COMMUNITY DENTAL HEALTH

1
COMMUNITY DENTAL HEALTH
2
STATISTICS

• Statistics is the field of study which concerns
  itself with the art and science of data analysis
• Planning, collecting, organizing, analyzing,
  interpreting, summarizing and presenting the data
• Statistics, when used in the plural form, refers
  to the specific bits of data which either have
  been or are about to be gathered.

3
STATISTICS

• Foreign Language
• Special meaning for words like mean, regression,
  normal, confidence, correlation, population,
  discrete, conditional, union, posterior,
  hypothesis etc., etc., etc.
• Logic related to statistics more than math. (H.S.
  Algebra) computers
• Complex and demanding subject area

4
INTRODUCTION TO BIOSTATISTICS

• Biostatistics
• The mathematics of collection, organization and
  interpretation of numeric data having to do with
  living organisms.
• Techniques to manage data
• Descriptive
• Inferential

5
INTRODUCTION TO BIOSTATISTICS

• Uses for data (To name a few)
• Designing a health care program or facility
• Evaluating the effectiveness of an ongoing
  program
• Determining needs of a specific population
• Evaluating the accuracy of a journal article

6
EPIDEMIOLOGY

• The scientific study of factors that influence
  the frequency and distribution of disease in a
  population.

7
METHODS OF MEASURING ORAL DISEASE

• Counts
• A simple number of cases of occurrence
• Useful when there is a low prevalence
• e.g. 12 cases of oral cancer
• Proportions
• A count can be turned into a proportion by adding
  a denominator thus determining prevalence
• e.g. 12 cases in a population of 1,500 students
• Does not include a time dimension thus includes
  new cases as well as longstanding ones

8
METHODS OF MEASURING ORAL DISEASE

• Rates
• A proportion that uses a standardized denominator
  and includes a time dimension
• Types of Rates (As applied to Biostatistics)
• Morbidity Rate
• The proportion of people ill with the disease
  over a specified time span
• formula of new cases /100,000 people / year
• e.g. 12 / 1,500 / 2000

9
METHODS OF MEASURING ORAL DISEASE

• Mortality Rate
• The proportion of people who die from the disease
  during a period of time
• formula of deaths / 100,000 people / year
• e.g. 8 / 1,500 / 2000
• Case Rate
• Frequency of occurrence of the condition /
  disease
• formula of occurrences / of births / year
• e.g. 1 / 700 / 2001 (Cleft Palatte
  Cases)
• n.b. rates can be converted into percentages
  

10
INDEXES (INDICES)

• An index is a measure of quantification of
  epidemiological data
• A numerical value on a graduated scale
• Scores correspond to specific criteria
• Have definite upper and lower limits
• Examples
• DMFTs caries activity best known
  irreversible
• RCI root caries - irreversible

11
INDEXES (INDICES)

• GBI Gingival Bleeding reversible
• CPITN Community Periodontal Index of Treatment
  Needs
• DFI Dental Fluorosis
• Note
• No generic, all purpose scale
• Depends on the reason for using that measure, how
  to handle it reliably and what you want to
  demonstrate

12
DENTAL HEALTH INDICES

• Dental conditions readily lend themselves to
  study because we have specific tools for speed
  and accuracy of measurement.
• Index Properties
• Clear, simple, objective
• Valid measures what it is supposed to
• Reliable consistent on repetition
• Quantifiable data can be analyzed
• Sensitive can detect small shifts in either
  direction
• Acceptable not painful or demeaning to the
  subject
• Clinically significant and meaningful

13
CARIES ACTIVITIES INDICES

• DMFT decayed, missing, filled permanent teeth
• deft primary teeth
• Each tooth must have a score but only one (DMF or
  sound)
• Recurrent caries decayed (D)
• Missing teeth extracted or due to be extracted
  due to caries
• Teeth not deemed as missing unerupted,
  congenitally absent, accidentally lost or
  extracted for ortho. Purpose
• Third molars not scored
• DMFT and deft scores are objective thus require
  high agreement between examiners.
• DMFS and defs (surfaces) are more subjective thus
  less reliable.

14
FACTS ABOUT DATA

• Two types of data
• Qualitative labels used to identify an item when
  it cannot be numerically identified.
• e.g. marital status, car colour, occupation
• (attributes)
• n.b. has absolutely nothing to do with the
  quality of the data
• Quantitative characteristics that can be
  expressed numerically. Any mathematical
  manipulation that is carried out on them will
  have meaning.
• e.g. height, length, volume, number of DMFTs
• (variates)

15
FACTS ABOUT DATA

• Data Set
• Relates to a given group of data
• Generally denoted with brackets
• e.g. Q 17, 15, 18, 13, 12
• Data Point
• A single observation in a data set
• e.g. 15 is the second data point in the above
  data set
• Data is Plural
• Datum is singular

16
FACTS ABOUT DATA

• Raw Data
• Data still in the form that it was when
  originally gathered.
• e.g. A 14, 11, 17, 9, 12
• Rank Ordering
• Rearranging data in order usually ascending
• e.g. A 9, 11, 12, 14, 17

17
DATA MANAGEMENT

• Grouping data to make it easier to understand.
• Descriptive Technique
• Used to describe and summarize a set of numerical
  data
• Tabular and graphical methods
• Apply to generalizations made about the group
  studied

18
DESCRIPTIVE DATA DISPLAY TYPES
An Array A group of scores arranged from lowest
to highest in value. e.g. Histology test results
24 students
19 28 30 44 41 41
25 33 39 49 42 38
26 35 41 38 33 40
30 38 44 31 36 46
Raw Data
Array 19, 25, 26, 28, 30, 30, 31, 33, 33, 35,
36, 38, 38, 38, 39, 40, 41, 41, 41, 42, 44, 44,
46, 49 / 50 total
19
DESCRIPTIVE DATA DISPLAY TYPES

• Arrays are bulky and hard to read, thus an
  alternative is
• Frequency Distribution
• An organization of scores from lowest to highest
  which includes the number of times each score
  value occurs in the data set.

