Clinical Epidemioogy lecture series 1-1.pdf

CLINICAL EPIDEMIOLOGY -1
Dr. Asiimwe (MPH), Public Health Department, KIU; +256750319914
1

CLINICAL EPIDEMIOLOGY
 Course Outline
1) Intro to Clinical Epidemiology
2) Epidemiological Studies & Study Designs/Approaches
3) Descriptive Clinical Epidemiology
4) Analytical Clinical Epidemiology
5) Applied Clinical Epidemiology
7) Epidemiology of Infectious/communicable diseases
8) Epidemiology of Non-Communicable Diseases (NCDs)
9) Genetic Epidemiology
2

Definition of Epidemiology & key terms
 Epidemiology stems from;
 Derived from Epidemos (Greek); “Epi” = Upon, “Demos” = People.
 Relates to study (logy), prevention & control of epidemics
 Epidemiology: is the study of the;
 Frequency,
 Distribution and
 Determinants of diseases (dx) and other health related conditions
(HRC)
 in a specified human populations,
 and the application of this study to the promotion of health, and to the
prevention and control of health problems/outcomes (dx/HRC).
5

Clinical Epidemiology
 Clinical Epidemiology – Is the study of the;
 Patterns,
 Determinants (Causes/Risk factors/associated factors),
 and Effects of disease & health related conditions (HRCs) in patient
populations,
 and the relations/associations between exposure or treatment and
health outcomes.
OR: Clinical Epidemiology is the application of principles & methods
of epidemiology to clinical setting i.e. to conduct, appraise/evaluate,
and apply clinical research/studies for purposes of prevention,
diagnosis , prognosis and treatment of diseases in patients.
AIM of Clinical Epidemiology: To inform clinical decision –
making & improve Patient’s care outcome
6

Distinguishing Between Classical Epidemiology &
Clinical Epidemiology
 Classical Epidemiology focus on frequency, distribution &
determinant of disease/HRCs at population level WHEREAS;
 Clinical Epidemiology is on frequency, distribution &
determinant of disease/HRCs at Patients level in a clinical
care setting.
Note: Clinical Epidemiology is a sub-field of Epidemiology that
focus on issues relevant to clinical medicine (a.k.a; Basic science of
clinical medicine).
First introduced By Jean Paul (1938); in his presidential address
to American society of clinical investigations.
7

Classical Epidemiology Vs Clinical Epidemiology:
plus Objectives of Epidemiology
8

Perspectives of Epidemiology
 Epidemiology Asks & Seek to Answer Five (5) “W” Important Questions;
 What? (Disease/ HRC),
 Who? (is Affected - Person),
 Where? (is the disease - Place),
 When? (does it occur - Time) &
 Why? (Determinants of disease & HRC; Host, Agent & Environment factors)
 These pertinent questions spark/trigger an epidemiological study /
epidemiological investigation & Disease Surveillance guided by a set
hypothesis/ theories about disease occurrence & causations/risk factors.
 Answering these questions provided valuable information & Guides on;
 Natural history /Symptomatology & Prognosis of disease
 Distribution of the disease & HRC
 Determinants of the disease & HRC (Causative/Risk factors; H-A-E Factors)
 How? (Disease Mechanisms/Mode/Process –Pathogenesis & Pathophysiology)
 Cases Management & Prognosis, Disease prevention & Health Promotion.
9

Purpose/Application of study- Epidemiology
 Epidemiology is Applied Science/ public health → Applied Epidemiology
 Applied in;
 Determining of frequency/ Magnitude/burden/extent of disease
 Establishing& describing natural history & prognosis of disease
 Establishing determinants (causes & risk factors) of health & disease
 Provides evidence based effective health interventions
 Prevention & Control of disease and health conditions
 Health promotion,
 Risk communication,
 Diagnosis and Case management
 Used in evaluation of effectiveness of public health & clinical interventions
 Guide;
 Health policy formulation, guidelines & SOPs
 Health planning & Decision making/prioritization of actions in health services
 Further Clinical studies/research & development.
10

Measure of Morbidity /Disease & Mortality / Death
 PART A: MEASURE OF MORBIDITY [Determine Frequency of
Disease]
 Aim: Determine frequency of disease
 Measures of Disease Frequency:
1) Prevalence - frequency of existing cases (Prevalent cases)
2) Incidence - frequency of new cases (Incident cases)
 PART B: MEASURE OF MORTALITY
 Aim: Establish Burden & impact of disease
 Tools: 1) Mortality Rate
2) Mortality Ratio
11

PART A: MEASURE OF MORBIDITY
 Prevalence (Proportion of existing cases of disease expressed in %)
 Point prevalence
 Period Prevalence
 Life Time Prevalence (LTP)
 Prevalence Odds
 Prevalence Ratio (PR)
 Incidence (Freq. new cases of disease expressed in Person-time i.e. Rate)
 Cumulative Incidence (Incidence Risk)
 Incidence Rate (Incidence Density)
 Attack Rate (incidence proportion)
 Incidence Odds
 Life time Risk (LTR)
 Relationship Between Prevalence (P) & Incidence (I): P = I x D
12

14
 Prevalence - is proportion of diseased cases in a specified population at
given point in time or during a specified time period.
 Prevalence represents the case load or assess burden of disease in an area.
 Measures of Prevalence:
 Point Prevalence =
(𝑁𝑜.𝑜𝑓 𝑑𝑖𝑠𝑒𝑎𝑠𝑒𝑑 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝑎 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑝𝑜𝑖𝑛𝑡 𝑖𝑛 𝑡𝑖𝑚𝑒)
(𝑇𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑡𝑖𝑜𝑛 𝑎𝑡 𝑡ℎ𝑎𝑡 𝑝𝑜𝑖𝑛𝑡 𝑖𝑛 𝑡𝑖𝑚𝑒)
𝑋100
 Period Prevalence =
(𝑁𝑜.𝑜𝑓 𝑑𝑖𝑠𝑒𝑎𝑠𝑒𝑑 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝑎 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑑𝑢𝑟𝑖𝑛𝑔 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
(𝑇𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑡𝑖𝑜𝑛 𝑑𝑢𝑟𝑖𝑛𝑔 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
𝑋100
 Prevalence Odds:
=
(𝑁𝑜.𝑜𝑓 𝑑𝑖𝑠𝑒𝑎𝑠𝑒𝑑 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝑎 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑝𝑜𝑖𝑛𝑡 𝑖𝑛 𝑡𝑖𝑚𝑒)
(𝑁𝑜.𝑜𝑓 𝑁𝑜𝑛− 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝑎 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑡ℎ𝑎𝑡 𝑝𝑜𝑖𝑛𝑡 𝑖𝑛 𝑡𝑖𝑚𝑒)
NB: Lifetime Prevalence (LTP) is the proportion of individuals in a population
that at some point in their life (up-to point of assessment) have ever experienced
a health event, risk factor or disease.

PART A: MEASURE OF MORBIDITY; PREVALENCE
Case1: There was 52 cases of Hepatitis B viral infection (HBV) in
Pregnancy at the beginning on a 3rd quarter FY 2023/24. If 80 new
cases of HBV were reported in the 3rd quarter and Health Facility ANC
attendance during this time composed of 525 pregnant mothers,
determine the prevalence of HBV in pregnancy.
Case 2: In 2022, about 15, 190 adults aged 30 – 69 years visited a
HTN clinic. If 4,268 adults had Isolated Systolic Hypertension (ISH),
determine the prevalence of ISH.
Case 3: Out of 400 patients who visited a clinic, 100 had a common cold
at the beginning of Nov.2021. The number of cases of common cold
raised to 130 by 20th Nov. 2021. a) What was the prevalence of
common cold at the beginning of Nov. 2021; b) Calculate the prevalence
of common cold in the a period of 1st – 20th Nov. 2021. c) Calculate the
Prevalence odds of common cold at the beginning of Nov. 2021.
15

