How-to-Conduct-Cross-sectional-Study by PAFP

Download as pdf or txt
Download as pdf or txt
You are on page 1of 71

Noel L.

Espallardo, MD, MSc, FPAFP

Current Affiliation
• Member, Board of Director Philippine Academy of Family Physicians
• Chair of Research and Publications Committee, PAFP
• Editor-in-Chief, Filipino Family Physician Journal

Education and Training


• Doctor of Medicine at the Faculty of Medicine and Surgery of the
University of Santo Tomas
• Master of Science in Clinical Epidemiology at the University of the
Philippines College of Medicine
• Residency in Department of Family and Community Medicine, UP-PGH
How to Conduct a
Cross-sectional Study and
Use Epi-info
Noel L. Espallardo, MD, MSc, FPAFP
Outline

• Define cross-sectional studies


• Purpose
• Advantages and Limitation
• Steps in conducting cross-sectional studies
• Developing questionnaire
• Demonstrating use of MS Excel and Epi-info
• Writing the results
Observational Studies

• Four types of study designs


• Series of cases
• Cross-Sectional
• Case-control
• Cohort Studies

• Most frequently used is the cross-sectional study

Zangirolami-Raimundo et al, 2018


Cross-sectional Study

Subjects (random selection) Data collection/Questionnaire Results

Information
at certain
point in time
Start Future

• Cross-sectional studies determine the frequency (or level) of a particular


attribute, such as a specific risk factor, disease or any other health-related event,
in a defined population at a particular point in time
o Can be done by using a survey or analysis of an existing data
Purpose of Cross-sectional Study
• Cross-sectional studies are • Relevance
typically used to determine the • Burden of disease for society
current status or “situation” • Establish utilization of health
services
• Prevalence - the proportion • Guide allocation of health
in a population that at the resources
time of the study has the
disease, risk factor etc.
Purpose of Cross-sectional Study
• Assessing practices, attitudes, • Measure the difference in status
knowledge and beliefs “before” and “after”
• Basis for designing
appropriate public health
measures
• Useful for planning health
interventions
Purpose of Cross-sectional Study
• Establish association between a
risk factor and a disease or
outcome of disease
Purpose of Cross-sectional Study
• Validation of diagnostic tests
• This can be done by comparing the result of an instrument with a
reference standard both done cross-sectionally in the same
population at the same time

Positive Malignancy Negative Malignancy


(Biopsy) (Biopsy)
Probably Malignant
DRE 45 10
Not Malignant
DRE 15 50
Descriptive vs. Analytical
• Descriptive - provide estimates • Analytical - assess associations
of prevalence of disease, traits between different variables i.e.
such as smoking behavior, association between a risk factor
people’s attitudes, knowledge or and disease outcome.
health behavior.
Advantages and Limitation
• Advantages • Limitations
• Faster and are inexpensive • Difficult to derive “causal
relationships”
• Conducted either before
• No temporal relation (reverse
planning a cohort study or a
causality)
baseline in a cohort study
• not sufficient to show disease trends
• impossible to apply if the disease or
risk factor or both are rare
Setia MS, 2016
Steps in Conducting Cross-sectional Study
• Step 1 - Establish the Purpose or Objective
• Step 2 – Develop your Objectives
• Step 3 – Develop a Hypothetical Results (Dummy Tables)
• Step 4 - Define the Population
• Step 5 - Develop the Data Collection Method
• Step 6 – Collect the data
• Step 7 – Analyze the Data
• Step 8 – Write the Report
Steps in Conducting Cross-sectional Studies
Step 1 - Establish the Purpose or Objective
• Before proceeding to conduct a cross-sectional study, start by asking
yourself a few important questions:
• What do I need to know?
• Why do I need to know it?
• What will happen as a result of this study?
• Can I get the information from existing sources instead of
conducting a survey?
Steps in Conducting Cross-sectional Studies

• Which of the following you are aiming to measure:


