Study_Designs in quantitative research res

Study designs

1. Study Designs

2. DEFINITION A study design is a plan to conduct a study which allows the researcher to translate a conceptual hypothesis into an operational one. It is the method of data collection with respect to time, exposure and outcome Study Designs

3. Conceptual Hypotheses .  Conceptual hypotheses  predict theoretical relationships or differences between two or more variables. They are the “educated guesses” people make about relationships between and among variables. Operational Hypothesis . An  operational hypothesis  in a research experiment clearly  defines  what the variables of interest are and how the different variables are related to each other. The  operational hypothesis  should also  define  the relationship that is being measured and state how the measurement is occurring.

4. The selection of a study design depends upon the research Objective and Hypothesis . The researcher should know and use the most appropriate study design that matches best with the objective . Study Designs

5. • Descriptive or Observational designs Experimental or interventional TYPES OF EPIDEMIOLOGICAL STUDY DESIGNS Epidemiological study designs are classified as follows:

6. • Descriptive or observational designs for generating hypothesis: – Case report – Case series – Cross-sectional studies. TYPES OF EPIDEMIOLOGICAL STUDY DESIGNS Epidemiological study designs are classified as follows:

7. Analytical or observational designs for generating/testing hypothesis: – Case control studies – Cohort studies. Analytical or experimental designs for testing hypothesis: – Randomized control clinical trials (Gold standard) – Quasi-experimental design. (Non-Randomized control clinical trials ) TYPES OF EPIDEMIOLOGICAL STUDY DESIGNS

8. Hypothesis generating study only “an association” between an exposure and an outcome can be established. Whereas on the basis of an hypothesis testing study one can say with confidence that a certain exposure causes a certain outcome. The experimental studies (randomized controlled clinical trial) are the most powerful of studies and the only hypothesis-testing studies. Hence Randomized controlled clinical trial are considered the gold standard. The observational studies are weaker studies and can only generate a hypothesis. Difference between hypothesis-testing and Hypothesis Generation

9. Epidemiological study designs are broadly divided into two main types, Descriptive Studies Analytical Studies . In Descriptive studies a researcher quantifies (in % or mean ± SD) the distribution of certain variables in a study population at a point of time. In Analytical studies (observational or experimental), the researcher tests the prior stated hypothesis. Types of Epidemiological Study Designs

10. In observational studies, the researcher merely observes what is happening or what has happened in the past and tries to draw conclusions based on these observations . In experimental studies, the investigator assigns an intervention to one of the groups. Another distinguishing feature of the experimental study is the process of randomization . Observational and Experimental Studies

11. The basic variable that defines a study design is time. If both the exposure and outcome are determined at one point of time, it is a cross-sectional (descriptive) study . If the outcome has occurred and researcher goes back from the outcome towards exposure, it is a case control study , If patients are followed from the exposure towards the outcome , then it is a cohort study or experimental study .

12. These studies may be carried out in one patient or in a group of patients. These studies describe an event or a problem with respect to time, place and person . The researcher usually does not have a hypothesis at the beginning of the study though one can formulate/generate a hypothesis based on the conclusion of the study. The three different types of descriptive observational study designs are case report, case series and cross-sectional studies DESCRIPTIVE OBSERVATIONAL STUDIES

13. It is report of a single case of disease, usually with an unexpected presentation , which typically describes the findings , clinical course and prognosis of the case. Writing of a case report is like writing a good clinical history of a patient that includes presenting features, clinical signs, lab investigations, and diagnosis after excluding a list of differential diagnosis. Case Report

14. A classical example of a case report from history is that of a congenital anomaly affecting limbs and digits from Germany in late 1959 ( Te Talidomide tragedy ). The world has never heard or seen such a unique congenital anomaly before. These are the type of cases which should be presented as a case report.

15. The thalidomide  disaster is one of the darkest episodes in pharmaceutical research history. The drug was marketed as a mild sleeping pill safe even for pregnant women. However, it caused thousands of babies worldwide to be born with malformed limbs. The damage was revealed in 1962. The thalidomide   Disaster

16. The Thalidomide Tragedy

17. When several unusual cases all with similar conditions are described in a published report, this is called a “Case Series”. A case series does not include a control group. Subsequently after the first case report of thalidomide tragedy a case series was published in 1961. The thalidomide was used for nausea and vomiting in pregnancy in that era , hence soon more such mal-developed children’ were identified becoming a basis for a case series . It was quite easy to identify the exposure now as thalidomide because all mothers with the outcome (mal-developed children) used this drug. Case Series

18. In a cross-sectional study, the data is collected at one point of time . The hallmark of such studies is that there is no follow-up. These studies are also called “ Prevalence Studies ” as they determine the burden of disease in a population, e.g . National Health Survey of Pakistan on the prevalence of hypertension in Pakistan or Pakistan National Diabetic Survey—shows Prevalence of Diabetes Mellitus in Pakistan . Cross-sectional Studies

19. A survey is a classical example of a cross-sectional study. These days surveys are also being carried out by people other than the health professionals, for example, the media . In a cross-sectional study, data on both the exposure and outcome are determined at the same time. Exposure rates are calculated in each group, thus a 2 × 2 table can be constructed. These exposure rates are compared. Cross-sectional Studies

20. Ideally, the exposure should always precede the outcome. For example, if the researcher is studying the association of uric acid level and hypertension, and on analysis fnds

21. If a cross-sectional study covers the whole population, it is called a census. A cross-sectional design is not suitable to study the association between an exposure and an outcome. While using this design it is difficult for the researcher to establish whether the exposure preceded the outcome or not.