 Incidence- number of new cases of a disease in a specific population
at risk during a given time period.
 Incidence is proxy for Risk (prob. of occurrence of disease in a
disease free population during a specified time period)
 Measures of Incidence:
 Cumulative Incidence (Incidence Risk) =
(𝑁𝑜.𝑜𝑓 𝑁𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 𝑑𝑢𝑟𝑖𝑛𝑔 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
(𝑁𝑜.𝑜𝑓 𝑝𝑒𝑜𝑝𝑙𝑒 𝑎𝑡 𝑟𝑖𝑠𝑘 𝑎𝑡 𝑡ℎ𝑒 𝑠𝑎𝑟𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
𝑋100
NB: Incidence Risk relates to popn. at risk at the beginning of time period.
 Incidence Rate ( Incidence Density)
=
(𝑁𝑜.𝑜𝑓 𝑁𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 𝑑𝑢𝑟𝑖𝑛𝑔 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
(𝑇𝑜𝑡𝑎𝑙 𝑃𝑒𝑟𝑠𝑜𝑛 −𝑡𝑖𝑚𝑒 𝑎𝑡 𝑅𝑖𝑠𝑘 )
𝑋100
Where : Person – time = No. of persons at risk during a time period.
 Incidence Odds:
=
(𝑁𝑜.𝑜𝑓 𝑁𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 𝑎𝑡 𝑎 𝑔𝑖𝑣𝑒𝑛 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
(𝑁𝑜.𝑜𝑓 𝑁𝑜𝑛− 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝑎 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑒𝑛𝑑 𝑜𝑓 𝑡𝑖𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑)
16

PART A: MEASURE OF MORBIDITY; INCIDENCE
Case 4: A study X enrolled 4340 sexually active HIV free Ugandans at the
beginning of 1999 and screened them for HIV after exposure at the end of the
year. Results of the study were as indicated in Tab. 01.
Determine the;
a) risk of acquiring HIV in Uganda (1999).
b) Incidence odds of acquiring HIV in Uganda using the case above.
Case 5: B1: During the first phase of COVID-19 in 2020, Bushenyi district
reported 26 new active cases of COVID-19 and a cumulative number of 264
cases of COVID-19 1st July – 31st Dec.2020 . If the mid-year district population
was 183,000, What was the; a) Number of existing cases of COVID-19, b) Risk
of contracting COVID-19; b) period Prevalence of COVID-19.
HIV +ve HIV –ve Total
Total 90 4250 4340
17

PART A: MEASURE OF MORBIDITY; Case Scenario
Case 6: Conrad et al studied liver disease among adults aged 40 years
and above for a period of 10 years from 1994 to 2003 in an urban
area in Michigan USA. The total population of adults aged 40 years and
above was 10,000 in the area. He enrolled 800 participants of whom
121 developed liver disease during the period of study. Before selecting
the participants, there were no live person/patient with liver disease in
the area.
a) What was the;
(I) point prevalence of liver disease at the start of study?
(II) Cumulative incidence of liver disease during the study period?
b) Analysis of his data showed that two thirds of the cases of liver
disease got the disease within the period 1994 -2000; Determine the
cumulative incidence of liver disease from 1994 to 2000.
18

PART A: MEASURE OF MORBIDITY;
Person – Time of Observation/ Study
 Suppose in a clinical epid. Study, a population at risk (N) is exposed
(E), and at any time (𝑡) the number of persons who developed the
disease [New diseased cases] is D. Then the No. of non-cases at the
end of the time period = [N – D] and Person – Time at risk (𝑃𝑇 =
𝑛 × 𝑡, where n = no. of person at risk at any time in years, months,
weeks, days, etc.).
 A Graph/plot of person –time Vs time
y = 𝑃𝑇
N
D = cases
n = [N – D] = non-cases
𝑡
19

Incidence Risk Vs Incidence Rate
 From the graph:
Incidence Risk (Cumulative Incidence) =
𝐷
𝑁
Whereas:
Incidence Rate ( Incidence Density) =
𝐷
𝑦
=
𝐷
𝑃𝑇
Case L: In a brief clinical study, 24 patients were enrolled and followed
for six days to observe development of H1N1 influenza. Details of how
they developed influenza were recorded; If eight(8) patients developed
flue on day 1, seven (7) on day 2, three (3) on day 3, two (2) on day
4, Two (2) on day five and only two(2) did not developed flue by the
6th day (Day 6). Determine the; (a) Risk of developing flue; (b) Rate
at which patients developed flue in the follow-up period.
20

Incidence Risk Vs Incidence Rate
 Case M: A person – time (years) at risk for 5 individuals in a
hypothetical cohort study between 2000 and 2004 is as indicated
in table below: where -start of study, X – disease, L – person Lost
to follow-up
 Determine the incidence rate of the disease in this case.
Year 2000 2001 2002 2003 2004 Years at Risk
Persons
1 5
2 X 4.5
3 X 3.5
4 X 1.5
5 L 3.5
Total 4 5 4.5 3 1.5 18
21

Attack Rate (Incidence Proportion)
22
 Attack Rate – the proportion of a population at risk (exposed people) that
contract / develop a disease in an outbreak after/following exposure in a
specified time interval.
𝐴𝑅 =
𝑁𝑜. 𝑁𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 𝑖𝑛 𝐸𝑥𝑝𝑜𝑠𝑒𝑑
𝑇𝑜𝑡𝑎𝑙 𝑃𝑜𝑝𝑜𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑟𝑖𝑠𝑘
× 100
 Used in acute diseases, esp. in disease outbreak investigations as a measure
of risk /risk of disease transmission in an outbreak & hypothetical predictions
 Types of Attack Rates:
1) Primary Attack Rate (PAR) =
𝑁𝑜. 𝑖𝑛𝑖𝑡𝑖𝑎𝑙(𝑝𝑟𝑖𝑚𝑎𝑟𝑦) 𝑐𝑎𝑠𝑒𝑠 𝑑𝑢𝑒 𝑡𝑜 𝑒𝑥𝑝𝑜𝑠𝑢𝑟𝑒
𝐼𝑛𝑖𝑡𝑖𝑎𝑙𝑙𝑦 𝐸𝑥𝑝𝑜𝑠𝑒𝑑 𝑃𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛
× 100
2) Secondary Attack Rate (SAR) =
[𝑇𝑜𝑡𝑎𝑙 𝑁𝑜. 𝑁𝑒𝑤 𝑐𝑎𝑠𝑒𝑠 −𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝐶𝑎𝑠𝑒𝑠]
[𝑇𝑜𝑡𝑎𝑙 𝐶𝑎𝑠𝑒 𝑐𝑜𝑛𝑡𝑎𝑡𝑠 ]
× 100
3) Overall Attack Rate (OAR) =
𝑇𝑜𝑡𝑎𝑙 𝑁𝑜.𝑁𝑒𝑤 (𝑝𝑟𝑖𝑚𝑎𝑟𝑦+𝑠𝑒𝑐𝑜𝑛𝑑𝑎𝑟𝑦) 𝑐𝑎𝑠𝑒𝑠
𝑇𝑜𝑡𝑎𝑙 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑟𝑖𝑠𝑘
× 100
4) Crude Attack Rate Vs Specific Attack Rates e.g. Food – specific Attack Rate,
And or Stratified Attack Rates e.g. Age/sex/food-type stratified Attack rate…

23
Case scenario: C1: In a family of 6, with 2 parents (already immune) and 4
children (all at risk of measles), if a child contracts measles from unknown
contact at school and two contacts among the siblings contract measles 3 days
following eruption of a skin rash in the index case within the family, determine;
a) primary attack rate & b) secondary attack rate of measles in this family.
C2: Disease Surveillance reports an outbreak of shigellosis in a community
with a population of 1000 people. If there were 18 index cases of shigellosis
from 18 different Households each housing an average number of 5 people,
determine; a) the overall attack rate of shigellosis in the community; b)
secondary attack Rate, if the total number of cases from the 18 H/Hs raised to
35 after an incubation period of shigellosis.
C3: Out of 750 new students who attended a university Bazaar, 540 ate ice
cream. If 430 students who ate ice cream and 30 who never ate ice cream
developed food poisoning, determine; a) Attack rates in the two strata of
students , b) Rate Ratio (Risk Ratio) and c) Rate difference (Risk difference or
attributable Risk). Comment on your answers.