• Presence of disease or risk factor
• Knowledge, attitudes, behaviors and practices
• Perceptions of knowledge, attitudes, behaviors and practices
Association of risk factors to development of disease
• Validation of test
Diem KG, 2016
Steps in Conducting Cross-sectional Studies
Step 2 – Develop your Objectives
• The general objective states the general purpose or goal of the study
i.e. a general statement of what the results of the study can
contribute to health care.
• Develop the specific objectives using SMART criteria
• Specific
• Measurable
• Attainable
• Relevant
• Time bound
Steps in Conducting Cross-sectional Studies
Step 3 – Develop a Hypothetical Results (Dummy Tables)
• Often it is useful to construct a hypothetical report or dummy table
before developing the data collection.
• This will make you focus on what you need to know and what data
you need to collect
• This will also help you develop your sampling frame, sample size and
method of sampling
Hypothetical Results (Dummy Tables)
Mean/ Percent/ Objectives
Frequency SD 1. Determine the prevalence of herbal
Mean Age medicine use in family practice.
Sex 2. Determine the difference in herbal
Male medicine use among hospital and
Female clinic-based family physicians.
Practice
Hospital
Clinic Herbal Non-prescriber
Herbal Prescriber Prescriber
Regular Clinic-based
Seldom Hospital-based
Steps in Conducting Cross-sectional Studies
Step 4 - Define the Population • Target or theoretical population
• Sample or study population
• Entire population from national
registers or on a sample
• Selection bias is an issue in
cross-sectional studies
Kesmodel, 2018
Population and Setting
• Define the study population
• What or who should be the study population
• Inclusion criteria
• Exclusion criteria – those who fit the inclusion but should be
excluded because the data they might contribute will not be
accurate

• When and where should the study population be recruited


• How should the study population be selected
Inclusion and Exclusion Criteria
• Inclusion Criteria
• “Patients consulting for hypertension at the Family Medicine Clinic of the
Philippine General Hospital between the period January 1 to December 31,
1998, will be recruited for the study. To be included they should fulfill the
following criteria:
• Age 25-55 years old
• Systolic blood pressure equal or more than 140 mmHg
• Diastolic blood pressure equal or more than 90 mmHg
• No history of myocardial infarction or stroke

• Exclusion criteria
• Those with psychological or mental health problems
• With history of allergy or any adverse reaction to the study drugs
Bias if Population is Wrong
• Accuracy is the very close • Precision is the variation that a
estimate to the TRUE value in a result would have if it were
population calculated from several samples
• It depends on • Confidence interval (CI) –
• Representativeness of the narrow interval more precise
sample population • It depends very much on the
• Reliability of the data size of the sample
collected

To avoid these biases, sample size computation and


random selection are very important!
Sample Size Computation Using Epi-Info
• A practice-based resident wants to
study herbal medicine use among the
members of the PAFP chapter in his
province.
• In the last QA committee activity,
based on review of medical records
herbal medicine was prescribed in
20% of patient with acute bronchitis.
Sample Size Computation Using Epi-Info
Sample Size Computation Using Epi-Info
Sample Size Computation Using Epi-Info
Sample Size Computation Using Epi-Info
Selection of Sample Population

• The best approach is to use random sampling.


• Chance determines who will be included in the sample
• Sampling frame, i.e., a complete list of the study population
• Simple random (pull names from a hat)
• Systematic random (i.e. every 5th name)
• Convenience sampling