22. Hence, cross-sectional studies are useful for determining the prevalence of a disease, but not recommended if the researcher wants to study an association between an exposure and an outcome.

23. Hence , temporal association cannot be established in such studies. Temporal association is one of the first criteria according to Hill’s Criteria to confirm an association between an exposure and an outcome. Temporal association simply means that there has to be a time period between the exposure and an outcome, and that the exposure should always precede the outcome. .

24. For instance, in the above example it has to be shown that a person had hyperuricemia initially and then after a period of time developed hypertension. Unfortunately, in a cross-sectional study the data is collected on hyperuricemia and hypertension at the same time and cannot establish which came first, the “chick” and “egg” phenomenon

25. not

26. • Multiple exposure for a single outcome can be detected • Inexpensive as compared to other analytical study designs • No need of follow-up • Takes lesser time as compare to other analytical study designs • Recommended for those problems which have a long incubation period as cancers. • Recommended for studies on rare diseases • Recommended for investigating a preliminary hypothesis. Advantages

27. Recall bias is the main problem as the “cases” will be more likely to recall the past exposure. Similarly, if the researcher is working on geriatric patients then recall bias can be problematic both in cases and controls as the respondents might not have good memory due to old age . For example, in a study looking at the association of being a cigarette smoker for ten years and development of lung cancer, some participants may have difculty in recalling whether they have been a cigarette smoker for ten years or not. • Selection bias is another problem if the cases and controls are not properly selected. Here are two examples of selection bias in two studies carried out at two leading tertiary care centers of the world by two very eminent researchers of the time. Disadvantages

28. Prospective Cohort Study In prospective cohort studies the investigators conceive and design the study, recruit subjects, and collect baseline exposure data from all subjects, before any of the subjects have developed an outcome of interest Cohort Studies

29. Multiple outcomes to a single exposure can be detected • Incidence rates are calculated • It helps in calculating the relative risk and the attributable risk • Temporal association is best studied in prospective cohort study • It allows the assessment of dose response relationship Advantages

30. Expensive • Time consuming • Strict follow-up is required • Not suitable for diseases that have a long incubation period • Not suitable for rare diseases • Attrition (loss to follow-up) due to migration or death of the respondents. Disadvantages

31. Retrospective studies are also called historical cohort studies. Sometimes in a prospective cohort study with a long outcome for example the cigarette smoking for ten years and lung cancer study loss to follow-up, long wait for the completion of the study and fnding a funding source are issues. In order to save time and money and to complete the study in a shorter time the retrospective study is an ideal situation Retrospective Cohort Study

32. Advantages • Less expensive • Less time consuming • Follow-up data is obtained through records so ‘follow-up time’ is saved • Other advantages of cohort studies are also there

33. Disadvantages Tere is no control over the data, whatever variable information is available is there. Nothing can be done about missing data. Sometimes information on a variable of interest is not available. In a prospective cohort study, the investigators are typically present from the beginning to the end of the observation period. However, it is possible to maintain the advantages of the cohort study without the continuous presence of the investigator, or having to wait for a long time to collect the necessary data, through the use of a retrospective cohort study .

34. INTERVENTIONAL/EXPERIMENTAL STUDIES

35. The aim of the experimental study designs is to provide scientific proof of the etiological factors/risk factors. There are three main types of experimental study designs: 1. Clinical Trial or Randomized Controlled Trial with patients as unit of study. 2. Field Trial or Community Intervention Studies with healthy people as unit of study. 3. Community Trial with entire community as unit of study

36. BLINDING Blinding represents an important, distinct aspect of randomized controlled trials. Te term blinding refers to keeping trial participants, investigators or assessors (those collecting outcome data) unaware of an assigned intervention. Blinding is of three types:

37. Single-blind Here the participants do not know whether they are assigned to the study or the control group. It means that they do not know whether they are getting the new drug which is under investigation or the old conventional drug. However, only the investigator knows who is getting which drug. Tis trial helps to overcome subject variation.

38. Double-blind Here neither the investigator (doctor) nor the participant (patient) knows the group allocation and treatment received. However, the statistician knows it. Te drug is coded before handing over to the doctor. Usually this trial is in practice.

39. Triple-blind It goes one step further. All the participants, the doctor and the statistician are unaware (blind) of the group allocation. Only the principal investigator is aware of the group allocation and the treatment allocation

40. Thanks Questions