research design

1. RAJESHWORI NGAKHUSHI RESEARCH DESIGN

2. Research? Research is the systemic collection, analysis and interpretation of data to answer a certain question or solve a problem.

3. Research design Decisions regarding what, where, when, how much, by what means concerning an inquiry or a research study constitute a research design pattern, scheme, or plan to collect evidence Depends on the objective of the proposed study

4. Function – To permit valid conclusion which should be justified and unbaised Blue print of study

5. Defines study type sub-type research question  hypotheses independent and dependent variables experimental design data collection methods and a statistical analysis plan

6. Research design have following parts Sampling design Observational design Statistical design Operational design

7. Sampling designs Which deals with the methods of selecting items to be observed for the study Observational design Which relates to the condition under which the observation are to be create

8. Statistical design Which concern the question of the of How the information and data gathered are to be analyzed ? Operational design Which deals with techniques by which the procedures satisfied in sampling .

9. O bservational studies Based on naturally occurring events Observes and measures the characteristics of interest to the study Record based studies

10. Does not intervene or control the factors related to exposure or outcome Descriptive studies Analytical studies

11. Descriptive study Study is designed primarily to describe what is going on or what exists Involves describing the characteristics of a particular situation, event and case in term of time, place and person

12. Does not seek explanation or causes nor tries to find which group is better relative to other Can also generate hypothesis regarding aetiology of health condition Can be designed to test a hypothesis

13. Helps to assess the type of diseases prevalent in various groups and their load in a community Provide baseline data to launch a programme and can measure achievements made

14. Also good to study relationships ( systolic and diastolic blood pressure) Can be carried out on a large or small scale Eg : Estimating prevalence of blindness in cataract cases

15. Descriptive study

16. Case study A case study is defined as the use of a single person in a research study . Generally describes features of a new disease entity In clinical medicine the characteristics of so far unrecognized illness may be documented as a case study

17. First step toward building up a clinical picture of that illness Can lead to a hypothesis

18. It allows to gain a lot of in-depth, detailed information due to the close examination of single case Disadvantage - might not be true for other

19. Case series Series of case study form case-series Objective and brief report of a clinical characteristics or outcome from a group of clinical subjects lead to the generation of hypotheses

20. Advantages: 1. Easy to write. 2. The observations can be extremely useful to other investigators.

21. Disadvantages: 1. Susceptible to many biases. 2. They are not able for conclusive decisions 3. Unrepresentativeness of subjects 4. Lack of control group

22. S urveys Community based investigation Easy, simple and inexpensive Large scale

23. Disadvantage: It's difficult to truly test the impact of certain variables when all people are doing is filling out a survey Some people might not be honest in their survey responses.

24. Analytical study Study designed to look at the relationships or associations between two or more variables EXPLANATORY STUDY To test hypothesis

25. Antecedent An antecedent is a precursor such as an exposure or a risk factor suspected to affect the disease Other terms - Cause , predisposing factor and determinant

26. Outcome Outcome could be a health state, recovery, side effect, death or any other event of interest This is a consequence and must necessarily occur after the antecedent Other terms - Effect and result

27. Prospective study Antecedents are assessed prior to the outcome Since outcome occurs after the antecedent, follow up is essential Umbrella term includes cohort, longitudinal and follow up study

28. PROSPECTIVE STUDY PRESENT Diabetes Mellitus Time Future Risk factors??? Retinopathy- Retinopathy+

29. Prospective study

30. Cohort study Word cohort has its origin in the Latin cohors Refers to a group of warriors and gives notion of a group of persons proceeding together in time( same statistical characteristic )

31. A cohort is a group of people who have something in common and who remain part of a group over an extended time Incidence study “ What will happen?”.

32. Cohort study (Follow-up Studies) Groups of people having similar attribute (cohort) with & without a particular exposure, or an entire community, or a random sample of a community are selected for the study, enrolled & followed up over time to determine their disease rates Disease rates of exposed subjects are compared with rates of an unexposed group or a group with low exposure or with registry rates

33. Advantages: 1. Opportunity to measure risk factors before disease occurs: evidence of causality 2. Can study multiple diseases outcomes 3. Can yield incidence rate as well as relative risk estimates. 4. Good when exposure is rare 5. Minimizes selection and information bias

34. Disadvantages: 1. Expensive and inefficient for studying rare outcomes 2. Often need long follow-up period and/or a very large population 3. Losses to follow-up can affect validity of findings 4. Ineffective for rare diseases 5. Expensive 6. Ethical issues

35. Longitudinal study Another version of prospective study When observation or measurements are repeatedly made at several points of time Repeated measures study

36. Longitudinal Study Multiple points of data collection from the same population Initial point of Data collection ( Present) Follow ups (Future)

37. Cross sectional May be descriptive or analytic Prevalence study/ Instantaneous study/ Simultaneous study Examine the relationship between a disease and an exposure among individuals in a defined population at a point in time. “What is happening?”

