Public Health Essay: A critical appraisal on ‘Burden of obesity in middle east’
You are suppose to prepare a public health essay by undertaking a critical appraisal of two original research papers (this needs to be a different area to what was originally done) in the same topic area. You need to choose one qualitative and one quantitative. They should be from a good quality journal. Use a critical appraisal tool to assist in the appraisal of the literature i.e. CASP http://www.casp-uk.net/criticalappraisal or CEBM
The first study conducted by Yazdani et al. (2019) considered in this public health essay illustrates that a design concerning exploration of quality of life among Iranian morbid obese people followed a qualitative approach to interpret the results in a meaningful manner. On the other hand, the second study by Khondaker et al. (2020) adopted a case-control study design for identifying associated risk factors concerning obesity in Qatar.
In the first study there was a clear statement narrating the aim of their research. Study emanated from the research lacunae that did not provide sufficient information regarding morbid obesity in Iran despite their rising trend and poor quality of living (QOL) and social isolation of the affected group of individuals. Therefore, the present study was directed specifically with the purpose of exploring the viewpoints of the morbid obese people about their QOL thereby rendering relevance to the research conducted. Moreover, the study was novel in their conception as it investigated a lesser-studied domain of morbid obesity thus adding a different perspective to the research carried out. Setting an aim of research is vital to recognize the objectives of the study being conducted (Doody and Bailey 2016). In the present study, the qualitative research methodology is appropriate as it adequately addressed the research goal. The research aim of the study under consideration was to explore the viewpoints of the morbid obese patients in Iran and hence in accordance with the qualitative study design they rightly sought to illuminate on the subjective experiences of the research participants, in this case, the morbid obese patients of Iran. It has been mentioned that semi-structured interviews were conducted with the research participants in order to answer top research question. Non-numerical data have been resorted to gather information in connection to their viewpoints regarding the QOL. In this study, qualitative approach was most suitable as it provided the best opportunity on the part of the researchers to gain a thorough understanding with respect to the innermost deliberation about lived experiences of the study respondents, thus adding rigor to the investigation (Alase 2017). In order to address the aim of the research, whether the research design adopted was appropriate has not been clearly mentioned. In the article, reference has been made to the research design that has been utilized to address the aim of the research. By means of following conventional content analysis, semi-structured interviews have been conducted to explore the research aim. Justification for the research design adopted has not been stated specifically as no discussion has been made with respect to how they decided on which method to use while carrying out the study. It has been held imperative to design a suitable method for the sake of procuring information about the question in hand in case of a research investigation through an objective, systematic and formal approach. Hence, insight into the research question or exploration of the depth, richness and complexity of the phenomenon may be understood better (Leavy 2017). In the appraised article, the recruitment strategy was appropriate to the aim of the research. The mode of selection of the participants has been mentioned. Persian speaking adult patients having BMI ? 40 kg/m2 or > 35 kg/m2 with co-morbidity was recruited in the study. Moreover, in the study, participants having serious chronic illness or injury (physical or mental) or those who were not willing to take part were excluded from participation. However, no explanation regarding selection of the most appropriate method for providing access to the type of knowledge sought by the study has been provided in course of the discussion in the article. It has been mentioned that recruitment was done across various social and cultural levels encompassing age, marital status, education, job and several BMI categories. In qualitative study, recruitment strategy plays an immense role in shaping the course of research conducted (Elliott et al. 2017). In the concerned article, overall the data were collected in a way that addressed the research issue. Nutrition and obesity clinics of Shiraz and Ahvaz Jundishapur University of medical Sciences were the setting for the data collection and considering the aim of exploring the viewpoint of the morbid obese patients about QOL, choice of setting is justified to carry out the research. Data were collected by virtue of conducting semi-structured, face-to-face interviews recorded by audio recorder and typed in MAXQDA software using open-ended questions of 1-3 hours duration. However, justification for the same has not been provided. Interview methods have been described explicitly with mention about data saturation in which the researcher asked the initial questions followed by the principal questions to the respondents. Now, as per the respondent’s answer, the researcher aided the respondents in answering queries about their experience in relation to morbid obesity by shooting probing questions. The relationship between researcher and participants has been considered adequately in the concerned study. Derivation of the coding categories was done directly and inductively from the participants’ quotations. Definite method namely Graneheim and Lundman’s method was adopted for data analysis by the research team followed by categorization of data by content meanings after identification of main categories. Again, there is no mention regarding the implications for any changes in the research design. However, nothing else has been described concerning the critical roles of the researcher for potential bias and influence while formulating research questions, collecting data, including sample recruitment and choice of location. Despite these, factors such as credibility, confirmability, dependability and transferability were taken into consideration whereby confirmation of codes from each respondent as well as two qualitative research experts was done alongside recording, word-by-word transcribing and saving of all interviews, detailed description and time-bound recording of researcher activity. Thus, overall the study may be considered valid.