Diem KG, 2016


Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Random Sampling Using MS Excel
Steps in Conducting Cross-sectional Studies
Step 5 - Develop the Data Collection Method
• Survey questionnaire
• Ask the same question in the same way to all respondents
• Can be face-to-face interview, telephone, email or web-based
• Challenge is to avoid information bias
• Pilot test your questionnaire
Tips in Developing Questionnaire
• Title the questionnaire
• Include simple instructions
• Use plain language and be direct
• Be brief, focus on “need to know” questions and minimize “nice to
know” information
• Start with non-threatening questions and most important questions
• Make sure answer choices correspond to the questions
• Balance the “negative” or “low” answer choices with “positive” or
“high” choices
Steps in Conducting Cross-sectional Studies
Step 6 – Collect the data
• Face-to-face survey data collection yield the best results
• Certain they are the actual respondent
• Explain the survey, the questions and choices
• Electronic or web-based surveys are also becoming popular
• Response rate of 60% to 80% should be the target
Steps in Conducting Cross-sectional Studies
Step 7 – Analyze the Data
• Descriptive statistics counts,
percent, means and standard
deviation
• Analysis of association can be
done if response rate is high
o Divide the subjects according
to the risk factor and outcome
into four distinct groups.
Zangirolami-Raimundo et al, 2018
Statistical Treatment for 2x2 Tables
• Chi-square and Fisher exact test
• Odds ratio
• Pearson or Spearman correlation
• Sensitivity, specificity, predictive
values and ROC curve

Zangirolami-Raimundo et al, 2018


Questions or Clarification
• Let’s entertain 2-3 questions or
clarifications
Statistical Analysis Using Epi-Info
• A practice-based resident wants to
study herbal medicine use among the
members of the PAFP chapter in his
province.
• He wants to see if members who are
hospital-based practice or clinic-based
practice increase or decrease the use
of herbal medicine
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Dummy Tables
Mean/ Percent/ Herbal Non-prescriber
Frequency SD Prescriber
Mean Age Clinic-based
Sex Hospital-based
Male
Female
Practice
Hospital
Clinic
Herbal Prescriber
Regular
Seldom
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Dummy Tables
Mean/ Percent/ Herbal Non-prescriber
Frequency SD Prescriber
Mean Age 48.63 SD 12.23 Clinic-based
Sex Hospital-based
Male
Female
Practice
Hospital
Clinic
Herbal Prescriber
Regular
Seldom
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Dummy Tables
Mean/ Percent/ Herbal Non-prescriber
Frequency SD Prescriber
Mean Age 48.63 SD 12.23 Clinic-based
Sex Hospital-based
Male 62 44.29%
Female 78 55.71%
Practice
Hospital 63 45.00%
Clinic 77 55.00%
Herbal Prescriber
Regular 77 55.00%
Seldom 63 45.00%
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Statistical Analysis Using Epi-Info
Dummy Tables
Mean/ Percent/ Herbal Non-prescriber
Frequency SD Prescriber
Mean Age 48.63 SD 12.23 Clinic-based 56 (72.73%) 21 (27.27%)
Sex Hospital-based 21 (33.33%) 42 (66.67%)
Male 62 44.29% *Odds ratio = 5.33; 95% CI, 2.58 and 11.01)
Female 78 55.71%
Practice
Hospital 63 45.00%
Clinic 77 55.00%
Herbal Prescriber
Regular 77 55.00%
Seldom 63 45.00%
Questions or Clarification
• Let’s entertain 2-3 questions or
clarifications
Steps in Conducting Cross-sectional Studies
Step 8 – Write the Report
Writing the Report
• Title – PIO and study design
• Abstract – Background, Objectives, Design, Results, Discussion, Conclusion
• Introduction
• Define disease
• Prognostic factors that affect the disease outcome
• Differences of the local setting from previous study
• Pre-specified hypothesis on the prognostic factor in the local setting
• Objectives - state general and specific objectives
Writing the Report
• Methodology • Results
• Study design • Narrative data
• Subjects and setting • Participants
• Instruments and • Outcome
Questionnaire • Main results
• Design and development
• Describe the questionnaire
• Pilot testing
• Sample size
• Statistical Method
Figures and Tables in Results Section
Writing the Report

• Discussion
• Summarize key results in relation to the study objectives
• Interpretation of results, implication to family and community
practice
• Discuss similarity or difference of the results to other studies and
generalizability of the results to other population group
• Discuss the limitation and potential bias of the results
• Conclusion and recommendations
Questions or Clarification
• Let’s entertain 2-3 questions or
clarifications
Thank You
• Provide feedback after the session to get your PAFP CPD Units by clicking the link
posted in the Chat box
Thank You

You might also like