38. Antecedent and outcome are observed at the same time Also appropiate to generate hypothesis regarding aetiology Subsequently tested by a case control or a prospective study

39. Cross sectional One time data collection Data collection time

40. Advantages: 1. Useful to know the burden of a disease in a group – prevalence rate can be obtained 2. Cheap and fast 3. Useful to evaluate diagnostic procedure 4. To study common risk factors and outcomes

41. Disadvantages: Population little willing to collaborate 2. Doesn’t tell the flow of events 3. Only shows association between factor and disease studied 4. It is not useful to search causes of the outcome

42. 5. It measure at a point of time therefore mostly it is useful to study chronic diseases 6. Confounders may be unequally distributed 7. Group sizes may be unequal 8. Recall bias

43. Retrospective study First assessed outcome and antecedents are subsequently assessed Outcome is already known and antecedent is obtained either from records or enquiry

44. Cases and control are assembled Information about their past exposure or risk factors is collected

45. RETROSPECTIVE STUDY Time PAST Life style Diet Medication Family History Genetics PRESENT Disease Determining differences in:

46. Case control design Dominant format of retrospective study Subjects with or without disease are investigated for past exposure

47. Case- suffering from disease or have interested health condition Control – without particular health condition “ What happened?”

48. Advantages: 1. Allows examination of several risk factors 2. Can study long-term effects of an exposure in short period of time 3. Use fewer subjects, 4. Relatively quick and relatively less expensive 5. Suitable for rare diseases

49. Disadvantages: 1. Selection of an appropriate control group can be difficult 2. Recall bias: retrospective nature 3. Cannot tell about incidence or prevalence 4. Difficult to establish time relationship between exposure and outcome

50. Choice of the wrong control group ~ selection bias Cases may over-report past exposures ~ information bias

51. Case control study 53 Comparative study done by involving two population, i.e., case and control This permits estimation of odds ratios (but not of attributable risks) Allowance is made for potential confounding factors by measuring them and making appropriate adjustments in the analysis.

52. Basic steps 54 Selection of cases and controls Matching Measurement of exposure Analysis and Interpretation

53. Case selection Clinic/hospital records or staff Death records Special clinical exams or symptom/disease questionnaire surveys Special disease registries

54. Control selection Other hospital/clinic patients (well or will) General population registers (lists of residents in a region) People living in the same neighborhood as cases People with “other diseases” in disease registries

55. Matching 57 Defined as the process by which we select controls in such a way that they are similar to cases with regards to certain pertinent selected variables which are known to influence the outcome of disease and which, if not adequately matched for comparability, could confound the results.

56. 58 A confounding factor is one which is associated with exposure and disease, and is distributed unequally in study and control groups It also independent risk factor to produce the disease. e.g smoking works as confounding in study of the role of alcohol in duodenal cancer.

57. Measurement of exposure 59 Obtain data related to the exposure with risk factors of the disease under study. The data can be obtaining by: Interviews Using questionnaire Studying past records of cases form hospital, other health institutions and employment records.

58. Analysis 60 Odds Ratio: The ratio of probability of getting disease with that of not getting disease is Odds ratio. Ratio of the odds of developing disease in the exposed and the unexposed = a/b  c/d = ad/ bc Odds ratio is an approximation of Risk Ratio in rare diseases.

59. Nested case control design Combine cohort and case control features Since case control setup is nested within a cohort Eg : cohort of persons of age 40-44 years who are followed up for 15 years for development of cataract Case- develop cataract Matched control – did not develop cataract

60. Cross sectional

61. Basis Cohort Cross sectional Case control Main antecedent Outcome Recruitment of subject Definition of case Known Elicited Basis of antecedent Subject with specified antecedent Elicited Elicited Neither outcome nor antecedent Any subject in defined population Elicited Known Basis of outcome Subject with specified outcome

62. Basis Cohort Cross sectional Case control Evaluation and control of confounders No of subjects Measure of disease frequency Poor large Incidence Fair Large Prevalence Good Small Odds obtained for antecedent

63. Basis Cohort Cross sectional Case control To determine whether exposure preceded disease If all factors are not known Time and money Best Best Most expensive Not appropiate Less appropiate In between Not appropiate Not Appropiate Least expensive

64. Experimental study Require deliberate human intervention to change the course of events Often infeasible because of difficulties enrolling participants, high costs, and big ethical issues, To prove hypothesis

65. Objectives To provide scientific proof of etiological factors which may permit the modification or control of those disease To provide a method of measuring the effectiveness and efficiency of health services for the prevention, control and treatment of disease and improve the health of the community

66. Usually compared two groups One group in which intervention takes place Another group that remained untouched

67. Clinical Trials Studies (experimental studies) Involve humans are called clinical trials studies because their purpose is to draw conclusions about a particular procedure or treatment used for evaluating the effectiveness of an intervention (therapy research questions).

