CONCEPT OF RISK IN

PHARMACOEPIDEMIOLOGY

RISK (1,2)
DEFINITION: Risk is a combination of the likelihood of an occurrence of a hazardous event
(or) exposure(s) and the severity of injury (or) ill health that can be caused by the event (or)
exposure(s)

Or

Risk refers to the probability of developing an outcome, regardless of severity.

RISK SCALE (1)

 According to the Council for International Organizations of Medical Sciences
(CIOMS).

1. VERY COMMON: is an event occurs ≥1 in 10 exposure.

2. COMMON: is an event occurs <1 in 10 but ≥1 in 100 exposure.

3. UNCOMMON: is an event occurs <1 in 100 but ≥1 in 1000 exposure.

4. RARE: is an event occurs <1 in 1000 but ≥1 in 10000 exposure.

5. VERY RARE: is an event occurs <1 in 10000 exposure.

FACTORS INFLUENCING RISK (2,9)

 Age

 Sex

 Physical condition
  Disease state

 Drugs

 Drug absorption pattern

 Medication adherence

 Medication error

 Lack of awareness about disease and drugs

 Social hobbits

MEASUREMENT OF RISK (1)

 RELATIVE RISK
 ATTRIBUTABLE RISK OR RISK DIFFERENCE
 ODDS RATIO
 TIME AND RISK RELATIONSHIPS

RELATIVE RISK (3,8,9)

DEFINITION: Relative risk is simply a ratio of the cumulative incidence in the exposed
group over the cumulative incidence in the unexposed group.

CALCULATION OF RELATIVE RISK (3,8)

RISK FACTOR YES NO TOTAL

EXPOSED A b a+b

UNEXPOSED C d c+d

It can be calculated by:

 RR =CI exposed / CI unexposed

RR = [a/ (a+b)] /[c/(c+d)]

INTERPRETATION OF RELATIVE RISK (1,3,8)

Relative Risk Association between Exposure and Outcome
=1 No association
<1 Negative association/Decreased risk
>1 Positive association/Increased risk

EXAMPLE: In a study (3)

LUNG CANCER YES NO TOTAL

SMOKERS 200(a) 1200(b) 1400(a+b)

NON-SMOKERS 100(c) 1200(d) 1300(c+d)

SOLUTION:

RR =CI exposed / CI unexposed

RR = [a/ (a+b)]/[c/(c+d)]

RR = 200/1400/100/1300

RR = 1.86

From the above case the frequency of lung cancer was definitely higher among smokers than
non-smokers.

95% CONFIDENCE INTERVAL (2,4,6,7,8,9)

 95% CI= [±z*SE]

95% CI= [± (1.96) b/a X (a+b) +d/c X (c+d)]

USES OF RELATIVE RISK (1, 2, 8, 11)

Relative risk can be used to measure the association between exposure and outcome in cohort
studies and clinical trials. The relative risk is not used in the context of a case-control study.

ATTRIBUTABLE RISK OR RISK DIFFERENCE (1, 3, 11)

DEFINITION: It is the difference in incidence rates of disease between an exposed
group and non exposed group.

It can be calculated by:

AR = {(Incidence of disease rate among exposed) - (Incidence of disease rate among

nonexposed)} X 100/ {Incidence of disease rate among exposed}

It indicates to what extent the disease under study can be attributed to the exposure.

EXAMPLE: In a study (3)

LUNG CANCER YES NO TOTAL

SMOKERS 200(a) 1200(b) 1400(a+b)

NON- 100(c) 1200(d) 1300(c+d)

SMOKERS

SOLUTION:
AR = {(Incidence of disease rate among exposed)- (Incidence of disease rate among
nonexposed)} X 100/{Incidence of disease rate among exposed}

AR = {(200/1400)-(100/1300)} X 100 / { 200/1400}

AR = 46.148%

 The figure in our example indicates that the association between smoking and lung
cancer is causal, 46.148% of the lung cancer among smokers due to their smoking.
 This suggests the amount of disease that might be eliminated if the factor under study
could be controlled or eliminated.

