Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Epidemiology > Epidemiology > Flashcards

Epidemiology Flashcards

1
Q

epidemiology

A

study of diseases in populations (compare between groups) (incidence and distribution)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

why do we study edipemiology

A

calculate mean of population so can infer what’s going on in patient by looking at population data - diagnosis, prognosis (how long live e.g. compare CD4 levels to population mortality)
identify risk factors
quantify incidence/prevalence so targeted intervention

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

epidemiologic approach

A

find association between exposure to variable and disease (risk factors)
is the difference real
why has is occurred (cause)
prevention and control

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

target population

A

population of interest, maybe who is at risk e.g. gender/age

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

study population

A

subset of target population that you can access e.g. women gone to doctors in 2009

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

study sample

A

subset of study population, actual number of people you study, needs to be representative

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

proportion vs ratio equation

A

proportion is a/(a+b)

ratio is a/b

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

incidence

A

number of new cases of disease in time period (rate)
(new cases)/(at risk in that time period) so measures risk

already infected or not in target population are not counted because not in ‘at risk’ group

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

prevalence

A

disease in population at any one time (snapshot), point prevalence

(cases at specific time)/(no. in population at time)

period prevalence is over defined time because hard to sample at 1 time

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

incidence vs prevalence

A

incidence is a measure of risk and prevalence isn’t

incidence is prevalence / duration of infection
prevalence is incidence x duration

(like speed distance time equation)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

risk ratio (relative risk)

A

risk of disease in exposed / risk in unexposed
(no. of disease in exposed / total exposed) divided by (no. of disease in unexposed / total unexposed)

RR 1 means same risk in both so no association
RR more than 1 means risk in exposed is higher
RR less than 1 means risk in unexposed is higher

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

odds ratio

A

odds of disease in exposed / odds in unexposed
(no. disease in exposed / no. of healthy in exposed) divided by (no. disease in unexposed / no. healthy in unexposed)

OR 1 means both same
OR more than 1 means odds higher in exposed
OR less than 1 means odds higher in unexposed

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

attributable risk

A

relative importance of different risk factors

risk in exposed - risk in unexposed
(no. disease in exposed / total exposed) take away (no. unexposed w/ diseases / total unexposed)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

population attributable risk (PAR)

A

amount of disease attributable to exposure in population, relevant to public health decisions because if risk is important or not and should intervene

AR x prevalence of exposure in population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

attributable proportion

A

proportion of disease that would be eliminated in population if risks reduced to that of unexposed, so number of lives saved

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

types of epidemiological studies

A

experimental - randomised controlled

observational - cohort, less case-control, cross-sectional, ecological (population based)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

experimental studies

A

clinical trials to observe effect independent of others, compare treatments experimentally

strongest inference because randomly assigned groups and control other risk factors
similar participants
look forward in time
conclude that difference is due to risk factor

some bias if not random groups

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

observational studies (descriptive vs analytical)

A

compare natural exposure so no randomisation so hard to identify cause/effect

descriptive - estimate w/o comparison so don’t identify cause e.g. case studies

analytical - statistical comparisons, cohort study, case-control study, cross-sectional study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

cohort study

A

best observational study
follow exposed/unexposed through time (prospective)
exposures not randomised so bias

can identify cause because temporal sequence (change occurs after risk)
no recall bias because data gathered as happens

expensive and time consuming and can’t use key tests to analyse

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

case-control study

A

compare with and without disease and who has exposed (retrospective so back in time) and weaker inference
non randomly assigned
use medical records/question
compare lifestyle and risk

cheap, quick
can’t identify cause, rely on recall

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

cross-sectional study

A

measured at 1 point in time like questionnaire/sample once

generate hypotheses but then need experiment
no time sequence so don’t know which cause/effect

quick, cheap

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

case studies

A

describes without comparison

understand disease/treatment before experiment

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

order of studies used

A

clinical observation then identify available data then case-control study then cohort study then randomised trial

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

order of studies in best inference

A

RCT (randomised controlled trial), cohort study, case-control study, cross-sectional study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

multiple causal pathway

A

risk factor may not be associated directly with disease but may need both risks or both cause but don’t need them together

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

guidelines to judging whether an association is causal (when we can’t do experiment) - Hills criteria