68. Animal studies Laboratory studies

69. Advantages: 1. Give a strong causality evidence 2. Less bias 3. Historic controls can be used in preliminary study

70. Disadvantages: 1.Expensive 2. Ethical issues 3. They need time 4.Participant compliance

71. Controlled trials are studies in which the experimental drug or procedure is compared with another drug or procedure as usually previously accepted or placebo treatment. Uncontrolled trials are studies in which the experimental drug or procedure is described being not compared with another treatment.

72. a) Randomized Controlled Trials Provides the strongest evidence for concluding causation It provides the best insurance that the result was due to the intervention only

73. Provide that each individual has the same chance of receiving each of the possible interventions, so allocation of subjects in experimental or control group is given by chance. Ensures that known and unknown confounding factors are equal in both group, (reduce bias)

74. b) Nonrandomized Controlled Trials Clinical trial or comparative studies with no mention of randomization as well Considered much weaker because they do nothing to prevent bias in patient assignment.

75. Field trials In contrast to clinical trials, involve people who are disease free but presumed to be at risk; data collection take place ‘in the field’, usually among non-institutionalized people in the general population Purpose is to prevent the occurrence of the diseases that may occur with relatively low frequency, field trials are often huge undertakings involving major logistic and financial considerations

76. Contd … Field trial method can be used to evaluate interventions aimed at reducing exposure without necessarily measuring the occurrence of the health effects E.g Blood lead levels in children has shown the protection provided by elimination of lead paint in home environment

77. Community trials In this form of experiment the treatment groups are communities rather than individuals Appropriate for diseases that have their origins in social conditions, which in turn can most easily be influenced by intervention directed at group behavior as well as at individuals

78. True experimental research designs where researchers have complete control over the extraneous variables & can predict confidently that the observed effect on the dependable variable is only due to the manipulation of the independent variable Essentially consist of the following three characteristics:  Manipulation  Control  Randomization

79. Quasi-experimental research design involves the manipulation of independent variable to observe to effect on dependant variable, but it lacks at least one of the two characteristics of the true experimental design;randomization or a control group

80. PRE –EXPERIMENTAL RESEARCH DESIGN This research design is considered very weak, because the researcher has very little control over the experiment Most suitable design for the beginners in the field of experimental research

81. Type of study Ability to prove causation Randomised controlled trials Cohort studies Case control studies Cross sectional studies Strong Moderate Moderate weak

82. Quantitative vs. Qualitative Quantitative Research is based on the measurement of quantity or amount. It is applicable to phenomenon that can be expressed in terms of Quantity. Qualitative Research is concerned with Qualitative phenomena related to quality of kind, human behaviour, attitudes, motives, etc.

83. Measurement in various study designs Prevalence Incidence Odds ratio Relative risk Attributable risk

84. Standard 2X2 contingency table disease Yes No exposure Yes a b No c d Total (a + c) (b + d) 88

85. Prevalence rate Measured by cross sectional study Prevalence rate = No of all cases under study at a specified time in specified population/ No of persons in the population at risk at a specified time X 1000

86. Incidence rate No of new cases of a disease under study during a specified period of time/ total susceptible population or population at risk of developing the disease under study during a specified period of time X 1000 Cohort study

87. Odds ratio A measure of effect size Measured in case control studies Ratio of the odds among exposed to the odds among unexposed OR of 1indicates that the condition or event under study is equally likely in both groups

88. According to standard 2X2 contingency table Odd of disease among exposed = a/b Odd of disease among unexposed = c/d OR = (a/b) / (c/d) = ad/ bc

89. Geater than 1 indicates that the condition or events is more likely in the first group Less than 1 indicates that the condition or event is less likely in the first group OR increases strength of association also increases

90. Relative risk / Risk ratio Risk of an event ( developing disease ) relative to exposure A ratio of the probability of the event occuring in the exposed group versus a non exposed group According to standard 2X2 contingency table Risk of disease among exposed = a/ a+b

91. Risk of disease among unexposed = c/ c+d RR = risk among exposed / risk among unexposed For null hypothesis, risk ration will be equal to 1

92. RR more than one shows an association Weak association RR >1 and < 2 Moderate to strong RR = 2 – 4 Very strong RR > 4

93. Attribute risk / risk difference Simply a rate of disease in the exposed people minus the rate in the unexposed people According to standard 2X2 contingency table Risk among exposed – risk among non exposed

94. Risk of disease among exposed = a/ a+b Risk of disease among unexposed = c/ c+d AR = risk among exposed - risk among unexposed For null hypothesis, AR = 0

95. References Basis methods of medical research A. Indranarayan , 3 rd edition Previous presentation Internet Research methodology for health professionals RC Goyal Park’s textbook of Preventive and social medicine, 22 nd edition

96. THANK YOU !!!