24
C4: During the 1st wave of COVID-19 in 2020, Mubende Municipality with a population 600
people ( all at risk of COVID-19) reported 66 cases of COVID-19, with 65 cases admitted
for case MGT at Mubende Regional referral Hospital (MRRH) and 1 case was referred to
Mulago National Referral Hospital (MNRH) ICU for advanced life saving MGT. If 22 of 110
health workers on COVID -19 case MGT team in MNRH contracted COVID-19 from the
referral case, What was the; a) Primary Attack Rate for COVID-19 in Mubende Municipality
b) Secondary Attack rate of COVID-19 at Mulago NRH.
C5: Following an outbreak of E. coli O157:H7 bloody diarrhea in Kinshasa DRC (2003), an
epidemiologist conducted a case study to establish the link between exposure to certain
foods and E. coli O157:H7 bloody diarrhea. Study results were as shown below:
a) Determine the Attack rates and Relative Risk for each food item (present your results in a
table) b) Which was the most likely cause of E. coli O157:H7 bloody diarrhea outbreak
Food Item
Ate/Drank Did not eat/Drink
P-Value
Sick Total Sick Total
Leafy Vegetable salads /fruits 179 264 22 45 0.024
Beef / Poultry ( raw/half cooked) 176 226 27 73 < 0.001
Beverages 5 24 95 249 0.051

Lifetime Risk (LTR) Vs Lifetime Prevalence(LTP)
25
 Lifetime Risk (LTR) – A measure of risk or likelihood that a certain
health event (disease/death) will happen/occur during a person’s life
time i.e. the risk (probability) of developing a disease or dying from a
disease across a person’s life time e.g. Life time risk of developing
diabetes or contracting HIV / dying from advanced cancer or Life time
risk of maternal death.
 Lifetime Prevalence (LTP) – is the proportion of individuals in a
population that at some point in their life (up-to point of assessment)
have ever experienced a health event, risk factor or disease. E.g.
Lifetime prevalence of depression/anxiety, Alcohol/tobacco smoking
and drug abuse, accident, violence or assault, disaster, etc.

Lifetime Risk (LTR) of Maternal Death
26

Lifetime Risk (LTR) of Cancer & HIV
27

Relationship Between Prevalence (P) &
Incidence (I)
 Suppose a disease X occurs at an incidence Rate (I) during
a duration (D), then Prevalence (P) of the disease is;
 𝑷 = 𝑰 × 𝑫
 Also; I =
𝑃
𝐷
and D =
𝑃
𝐼
28
High Incidence Low Incidence
High
Prevalence
Common
Acute conditions
Ex: Endemic diseases
Long term / chronic conditions
Ex: DM, HTN, Arthritis, IBD, HIV.
Low
Prevalence
Acute Diseases
Very brief conditions
Ex: acute Diarrhea, ARTI,
Common cold, Epistaxis, ..
Rare
Short term condition
Ex: Pancreatic Cancer,
fulminant liver disease, etc.

Factors affecting Incidence(I) & Prevalence (P)
Factor /situation Incidence Prevalence
↑Survival rate / Chronicity of disease and or recurrence - ↑
↑Mortality - ↓
Cure Treatment/ therapy →↑Cure rate / recovery - ↓
Chronic care treatment →↑survival rate - ↑
Preventive treatment ↓ ↓
Vaccination → Herd immunity ↓ ↓
Prevention (↓ cause/Risk factors, interrupt transmission, protect host) ↓ ↓
Note: 1) For Chronic diseases, P > I since P increases with New cases adding on Existing cases.
2) For diseases that is rapidly/highly Fatal or quickly cured, P ≅ I
Points to Ponder:
 Cure/Recovery Factors ???
 Recurrence / Relapse Factors ???
 Mortality Factors ???
30

Relationship Between Prevalence (P) &
Incidence (I)
31
Case scenarios:
1) The incidence rate of Frontal temporal dementia is 8.9 cases
per 100,000 person per years, if the mean duration of the
disease is 7.5years determine the prevalence of the disease?
2) As per GBD (2021), More than 6 million people world over
are affected by Inflammatory Bowel disease (IBD). If the
incidence rate of IBD is 2.33 (95%CI; 1.4 – 3.4) per 1,000
persons per year in Arab World, determine the expected
prevalence of IBD in 5years time?.
3) Assuming an equal incidence of disease, which of the cancer
following is more prevalent? A) Ca. Pancreas with an average
duration of 3months B) Brain cancer with average duration of
1.5years.

Ice Burg Phenomenon
 This is a theory / biological spectrum of diseases in the community
 With very small portion of diseases (tip of an ice berg) presenting as
clinically apparent/diagnosed cases ( often symptomatic / severe conditions
~ 10%) and
 A large chunk/mass/ proportion of disease (~90%) remaining
undiagnosed/inapparent in undiagnosed clinical, latent phase, sub-
clinical/preclinical stage, asymptomatic phase and carrier state,
 This is due to variations in a series of cellular and host responses confounded
by severity of disease, health seeking behaviors and access to health care.
 Examples of diseases that show Iceberg phenomenon; HIV, TB, Hepatitis,
HTN, Diabetes,
 Noted Iceberg phenomenon is not shown by Measles, Rubella, Rabies, &
Tetanus.
 Thus, it necessitates advanced diagnostics & clinical competence (Clinicians),
disease screening, active disease surveillance (epidemiologists), case
finding, contact tracking & line listing to inform health authorities on the
exact picture and burden of the disease & health conditions.
32

Point to ponder: Does these also contribute
to ice-burg effect?
36

Natural History of Disease
This is evolution & progression of disease in a susceptible host over time –
showing uninterrupted course of a disease process in an individual (
without clinical intervention/Treatment) from exposure, through
pathological changes leading to onset of symptoms to complete recovery
with/without disability or death.
37

PART B: MEASURE OF MORTALITY
Mortality – refer to occurrence of death in a given population over a given
period of time.
 1) Mortality Rate (Refers to incidence/number of deaths per 1,000
individuals (people) in a given area & time)
 Crude Mortality Rate (CDR)
 Specific Mortality Rate (SMR)
 Age specific Mortality Rate (ASMR)
 Sex-specific Mortality Rate (SSMR)
 Cause Specific Mortality Rate (CSMR)
 Maternal Mortality Rate (MMR)
 Case – Fatality Rate (CFR)
 Proportionate Death Rate (PDR)
 2) Mortality Ratio
 Maternal Mortality Ratio (MMR)
 Standardized Mortality Ratio (SMR)
38

MORTALITY RATE
 Crude Mortality Rate (CDR)
𝐶𝐷𝑅 =
𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐷𝑒𝑎𝑡ℎ𝑠 𝑖𝑛 𝑎𝑔𝑖𝑣𝑒𝑛 𝑝𝑒𝑟𝑖𝑜𝑑 & 𝑎𝑟𝑒𝑎
𝑀𝑖𝑑−𝑦𝑒𝑎𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛
× 1000
Note: Simplest mortality indicator, with less sensitivity due confounding
factors such as age-structure, cause of morbidity (infectious /NCD), etc.
Notable disparity: CDR developed world > CDR developing countries i.e.
10 > 8 per 1000 people per year irrespective of a reversal picture in
life expectancy at birth i.e. 76 & 66 years respectively (UN, 2011).
 Specific Mortality Rate (SMR)
 Age specific Mortality Rate (ASMR)
=
𝑁𝑜.𝑜𝑓 𝐷𝑒𝑎𝑡ℎ𝑠 𝑜𝑓 𝑝𝑒𝑟𝑠𝑜𝑛𝑠 𝑜𝑓 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝑎𝑔𝑒 𝑖𝑛 𝑎𝑔𝑖𝑣𝑒𝑛 𝑝𝑒𝑟𝑖𝑜𝑑 & 𝑎𝑟𝑒𝑎
𝑀𝑖𝑑−𝑦𝑒𝑎𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑠𝑎𝑚𝑒 𝑎𝑔𝑒
× 1000
Forms of ASMR: 1) Child MR [1 -4years] 2) Under five Mortality Rate (U5MR)
3) Infant MR (IMR), 4) Post Neonatal MR, 5) Neonatal MR [Early NMR + Late
NMR] 6) Peri-natal MR 7) Still Birth Rate (SBR), 8) Fetal Death Rates (FDR)[
Early, intermediate, Late FDR], 9) ASMR for people aged 60yrs & above, …
39

