Health Statistics Course

Generality
Statistics is the science of synthesizing and analyzing data that exhibits random variations.
(Last 1988). Obviously, statistics is a very important tool in epidemiology. It allows:
to organize the disparate data from individual observations,
- to clearly describe the phenomena using parameters in the populations from which they come
observed samples,
to compare these parameters among several populations,
to predict the likelihood of events occurring.
The set of reasoning methods that deals with enumeration, organization, and representation.
class graph, to interpret them.
Biostatistics is a "method of reasoning that allows for the interpretation of the kind of data very
particular, which are notably encountered in the life sciences, whose essential character is variability
(D. Schwartz).
There are several types of statistics:
Health statistics,
Demographic statistics
AVP Statistics
Financial statistics
Biostatistics
Agricultural statistics, livestock, tourism.......

1.1: Health statistics:

It is a branch of statistics that applies to issues related to public health. It focuses on
to all the data necessary for health administration and includes not only the Statistics of the
morbidity (diseases) but also demographic statistics (Population), as well as information
digital and administrative related to health services.
1-2: Health indicators:
These are variables that help to measure changes, they provide an idea about a given situation.
Indicators help us in planning, programming, and decision-making.
They are also a parameter that provides one or more pieces of information about the levels, the trends,
progress, the evolution... of a situation in relation to the objectives and/or targets set.
The indicators are developed once the objectives and activities of a program are defined.
Generally, it is preferable to have several indicators to capture the multiple dimensions of programs and
accurately reflect the objectives and activities implemented.
Indicators can be expressed in different ways: Quantitative or qualitative.
Selection criteria for an indicator:

Corresponds to the objectives of the program,

Possibility to collect it
Measurable in time
Simple and easy to interpret
Compatibility with other national and international indicators
Acceptable cost for measuring the indicator
 Characteristics of an indicator

Validity: it must measure the state or phenomenon for which it was chosen
Reliability: it must produce the same result when measured under the same conditions. It is based on the
measurement method used and on the person
Specificity: it must exclusively measure the state or phenomenon for which it was chosen.
Sensitivity: it must reflect the changes in the state or phenomenon being studied.
Compatibility: it must be measurable using standard definitions and references.
Economy: the costs of its measure must remain reasonable.
Feasibility: it must allow for the effective calculation of the desired measurements..

II- INTERESTS OF STATISTICS

Health statistics is an essential discipline in health planning. It is an essential means.

to understand the health situation and track its evolution; indeed, knowledge of the importance and of the
distribution of various health issues allows for better organization of health policy and to use effectively
rational resources that are most often limited.
2-1: Objectives of health statistics:
Health statistics allows to:
Identify the main health issues,
Establish the prioritization of issues and the actions to be taken,
Assess the level of health,
Determine the needs of the population,
Programmer, implement and evaluate health programs,
Evaluate the organization and functioning of health services,
Allows for the development of medical and health research,
Allows for the dissemination of information to the public and decision-makers.

2-2: Role of health statistics:

Health statistics allow:
the collection of health information,
the analysis and interpretation of the results.

III - STATISTICAL VARIABLES

3-1: Variability:
Variability is an essential characteristic of living beings and therefore of human beings in particular. For example,
Certain characteristics of the human being, such as weight or blood sugar levels, vary from one individual to another or from
childhood to adulthood, sometimes even from one moment to another during the day. The presence of an illness can
also explain the variability of a characteristic. For example, during certain diseases, the number of
Red blood cells circulating per unit of blood may be lower, indicating anemia. Nevertheless, the subjects
Those affected by this disease will not all have anemia, while some individuals free from this disease may have it.
to have anemia
3.2 : Variable :
A variable is a characteristic for which different values can be observed within a group of subjects.
A variable can be of 2 types:
* Quantitative values: What we measure
 Qualitative values: What we observe
3-2-1: Quantitative variables represent the measurement of a quantity. We distinguish:
3-2-1-1: continuous quantitative variables:
The variable can take several values, potentially an infinity, upon which one can calculate an average.
When rounding the obtained value, we say that we discretize this continuous variable, because we impose on it to
take certain values
Examples:
Variable Value
Weight 56.3
Size 1.76
T. A 14.9

