Roles of Colonoscopy - Essay Example

In the subsequent discussions we will use the example of a clinical study into the technique of double contrast barium enema. In 2000, Winawer and colleagues published the results of a randomized controlled trial on the relative roles of colonoscopy and double contrast barium enema in detecting recurrent polyps in the post-operative follow-up of patients with colonic polyps. DATA COLLECTION Data Collection for RCT and Phenomenology are very different e.g. in RCT are usually in the form of measurements and all data generated is numerical (statistical); the statistics used are descriptive to describe the sample. On the other hand, in Phenomenology, spoken words are reported verbatim, because one is looking for the interviewee's in-depth experience. Data Collection using the phenomenological approach can take many forms. The most commonly used is dialogue with open interviews. Methods of data collection could be written notes, video and audio recordings. Non verbal communication is body language which can be taken into account. These techniques are somewhat subjective; this is because the principal aim here is to explore the problem and get a deeper understanding so that some hypothesis can be formulated. In an RCT, the interview has to be highly structured, because quantitative data has to be generated. In a structured interview, the questions have set answers, such as in multiple-choice questions, so that a limited choice is provided to the responder, and one question leads to another; for example, if the answer to above question was 'Yes' then please answer the next 3 questions. Unstructured questions are more open-ended, allowing the responder a lot more leeway in terms of answers that he can provide in his/her own language. Further, as RCTs usually entail some form of intervention/procedure/treatment, there are many more potent means of data collection. In the example cited above, there can be two phases of data collection; in the first phase, qualitative data can be obtained through unstructured interview of endoscopists to find out details about the two procedures (colonoscopy and double contrast barium enema) which can establish the controversy that exists. However, the essential question as to which of the two modalities is most sensitive and specific, and therefore more reliable, requires quantitative data of the type obtained through RCTs. This depends on blinded observations by the endoscopists who initially perform colonscopy without knowing the results of the Barium studies; if no polyp was found during colonoscopy, then the barium enema result was revealed to them, and if a polyp had been found in the latter, they re-examined the segment. Data collection here depends on highly structured and rigid enquiry of the endoscopists, and numerical data is generated. Compared to this the phenomenological approach to the interview is unstructured, flexible and open-ended, using audio tapes and verbatim reports so that more detailed and in-depth, though descriptive information can be obtained. DATA ANALYSIS Data analysis is different in both methods. Among the many different means of analysing data available in phenomenological research, the method propounded by Glaser and Strauss (1967) is the one most utilised, namely the Grounded theory approach. Data from RCT is analysed statistically, whereas data from phenomenology is analysed for codes, themes and patterns; for example in the study on double contrast enema, the initial phenomenological research data will be analysed to establish whether there is any pattern, or if a theory can be generated which can then be studied in the RCT phase. In RCT, data is analysed using tables and graphs, whereas in Phenomenology, data is analysed more subjectively. The theory generated in phenomenological research depends on, emerges from and is grounded in data, which is why it tends to be a cyclical process, as collection and analysis can go hand-in-hand, unlike in RCTs where data collection is based on decisions taken during the design stage, and analysis proceeds discretely. This is because of the demand for more accurate and precise information from quantitative research methods like RCTs. In the above example, initial in-depth interviews of various physicians which are unstructured and open-ended establish the general opinion that Double Contrast Barium Enema is safer, more convenient, more readily accepted by patients, costs less, is more cost effective, etc. Since at this stage nothing is known about the phenomenon, the interviews are open-ended and analysis running simultaneously informs the process of data collection. However, in the second phase the question is more specific, namely which of the two techniques is more reliable; here the endoscopists are asked fixed questions, quantitative data is generated, and data analysis is done discretely and is not allowed to contaminate the data collection phase. For this very reason, RCTs need to be well-planned and the interviews need to be formulated keeping in mind the various answers that are possible. It is not possible to change the format of the questions or answers at a later date, and in fact data collection and analysis are rigidly compartmentalized. RESEARCH OBJECTIVITY Qualitative research, by its very nature, tends to be non-objective; this was the main reason why in the past it was never given the importance it deserved. RCTs on the other hand are pre-designed to be maximally objective. In RCTs the research process entails blinding which ensures there is no bias during data collection, whereas this bias-elimination is not possible in phenomenological research. For example, in the study on double contrast barium enema, the initial pilot project which is a research looking into the various phenomena, the approach is qualitative, trying to establish the opinion of various physicians by using open-ended questions and the answers given to some questions lead to further questions, so that there is no blinding of the observer. In the second RCT phase, there is an elaborate mechanism of blinding; the endoscopists first perform the procedure without knowing the results of double contrast barium enema. This makes the project highly objective and free of bias. At the next stage, however, ethics dictate that if the barium enema has failed to detect any lesion, then the endoscopist needs to be aware of this finding so that he can repeat the procedure, and an unblinding step is taken, and this part of the study is then not objective, and the information so obtained then decreases in its reliability. In recent times there has been a move towards making phenomenological research more objective, for example in the FRAMEWORK approach described by Ritchie and Spencer (1974) in which interviews tend to be highly structured, data collection is targeted and focused, and some degree of observation (participant or non-participant) is also included in the data collection process. However, bias is not entirely eliminated, making the research less objective. EXTERNAL VALIDITY A test is said to be valid when it actually measures what it was set out to measure; if the results of a research can be extrapolated and applied to populations outside those which were included in the study, then it is said to be externally valid. In PHENOMENOLOGY the external validity is low, because the sample size is small, whereas in RCTs the external validity is high because the sample size is large. In qualitative approaches like phenomenological research, external validity is not possible since the population studied is very homogenous and the sample is small and unlikely to reflect other groups; they may even be unique groups. RCTs, on the other hand, are conducted for the sole purpose of extrapolating to a larger population which resembles the study population except in the independent variable being tested. A large sample size has to be used, and that is why in RCTs the research method and design are such that the study has external validity. In the example above, the qualitative phase, where in-depth physician interviews are conducted to identify what is the general belief among physicians regarding the advantages of double contrast barium enema, gives us data which is strictly applicable only to the group studied. The opinion of 100 physicians does not give us the ability to say what the opinion of the 101st is going to be since the sample is very small. However, in this study, the essential question of reliability of the procedure (colonoscopy versus double contrast barium enema) in accurately diagnosing recurrent polyps can only be answered by an RCT; In the RCT phase, a large sample is used and this gives the study external validity so that the results can be extrapolated to apply to all endoscopists who perform the procedure. Read More