Bias means deviations of outcome or conclusions from the reality or lack of internal authenticity in epidemiological studies. Selection, information and confounding are three types of biases (Recall bias can be a threat to Retrospective and Prospective Research Designs, 2011). Recall bias is a variety of information bias that threatens the internal authenticity of the investigation using self-reported numbers, and can be introduced during the data compilation stage of the exploration (Bias in analytic research, 2011).There may be premeditated or unintended discrepancy in recall of information regarding the conclusion leading to misclassification of the subjects under study with respect to outcome variables. Recall of information is memory based, so faulty and untrustworthy. Retrieving information from old memories can be difficult. The longer the time spans, the higher the chances of incorrect recalls, thus creating errors in epidemiological inquiries (Bradburn, Rips, & Shevell, 1987). This review aims to discuss recall bias in epidemiological studies, possible ways to minimize them, and an analysis of paper methodology for recall bias, report findings and impact.
Recall bias in epidemiological studies. Recall bias is a chief anxiety in case control studies, but may be found in certain circumstances of prospective cohort and clinical trial designs (Bias and causal associations in observational Research, 2011). These may occur in examples like that of occurrence of important disease as cancer or genetic malformation; a particular exposition as an important factor to the causation of the disease, media publicity for a unsure connection of a factor with a disease, and exposure to the factor under study is publicly unlikable such as illegal drug abuse (Methodologic problems and standards in case control research, 2011).
To minimize recall bias one has to clearly define and express the research problem irrespective of the design used. The planning stage must be well refined with respect to the feedback form for a response, interview process and personal qualities (Design and analysis of case control studies, 2011). In case control inquiries, recall bias may be minimized by adopting various strategies like nested case control design, choosing new diagnosed cases to respond, the correct control group, standardized data collection methods, using a validated instrument for exposure review, conduct a sub-group analysis, allowing time for response, deleting proxy answers and verification of the information with a reference (Investigator bias and Interviewer bias: the problem of reporting systematic error in epidemiology, 2011).
Analysis of paper methodology for recall bias, report findings and impact. Rockenbauer studied for cancer or gene defects in infants, found recall bias after analyzing reported figures on drug intake during pregnancy by mothers questioned few months after birth than when compared to drug intake data recorded in a log-book during pregnancy. The sensitivity of disclosure reporting was higher for cases than for controls when there was a better recall by mothers of cases, so higher number of truly exposed mothers in cases than controls. The lower specificity of self reported experience for cases than controls indicated over reporting and the proportion of truly unexposed mothers were lower in cases than controls. Parental reporting may nullify recall bias in childhood diseases (Father’s occupational exposure to carcinogenic Agents and childhood acute leukemia: a new method to assess exposure (a case control study), 2011). Recall bias cannot be ignored and the results should be carefully analyzed as there is possibility of inflating the expected risk ascribed to the contact under investigation (King, 2010).
References
Bias and causal associations in observational Research. Web.
Bias in analytic research. (2011). Pubmed.gov. Web.
Bradburn, N. M., Rips, L. J., & Shevell, S. K. (1987). Answering autobiographical Questions: the impact of memory and inference on surveys. Science, 236, 157-161.
Design and analysis of case control studies. (2011). Pubmed.gov. Web.
Father’s occupational exposure to carcinogenic Agents and childhood acute leukemia: a new method to assess exposure (a case control study).
Investigator bias and Interviewer bias: the problem of reporting systematic error in epidemiology (2011). Pubmed.gov.Web.
King, G. (2010). A Hard Unsolved Problem? Post-Treatment Bias in Big Social Science Questions. Web.
Methodologic problems and standards in case control research. Web.
Recall bias can be a threat to Retrospective and Prospective Research Designs. Internet Scientific Publications. Web.