3-2-1-2: Discrete quantitative variables:

These are discrete numeric values (which can only take integer values).
Examples:
number of children in a family.
Parity: 5 deliveries
Vaccination reminder: 3
2-2-1-3: Temporal quantitative variables:
These are particular quantitative values that use units of time measurement:
Exemples : - Seconde, minute, mois, type date ……
3-2-2: Categorical or qualitative variables:
These are variables that do not have numerical values. A so-called categorical or qualitative variable is a
characteristic having a certain number of exhaustive and mutually exclusive categories or modalities:
exhaustive because all possible modalities are mentioned, mutually exclusive because each individual can be
classified in one and only one category. We distinguish:

3-2-2-1: Ordinal variables:

They express themselves in classes that can be ordered according to a scale of values... They are variables whose
Classes cannot be hierarchized: They are named but not ordered.
Examples:
Variable Class
Blood type A, B, O, AB
Civil status Single, married, divorced ...

3-2-2-3: The binary variables

This is a particular type of nominal qualitative variables that can only take two values; they
are also called:
Dichotomous variables: because they divide the population into two parts
Boolean variables that can take the value True or False
Bernoulli variable coded as 1 and 0 respectively
 Examples:
Variable Class
State of health Sick hello
Sex Woman Man
Smoking Smoker Non-smoker
Vaccination status Vaccinated Unvaccinated

IV - Methods of developing statistics:

4-1: Collection and organization:
To plan, schedule, manage training on an ongoing basis, and implement health policies
to proceed with the integration of the various programs into the overall health system, it is necessary to:
Access to information: The methods used allow for gathering and analyzing the various
data
Routine system: It is the set of informational supports (registers, reports) that allows
collection of data available for management, planning, monitoring, and performance evaluation
care activities and health programs
4-2: Basic documents:
The results of the observations are recorded on different documents from which the data is developed.
Statistics; these documents are indeed written information. They are of different kinds:
Individual form
Collective record
Register
4-3: Processing of statistical data:
It is an operation that involves distributing the characteristics of the different statistical units studied in a certain
nombre de classes ; préalablement dans chacune des classes considérées, les différents types de dépouillement :
Manual and electronic.

V- Representation of results:
In order to easily interpret previously collected data, we need to organize it under the
in the form of tables, graphs, or diagrams. This visual representation of data must not only
facilitate the analysis of results but also enable their communication.
5-1: Tables:
A table is a representation of data, usable regardless of the nature of the variable to be represented,
quantitative or categorical. The construction of a table allows for the availability of all the data.
Presenting data in table form is particularly suitable for repeated and precise data.
The tables must include all the necessary information for their understanding without the reader having to
need to consult an explanatory text. In a table, it is necessary to indicate:
-The title,
The source of information,
Document date, row and column titles
Bibliographic references
5-2: Graphics:
The graph essentially allows you to visualize a remarkable phenomenon: contrasts or trends. The eye must
to be able to observe the changes in values on the y-axis (vertical scale) for a unit change, from
class or modality on the x-axis (horizontal scale). The choice between the possible graphs relies on
essentially on the type and number of variables to be represented.
The objective of a graph:
 Summary: It allows you to quickly capture the richness of a statistic at a glance,
Discovery: from the trend, shows accidental facts, specifies the minimum and maximum values,
Control: It reveals anomalies that quickly stand out.
Comparison: The comparison that can be made of several graphs,
Allow for instructive confrontations.
There are several types of graphic presentation.
5-2-1: Histograms:
A histogram represents a frequency distribution using contiguous vertical bars. This graph does not
should not be confused with a bar chart for which space is kept between the bars; the area
Each bar of the histogram is proportional to the value it represents.
This representation is used for measurable values that can be discrete or continuous.

5-2-2 : Frequency polygons :

The frequency polygon is the curve that connects the midpoints of the tops of the rectangles of the histogram. The
the term is general, as we can represent each age category either based on its size or
based on its frequency.
The main interest of this type of graph is to be able to represent multiple data on the same graph.
distributions.