A 
strength of association
dose-response relationship
correct temporal relationship
independent of recognised confounders
consistency with other knowledge
biologically plausible
reversible
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

strength of association

A

high relative risk/odds ratio in a cohort/case control study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

dose-response relationship

A

disease increases as exposure increases

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

correct temporal relationship

A

if exposure before disease then can establish sequence of events from prospective study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

independent of recognised confounders

A

if known causes are accounted for and there is still a significant association

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

consistency with other knowledge

A

if same pattern in many studies

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

biologically plausible

A

coherence with biological literature and makes sense

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
33
Q

reversible

A

disease goes away if remove risk factor

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
34
Q

3 types of error

A

random error
measurement error
systematically (bias)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
35
Q

random error

A

mean not representative of population because of sampling

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
36
Q

measurement error

A

type of random error from data collection

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
37
Q

systematically (bias)

A

if allocations not random it lowers significance of associations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
38
Q

random vs bias error

A

random decreases chance of detecting difference while bias can modify association so get a wrong conclusion

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
39
Q

type I error (alpha)

A

say significantly difference but truly they are not
happens by chance
allow 5% chance they are false

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
40
Q

type II error (beta)

A

don’t detect association where there truly is one
dependent on ability of study to detect association
aim for 20% chance of error

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
41
Q

types of bias

A

selection bias
recall bias
funding bias
bias due to confounders

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
42
Q

selection bias

A

individuals not representative of population
concern for all types of studies

3 types:
case-control studies - if select on basis of exposure
clinical trials - allocation like double blind
cohort studies - don’t follow up, group less likely to come back

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
43
Q

what is done to control errors and bias

A

hypothesis and alternative hypotheses
appropriate study design and good sample
avoid confounders

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
44
Q

recall bias

A

rely on memory
may recall differently with disease
some groups less likely to report

try to ask Qs in multiple ways

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
45
Q

funding bias

A

sponsored by bias company

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
46
Q

bias due to confounders

A

another variable associated
more likely in observational studies
e.g. sex, age, race

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
47
Q

biological variation

A

diagnosis involves comparing to what know at population level - with distributions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
48
Q

types of distribution

A

bimodal distribution - 2 peaks so 2 means

unimodal distribution - 1 peak, harder to say if patient infected/not

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
49
Q

case definition (+limitations and good enough)

A

criteria used to decide if individual has disease
clinical signs, diagnostic tests, history

tests often not 100% accurate
not 1 unique criteria and sometimes no known cause
some can only diagnose after death
definitions and symptoms can chnage

but epidemiology is based on groups so some error is okay if we recognise and account for them

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
50
Q

dichotomous results

A

only positive or negative and no distribution

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
51
Q

sensitivity (Se)

A

probability of +ve test result if have disease so how good at identifying disease
(true positive / total with disease)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
52
Q

specificity (Sp)

A

probability of -ve test result if don’t have disease

true negative / total without disease

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
53
Q

test accuracy

A

probability of correct diagnosis provided by test

varies with prevalence and sensitivity and specificity

54
Q

true prevalence

A

proportion that truly infected

total w/ disease (true positive + false negative) divided by total tested

55
Q

apparent prevalence

A

proportion diagnosed with test

total w/ disease according to test (true positive + false positive) divided by total tested

56
Q

positive predictive value (PVP)

A

probability that have disease when tested +ve

true positive / (total w/ disease according to test so true positive + false positives)

57
Q

negative predictive value

A

probability that don’t have disease when -ve results

true negative / (total w/o disease according to test so false negative + true negative)

58
Q

how does prevalence link to predictive value

A

as prevalence increases, PVN decreases

59
Q

tests with continuous results

A

bimodal distribution, need to establish cut-off level to define +ve and -ve result
sometimes overlap so several cut-offs

60
Q

choosing cut-off point for tests with continuous results

A

1) test with high sensitivity - few false negatives, use when consequence of false negative worse than false positive (eradication needs to detect all disease
2) test with high specificity - few false positives, if consequence of false positive worse e.g. +ve for TB means farm depopulated
3) parallel testing - simultaneously apply both tests, considered +ve to any, increases sensitivity but decreases specificity
4) sequential (serial) testing - run 2nd test if +ve first test, consider +ve if BOTH +ve so increases specificity, decreases sensitivity

61
Q

why model?