USES OF ATTRIBUTABLE RISK (1,3,8)

 Attributable risk allows the investigator to determine how morbidity and mortality are
affected by removing the exposure.

 The attributable risk becomes important for agencies attempting to appropriate

monies to reduce exposures.

 The attributable risk provides information on the type of effect that can be achieved
by decreasing or eliminating the exposure.

POPULATION ATTRIBUTABLE RISK (3)

DEFINITION: It is the total population minus the incidence of disease (or death) among
those who were not exposed to the suspected causal factor.

It can be calculated by:

Population AR = (c-b) x100/c

EXAMPLE: Lung cancer death rates among smokers and non-smokers (3)

Deaths per 100,000 person-years

 Heavy smokers 224 Exposed to suspected factor (a)

Non-smokers 10 Nonexposed to suspected causal factor(b)

Total population 234(c)

Population AR (c-b)x100/c =86%

From the above case one might expect that 86% of deaths from lung cancer could be avoided
if the risk factor of cigarettes were eliminated.

It is useful in that it provides an estimate of the amount by disease could be reduced in that
population if the suspected factor was eliminated or modified.

ODDS RATIO (1,2,3,4,8,11)

DEFINITION: It is a measure of strength of association between the risk factor and
outcome.

CALCULATION OF ODDS RATIO (3)

RISK FACTOR YES NO TOTAL

EXPOSED a b a+b

UNEXPOSED c d c+d

From the above table:

ODDS RATIO (OR) = ad/bc

ASSUMPTIONS (8)
 The disease being investigated is relatively rare.

 The case must be representative of those with the disease.

 The controls must be representative of those without the disease.

INTERPRETATION OF ODDS RATIO (8)
Odds Ratio Association between Exposure and Outcome

=1 No association

<1 Negative association/Decreased risk

>1 Positive association/Increased risk

EXAMPLE: In a study (3)

LUNG CANCER YES NO TOTAL

SMOKERS 200(a) 1200(b) 1400(a+b)

NON- 100(c) 1200(d) 1300(c+d)

SMOKERS

From the above study

 Odds ratio = ad/bc

 Odds ratio = 200x1200/100x1200

 Odds ratio = 240000/120000

 Odds ratio = 2
  People who smoke cigarettes showed a risk of having lung cancer 2 times higher as
compared to non-smokers.

95% CONFIDENCE INTERVAL (2,6,7,11)

95% CI OR= (OR) e [±z X (1/a+1/b+1/c+1/d)]

95% CI OR= (ad/dc) e [±1.96 X(1/a+1/b+1/c+1/d)]

TIME AND RISK RELATIONSHIPS (1)

 The risk should ideally be expressed as function of time

 The risk must at least be estimated for different time segments.
 In practice, this may be difficult, since the type of function may not be used, the
exposure duration in a large number of cases can be investigated
 The risk is independent of time i.e., the risk is the same for the first week, second
week etc., but this is rarely true in pharmacoepidemology.
 The risk varies with time in a number of different ways that are dependent on the drug
and/or the type of ADR that it produces.
 There are several mechanisms for the different shapes of the time risk curves.
 Early adaptation of homeostatic mechanisms may explain first –dose or early
symptoms.
 Immunological mechanisms: it takes a certain amount of time for the immune system
to become activated and to synthesis antibodies. Depletation of these individuals will
be seen as a decrease in population risk with increased exposure duration.

EXAMPLES (1)
 The use of oral contraceptives has been measured in person-year.

 The relative risk for developing breast cancer was found to increase with the duration
of exposure.

 Fibrotic reactions: liver fibrosis with methotrexate.

Risk Management Guidance's (10)

(1) Assessing a product’s benefit-risk balance.