5-2-3: Diagrams:
The diagram is made up of different bars that have the same width, but unlike a histogram, it
there are spaces between the bars.
This type of diagram is used to represent comparative data.
The bars can be horizontal or vertical, and they can be arranged in ascending or descending order to...
facilitate interpretation.
5-2-4: Pie charts:
It is a circle divided into sectors; each sector represents a class of the variable.
The area of the sector is proportional to the frequency of the class size; the number of classes must be
low (maximum of 6).

5-2-5: Pyramid:
This particular type of graph is used to show the age and gender distribution of a population.

VI - Calculation of statistical indices

6-1: Basic measures:
6-1-1: Reports (Ratio)
A ratio is the expression of the relationship between two quantities: X/Y
The quantity X does not necessarily represent a fraction of the quantity Y that constitutes the denominator of the ratio.
Example.
In a sample of 4000 men and 2000 women
Report
M/F = 4000/2000 = 2/1 = Male to Female sex ratio = 2

F/M = 2000/4000 = 1/2 = 0.5 sex ratio F/M = 0.5

 6-1-2 : Proportion :
In a proportion, the numerator is a part of the denominator. P = a / a + b the numerator and the denominator are
of the same nature. It is expressed as a number between 0 and 1 or as a percentage
Example: In a population of 7500 children under 5 years old, it is observed that 5300 are correct
vaccinated against measles; the proportion of vaccinated children is 5300/7500 = 0.707 = 70.7% = Coverage
vaccine
6-1-3 : Taux :
A rate measures the probability of the occurrence of a given event over time. It is the expression of the form
X/Y multiplied by a constant K. A rate must always be expressed over a certain unit of time,
for a given geographical location and for a well-defined group of people.
The numerator = Number of events (deaths, illnesses, etc.) that occurred over a certain period,
The denominator = The population exposed to the risk during this period,
The constant K is a power of 10 that depends on the relative size of the quantities X and Y.
Example:
If we have a rate = 0.00036 per year K= 10000
The rate of disease A is 3.6 per 10,000 per year.
3 categories of rates are used in epidemiology:
The morbidity rates
Mortality rates
The birth rates
6-1-4: Indices:
An index is a measurement parameter used to estimate a rate in cases where the denominator of the latter does not
may be measured accurately. Thus, what is commonly referred to as a maternal mortality rate (MMR)
is actually just an index.
TMM= Nombre de décès maternels attribuables à la grossesse, Grossesses survenues au cours de la période
in the same population
In practice, it is difficult to calculate the denominator; that is why the latter is replaced by births.
declared living beings.
6-2: Health program indicators: Maternal and child health
6- 2 -1: Family planning:
Recruitment rate:
Proportion of FMAR who are opting for the first time for the use of a given contraceptive method
course of a given period.
Recruitment rate = New acceptances of a given method X 100 FMAR 15 - 49 years over a given period
UTILITY:
This rate shows the results of the efforts made in terms of motivation for recruitment.
new clients for the use of a contraceptive method during a given period.
EXAMPLE:
In 2000, FMAR number = 3540783
New acceptances of a given method = 407190
Taux de recrutement = (407190 /3540783) x 100 = 11.5%
In 2005 according to health in figures 2006 DPRF

Recruitment rate =
Couples Years of Protection (CYP):
The number of protective year couples is a theoretical measure that informs about the number of couples.
theoretically covered against the risk of pregnancy during a given year. The CYP is a number in
absolute value obtained by weighting the quantities of each contraceptive product distributed to clients
year course by a conversion factor.
Calculation formula:
CYP= Pill/ 15 + (DIUx 3.8) + (LTx10) + Condom / 150 + Injectable/4 + (Norplant x 3.5) It should be noted that in
this formula only takes into account the quantities of contraceptive product distributed
Utility: It is used to estimate the number of couples protected annually by contraceptive methods.
The contraceptive prevalence:
It is the percentage of FMAR who use a contraceptive method at a given time; this indicator
can be calculated in 2 ways depending on the data used:
Effective prevalence: It is representative of the entire population; calculated from surveys Number of
Contraceptive prevalence rate at time t x 100, contraceptive prevalence rate at time t

