV and V evidence reflects decision-making that is necessary but in the absence of published evidence.
Figure 14. Levels of evidence classifications
Note: This is only one method of classifying evidence.Various systems exist,but operate within the same premise that certain types of evidence carry
more weight than others
ct the nature of the evidence. While
Health System Planning and Quality
Continuous Quality Improvement
Quality Improvement (Ql)
• a means of evaluating and improving processes;focusing more on systems and systematic biases,
which are thought to cause variation in quality
• measures to increase efficiency of action with the purpose of achieving optimal quality
Quality Assurance
• process to guarantee the quality of health care through improvement and attainment of set standards
•
“five-stage process of quality assurance”
Source;Shah. CP.Public Healthand Preventive Medicine in Canada, 5e.Toronto:Elsevier. 2003.
1.formulation of working goals
2.procedural changes to implement those goals
3.regular comparison of current performance with original goals
4.development ofsolutions to bring performance closer to goals
5.documentation of quality assurance activities
Quality Control
• a process of surveying the quality of all factors involved in the process to maintain standards
Continuous Quality Improvement
• the process of ongoing service/product refinement via the vigilant review of expectant issues
detrimental to the system and regular incorporation of improvements
Quality Management
• combination of several processes (assurance, control, improvement) to maintain consistent quality
Total Quality Management
• management principle for advancing quality while minimizing additional expenditures
• focuses on the entire system rather than discrete elements
Audit
• methodical analysis of a quality system by quality auditors
• to determine whether quality processes and results comply with goals and whether processes have
been implemented effectively
Systems Analysis Tools
1. 5 Whys: brainstorming to simplify the process of change; continue asking 'why'
until the root of the
problem is discovered
2. Ishikawa Diagrams (i.e.fishbone Diagrams):identify generic categories of problems that have an
overall contribution to the effect
3. Defect Check Sheets; consider all defects and tally up the number of times the defect occurs
4. Pareto Chart: x vs. y chart; x-axis = defect categories, v-axis = frequency; plot cumulative frequency
on the right y-axis; purpose is to highlight most important among large set of factors contributing to
defects/poor quality
r n
L J
+
PH23 Public Health and Preventive Medicine Toronto Notes 2023
Decreased Handwashing j (Improper Precautiunsj
Lack of time Lack of training
Lack of hand sanitizer Improperly assigned
- •Nosocomial Infections 7 Common waiting /
rooms
Not sterilizing stethoscope
between patients
Not changing uniforms
between shifts ,
More thanI— 7
patient per room7
^
Healthcare Team j
^
Patient Exposures j
Figure 15.Ishikawa diagram
Precede-Proceed Model
• tool for designing, implementing, and evaluating health interventions/programs
Table 9. Precede-Proceed Model
PRECEDE Phase PROCEED Phase
Phase1-Identify the ultimate desired result
Phase 2 -Identity health issues and then behavioural and
environmental determinants. Set priorities among them
Phase 3 -Identify the predisposing, enabling, and reinforcing factors
that ailed the behaviours and environmental delciminanls
Phase 4 -Identify the administrative and policy factorslhal influence
what can be implemented
Phase 5-Implementation (design and conduct the intervention)
Phase 6 - Process evaluation (determine II the piogiam Is implemented
as planned)
Phase 1 -Impact evaluation (measure intermediate effects on the
target population )
Phase 8- Outcome evaluation (determine whether the original desired
result was achieved)
Planning Cycles/Models
1 . APIE Planning Model: Assessment, Planning, Implementation, Evaluation
2. PDSA Planning Cycle: Plan, Do, Study, Act
Economic Evaluation
Cost Benefit Analysis (CBA)
• an analysis which compares the total expected costs with the total expected benefits of actions in
order to choose the most profitable or beneficial option(s)
• costs are controlled for inflation and market changesso that the effect of the change is evaluated over
a consistent, preset financial value
Cost Effectiveness Analysis (CEA)
• ratio of change in cost (numerator) to change in effect (denominator) in response to a new strategy or
practice
the numerator highlights the cost of the health gain
• some examples of changes in effect (denominator) could be years of life gained or sight-years
gained
• the most commonly used outcome measure is quality-adjusted life years (QALY) (see Quality
Adjusted Life Year, Pill 3 )
• can be used where an extensive cost benefit analysis is not applicable or appropriate
Cost Utility Analysis (CUA)
• special case of CEA where effectiveness is measured in utility, commonly in quality-adjusted life years
(QALY)
• Note: term issometimes used interchangeably with CEA
+
PH21 Public Health and Preventive Medicine Toronto Notes 2023
Managing Disease Outbreaks
COVID-19 precautions
Precautionsinclude hand hygiene,
gown, eye protection, and wellfitting masls(e.g.surgical mask).
N95 respirators are reserved for
aerosol-generating procedures,
such as endotracheal intubation and
bronchoscopy
For specific examples,see
“Communicable Diseases" section in:
Shah CP. Public health and preventive
medicine in Canada.5th ed. Toronto:
Elsevier:2003
Source:Pubic Health Ontario:
hltpsu'
iWww.pubikheallhontaiio.u
'
-i
'medidi'docunert
^
1
No comments:
Post a Comment
اكتب تعليق حول الموضوع