MORTALITY RATE, CONT
 Sex Specific Mortality Rate (SSMR)
=
Number of deaths among a specific sex group Male or Female
Mid−year population of the same sex group in a given period of time
× 1000
 Cause Specific Mortality Rate (CSMR)
CSMR =
𝑁𝑜.𝑜𝑓 𝐷𝑒𝑎𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑎 𝑐𝑒𝑟𝑡𝑎𝑖𝑛 𝑐𝑎𝑢𝑠𝑒
𝑀𝑖𝑑−𝑦𝑒𝑎𝑟 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑎𝑡 𝑟𝑖𝑠𝑘 𝑜𝑓 𝑠𝑎𝑚𝑒 𝑐𝑎𝑢𝑠𝑒
× 100
Forms of CSMR:
a) Case – Fatality Rate (CFR)
CFR =
𝑁𝑜.𝑜𝑓 𝐷𝑒𝑎𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑎 𝑐𝑒𝑟𝑡𝑎𝑖𝑛 𝑑𝑖𝑠𝑒𝑎𝑠𝑒
𝑇𝑜𝑡𝑎𝑙 𝑁𝑜.𝑜𝑓 𝑝𝑎𝑡𝑖𝑒𝑛𝑡𝑠 𝑤𝑖𝑡ℎ 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑑𝑖𝑠𝑒𝑎𝑠𝑒
× 100
b) Maternal Mortality Rate (MMR)
MMR =
𝑁𝑜.𝑜𝑓 𝑀𝑎𝑡𝑒𝑟𝑛𝑎𝑙 𝐷𝑒𝑎𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑑𝑢𝑒 𝑡𝑜 𝑝𝑟𝑒𝑔𝑛𝑎𝑛𝑐𝑦 𝑜𝑟 𝐶ℎ𝑖𝑙𝑑 𝑏𝑖𝑟𝑡ℎ
𝑀𝑖𝑑−𝑦𝑒𝑎𝑟 𝑁𝑜.𝑜𝑓 𝑊𝑜𝑚𝑒𝑛 𝑜𝑓 𝑐ℎ𝑖𝑙𝑑 𝑏𝑒𝑎𝑟𝑖𝑛𝑔 𝐴𝑔𝑒 (𝑊𝐶𝐵𝐴)
× 100
 Proportionate Death Rate (PDR) =
𝐷𝑒𝑎𝑡ℎ 𝑑𝑢𝑒 𝑡𝑜 𝑎 𝑐𝑒𝑟𝑡𝑎𝑖𝑛 𝑐𝑎𝑢𝑠𝑒
𝑇𝑜𝑡𝑎𝑙 𝐷𝑒𝑎𝑡ℎ𝑠 𝑤𝑖𝑡ℎ𝑖𝑛 𝑡ℎ𝑒 𝑠𝑎𝑚𝑒 𝑝𝑒𝑟𝑖𝑜𝑑
× 100
40

MORTALITY RATIO
2) Mortality Ratio
 Maternal Mortality Ratio (MMR) =
𝑀𝑎𝑡𝑒𝑟𝑛𝑎𝑙 𝐷𝑒𝑎𝑡ℎ
𝑁𝑜.𝑜𝑓 𝐿𝑖𝑣𝑒 𝐵𝑖𝑟𝑡ℎ𝑠
× 100,000
 Standardized Mortality Ratio (SMR) =
𝑂𝑏𝑠𝑒𝑟𝑣𝑒𝑑 𝐷𝑒𝑎𝑡ℎ
𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝐷𝑒𝑎𝑡ℎ
× 100
Note: If SMR > 1, then Observed Death > Expected Death and
Excess Deaths = (Observed Death – Expected Death)
Case scenario: A study indicates cancer Mortality analysis for all Males in
a give city between1975 – 1985: Determine the SMR & interpret results
Cause of Death Expected Death Observed SMR
Ca. stomach 54.58 72
Genitourinary cancer 168.90 162
Lymphoma 44.57 47
41

CASE SCENARIO: MEASURE OF MORTALITY
42
HMIS Report K: The National Referral Hospital HMIS performance report
for FY 2022/23 indicated the following health indicators. Use this report
to answer Qns below;
Determine the; a) Still Birth Rate; b) Neonatal Mortality c) Under five
Mortality Rate; d) Maternal Mortality Rate & e) Maternal Mortality Ratio
Indicator Number
OPD Attendance for;
 Children Under 5years
 Women of Child Bearing Age (WCBA)
98,765
56,789
Total Live births 8,765
Still Births (fetal death ≥ 28WOA, or ≥ 1000g) 765
Deaths of children aged < 28days 678
Maternal Deaths relating to pregnancy & child birth 65
Total Deaths for children Under 5years 567

43
NIRA Statistical Report L: The National Identification & Registration
Authority statistical report for FY 2021/22 indicated the following vital
statistics. Use this report to answer Qns below;
Determine the; a) (i) Crude Birth Rate & (ii) Crude Death Rate; b) Infant
Mortality Rate; c) Aged- specific Mortality Rate for people aged 65yrs
& above d) Cause –specific Death Rate for (i) Cancer & (ii) CVD and e)
Vital Indicator Number
Total Mid-year population 100,000
Population size aged;
 65years and above
 Infants born alive
25,000
3,000
Total Deaths (all causes) 1,500
Deaths of;
 Infants under 1 year age
 Persons aged 65years and above
50
1000
Death due to;
 Cancer
 Cardiovascular Disease (CVD)
100
300

44
Vital statistics Data Base of a given Country Q provides the following
annual Data; . Use this report to answer Qns below;
Determine the; a) Still birth rate b) Perinatal death rate c) Compare Early
neonatal mortality rate and Late neonatal mortality rate d) Post-neonatal
mortality rate e) Child mortality rate for children 1 – 4years.
Vital Indicator Number
Total Mid-year population
Children aged 1 – 4years (12 – 59mo)
1,250,000
5,000
Births;
 Live Births
 Still Births
8,900
350
Total Deaths (all causes) 1,500
Deaths of;
 Neonates < 7days of life
 Neonates 7 – 28days of life
 Infants aged 4 – 52weeks
 children 1 – 4years
890
260
252
240

Distribution of Maternal Death & Still Birth Death Rates
45

1) A Half (½) of under 5s deaths in 2019 occurred in 5 countries;
Nigeria, India, Pakistan, DRC and Ethiopia
2) Only Nigeria & India accounted for a third (1/3 ) of all deaths.
858
824
399
291
178
132
115
103
93
90
Nigeria
India
Pakistan
DRC
Ethiopia
China
Indonesia
Tanzania
Angola
Bangladesh
Top 10 Countries with the Highest Number (1000s) of
Under 5 years Mortality Globally (WHO, 2019)
46

47
WHO/CDC (May 2016): An outbreak of Chikungunya viral fever (CHIKV) was
confirmed in Mandera County, Kenya; 85% of CHIKV patients are symptomatic
and presented with sudden onset of fever in 3-7days, headache, Joint
pains/swelling, vomiting & Skin rash . A total of 260 cases were reported in
Mandera county and 71 cases admitted in various pubic Hospitals which were
then battling to control already exiting cholera outbreak. Below was the status
report of this CHIKV outbreak (1st - 31st May 2016).
Determine; a) Primary attack rate for CHIKV in Mandera county.
b) Case fatality rate (CFR) of CHIV in (i) Mandera county & (ii) Public Hospitals assuming
the only fatalities were those who had already died; and all infected cases recovered.
c) CFR of (i) CHIV in Mandera county assuming outcome of all infected cases is unknown,
40 in 189 infected cases died, out of 149 remaining cases 119 recovered and 30 were
still ill with unknown outcome by 31st May 2016; (ii) CFR of CHIKV in Public Hospitals
assuming outcome of all infected cases is unknown, out of 71 admitted cases, 11died,
and 10 in the remaining 60 cases were still sick with unknown outcome by 31/5/2016.
Morbidity/Mortality indicators Mandera county Public Hospitals
Number of people at risk 1438 254
Number of CHIKV cases 189 71
Number of Deaths 40 11
Number of alive/ill 149 60