Approximate prevalence: Represents only the performances of the public sector calculated at
These prevalence rates allow us to provide the % of women using modern contraception among
all married women at one point.
Example.
The results of a PANEL survey showed:
FMAR= 2481, 1247 reported that they are on contraception.
Effective prevalence = (1247/2481) x 100 = 50.3%
Population des FMAR en 2000 = 3540783, nombre de CYP= 932238
Approximate prevalence = (932238/3540783) x 100 = 27%
According to the Population and Family Health Survey (EPSF 2003-2004)
The contraceptive prevalence is 63%
The pill is the most used method 40.1%
* DIU 5.4%
Other modern methods 1.6%
In 2005 according to health in figures 2006 DPRF
Prevalence rate = 63%
6- 2 -2: Pregnancy and childbirth monitoring program:
• Example: In 2000, the total new registrations = 257527, Other CPN = 249315
Average number of prenatal visits per pregnant woman = (257527 + 249315) = 1.97257527% screening of
high-risk pregnancies (HRP):
It is the ratio between the number of screened GAR and the number of new women registered in prenatal care.
GAR detected at time t x 100
The number of new women registered for prenatal care at time t
Usefulness: It provides information on the proportion of pregnancies identified as at risk compared to all pregnancies.
consulted.
Exemple. : en 2000, Le nombre des GAR = 24350, Nouvelles inscriptions = 257527
% of screening (GAR) = (24350/257527) x 100 = 9.4 %
percentage of childbirth in monitored environments:

It is the ratio of the number of births performed in a monitored environment to the number of expected births.
NAEMST
%d' supervised childbirth= 100
NNA
NAEMST: Number of deliveries performed in a monitored environment at a time t
NNA: expected number of births
Utility:
This indicator informs us about the proportion of deliveries performed in monitored settings.
Example: In 2000
The number of childbirths performed in a monitored environment is 243,675. The expected number of births =
694539
% of births in monitored environment = (243675/694539) x 100= 35%
In 2005 according to health in numbers 2006 DPRF

% of childbirth in monitored environments = 60.8 %

According to the Demographic and Health Survey (DHS 2003-2004):
– Taux de recrutement = 69%
– % of delivery in monitored environment = 60.8%
% of delivered by a health worker = 62.6%
% of women who received at least one tetanus injection = 28%
Monitoring indicators for emergency obstetric and neonatal care (SONU)
See tables
INDICATOR
NAME OF
STRUCTURES
SOU
EFFECTIVE FOR 125,000
HABITANTS
NAME OF
STRUCTURES
ACCEPTABLE LEVEL
At least ONE SUB for 125,000 inhabitants: Total population
------------------------------- = n
125,000 H
At least ONE CENT for 500,000 inhabitants: Total population
VALUE/SITUATION
CURRENT
Total number of SOUB staff = n 1
 DEFICIT
n2 = n - n1
SOUC
EFFECTIVE FOR 500,000
HABITANTS
----------------------------- =
N 500,000 H
Total number of SOUB staffN2 = N - N1 = N 1
…
INDICATOR
AVAILABILITY
USE
QUALITY
Besoins en SOU satisfaits : Proportion de cas compliqués pris en charge dans les structures susceptibles
to ensure reference rates for obstetric cases
Reference rate of NNes
Proportion of births occurring in facilities capable of providing maternal and child health services
SOU needs met: Proportion of complicated cases handled in likely facilities
to ensure SOUC
Satisfaction of needs in cesarean sections: Cesarean section rate

Reference rate of obstetric cases

15% of births expected
Level urban national …… 52%
15% of births expected.
Cesarean deliveries account for no less than 5% and no more than 15% of expected births.
…
6- 2 -3: Program for the fight against deficiency diseases (PLMC):
Vitamin D coverage rate (first intake):
It is the ratio of the number of children under one year old who received the first dose of vitamin D to the total number.
of children under one year old
NE−1anrdc
TCVd¿ 100
NTE−1 a n

TCVd: Vitamin D coverage rate

NE-1an rdc: Number of children under one year who received the 1st dose of vitamin D
NTE-1an : Nombre total d’enfants âgés de moins d’un an
Usefulness:
This indicator informs us about the proportion of children who received the first dose of vitamin D by
report on the total number of children under one year old
Example: In 2000, the number of children under one year old who received the first dose of vitamin D = 508390
total d’enfants âgés de moins d’un an= 652817
Vitamin D coverage rate (first intake) = (508390/652817) x 100 = 78%

% of weights of children aged 12-23 months (TPE 12-23):

It is the ratio of the number of children aged 12 – 23 months weighed to the total number of weighings.