In contrast, the second study addressed a clearly focused issue. The studied population was that of obese people living in Qatar and it tried to reveal potential risk factors associated with obesity prevalent in that region. The risk factors include measurements related to spirometry, physio-clinical biomarkers and DXA body composition related parameters. In this case-control study, the one having a health issue of obesity like condition is termed as case group. Contrarily, it is matched with another group that do not have obesity condition. For this particular instance, it has been done to confirm the presence of the risk factors associated with the condition of obesity in the studied population. It is mainly an observational retrospective study whereby the extensive clinical measurements data have been obtained from Qatar biobank (QBB) repertoire for thorough, insightful investigation. The authors have used appropriate methods to answer their questions. The study highlighted that efforts were streamlined to develop certain machine learning models for distinguishing healthy from obese persons and then bring to the forefront potential risk factors associated with obesity. In order to answer the question under circumstances, a case-control study was the appropriate one as it focused on revealing the potential risk factors associated with obesity in the Qatari population. Further, it addressed the study question well as the cohort was studied with the aid of data collected from the QBB repertoire that housed data and biomedical samples belonging to adult Qatari population. Thus, exposures to risk factors relevant to Case group have been looked into and not the Control group by looking into their case histories (Schulz and Grimes 2002). The cases were recruited in an acceptable way. The cases have been defined precisely as it has been mentioned that adult (above 18 years of age) obese individuals having BMI ? 30 kg/m2 were enrolled as cases for this particular study. Moreover, another inclusion criterion for the cases was that they did not have ailments such as cancer, cardiovascular disease and diabetes and were Qatari citizens. Number of cases recruited was 250 and thus may be said to be representative of the Qatari obese population. Further, an established reliable system was appointed for recruiting all the cases through adoption of machine learning strategy that distinguished obese from non-obese individuals. However, if power calculation was done or not has not been referred to anywhere in the article. Similarly, the controls were also recruited in a way that is acceptable. Number of controls appointed for the study was 250 that again may be considered as representative of the non-obese population in Qatar. Apart from being non-obese, the controls were those who did not have any history of diabetes, cardiovascular disease, cancer, hypertension, sleep disorder or stroke. Selection was done on a random basis, although their recruitment was matched with the controls after distinguishing them from obese persons following machine learning strategy. Thus, it may be interpreted that selection bias was done away with that might have compromised with the generalizability of the findings thereby diminishing the credibility of the research conducted (Munafò et al. 2018). The exposure whether measured accurately or not in order to minimize bias, cannot be told. The authors have resorted to subjective measurements of the physio-clinical parameters to identify the risk factors associated with obesity and they were similar in case of both cases and controls. Blinding whether applied wherever feasible and temporal relation was correct or not, cannot be stated correctly as no reference about these have been anywhere related to them. Therefore, there remains a possibility of classification bias to creep in that may undermine the results owing to inadequate, improper or ambiguous documentation of the individual factors related to outcome variables (Li et al. 2020). The groups whether have been treated equally or not have not been mentioned clearly in the article. It has been mentioned that the study cohort comprising of cases and controls had equal number of males and females, however whether they were age-matched have not been mentioned as their average age was quite different from one another. Further, whether the cohort belonged to similar socio-economic background and share similar genetic lineage is not mentioned. The authors have also taken into account of the potential confounding factors in the design and analysis of the research undertaken. Data obtained have been pre-processed, followed by feature subset selection apart from development of machine learning model following in-depth evaluation of that model to interpret the results in a meaningful way. Statistical operations for checking normal distribution of data, followed by t-test or Mann-Whitney test to check the statistical significance of the data thus obtained was done. All these procedures facilitated the data analysis process thereby aiding in accurate interpretation of data. Hence, it may be said that the results of the trial are more or less valid as it followed a systematic approach in analyzing the data.
Out of the two studies, the first one may be considered better in terms of study validity as it ensured all the pre-requisites of a qualitative research is followed correctly without compromising the credibility of the study. Contrarily, in the second study factors related to blinding, classification bias was not attended satisfactorily thereby hampering the rigor of the study conducted. Methods employed in the first study was more appropriate to answer to the research question compared to the second one and therefore contributed to more reliable knowledge related to obesity in middle east.