Attributable Mortality Vs Attributable Morbidity
48
 Attributable Mortality (AM): refers to the death attributable or due to
an exposure; Excess deaths due to exposure
i.e. AM = Mortality Difference (MD) = (Me – Mo)
Examples: MRSA attributable mortality, Nosocomial sepsis attributable
mortality, CVD attributable mortality, Alcohol attributable Mortality,
Tobacco smoking-attributable Mortality, Attributable mortality of
ventilator-associated pneumonia (AM VAP), etc.
 Attributable Mortality Fraction (AMF) = % AM =[AM/Me] x 100:
refers to the proportion of death attributable or due to an exposure.
 Key terms/indicators:
 Attributable Mortality Rate (AM Rate)
 Attributable Mortality Risk (AM Risk)
 Attributable Mortality Risk Ratio or Rate Ratio (AM RR)
 Residual Attributable Mortality (RAM)
TASK: With Suitable Examples Discuss, Attributable Morbidity

Premature Mortality Vs Potential Years of Life Lost
49
(Refer to Group B4/A4 course work)
 Measures Potential Years of Life Lost (PYLL) before 70years of age.
 Focuses on deaths among young age structure of a population; required age
– adjusted mortality measures.
 Determinants of PYLL: Infant Mortality Rate, Mortality from diseases &
Injuries / violence/ civil unrest
 Indicators Premature Mortality
 Premature Mortality Rate
 Age – standardized Premature Mortality Rate (30 – < 70yers) (esp. due to NCDs)
 Premature Mortality Risk
 Attributable Premature Mortality; Rate & Risk
 Diseases that Leads to the highest Premature mortality Rate
 Trends & Disparities in Premature mortality Rate
 Varies with sex (more in male), income level/GDP per capita, advance in
medical research & technology, life style, occupation, etc.
TASK: Read & Write Short Notes on:(i) Early Mortality (ii) Late Mortality

Survival Rate; Survival Analysis→ Predict Prognosis
50
 Survival Rate: proportion of people/patients in a study or treatment group who are
still alive for a certain period of time after they were diagnosed with or started on
treatment for a disease or HRC.
 𝑆𝑅 =
𝑁𝑜.𝑜𝑓 𝑝𝑎𝑡𝑖𝑒𝑛𝑡𝑠 𝑎𝑙𝑖𝑣𝑒 𝑜𝑣𝑒𝑟 𝑎 𝑝𝑒𝑟𝑖𝑜𝑑
𝑇𝑜𝑡𝑎𝑙 𝑁𝑜.𝑝𝑎𝑡𝑖𝑒𝑛𝑡𝑠 𝑑𝑖𝑎𝑔𝑛𝑜𝑠𝑒𝑑 𝑜𝑟 𝑡𝑟𝑒𝑎𝑡𝑒𝑑
× 100 or =[ 100 – Death Rate]
 Survival rate indicators:
 Overall (Observed) survival rate Vs Net Survival Rate
 Cause specific survival rate or Disease specific Survival rate e.g. cancer survival rate
 Time specific survival rate e.g. 5 – year survival, 10 – year survival rate, etc.
 Relative survival rate (Overall SR in diagnosed : Observed SR in identical popn not
diagnosed with that disease)
 Perceived Survival Rate Vs Lead Time Bias
 Others; Median survival rate, disease-free survival rate, progression survival rate,
metastasis-free survival rate, etc.
Uses: 1) Describes prognosis of a disease
2) Yard stick for assessment/evaluation of therapy/clinical intervention/techn.

Survival Rate; Survival Analysis
51

IMPACT/BURDEN OF DISEASE
52
 Health impact; complications/Dysfunctions/ disability, ↓QALYs,
↑DALYs, ↑ Health care costs / Catastrophic health expenditure,
Hospital stay/ Hospitalization, pain/suffering/stress, psychological
trauma, Death.
 Socio-economic impact; ↓ productivity (Vs ↑Presenteeism), ↑ sick
absenteeism, loss of income/job, unemployment, poverty, ↓ academic
grade, destitution, ↑National Total Expenditure on Health care, civil
unrest, etc.

Assignment:
Case P: The mortality rate for stomach ulcers in males in the UK was 96 per million
population per year in 1950 and 31 per million in 1980. Determine the Mortality Rate
Ratio of UK from 1950 to 1980.
Case Q: In a country R with a mid-year population 6 million people, 60,000 deaths
occurred during the year ending 31st Dec. 2020. These included 30,000 deaths due to
cholera in 100,000 cholera cases reported in the country amidst the first wave of
COVID-19 pandemic. If COVID-19 alone attributed to 3% of the death toll in this
country, (A) What mortality indicators does the 3% of death toll in this country
represent? (B) Determine COVID-19 deaths in this country (C) Define (i) Case fatality
rate (ii) Cause Specific Mortality Rate (D) Determine the (i) Case fatality rate (ii) Cause
Specific Mortality Rate (iii) Proportionate death rate of Cholera in Country R (2020).
READING ASSIGNMENT: Read & write Short Notes on:
A) Age Adjustment / Standardization : Direct & Indirect standardization
B) Age Adjusted Prevalence and Age Adjusted Incidence Rate
C) Age adjusted Abortion rate for WCBA 15 – 49years & Abortion Mortality Rate
D) Age Adjusted Mortality /Death Rates
E) Life Expectancy; Life expectancy at Birth (LEB), Period Life Expectancy vs Cohort Life
Expectancy, Health adjusted LE (HALE), Disability adjusted LE (DALE) and Sexually Active
Life Expectancy.
F) Life Table
G) Quality Adjusted Life Years (QALYs)&Disability Adjusted Life Years(DALYs=YLD+YLL)
53

Life Expectancy Vs Longevity
54

Distribution of Disease & HRCs
 Descriptive Epidemiology describes disease occurrence in terms of Persons, Place &
Time; It asks & seek to answers three (03) questions;
 Who is affected /getting disease? – Person(s) (?? Patient’s Biodata- Age, Sex, …)
 Where is disease occurring? – Place (?? Patient’s address/residence– Rural/urban)
 When is disease occurring? – Time (?? Onset, duration, day/night, admission, …)
 It Use Descriptive studies (e.g. Descriptive Cross sectional study, clinical/case studies)
& Descriptive statistics including; Frequency (prevalence, incidence) , Measures of
central tendencies, and measures of dispersion to describe distribution of disease &
HRCs in terms of Person, Place & Time.
 Disease & HRCs are described in terms of;
 Demographic distribution using (Tools): descriptive statistics, tabulation, Graphs &
pictorials/ info graphics.
 Spatial distribution using (Tools): Spatial mapping, spatial analysis, remote sensing &
geospatial technology/referencing with GIS & GPS.
 Temporal distribution using(Tools): Disease-Time plots/curves & Temporal analysis
(Bubble plots, disease timelines, moving point averages, exponential smoothening, ...).
55

Determinants of Disease & HRCs
 Epidemiology assert that disease & HRCs do not occur at random,
rather due to causative agents/risk factors referred to as
determinants of the disease & HRCs.
 Analytical Epidemiology seeks to establish/determines causes and
risk factors for the occurrence of a disease & HRC so as to prevent and
control diseases & improve/promote health.
 Disease determinants include;
 Individual; biological (age, sex), genetic, immunological, and
behavioral & lifestyle factors,
 Familial, socio-demographic, Socio-economic and community factors,
 Environmental & Ecological factors; as well as Health care factors.
 The Association between Exposure (E) to the determinants and health
Outcome (O = Disease) is measured by Measures of
Effect/Association using Analytical statics,
 Guided by Theories & Models of disease causation & prevention
 Analytical epidemiology Set & test hypothesis between E & O
56

Determinants of Disease &Health Related Conditions
 Theories Disease Causations & Prevention
(A) The 19th century model
 Contagion theory
 Supernatural theory
 Personal behavior theory
 Miasma theory
(B) The 20th century models
 The Germ Theory
 The Life Style Theory
 The Environmental Theory
 The Multi Causal Theory
57

Determinants of Disease &HRCs
 Models of Disease Causation
 Epidemiological Triad ( Prof. Wade Hampton Frost, 1928)
 Web Model ( MacMahon, Pugh & Ipsen, 1960)
 Rothman’s component causes model (The Sufficient cause & Components
causes Model) ( Kenneth Rothman, 1976)
 Socio-Ecological Model (Urie Bronfenbrenner, 1977)
 Wheel Model (Mausner & Kramer, 1985)
 Types of disease causation
 Direct Causation
 Indirect Causation
58