A

prediction
put complex disease into simple components
predict size of outbreak and time-course

62
Q

heterogeneities

A

groups that don’t conform to the norm

63
Q

differentials

A

consider rate of change to calculate future scenarios
e.g. ds/dt change in S over time
if S and I rate of change both 0 then reached endemicity

64
Q

SIR

A

susceptible, infectious, recovered immune (non-overlapping and have to fit into 1 group)

dS/dt - S decreases when infection so -ve rate
dI/dt - move from S to I so +ve infection but move from I to R so minus this
dR/dt - I to R so +ve rate

S+I+R = N constant

65
Q

gamma

A

rate of infected that recover

66
Q

1/gamma

A

average infectious period

67
Q

lambda

A

rate of susceptibles getting infected
force of infection

no. contacts x probability of transmission x probability that contact is infectious (number infected / total population)

68
Q

equations for SRI differentials (rates)

A

lecture 5 top 2nd page

69
Q

beta

A

number of contacts x probability of transmission

70
Q

Ro

A

basic reproductive ratio
average number of 2nd cases produced by infected in a totally susceptible population

rate infection x infectious period
Ro = beta/gamma
needs to be more than 1 for infection to succeed in population

71
Q

Rinfinity

A

no. infected in outbreak

rise with Ro

72
Q

SIR with births

A

usually birth equals death so population size constant
slow compared to infection and recovery

Ro = beta/gama+deaths

but birth and death usually so low that get same value without including them

73
Q

asynchronous S and I graphs

A

peaks out of phase because S increased then can infect them

I lags behind S graph

74
Q

endemic

A

all rates of change are 0
so S will = 1/Ro
I will = Ro - 1

75
Q

SIS model

A

2 states - susceptible and infectious because immune for very short time so don’t count
don’t need births because susceptibles replenish themselves so endemicity
Ro = beta/gamma (same as SIR)
S* = 1/Ro
I* = 1 - 1/Ro

76
Q

S*

I*

A

during endemicity

77
Q

SEIR model

A 
4 states (added exposed - infected before infectious) so more realistic
extra dE/dt and aE rate for movement from exposed to infectious
78
Q

frequency dependent transmission

A

more people means more contact

79
Q

more complex models with heterogeneities

A

by age, risk
population behaviour determines risk
age gives time to acquire disease

80
Q

aging

A

if cohort born same time and joined endemic, can see how susceptibility changes through time

calculate average age of infection, A = L/(Ro-1)
increases with life expectancy, decreases with Ro (slower spread so older when infected)

but assortative mixing - interact with same age so need to split heterogeneity into age groups (children and adults) so add extra group to equations

81
Q

WAIFW

A

who acquires infection from whom

betaAC child infects adult
betaCA adult infects child
beta to from

assortative - diagonal terms CC and AA dominate so mix most with same age
symmetric - AC and CA usually same

total Ro calculated with different Ro’s for each transmission

term times can drive infection

82
Q

risk-structured models

A

sexually transmitted - spread through networks of contacts so risk in terms of partners
but probs not random but assortative mixing so Ro higher

83
Q

immunisation

A

only when immune response, not same as vaccines

but can’t tell who is immunised so assume when vaccinated

84
Q

basic reproductive ratio

effective Rt

A

when entire population is susceptible

Ro = beta / (g+m+d)

g=recovery
m=disease mortality
d=birth/death rate

effective Rt = bete/gmd x S/N

S=how much of the population is susceptible
N=normally ignore because look at proportions

85
Q

types of control to reduce Rt

A 
beta - social distancing and awareness
g/B/S - antivirals/lower shedding/prevent susceptible
immunisation - out S group
culling (livestock) - increase d
isolation/quarantine - increase g
treatment - increase g
86
Q

vaccination

A

elicit immune response on host

e.g. seasonal flu (effectiveness changes each year with new strain)

87
Q

critical levels of vaccination

A

level needed to eliminate/stop invading infection

if vaccination constant and same proportion of births vaccinated then proportion in recovered is R=p (p births added to recovered) and S=1-p
so if Rt=Ro x S then Rt=Ro x(1-p)
and pc = 1 - 1/Ro

vaccination threshold % depends on Ro

88
Q

what if can’t reach vaccination threshold?