20
DESCRIPTIVE DATA DISPLAY TYPES

• Frequency Distribution 3 Types
• Ungrouped
• Each possible score value of the variable being
  measured is represented in the display and the
  frequency of occurrence of the value is recorded.
  Sample

21
DESCRIPTIVE DATA DISPLAY TYPES
Frequency Distribution Ungrouped
Score F Score F Score F
50 40 1 30 2
49 1 39 1 29
48 38 3 28 1
47 37 27
46 1 36 1 26 1
45 35 1 25 1
22
DESCRIPTIVE DATA DISPLAY TYPES
Scores Grouped Cumulative
16-20 1 1
21-25 1 2
26-30 4 6
31-35 3 9
36-40 6 15
41-45 7 22
46-50 2 24
2. Grouped Frequency Distribution When a broad
range of values on the measurement is possible
(i.e. gt 30), the range is collapsed by grouping
scores together into smaller value ranges.
23
DESCRIPTIVE DATA DISPLAY TYPES
Scores Grouped Cumulative
16-20 1 1
21-25 1 2
26-30 4 6
31-35 3 9
36-40 6 15
41-45 7 22
46-50 2 24
3. Cumulative Frequency Distribution Used with
score groupings where the frequency of any one
group includes all instances of scores in that
group plus all the groups of lower score values.
24
GRAPHS AND TABLES

• Histograms
• Polygons most frequently used
• Bar graphs
• Pie charts

25
PRINCIPLES FOR CONSTRUCTING GRAPHS AND TABLES

• (Course supplement Pages 6, 7, 8)
• Items in separate columns should be clearly
  defined and the units of measure of the
  observation included
• A suitable descriptive title should define the
  contents as a whole
• Rate statistics clearly stated (per 100 or per
  1,000)
• When possible and practical, frequency
  distribution should be in full

26
PRINCIPLES FOR CONSTRUCTING GRAPHS AND TABLES

• 5. When using rates or proportions, include
  numbers of observations
• 6. Clearly state when using percentage
• 7. Do not include too much on the same table
• 8. If observations are excluded, give reason and
  criteria

27
GRAPHING TECHNIQUES

• Descriptive data in pictorial fashion as a graph
• Y Axis (Ordinate) vertical axis
• Represents frequency of occurrence
• Represents score value
• X Axis (Abscissa) horizontal axis
• Represents scale of measurement of the
  characteristic of the sample
• Indicates the variable or group studied

28
FREQUENCY HISTOGRAM

• See course supplement page 8.
• A histogram is a graphical method for variate
  (quantitative characteristic) data. Note that
  there is no space between the vertical bars.

29
FREQUENCY POLYGON

• See course supplement page 9.
• A line graph created by joining the frequency /
  scale value coordinate points for each value in
  the scale represented. Used for variate data.

30
BAR GRAPH

• See course supplement page 10.
• 2-dimensional pictorial display of attribute data
  that are discrete in nature
• Bars do not touch

31
CENTRAL TENDENCY

• Term in statistics that describes where the data
  set is located.
• Measures of Central Tendency
• Used to describe what is typical in the sample
  group based on the data gathered.
• Three Main Indicators
• Mean
• - Median
• - Mode

32
CENTRAL TENDENCY

• Mean arithmetic average of scores
• Mean symbol is ( x )
• Scores are all added then divided by the number
  of scores.
• The most common measure
• Data set 3, 7, 9, 4, 9, 16 48 / 6 8

33
CENTRAL TENDENCY

• Median
• Is the point that divides the distribution of
  scores into 2 equal parts 50 / 50
• With odd set of numbers, median is the datum in
  the middle
• i.e. 3, 7, 2, 5, 9 rearranged to 2, 3, 5, 7,
  9
• median 5
• With even set of numbers, median is the average
  of the two middle values
• i.e. 4, 7, 1, 3, 8, 2 rearranged to 1, 2, 3,
  4, 7, 8
• 3 4 7 / 2 median 3.5

34
CENTRAL TENDENCY

• Mode
• Is the most frequently occurring score in a
  distribution
• i.e. 4, 3, 4, 9, 7, 2 mode 4
• i.e. 3, 8, 4, 2, 4, 9, 7, 4, 9, 1, 9
• bimodal data set 4 and 9

35
QUESTIONS

• Determine the mode, mean and median for
• Survival time, in months, for 10 patients
  following a new cancer treatment
• 24, 8, 12, 3, 20, 18, 24, 19, 27, 25
• Salaries of 7 dental hygienists and 2 dentists in
  a productive office
• 88,500 36,500 28,300
• 80,000 34,000 28,300
• 41,000 32,000 28,300

36

• If a statistician had her hair on fire and her
  feet in a block of ice, she would say that on
  the average, she felt good.
• What is she referring to?
• What is she ignoring?