NE12−23 months
TPE 12-23¿ 100
NTp
TPE 12-23: % of weights of children aged 12 to 23 months
NE12−23 months Number of children aged 12 – 23 months weighed
NTp Total number of weighings
Usefulness:
It is used to assess the number of children aged 12 to 23 months weighed and their nutritional status.
Example: In 2000, the number of children aged 12-23 months weighed = 1123211 the total number of weighings =
5048500
TPE 12-23 = (1123211/5048500) x 100= 22.3 %

Rate of detected malnourished cases (RDMC):

It is the ratio of the number of children screened for malnutrition to the number of children under one year of age.
N CDM
TCMD¿ 100
NE−1 a n
TCMD : Taux des cas malnutris dépisté
NCDM: Number of detected malnutrition cases in one
NE -1 year: Number b of children under one year
Utility: It is used to assess the nutritional status of children under one year old and allows for regular monitoring of
program.
Example: In 2000, the number of children under one year old = 652817 number of children screened.
malnutri =35798
TCMD= (35798/652817) x 100= 6%

6- 2 - 4: Diarrheal Disease Control Program (DDCP)

Acute dehydration rate (ADR):
It is the result of dividing the number of cases B+C by the total number of diarrhea cases Seen (A+B+C):

NCB+C
DHA¿ 100
NCTDA+ B+C
DHA: Acute dehydration rate
NC B+C: Number of cases B+C at time t X
NCTDA+B+C: Total number of diarrhea cases observed (A+B+C)
Utility: This indicator provides information on the degree of severity of the disease in the affected individual as well as the % of
children requiring effective care.
Example: In 2000, the number of cases of DHA (B+C) = 23155, the total number of diarrhea cases seen (A+B+C) =
922035
DHA= (23155/922035) x 100= 2.5 %
Reference rate of diarrhea cases (RRDC):
It is the ratio of the number of referred diarrhea cases to the total number of diarrhea cases seen.
NCDR
TRCD¿ 100
NCDV
TRCD: Reference rate for cases of diarrhea
NCDR: Number of diarrhea cases referred at time t
NCDV: Total number of diarrhea cases seen
Utility: It provides information on the proportion of cases referred to a higher level of care.
6- 2 - 5: National Immunization Program (NIP):
BCG Recruitment Rate (TRBCG): It is the ratio between the number of BCG doses administered.
to children under one year old and the total number of expected births:
NDBCGA−1 a n
TRBCG¿ *100
NTNA
NDBCGA: Number of BCG doses administered to children under one year old at time t
NTNA: Total number of expected births
Utility: This indicator shows the percentage of children who received the BCG vaccine before reaching the age of 12 months.
Example: In 2000, the number of BCG doses administered to children under one year old = 583548
nombre total des naissances attendues = 694539
AN : RBCG= (583548/694539) x 100= 84%
In 2005, according to health statistics 2006 DPRF, BCG recruitment rate = 98.4%
Vaccination coverage rate of fully vaccinated children (TCECV):(completion) : It is the ratio of
children under one year old who have received all vaccinations and the total number of children under one year old