A

levels of infection drops linearly (1/2 vaccinated means reduce 1/2 infection) so tells if worth vaccinating

vaccination honeymoon - goes really low after vaccine so allows S to grow so infection returns and further peaks

89
Q

targeted vaccination

A

target diff proportion of low/high risk people
example optimum 46% of population vaccinated and only 74% high risk and 40% low risk
can always at least do random vaccination but better to target
and better to over target if uncertainty

90
Q

vaccination can increase disease (not infection)

A

Rubella severe in babies of pregnant women
if vaccinate, more population is susceptible at childbearing age than w/o vaccine (curve of susceptibles drops slower with vaccine)

91
Q

isolation

A

pathogen independent (don’t need to know what sick with)
requires sufficient resources - not if high Ro
need space
need other measures

92
Q

culling

A

plant and livestock to stop transmission (some dead can transmit)
don’t cull enough - no difference
too much - worse than disease
cull infected or radial culling - larger radius more likely to kill all disease
but want to minimise farms lost so need balance

93
Q

stochasticity

A

infection is often a random chance event so look at individual level
simulation won’t provide exact answers because no 2 epidemics are the same

94
Q

2 forms of stachasticity

A

observational - noise, mis-read output, error term, no effect on dynamics of disease

demographic - all events by chance, unlikely events can happen, random noise to next generation (e.g. random rise in Ro)

95
Q

failure to invade

A

even when Ro>1 it can fail to invade by chance
Ro can randomly change even if average Ro stays the same - geometric distribution with mean Ro

possible that first one didn’t infect anyone - probability recover before infect is 1/1+Ro
or make it to first generation then fail

chance that infection dies out before epidemic is (1/Ro)^n
starts with n cases - less likely to fade out if n higher

96
Q

extinctions

A

chance events in a row

only if small population, low birth rates, low Ro

97
Q

critical community size

A

certain size to not persist and cause extinction
larger size more infected so less risk of chain of transmission being broken
also more interactions and imports of infection

98
Q

complex models of stochasticity

A

add heterogeneity for more realistic

metapopulation - diff towns with link so movement between 2 diff population

individual level - model every individual so contacts and probability of each individual getting infected

99
Q

demography

A

study of population
size dependent on birth, death, immigration, emigration
record so know size

100
Q

rise of older mother

A

increasing over 40 in UK
declining under 18
increase by 2 in 16 years

101
Q

deaths

A

death certificates with ICD-10 codes by WHO
young adults mostly from external causes so most healthy
cause of death for age group changes over time
diff in men and women

mortality measure focus on early deaths and loss of life years
LYG life-years gained
PYLL potential years of life lost
assume all years are equal

102
Q

life expectancy in UK

A

77 male

82 female

103
Q

morbidity

A

loss of healthy life

decreased quality of life

104
Q

adjusted life years

A

QALY quality adjusted life years

DALY disability adjusted life years

105
Q

DALYs

A 
disability adjusted life years
life lost from death/disease/injury
mortality stays same for DALY and LYG
disability weighting - blind life id 60% healthy so not whole year lost but 0.4
although some things worse than death

perfect healthy year reduces with age so young worth more

can’t use with children because need to ask how look after themselves and measures independence

106
Q

age weighting

A

age should be valued equally but value of life decreases with age because e.g. want to save life now because immediate impact and not in future when uncertain

107
Q

PYLL

A

rank order of causes of death

biggest impact of things that make you die early

108
Q

how decide death interventions

A

resources are limited so..
save more lives if cost the same
ones that buy most health (LYG/£ or DALY/£)
plot cost per time (lowest on right) against increase in DALYs (want top right)

109
Q

NICE

A

UK national institute of health and clinical excellence
guidelines to avoid difficult decisions

intervention less than £30,000 per DALY is cost effective

tension between NHS and pharmaceutical industry

110
Q

public health care functions

A 
health monitoring/analysis
investigate outbreaks
prevention
improve health
ensure compliance with laws
good health workforce
111
Q

immunisation in the UK

A

by JCVI (drugs by NICE)
drugs work on individual patients, vaccines protect population
Green Book - schedule for UK routine vaccinations

112
Q

aims of interventions

A

control - reduce morbidity/mortality
elimination of disease - reduce incidence to 0 in defined area
elimination of infection - reduce incidence of infection (not just diease) to 0 in defined area
eradication - permanent reduction to 0 worldwide, no transmission
extinction - none in nature or lab