Models of Disease Causation
 Epidemiological Triad
59

 Ecological Model
60

 Wheel Model (Mausner & Kramer, 1985)
62

EX: Wheel Model (Mausner & Kramer, 1985)
63

 Web Model of Disease Causation
64

 Rothman’s component causes model (The Sufficient cause &
Components causes Model) ( Rothman 1976)
Where: 1) A, B &C are known components (causes).
2) U is unknown component/ cause (s)
3) N is a Necessary Cause/Factor
For disease Causation:
Known components/causes [A,B,C]
+
Unknown component (s)/Cause(s)[U] = Sufficient cause
+
Necessary Cause/factor[N]
65

Ex: Rothman’s component causes model
E.g. HIV infection Causation
66

Types of disease causation
 Direct Causation
 Indirect Causation
67

Direct Causation
Types of Direct Causation:
1) Mono-causal Association (One Causal/Risk Factor → One disease)
2) Multi-Factorial Causation
2.1. Independently 2.2. Cumulatively/ concertedly
68

Indirect Causation
Examples:
1) Unprotected Sex → HIV → HIV/AIDs
2) Malaria → Severe Anemia → Heart Failure
4) Poor Diet → HTN → Heart Disease
5) Obesity/ Hyperlipidemia → Atherosclerosis → CHD
6) Others
69

Causal Relationships
 Types & Levels Casual Relationship
1) Necessary and Sufficient (Absolute Causality)
2) Necessary but Not Sufficient (Conditional Causality)
70

Causal Relationships, CONT
3) Sufficient But not Necessary
4) Neither Necessary nor Sufficient ( Contributory Causality)
71

Evidence For Causal Relationship
 Koch’s Postulates on Causal Relationship (Robert
Koch. 1840)
 The microorganism is always found with the diseased But not healthy
individuals.
 The microorganism is not found with any other disease.
 The microorganism isolated from the diseased and cultured through
several generation, if inoculated in healthy individual(experimental
animal), produces/cause a disease.
 The microorganism must be re-isolated from the inoculated, diseased
individual and matched to the original microorganism.
72

Bradford Hills criteria for Causal Relationship (Austin
BH, 1965)
 Strength of Association [?P-value <0.05, then OR/RR/AR & 95% CI
(Confidence interval) does not cross the null hypothesis Ho]
 Consistency of findings (Reliability; Reproducibility- in repeated exposure
settings)
 Specificity (casual agent /factor almost exclusively associated with outcome)
 Temporality (sequence of events; Exposure should precede Outcome)
 Biological gradient (dose –response relationship)
 Biological plausibility (mechanism/process – pathogenesis/ pathophysiology)
 Coherence (btn epidemiological & laboratory findings/evidence)
 Interventional/experimental evidence – Occurrence & Reversibility of Outcome
 Analogy - Absence of similar/alternative explanation for the outcome
73

EX1: MATCH the Following Bradford Hill’s Criteria of Causal
Relationship with a correspondingly appropriate statement
74

Association BTN Exposure & Outcome [E → O]
 Associations Types/Pyramid of Association;
Recall Bradford Hill’s Criteria for causal relationship; 1) For a causal
Association ((E → O), Exposure(E)/ Factors should be strongly ( P<0.05)
associated to Outcome (O); with a 95% Confidence Interval (C.I).
75

Measure of Effect / Association between Exposure &
Health Outcome
 Aim: Establish determinants of disease (since diseases do not occur at
random) i.e.
 Aspects: 1) Determine Association ( i.e. P-value)
2) Measure Strength of Association ( i.e. OR, RR, PR).
3) Measure Potential of disease prevention (RD or AR, PAR).
 Methodology: Association between Exposure(E) and Outcome(O) is measured
by Analytical studies i.e.
1) Analytical Cross-sectional studies,
2) case-control studies and
3) cohort studies
 Recall: Outcome (O) following exposure/ clinical intervention (E) can be;
 Disease/injuries, recovery, disability/complications/defects & death,
 ↑ Immunity/ Antibodies, ↑Nutritional status, ↓pregnancy.
 ↑Acceptance & Utilization of health services/products, ↑Improved Health &
↑Quality of life (QALY=Quality Adjusted Life Years ).
76

Measure of Association/Effect
 Basic/Common Measures of Association
 Odds Ratio (OR) – compare odds of disease occurrence in exposed & non-
exposed or Cases & non-cases. Ex: Prevalence odds ratio, Incidence odds ratio.
 Rate Ratio (RR) – compare rates of disease occurrence in exposed & non-
exposed or cases & non-cases. Ex: Prevalence rate ratio, Incidence rate ratio.
 Risk Ratio (RR) – compare risks/incidences of disease occurrence in exposed
& non-exposed or Cases & non-cases. Ex: Incidence risk ratio
 Relative Risk (RR) – compare risks/incidences of disease occurrence in
exposed & non-exposed in prospective studies (only/exclusively).
 Attributable Risk (AR) [Risk Difference (RD) & %RD]
 Population Attributable Risk (PAR)
 Other Measures: Chi square [𝑥2] and Correlation coefficients [r]
NB: P-Values and Measures of Effect/Association can be obtained by Logistic
regression Analysis (Univeriate, Bivariate & Multivariate analysis) using
Analytical software/computations or Z-distribution Tables using Z-critical values.
77

Measure of Association/Effect
 Odds Ratio
OR =
𝑂𝑑𝑑𝑠 𝑖𝑛 𝑒𝑥𝑝𝑜𝑠𝑒𝑑
𝑂𝑑𝑑𝑠 𝑖𝑛 𝑢𝑛𝑒𝑥𝑝𝑜𝑠𝑒𝑑
Consider, a retrospective study that enrolled (a+c) cases and (b+d)
identical controls. The cases & controls were assessed for history of
exposure, and results summarized in a 2X2 contingency table below:
From tab; OR =
𝑎/𝑏
𝑐/𝑑
=
𝑎𝑑
𝑏𝑐
with a 95% CI = 𝑂𝑅 × 𝑒
[±1.96 (
1
𝑎
+
1
𝑏
+
1
𝑐
+
1
𝑑
)]
Exposure Outcome (i.e. Disease) Total
Diseased (Cases ) Non-diseased (Control)
Exposed a b (a+b)
Unexposed c d (c+d)
Total (a+c) (b+d)
78

Odds Ratio (OR)
 Interpretation of OR:
 OR > 1: Increased Odds of disease with Exposure/Positive Association
 OR = 1: No Association Btn Exposure & disease
 OR < 1: Decreased Odds of disease with Exposure/ Negative Association
 OR is Determined/Measured in Analytical study designs below
Case Scenario: Interpret findings of study on HTN associated Factors:
Study design Numerator Denominator
Cross sectional Odds of outcome in Exposed Odds of outcome in unexposed
Case – control (unmatched) Odds of exposure in Cases Odds of exposure in Non-cases
Cohort Odds of outcome in Exposed Odds of outcome in unexposed
Variable Category OR 95% confidence Interval P-value
Sex Male 3.12 1.04 7.45 0.02
Female 1.0 Ref. Ref.
79

Odds Ratio (OR)
 Case P: Ensekiriyo et al conducted a study to on 800 individuals to
determine association between exposure to asbestos & Malignant
mesothelioma. If out of 300 exposed individuals, 250 developed
Mesothelioma and of the unexposed individual, 100 developed the
cancer. Determine the odds of developing cancer & comment on your
answer.
 Case Q: Kemigisha conducted a cross-sectional study to assess the
effect of smoking on Pre-eclampsia in pregnant mothers. Interpret her
logistic regression analytical findings.
80

Measure of Association in Matched Case-control Studies
81
 Consider a matched case control study where equal pairs (n) of
matched cases (D+) and Controls (D-) are enrolled i.e. Each cases is
matched by person factors such as age, race, weight, etc. to a control.
 Study results may be presented as:
 Exposure in cases and Control in particular pairs can yield: A -
Concordant (Exposed cases & Controls), B- Discordant
Pairs Exposure status in
Cases (D+) Control (D-) Concordance/Discordance
1 Exposed Exposed Concordant
2 Exposed Not-exposed Discordant
3 Not-exposed Exposed Discordant
4 Not-exposed Not-exposed Concordant
n ……. …….. ……..