NE−1an ARV
TCECV¿ *100
NTE−1an
NE-1an ARV : Nombre d’enfants moins d’un an ayant reçu tous les vaccins en t X
NTE-1an: Total number of children under one year old in t
Utility: It provides the percentage of fully vaccinated children, it informs about the effectiveness of the program.
Example: In 2000, children under one year who received all vaccines = 536392, the total number of children of
less than a year = 652817
AN: TCECV= (536392/652817) x100= 82.2%
Tetanus vaccination coverage rate (TCVAT): It is the ratio of women
pregnant women who have received at least the third dose of the VAT and the total number of pregnant women
NFEAR3VAT
TCECV¿ *100
NTFE
NFEAR3AVT: Number of pregnant women who received at least the third dose of VAT in t X
NTFE: Total number of pregnant women (Expected births) at t
Example: In 2000, Expected births = 694539, pregnant women who received at least the third dose of
VAT= 166519
AN : TCVAT = (166519/694539) x100 = 24 %
6- 3: Hospital indicators:
 Average occupancy rate (AOR): It is the ratio between the number of days achieved compared to the
theoretical days (litter capacity x period t in days). It informs us about the % occupancy of the
health training
Average length of stay (DMS): It is the ratio between the number of days spent and the entrants.
(in days) DMS = Days completed Entrants. It provides information on the average length of stay of patients in the
health facilities
Average of entrants: It is the ratio of entrants relative to a given period.
6- 4: Epidemiological indicators:
6- 4- 1 : Prevalence :
Prevalence is the measure of the frequency of a disease (new and old cases) at a given point in time:
Instantaneous prevalence, that is during a time interval t1-t2: Period prevalence.
Prevalence is not strictly a rate but only a proportion.
Prevalence rate = (Old and new cases in t / Exposed or receptive population) * 100
6- 4- 2 : Incidence rate :
The incidence rate measures the speed of occurrence (or the strength) of a disease in a population.
Numerator: new cases that appeared during a time interval t1-t2
Denominator: Estimation of the population at risk of contracting the disease at the beginning or at
middle of the time interval t1-t2
Incidence rate = (New cases at time t / Exposed or receptive population) * 100
6- 4- 3: Attack rate:
The attack rate is a specific form of the incidence rate calculated over a short period of time (less than
of a month). Example: - TIAC, - Epidemic episode
6- 4- 4 : Mortality rate :
It is the ratio of the number of deaths due to disease X compared to the total cases of this disease X.
Case fatality rate = (Number of deaths due to disease X at time t / Total number of cases of this disease
X at time t)*100
VII- Notions of statistical interpretation:
7-1: Measurement of central tendency:
7.1.1 : Objective :
The measures of central tendency are the values that allow us to locate the center of a
frequency distribution. We seek to identify a central tendency to describe or represent simply
and concise a set of individual values. This description allows for a quick understanding and grasping of the
position of individual observations within the group. A measure of tendency is meant to be representative
from the frequency distribution.
7-1-2: Main parameters for measuring central tendency:
The average: The average is the sum of the measured quantities divided by the number of individuals.
Example: Let there be a sample of 9 people whose weight is:
The median: the median is the value for which there are as many individuals on the left as on the right in
the sample. To determine the median of a sample or a population: One can also limit oneself to
talk about median interval.
Example: Let there be a sample of 9 people whose weights are:
45 – 68 – 89 – 74 – 62 – 56 – 49 – 52 – 63 kg
Sorted in ascending order:
45 – 49 – 52 – 56 – 62 – 63 – 68 – 74 – 89 kg
 If the number of individuals is even, we take the average of the two central values:
Generally, if n is the number of individuals in the sample, the median carries the order number in the
sequence of individuals ranked in ascending order.
The mode: It is the value of the distribution that is most frequently observed.
Example: Let the distribution of the observations xi = 3, 3, 7, 8, 12, 16. The mode of the distribution is 3.
7.2: Measures of dispersion:
The range: it is the difference between the largest value and the smallest value. But, the range is
often insufficient to summarize the dispersion of a distribution, as extreme values are undoubtedly
quite particular.
The variance
The standard deviation

Table of Contents:
OBJECTIVES
I- DEFINITIONS
1- Theoretical objectives:
Define: - Statistics
Health statistics
List the different sources of statistical data
Explain the general interest of health statistics
2- Practical objectives:
Collect and present data from a hospital and prevention service.
Analyze and interpret obstetric statistical data, proposing corrective measures
Calculate the statistical indices related to service activities
3- Course outline:
Definitions: - Statistics
Health statistics
Health indicator
Interest in statistics
Statistical variables
Methods of developing statistics
Calculation of statistical indices
Representation of results
Notion of statistical interpretation