For the first study, ethical issues have been taken into consideration properly. Ethical standards were properly maintained whereby participants were made aware and informed regarding aspects such as ethics, anonymity while publishing and principles of voluntary participation besides recording of the interviews conducted. Thus, it is evident that ample efforts were taken to explain to the participants about the research methodology and objective from the very outset. Moreover, it is mentioned in the concerned article that signatures of the participants were sought in the form of written informed consents. Furthermore, the ethics committee of the Shiraz and Ahvaz Universities of Medical Sciences gave ethical approval for the study. Ethics in a scientific investigation plays a pivotal role to disseminate credible, evidence-based information about the topic under exploration. It implies certain standards of conduct in carrying out research work so that welfare, dignity and rights of the research participants may be kept intact. Following informed consent protocols as well as respecting confidentiality and privacy of the research participants is imperative in such circumstances (Resnik 2018). The data analysis was not sufficiently rigorous in the study under consideration. It has been mentioned that for data analysis post each session of interview, the research team utilized Graneheim and Lundman’s method until emergence of categories and sub-categories. This was followed by identification of the main categories and categorization of data as per content meanings. However, mention there is no mention regarding the mode of derivations of these categories from the retrieved data. The analysis process is not demonstrated explicitly in the article, although sufficient data has been presented to support the findings. Contradictory data whether has been taken into consideration or not has not been mentioned properly anywhere. Moreover, critical role, potential bias, influence during analysis and data selection for presentation on the part of researcher has not been addressed sufficiently. For research rigor, some of the attributes such as research reflexivity that renders insight into researcher’s own biases and rationale while decision-making with research progression are considered important (Johnson et al. 2020). The concerned study has a clear statement of findings. Findings have been documented explicitly under six main categories with adequate discussion for both and against the researcher’s arguments. Further, in order to ensure credibility of the findings, both the research participants as well as two qualitative research experts confirmed the codes. Moreover, all the findings have been discussed with respect to the original research question that aimed to explore the viewpoints of the morbid obese patients concerning their QOL. Different dimensions while exploring the viewpoints in relation to QOL for the morbid obese patients was revealed through the study that added to the knowledge bank regarding morbid obesity. Specific information were acquired concerning six principal categories including physical changes, negative body image, psychological experiences, financial pressure, socio-personal dysfunction and change in spirituality. Totality of outcomes in study depicts the direction of research thereby indicating trends (de Boer et al. 2018). Hence, results obtained from the study may be considered both valid and correct.
The results of the second study revealed that the analysis was more or less appropriate to the study. According to the outcomes of the ablation study it was determined that the physio-chemical measurements served as the most influential risk factors in differentiating between healthy and obese persons. Moreover, several feature ranking techniques lead to the inference that there are certain known obesity risk factors apart from potential risk factors, attributed to certain obesity related co-morbidities that were further identified through utilization of such techniques. However, lack of mention about odds ratio makes it difficult to gauge the association between exposure and outcome. The results whether have been adjusted for confounding is also not clear, although it has been mentioned that cohort was recruited through randomization. Confounding in research often lead to situations where contribution of the causative factors cannot be segregated (Nørgaard et al. 2017). The estimate of treatment effect was precise fairly. The significance level for each variable that was distributed normally was set at p < 0.05. All the important variables consisting of spirometry assessments, VICORDER evaluation, physio-clinical biomarkers, DXA body composition, DXA densitometry was taken into consideration. However, no mention has been made regarding evaluation of the effects of subjects who refused to participate, as it was not applicable in the concerned study. Estimate of treatment effect is an important criterion to measure the difference in outcomes due to an intervention and helps to provide an idea about the direction and magnitude of that effect by virtue of an interpretable value. It typically accounts for comparison of results for the studied variables (Concato et al. 2000). The results of the study undertaken are believable in the sense that it has emanated out of rigorous machine learning strategy induced analysis and statistical operations. Findings based on machine learning modelling have been found to be consistent with previous reports that identified albumin, insulin, uric acid, c-peptide as potential risk factors for obesity apart from additional biomarkers relevant to morbidities such as liver function, lipid profile, bone joint function, diabetes.