Measure of Association in Matched Case-control Studies
82
 From above, Exposure in cases and Control in particular pair(s) can yield:
 A -Concordant (Exposed cases + exposed Controls),
 B- Discordant ( Exposed cases + unexposed controls)
 C – Discordant (Unexposed cases + exposed Controls),
 D - Concordant (Unexposed cases + Unexposed Controls),
i.e. Presented in a 2x2 contingency table as
Exposure in case in particular pair Exposure in control in particular pair
Exposed Unexposed
Exposed Concordant (A) Discordant (B)
Unexposed Discordant (C) Concordant (D)
Odd Ratio (OR) =
𝐷𝑖𝑠𝑐𝑜𝑟𝑑𝑎𝑛𝑡 𝑝𝑎𝑖𝑟𝑠 𝑖𝑛 𝐶𝑎𝑠𝑒𝑠 𝑤𝑖𝑡ℎ 𝐸𝑥𝑝𝑜𝑠𝑢𝑟𝑒
𝐷𝑖𝑠𝑐𝑜𝑟𝑑𝑎𝑛𝑡 𝑝𝑎𝑖𝑟𝑠 𝑖𝑛 𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑠 𝑤𝑖𝑡ℎ 𝐸𝑥𝑝𝑜𝑠𝑢𝑟𝑒
=
𝐵
𝐶

In a small pilot Study M, Uterine Cancer Vs Estrogen Use
83
In this study,12 women with uterine cancer and 12 with no apparent
disease were contacted and asked whether they had ever used estrogen.
Each woman with cancer was matched by age, race, weight, and parity to
a woman without disease. The results are shown below:
Determine the odds of estrogen use among cancer cases compared to
non-cancer matched controls in;
(i) unmatched case-control study design?
(ii) a matched case-control study design? Comment on your Results

Relative Risk (RR)[Rate Ratio/Risk Ratio]
 Relative Risk ( Rate Ratio or Risk Ratio)
RR =
𝐼𝑛𝑐𝑖𝑑𝑒𝑛𝑐𝑒 𝑅𝑎𝑡𝑒 𝑖𝑛 𝑒𝑥𝑝𝑜𝑠𝑒𝑑
𝐼𝑛𝑐𝑖𝑑𝑒𝑛𝑐𝑒 𝑅𝑎𝑡𝑒 𝑖𝑛 𝑢𝑛𝑒𝑥𝑝𝑜𝑠𝑒𝑑
or =
𝐼𝑛𝑐𝑖𝑑𝑒𝑛𝑐𝑒 𝑅𝑖𝑠𝑘 𝑖𝑛 𝑒𝑥𝑝𝑜𝑠𝑒𝑑
𝐼𝑛𝑐𝑖𝑑𝑒𝑛𝑐𝑒 𝑅𝑖𝑠𝑘 𝑖𝑛 𝑢𝑛𝑒𝑥𝑝𝑜𝑠𝑒𝑑
Consider, a prospective study that enrolled a sample n = (a+b+c+d), with ne =
(a+c) exposed group and no = (b+d) unexposed controls. The exposed group
& controls were then followed up for period and assessed for a desired
outcome (D+, D-), and results summarized in a 2X2 contingency table below:
From tab; RR =
𝑎/(𝑎+𝑏)
𝑐/(𝑐+𝑑)
Exposure Outcome (Disease/Case) Total
Diseased (D+) Non-diseased (D-)
Exposed a b (a+b)
Unexposed c d (c+d)
Total (a+c) (b+d) (a+b+c+d)
84

Relative Risk (RR)
 Interpretation of RR:
 RR > 1: Increased Risk of disease with Exposure/Positive Association
 RR = 1: No Association Btn Exposure & disease
 RR < 1: Decreased Risk of disease with Exposure/ Negative Association
 RR is Determined/Measured in Analytical study designs below
 Conditions: Odd Ration (OR) to approximate Risk Ratio(RR)
OR ~RR, if;
1) The outcome is rare i.e. Prevalence < 10%
2) The cases & controls are from the same population (identical)
3) The cases in a sample are representative of the cases in the population
4) The controls in the same are representative of the controls in the population
Study design Numerator Denominator
Prospective Cohort Incidence of outcome in Exposed Incidence of outcome in unexposed
85

Measure of Effect/Association
86
 Case R: A Prospective cohort enrolled 500 individuals to study a effect of
Alcohol on Liver cirrhosis. If 80 cases of Liver cirrhosis were diagnosed in 200
alcoholics and only 20 cases of the disease diagnosed in the non-alcoholics.
Determine the Relative risk of liver cirrhosis in alcoholics compared to
counterparts and comment on your results.
 Case S: A longitudinal study enrolled 580 study participants in South-
Western Uganda to assess effect of exposure to Aflatoxins on Hepatocellular
Carcinoma (HCC). About 30 cases of HCC were identified in 230 adults who
were exposed to Aflatoxins and only 15 counterparts had HCC.
a) What study design was used in this study?
b) Draw a 2 X 2 contingency table to represent this epidemiological data.
C) (i) What is the suitable measure of effect of aflatoxins on Hepatocellular
carcinoma?
(ii) Measure the effect of aflatoxins on Hepatocellular carcinoma & comment on
your answer?

Measure of Effect/Association
Case T: Below are results of a study to determine effect of menopause on CHD
in women: Determine the Rate Ratio & interpret your results.
Case U: A study Q was conducted to test a hypothesis; “There is no association
between Safe Male Circumcision (SMC) and Incidence/prevalence of HIV”. (A)
What type of hypothesis was used in the study? (B) Suggest a study design that
can be used to test the study Hypothesis? (C) Propose a suitable measure of
Effect of SMC on Incidence/prevalence of HIV, (D) If a cross sectional study
design was conducted and showed that the prevalence of HIV in circumcised
men is 2.5% and that of the uncircumcised counterparts is 6.7% (Malemo,
2018), what was the prevalence ratio? Comment on your answer.
87

Attributable Risk (AR)[Risk Difference-RD]
Attributable Risk
AR =[ Incidence Risk in Exposed – Incidence Risk in unexposed]
AR = I (exposed) – I (unexposed) = (Re – Ro)
Risk Difference (RD) estimate the excess risk of disease caused by/
attributable to a risk factor among the exposed group.
Thus, AR or RD measure of the potential for prevention of disease if
exposure to the risk factor could be eliminated i.e. Guide Health policy
& decision on disease prevention
Percentage Attributable Risk ( % Risk Difference)
% AR or %RD = [
𝑅𝑒 –𝑅𝑜
𝑅𝑒
] × 100 = [
𝑅𝑅 –1
𝑅𝑅
] × 100
i.e. Attributable fraction in exposed
88

Attributable Risk (AR)[Risk Difference-RD]
Case X: A study on Mother to child transmission (MCT) of HIV showed
that the risk of HIV among breast feed babies was 280 per 1000
children and the risk in the non-breast feed babies was 150 per 1000
children. Determine the;
a) Risk of transmission of HIV attributable to Breastfeeding
b) % Risk Difference & interpret your results
Case Y: A prospective Cohort study to determine the association of
smoking on Coronary Artery Disease (CAD) enrolled 8000 study
participants. If 84 in 3000 smokers of tobacco developed CAD and only
4913 in the non-smokers did not at the end of the study period.
Determine;
a) Risk of CAD attributable to tobacco smoking
b) % Attributable Risk and comment on your answer.
89

Population Attributable Risk (PAR)
Recall; Attributable Risk (AR) applies to exposed persons, and guides
disease prevention to exposed individuals only, yet a holistic disease
prevention program requires targeting the entire population at risk
[Exposed + Unexposed persons]. Thus, need to determine Population
Attributable Risk.
Where:
PAR =[ Incidence Risk in total population – Incidence Risk in unexposed]
PAR =[RT – Ro] i.e. Total = Exposed + unexposed.
And % PAR = [
RT – Ro
RT
] × 100 = [Prev(exposed) ×%AR]
Application: Used by Health policy makers and Funders in
Planning/Prioritizing & funding disease prevention and Health promotion
program/strategies.
90