(2) Developing and implementing tools to minimize its risks while preserving its benefits.
 (3) Evaluating tool effectiveness and reassessing the benefit-risk balance.

(4) Making adjustments, as appropriate, to the risk minimization tools to further improve
the benefit-risk balance.

 This four part process should be continuous throughout a product’s lifecycle, with the
results of risk assessment informing the sponsor’s decisions regarding risk
minimization.

THE ROLE OF PHARMACOVIGILANCE AND

PHARMACOEPIDEMIOLOGY IN RISK MANAGEMENT (10,11)
 Pharmacovigilance: It is defined as “the science and activities relating to the
detection, assessment, understanding and prevention of adverse drug reaction or any
other possible drug related problems". This includes the use of
pharmacoepidemiologic studies.

 Essential to identify and to measure ADR.

 To identify early detection and early prevention of occurrence of of ADR.

 To improve drug safety monitoring.

 To inform prescribining physicians.

 To ensure better patient care.

 Signals can arise from post marketing data and other sources, such as preclinical data
and events associated with other products in the same pharmacologic class.

 It is possible that even a single well documented case report can be viewed as a
signal, particularly if the report describes a positive rechallenge or if the event is
extremely rare in the absence of drug use.

 After a signal is identified, it should be further assessed to determine whether it

represents a potential safety risk and whether other action should be taken.

 Good Pharmacovigilance program will identify the risk and risk factors in the shortest
possible time, so that harm can be avoided or minimized.
CONCLUSION
 Risk ratio: Relationship between two risks, generally estimated in different
populations. The ratio of the risk estimated in exposed (absolute risk) to the risk
measured in the non-exposed (reference risk) represents the particular case of the
relative risk

 Relative risk is important in etiological enquiries.

 Attributable risk indicates impact of successful preventive or public health program

might have in reducing the problem.

 Odds Ratio is a measure of strength of association between risk factor and outcome,
which is closely related to relative risk.

 Time-Risk relationship demonstrates that exposure time must always be considered

and the risk should ideally be expressed as a function of time.

REFERENCES
1. Trevorm .Speight, Nicholas, T.R.Einarson, U.Bergman, B.E.Wiholm. Avery’s Drug
Treatment: Principles and Practice of Phamacoepidemiology.4th; 1997.378-381.

2. G.Parthasarathi, Karin Nyfort-Hansen, Milap C Nahata.A Text Book of Clinical Pharmacy

Practice: Phamacoepidemmiology.Chenni:OrientLongman;1st; 2004.413-415.

3.Park, J.E.Pakr’s Text book of preventable and social medicine: principles of epidemiology
and epidemiologic methods.13th; 1991.64-69.

4.Brian L.Strom.Pharmacoepidemiology: Study Designs Available for Phamacoepidemiology

Studies, Case Control Surveillance. Philadelphia: Wiley; 4th; 2004.24-25,191-195.

5.Sheehe PR.Combination of log relative risk in retrospective studies of diseases, AmJ Pub
Health 56: 1966, 1745-17450.

6.Morris JA, Gardner MJ.Statistics in Medicine: Calculating Confidence interval for relative
risks (odds ratio) and Standardised ratios and rates.1st; 1998.1313-1316.

7.Douglas G.Altman, Chapman, Hall. Practical Statistics for Medical Research. Comparing
Groups-Categorical Data. 1th; 1991.266-269.

8.Bends Waning, Michael Montagne. Phamacoepidemiology Principles and Practice: Risk

Assessment. Wiley; 1st; 65-74.

9.Bernard Begand.Dictionary of Phamacoepidemiology: Concept of Risk in

Phamacoepidemiology.London:Willy;1st ; 2000.645-678.

10.http://www.fda.gov/cder/guidance/index.htm <Good Pharmacovigilance Practices and

Phamacoepidemiology Assessment> acessed on 10-10-2009
11.www.pharmacoepi.org <International Society for Phamacoepidemiology.> accessed on
12-10-2009