The first study lead to generation of better results following their qualitative approach to understand the viewpoints of morbid obese people in Iran regarding their QOL thereby aiding in recruitment of definite interventions to improve their condition. However, the second study in their quantitative approach through case-control study design emanated results that may only indicate potential risk factors associated with Qatari obese persons without any specific recommendations.
Study usefulness to practice
The research conducted in first study is valuable in the sense that it contributes significantly to the existing knowledge of morbid obesity in relation to QOL. The study may be considered vital whereby thorough understanding of the studied phenomenon by the healthcare providers may lead to implementation of community-based approaches in rendering the desired QOL to the target group of morbid obesity patients. Participants being selected across different social classes and diverse levels of cultural groups added to the strength of the study, although the results may not be generalized across other socio-cultural groups and age spectrum. However, clear statements regarding applicability as well as usefulness of the research in specific settings have been referred to in fair details thereby making its repeatability easier in future. Therefore, the study may be considered useful for public health practice and it adds to the evidence base regarding morbid obesity.
The results of second study may be well applied to the local population. This study has imparted significant knowledge in relation to obesity in Qatar thereby may be used to mitigate chances of risks associated with co-morbidities in obese individuals. Improvement in healthcare management for patients and physicians may be derived by means of paying attention to the study results. Thus, holistic improvement of obese persons may be brought about through proper implementation of healthcare practices that accord to the study findings. The results of the study are both unique and novel as they predict potential risk factors for obese people thereby aiding in their clinical management by identifying them in relation to co-morbidities. Hence, the findings cannot be corroborated with previous findings that provide ample evidence to support the claims made in the study. Thus, although the results of the study may help locally, however cautions must be adopted while extrapolating the findings in other settings.
Alase, A., 2017. The interpretative phenomenological analysis (IPA): A guide to a good qualitative research approach. International Journal of Education and Literacy Studies, 5(2), pp.9-19.
Concato, J., Shah, N. and Horwitz, R.I., 2000. Randomized, controlled trials, observational studies, and the hierarchy of research designs. New England journal of medicine, 342(25), pp.1887-1892.
de Boer, A.W., Drewes, Y.M., de Mutsert, R., Numans, M.E., den Heijer, M., Dekkers, O.M., de Roos, A., Lamb, H.J., Blom, J.W. and Reis, R., 2018. Incidental findings in research: A focus group study about the perspective of the research participant. Journal of Magnetic Resonance Imaging, 47(1), pp.230-237.
Doody, O. and Bailey, M.E., 2016. Setting a research question, aim and objective. Nurse researcher, 23(4).
Elliott, D., Husbands, S., Hamdy, F.C., Holmberg, L. and Donovan, J.L., 2017. Understanding and improving recruitment to randomised controlled trials: qualitative research approaches. European urology, 72(5), pp.789-798.
Johnson, J.L., Adkins, D. and Chauvin, S., 2020. A review of the quality indicators of rigor in qualitative research. American Journal of Pharmaceutical Education, 84(1).
Khondaker, M., Islam, T., Khan, J.Y., Refaee, M.A., Hajj, N.E., Rahman, M.S. and Alam, T., 2020. Obesity in Qatar: A Case-Control Study on the Identification of Associated Risk Factors. Diagnostics, 10(11), p.883.
Leavy, P., 2017. Research design: Quantitative, qualitative, mixed methods, arts-based, and community-based participatory research approaches.
Li, Y., Fan, W., Zhang, M., Liu, L., Bao, J. and Zheng, Y., 2020. Measurement bias: a structure perspective. arXiv preprint arXiv:2012.10980.
Munafò, M.R., Tilling, K., Taylor, A.E., Evans, D.M. and Davey Smith, G., 2018. Collider scope: when selection bias can substantially influence observed associations. International journal of epidemiology, 47(1), pp.226-235.
Nørgaard, M., Ehrenstein, V. and Vandenbroucke, J.P., 2017. Confounding in observational studies based on large health care databases: problems and potential solutions–a primer for the clinician. Clinical epidemiology, 9, p.185.
Resnik, D.B., 2018. The ethics of research with human subjects: Protecting people, advancing science, promoting trust (Vol. 74). Springer.
Schulz, K.F. and Grimes, D.A., 2002. Case-control studies: research in reverse. The Lancet, 359(9304), pp.431-434.
Yazdani, N., Sharif, F., Elahi, N., Hosseini, S.V. and Ebadi, A., 2019. Exploration of the quality of life in iranian morbid obese people: a qualitative study. International journal of community based nursing and midwifery, 7(2), p.138.