Case K: A study in Sub-Saharan Africa to determine effect of truncal
obesity on metabolic syndrome, enrolled 10,000 adults aged 18 –
59years. About 850 cases of metabolic syndrome found in 4500
obese adults and only 250 adults of normal BMI had metabolic
syndrome . Determine the;
a) Relative Risk of developing Metabolic syndrome in obese adults
compare to individuals with normal BMI.
b) Risk of developing Metabolic syndrome attributable to Obesity
c) % Population Attributable Risk (PAR) & make a policy statement on
prevention of Metabolic syndrome with evidence from this study.
91

AR/RD & PAR, cont.
 Case L: Table x shows risk of backache in KIU female workers aged
20 – 44years due to work posture.
Determine,
a) % Risk difference of backache in standing female workers at KIU
b) Proportion of risk of backache in KIU female workers
c) From the above findings, make an occupation health policy statement
to guide working conditions of female workers at KIU.
Work posture Risk of backache per 100 female workers
over two years
Standing female workers 12.3
Other Female workers (not standing at work) 7.7
All Female workers 8.3
92

Measures of Effect of Clinical Intervention in Clinical
Experimental/Interventional Studies
93
 Like Longitudinal prospective analytical studies, Experimental studies
use the following inferential statistical measures of effect;
1) Risk Ratio or Relative Risk (RR) or Hazard Ratio (HR) = Re/Ro
2) Risk Difference (RD) = Absolute Risk Increase (ARI) = Re –Ro or
EER – 𝐶𝐸𝑅 or Absolute Risk Reduction (ARR) or = Ro –Re or CER – EER
3) Relative Risk Reduction (RRR) = ARR/Ro or (1 – RR) x 100
or Relative Risk Increase (RRI) = ARI/Re = (RR - 1) x 100
or Relative Effect (RE) for unknown direction of effect of the clinical intervention
4) Number Needed to Treat (NNT) = 1/ARR
Interpretation: No. patients needed to treat with clinical intervention to prevent/protect
against/cure 1 patient/case/clinical event.
or Number Needed to Harm (NNH) = 1/ARI
Interpretation: No. patients needed to treat with clinical intervention to harm 1
patient.

Measure of Effect in Randomized Control Trial (RCT)
94
RCT case M: A study recruited 1000 patients to test the effect of a new drug on
treatment of TB at Mbarara regional referral hospital. A half of the patients
were allocated to receive the drug on a daily basis and the other ones did not.
The patients were followed over a period of six months?
(a) What type of study is this?
(b) How were the patients allocated to the two groups?
(c) How do we call the group that:
(i) Received the new drug ………………...
(ii) Did not receive the new drug …………
(d) What can we possibly give the group that did not receive the drug?
(e) At the end of the six months, 450 and 200 patients in the new drug group
and the no new drug group respectively had recovered from TB. Determine the
(i) Absolute Risk Reduction (ARR) and Number Needed to Treat (NNT),
(ii) Risk Ratio (RR) and Relative Risk Reduction (RRR) comparing recovery among
the two groups. Comment on your results.

95
Case N: A Randomized Control Trial R enrolled 455orthopedic patients with femoral
neck fractures to compare the effectiveness and safety of Hemiarthroplasy (HA) to that
of Open Reduction & Internal Fixation (ORIF). Two hundred twenty nine (229) patients
were subjected to HA and the rest to ORIF and their post-operative outcomes were
observed and noted. Only twelve (12) post-operative complications were reported in
patient that underwent HA and ninety (90) cases of post-op complications were reported
in their counterparts.
a) Draw a 2 X 2 contingency table to represent results of the study.
b) Determine the risk of post-operative complications among patients that underwent; (i)
Hemiathroplasty (HA)? (ii) ORIF?
c) What is the Hazard Ratio (HR) of post-op complications among patients that
underwent HA compared to their counterparts and comment on your result?
(d) Calculate the relative risk reduction (RRR) of post-op complications among patients
that underwent HA compared to their counterparts and comment on your answer?
(e) Estimate the Absolute Risk Reduction (ARR) of post-op complications among patients
that underwent HA compared to their counterparts?
(f) What is the number of orthopedic patient needed to Treat (NNT) with HA instead of
ORIF to prevent one patient from post-op complications?

96
RCT case O: (a) In a randomized control Trial (RCT) comparing two surgical
techniques, 200 surgical candidates were enrolled and 100 patients were
randomly assigned to each group of the two surgical groups. Ten (10) patients
in the first surgical group experienced post-operative complications while 15
cases of post-operative complications were reported in the second
group/counterparts. Determine the;
(i) Relative Risk (RR) or Hazard Ratio (HR) or Risk Ratio (RR)
(ii) Absolute Risk increase (ARI) or Risk Difference (RD) & Number Needed to
Harm (NNH). Comment on your Results
(iii) Relative Risk Increase (RRI) for the second surgical technique compared to
the first one. Comment on your Results
(b) Describe a well designed Randomized control trial to assess the New
vaccine’s Efficacy in comparison to the Existing one. Describe the key components
of your study design, including Study population-Selection & sampling,
Randomization, allocation of intervention & control, Blinding, Outcome measures,
and ethical considerations.

Other Measures of Effect/ Association
97
1) Chi square [𝑥2]; 𝑥2 =
( 𝑏−𝑐 −1)2
(𝑏+𝑐)
, or
(𝑂−𝐸)2
𝐸
; where O - Observed Value,
E –Expected Value; with a df = (n -1) = (r -1)(c -1) and
critical value 𝑥2
≥ 3.84 at P< 0.05 statistical significance level
2) Correlation coefficients [r]; i.e. Pearson corelation coefficient
𝑟 =
𝑛( 𝑥𝑦)−( 𝑥)( 𝑦)
𝑛 𝑥2 − 𝑥 2 [𝑛 𝑦2 −( 𝑦)2]
; with a Positive Correlation ( r = +ve; 0 ≤
r ≤ +1) or Negative Correlation (0 ≥ r ≥ -1) at P< 0.05 statistical significance
level or No Correlation (r =0)
0 ≤ r ≤ 1 Interpretation -1 ≤ r ≤ 0 Interpretation
0 ≤ r < 0.2 No or negligible correlation 0 ≤ r < -0.2 No or negligible correlation
0.2 ≤ r < 0.4 Weak positive correlation -0.2 ≤ r < -0.3 Weak Negative correlation
0.4 ≤ r < 0.6 Moderate positive correlation -0.4 ≤ r < -0.6 Moderate Neg. correlation
0.6 ≤ r < 0.8 Strong positive correlation -0.6 ≤ r < -0.8 Strong Neg. correlation
0.8 ≤ r < 1.0 Very strong positive correlation -0.8 ≤ r < -1.0 Very strong Neg. correlation

Study H: COPD Vs Exposure to Biomass Smoke
98
A MATLAB Output of Data on Adjusted prevalence of Chronic Pulmonary
Obstructive Disease (COPD) and % proportion of population exposed to
Biomass Smoke in various Regions within the catchment area of a National
Referral Hospital indicated the following with r = 0.71 and P – value <
0.001.

Study H: COPD Vs Exposure to Biomass Smoke, cont;
99
From the Study above:
a) What was the study design used in this study?
b) What is the suitable measure of effect of exposure to Biomass
Smoke on adjusted prevalence of COPD?
c) With evidence from this study, describe the effect of exposure to
Biomass Smoke on adjusted prevalence of COPD.
d) (i) What term is given to the logical error of ascribing the effect
of exposure to Biomass Smoke on adjusted prevalence of COPD in
the study to an individual community or person in the community?
(ii) State the possible logical error that could be made in this case.
e) (i) State a hypothesis that could be generated from this study.
(ii) Propose a study design that can be used to test the hypothesis in
e (i) above.

END.
 Next:
1) Natural History of Disease & Disease Prevention
2) Disease Screening
4) Epidemiology of Infectious/Communicable and Non-communicable diseases
3) Epidemiological Study Designs
4) Chance/Errors, Bias & Confounding
Exploring Greater Height
For God & My Country
Salaam Alaikum
100

Clinical Epidemioogy lecture series 1-1.pdf

Recommended

Recommended

More Related Content

Similar to Clinical Epidemioogy lecture series 1-1.pdf

Similar to Clinical Epidemioogy lecture series 1-1.pdf (20)

Recently uploaded

Recently uploaded (20)

Clinical Epidemioogy lecture series 1-1.pdf