113
Q

example of control intervention

A

HiB (haemophilus influenzae B) sharp decline after vaccine in under 5 yr olds, catch-up increase so introduced booster vaccine

114
Q

example of elimination of disease

A

Diptheria - reduction after vaccine, still need vaccine in case circulating because in wild and may reintroduce

Measles - MMR vaccine, 2nd dose in case not immunised, high Ro so need to vaccinate a lot (reduction in vaccines when autism link)

Polio - Sabin vaccine drops cases, target global elimination but hard, low Ro but resistant pockets

115
Q

examples of elimination of infection

A

2002 eradicated Polio in EUR (3rd region, America, West pacific)
22 cases reported worldwide in 2017

116
Q

example of eradication

A

smallpox
low Ro so less vaccination needed
ring vaccine around last few cases and 4 weeks enough time to stop transmission
no hidden infection because 80% symptoms

Dracunculiasis (Guinea worm)
dogs as reservoir so another pool of infection could recolonise humans

117
Q

HBV vaccine

A

hepatitis B
childhood infection in developing countries, adult in developed
targeted immunisation is more cost effective
screening and targeting has greatest impact
current HBV is low but not 0 so not eradicated

118
Q

HPV

A

human papillomavirus
genital warts/cervical cancer
vaccine in mid 2000s for girls

4 main strains - can’t vaccinate against all
Cervarix vaccine is bivalent
Garadasil is quadravalent
Garadasil 9 against 9 strains

now vaccinate boys so herd immunity

119
Q

mass vaccination problems with not vaccinating everyone

A

reduce in population so rarer but those not vaccinated could get infected later in life with more disease
e.g. Rubella is fine in children but problem when pregnant

120
Q

STI patterns in the UK

A 
almost everyone on average has 1 in life
Chlamydia most concern
HPV vaccine 2008
Syphilis virtually eliminated with antibiotics
Gonorrhea/Syphilis spike in 1940s
herpes increasing the most
121
Q

chlamydia

A

3.2 % females 16-24yrs
asymptomatic 70% women and 50% men so transmit w/o knowing so introduced National Chlamydia screening
PID (pelvic inflammatory disease) in 10-40% untreated causes 20% infertility and 47% ectopic pregnancy (not just in humans)

122
Q

measurement of sexual behaviour

A

interview/questionnaire subject to non-disclosure and definitions of sexual act
reported is not actual and can used PSA (prostate-specific antigen) to test semen in women in past 2 days

123
Q

risk distribution in sexually transmitted diseases

A

mean is about 1 partner change a year but variance is 5
Ro = mean + variance/mean

extremes - core group, small proportion have disproportionate effect on epidemic

124
Q

mixing patterns (who has sex with who)

A

big influence on dynamics

random - doesn’t matter how many partnerships
assortative - people with more partners sex with people with more (transmission goes up)
disassortative - people with many sex with people with few
concurrency - partnerships in parallel biggest risk (2 at same time)

125
Q

core group

A

most likely to infect and get infected

modifying their behaviour is the most effective

126
Q

frequency distribution of risk on populations

A

normal distribution

small group taking high risk=large group taking small risk (linear relationship)

127
Q

prevention paradox

A

more cases from common small risk than large rare risks

128
Q

relative risk

A

exposure and disease relationship at individual level

low risk around average (0 or below) so non-linear increase in risk with exposure

129
Q

PAF (equation lecture 12)

A 
population attributable fraction - how much disease attributable to a risk
2 curves (relative risk and normal distribution curves) multiplied to make 1 curve

most disease by exposure is just above average so pick on these people because largest group

shows how much reduction in risk if exposure reduced to ideal (but usually lots of risk factors)

130
Q

PAF equation

A

(% of population with low risk) x (relative risk at exposure (RR))
+ (% population with high risk) x (RR of this exposure)
- 1

all divided by
(% of population with low risk) x (relative risk at exposure (RR))
+ (% population with high risk) x (RR of this exposure)

131
Q

alcohol

A

common ICD code
deaths going up
mean units per week only tells average not distribution
consumption distribution graphs if max daily more than 4 for men or 3 for women so shows high exposure group
risk higher in 16-24 males

132
Q

alcohol patterns of consumption affecting risk

A

for same amount, heart risk increases by 45% if consume in less time (doesn’t apply when light-moderate is mixed with irregular heavy)

Epidemiology (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions