Professional Documents
Culture Documents
(EPI) 5.01 Overview On Descriptive Studies - Dr. Sison
(EPI) 5.01 Overview On Descriptive Studies - Dr. Sison
II. DESCRIPTIVE STUDIES ● 44-year-old patient with a new-onset seizure disorder after
● A descriptive study does not explain associations or vaccination against Japanese encephalitis: a case report
relationships, but seeks knowledge for better understanding ● Intratracheal keloid in a 21-month-old African American child.
→ Keloid scar formation led to the blockage of the airway
of the nature of the subject of the study to serve as basis for
some future actions, including formulation of hypothesis of ● 51-year-old woman crushed by an elephant trunk.
→ Woman rushed to the ER with multiple fractures, internal
relationship
bleeding, and injuries to the face
→ It is an inquiry into the nature of an unknown phenomenon
→ Author states hospitals and emergency physicians should be
or the occurrence of an event
equipped and ready for the most bizarre cases to manage
→ It seeks to know the characteristics of the phenomenon or
them well
event and to categorize it into some descriptive variables
● The aim of a descriptive study is to infer findings from a ● Unani treatment herbal and leech therapy saved the diabetic
sample from a reference population foot of a patient from amputation.
→ Concoction from herbal sources and leeches placed around
● “Everything starts with descriptive studies”
the wound of the patient to suck blood and necrotic tissue.
LE 2 TRANS 1 TG: Alvarez A, Alvarez P, Anonas, Apugan TH: Arboleda VPAA: Bermejo (09494741543) and Cobol (09157374720) Page 1 of 5
B. CASE SERIES → If you spent more hours studying, your GPA is also expected
● Similar in structure and form as case reports, but describes a to increase
group or set of cases, instead of a single patient
→ Case series – is an objective report of a clinical characteristic
or outcome from a group of clinical subjects with the same
or similar manifestations
→ Description
▪ A brief objective report of clinical characteristic or
outcome from a group of clinical subjects
→ Study question
▪ Report new disease or health related problem
● May be retrospective or prospective Figure 1. GPA and Hours of Studying Correlation
→ In prospective studies, individuals are followed over time and Negative Relationship
data about them is collected as their characteristics or
circumstances change ● Increase in the values of one variable is associated with
▪ Birth cohort studies are a good example of prospective decrease in the second variable
studies ● E.g. GPA (Variable 1) and alcohol intake (Variable 2)
→ In retrospective studies, individuals are sampled and → Increase in alcohol intake will show decrease in your GPA
information is collected about their past
● Usually requires a relatively small sample size
→ Usually 2 or more
▪ 10 or more according to the Powerpoint
● Reports the frequency of events or outcomes of a disease, but
does not show temporal relationships between events and
outcomes
→ Temporal relationship – “Which causes which”
Strengths and Limitations
● Control or comparison group is not included Figure 2. GPA and Alcohol Consumption Correlation
● Generalization of the results is limited because the selection of No Relationship / Non-relationship
study subjects is unrepresentative
● There is no relationship between the two variables
● This study design has case selection bias and lacks
● E.g. Socioeconomic status and honesty
statistical validity
→ Case selection bias – depends upon the investigator of
which case to study or to include in his case series
Case: Pneumocystis carinii pneumonia (PCP)
● Logical error in interpretation of statistical data, whereby Example: Correlations between meteorological parameters
inferences about the nature of specific individuals are based and prostate cancer
solely upon aggregate statistics collected for the group to which
those individuals belong
● Fallacy assumes that individual members of a group have the
average characteristics of the group at large
→ Question posed: Are these factors really affecting people?
● Statistics that accurately describe group characteristics do not
necessarily apply to individuals within the group
→ E.g. Falsely assumes that every individual´s IQ is high, since
the class average is high
Analysis of Ecological Data
● Pearson's Correlation Coefficient (r)
→ Evaluate the linear relationships between two continuous
variables
→ Provides information about the magnitude (r2) and direction
(r) of the relationship Figure 6. Relationship between heating degree days and prostate cancer
▪ Observations are collected, then Pearson's coefficient is incidence
computed to specify the nature of the relationship ● The colder the temperature, and the drier the climate in a
between the variables country, the higher the incidence of prostate cancer.
→ Use of interval variables and ratio
→ Values range from +1 to -1
Table 1. Pearson’s Correlation Coefficient Values
Pearson’s Correlation
Indication
Coefficient (r)
+1 Perfect positive relationship
-1 Perfect negative relationship
0 No relationship
Table 2. Degrees of Correlation
Pearson’s
Degree of Indication
Correlation
Correlation
Coefficient (r)
±1 Perfect correlation
Figure 7. Showing correlation denoted between temperature and incidence of
±0.50 - ±1 High Strong Correlation prostate cancer and amount of land used to grow crops and incidence of
Medium prostate cancer
±0.30 - ±0.49 Moderate
Correlation
● Furthermore, in counties with high average annual snowfall
< ±0.29 Low Small Correlation (>75 cm/yr), the amount of land used to grow crops (a proxy for
b No correlation pesticide use) was positively correlated with the incidence of
→ Measures the degree of correlation of two continuous prostate cancer
variables in an interval scale → Data analysis is done through correlation and regression
Table 3. Pearson’s Correlation Coefficient analysis
Interval Degree of Correlation ● Question generated: Do the people who have prostate cancer
0.00 – 0.25 Little or no relationship in this survey really live in those areas? They might only be
0.26 – 0.50 Fair degree of relationship there because the place has specialty hospitals
0.51 – 0.75 Moderate to good relationship → Shows that studies like this are for hypothesis generation.
There is a need to go for higher research design to know the
> 0.75 Good to excellent relationship
association
E. CROSS-SECTIONAL STUDIES
EPI Overview of Descriptive Studies Page 3 of 5
● In a cross-sectional study, the variables are all measured at ● Not useful for establishing casual relationships
a single point in time, with no structural distinction between → It does not allow us to answer the question, which came first
predictors (exposure) and outcomes (which caused which)
→ Therefore, there is no attempt to establish a temporal ● Data are particularly susceptible to distortion through the
relationship introduction of bias into the research during sampling
● It may involve the measurement of current exposure and/or questionnaire and interviewing
historic exposure → Prevarication bias
→ Example: COVID-19 ▪ Occurs when a study subject over- or under-estimates
▪ Cross-sectional analytical study outcome because of knowledge of the kind of treatment
− Distribution of people, prevalence they had received [Hiebert and Nordin, 2006]
− Associations, odds-ratio, relative risk ▪ Subject may subconsciously feel the need to provide
▪ Cross-sectional descriptive study answers pleasing to the interviewer, or may feel that there
− No associations made is some secondary gain from consciously misreporting
− 2x2 table, compute for specificity and sensitivity their clinical status [Hiebert and Nordin, 2006]
● The design is valuable for providing descriptive information ● Needs a large sample size
about prevalence – it is well suited to the goal of describing I. REVIEW QUESTIONS
variables and their distribution patterns For numbers 1 to 5, choose the appropriate RESEARCH
→ How many of the population really have the disease? STUDY DESIGN from the choices below for the given
→ Survey of the people who have the disease situations. Choices may be used more than once or not at all:
● This design may also be used for examining associations, A. Case Study
although the choice of which variables to label as predictors B. Case Series
and which as outcomes depends on the cause-and-effect C. Cross- Sectional Study
hypotheses of the investigators
D. Ecological Study
→ Example: Which is the risk factor and which is the outcome,
diabetes or hypertension?
1. In 1981, a previously healthy homosexual man presented with
▪ Depends on the researchers, must be backed up by
related literature serous fluid build-up in his lungs and recurrent episodes of
● Morbidity and mortality studies are cross-sectional studies Pneumonia. He was eventually diagnosed to have rare
Pneumocystis carinii pneumonia, a fungal infection not
Table 4. Morbidity and Mortality Studies
commonly seen in immunocompetent individuals.
Morbidity Studies Mortality Studies
Prevalence Studies Mortality Surveys 2. Also in 1981, a group of young men having sex with other men
Case series in health centers, Analysis of official mortality (MSM) in their second and third decades of life from New York
clinics, and hospitals statistics and hospital records City were suffering from Kaposi’s sarcoma, an aggressive rare
form of skin cancer. All men shared similar high-risk lifestyle and
● Cross sectional studies measure existing disease and current comparable clinical presentations.
exposure levels
3. Now in 2018, people having Pneumocystis carinii (i.e.,
● The study analyzes data collected on a group of subjects at
taxonomic reclassification as Pneumocystis jiroveci) pneumonia
one time rather than over a period of time
have been more common; thus there is a need for a study of the
→ If it is over a period of time, it is a case control study
true prevalence rate of the said disease.
General Steps in Doing a Cross-Sectional Study
4. A province found to have high consumption of fat also had a high
1. Select a sample from the population prevalence rate of cancer. Which uses the population of a
2. Collect data using standardized data collection method (e.g. specified geographic area as the subject?
questionnaire) 5. Which of the following research designs could study the
3. Analyze the collected data association of multiple exposures and multiple outcomes at the
same time?
6. The correlation between consumption of carbohydrates
and prevalence of diabetes mellitus (DM) was calculated to
have a Pearson’s r value of 0.88 with a p-value of 0.01. How
should these VALUES be INTERPRETED?
A. Carbohydrate consumption and DM have an excellent
correlation that is statistically significant
B. Carbohydrate consumption and DM have an excellent
correlation that is not statistically significant
C. Carbohydrate consumption and DM have little or no
correlation that is statistically significant
D. Carbohydrate consumption and DM have little or no
correlation that is not statistically significant
Figure 8. Schematic diagram for cross sectional studies
7. Which of the following can be a CASE REPORT?
● Identify the risk factors and diseases, not which came first A. Investigating the efficacy of Moringa oleifera
Strengths and Limitations (malunggay) cream in resolving bacterial skin infections
as compared to mupirocin cream
● Quick and easy B. Adverse events of glutathione capsules in five adult
● True rates are determined (the prevalence) women residing in an urban community
● Can study multiple exposure and multiple diseases C. The onset of generalized acneiform eruption on a
● Impractical for rare diseases teenage male given oral corticosteroids for optic neuritis
→ E.g. Doing a cross-sectional study for hypertension, a D. All of the above
prevalent illness, is more practical than doing a cross-
sectional study for penile cancer wherein only 1 out of 1000
men have it. It will be difficult to find respondents.
EPI Overview of Descriptive Studies Page 4 of 5
8. Which of the following does NOT belong to the category of
ANALYTICAL STUDIES?
A. Cross-sectional study
B. Ecological study
C. Case control study
D. Non-concurrent cohort study
1A 2B 3C 4D 5C 6A 7C 8B
II. REFERENCES
Dr. Sison’s Lecture and Powerpoint
2022A and 2022C Trans
2021B Trans
Hiebert R, and Nordin M. (2006). Methodological aspects of outcomes research.
Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3454552/.