Help regarding study design in my research proposal

[memadridista]

---

Members don't see this ad.

Hello everyone !

I am a 4th year medical student I am doing a research as part of my university requirement, and I needed help.

My research topic is : To evaluate the role of X-Ray in patients presenting with Acute abdomen at our medical hospital.

I wanted an elaboration of the study design that I am going to use , i will be collecting data using a questionnaire that will be filled by myself based on the patients case notes , the data that I will be collecting will be from the hospital and my participants will be those who will be admitted at the surgery ward in the hospital.

The question now is ... I was thinking of this as a hospital based prospective based study design. But the professor supervising me denied this and so I was left wandering about what can I modify regarding the study design and what else could it be ?

Thanks in advance.
If you need to see my research proposal so that you can evaluate it more please feel free to comment so i can email you the proposal or i can send a link for anyone who needs so he/she can access and help me out

Any help will be really appreciated.

Regards,
Mohammad Omar.

 

[idkididk]

Cross-sectional study?

 

[751866]

Why don't we start this way:

1) What precise question(s) are you trying to answer with your study? I read your post, but there isn't a clear objective.
2) What advice or feedback did your supervisor give you with regards to why the study cannot proceed?

 

[memadridista]

Why don't we start this way:

1) What precise question(s) are you trying to answer with your study? I read your post, but there isn't a clear objective.
2) What advice or feedback did your supervisor give you with regards to why the study cannot proceed?

Hey thanks for the reply

These are the comments from him:

This is to acknowledge receipt of your elective research proposal titled: The role of abdominal X-ray in evaluating patients admitted with acute abdomen in surgery department at MNH.
Your proposal is important and has a potential to influence doctors to make better decisions There is no doubt that the research will add data from Tanzania to existing knowledge.
After reading your proposal I found that the proposal does not read well (Poor language /grammar) and does not give a clear message. The objectives stated are not specific, measurable realistic and time bound. You need to sit down , reflect and ask your self what the research question is. I was left asking myself whether the research you are intending to do will add very much knowledge to what is already published in literature that you cited but this should not discourage you from doing the research.
Scientific aspects
o Research question — This is not clearly defined so that it can be answered appropriately
o Overall design of study — What you have is not detailed or appropriate to answer the research question. Can you please think and come up with a clear research question
o Participants —They are adequately described, their conditions defined, inclusion and exclusion criteria are also adequately described.
o Methods — More details are needed and the main outcome measure should be clearly stated.
o Ethical issues – What are they ? You should know that these issues may go beyond just getting consent from patients and simply whether the study was approved by an ethics committee or IRB.
o Data collection instruments such as the questionnaire you will be using is missing. Make sure the data is collected to answer the research question.
o References — There are no glaring omissions.You have adequate and relevant references
Finally I advise that you take time to reflect accurately what want to get out of this research and consider all the comments raised above.
Please get back to me with a revised version.

Here are the objectives.


I would like to share with you my proposal so that you can view and tell me what you think based on those comments.
Here is the link Shared proposal.docx

Please feel free to help me !

 

[memadridista]

Cross-sectional study?

Hello please review the proposal I attached in the above reply to @delimitedtab. And let me know your view. Thanks

 

[751866]

I skimmed it very briefly and I read the feedback you posted, and I believe it's congruent with my initial statement. It seems you have a general idea for a project without a specific question or set of questions you want to answer. I notice the proposal says you want to determine the role of...my question for you is can you more clearly define what that means?
I also see that most objectives are "to determine..." which are pretty general; this isn't necessarily bad if you are trying to get some descriptive information about your patient population and the cases you're seeing, but "to determine the correlation between..." isn't really a defined question. That goal is literally saying you want to calculate a correlation coefficient or some measure of association between two variables. The issue with that is that there is no stated goal or reason for calculating the correlation. If the study is exploratory, it should be stated as such, but if it is specifically testing a hypothesis, it should be clearly stated. Once that question is posed, then you can determine what you need to answer that question including the appropriate statistical approach. In short, what would you do with the calculate correlation, for example? How what are you trying to say from it? This would help you better formulate a specific set of questions and then you can decide on the appropriate way to answer them. I would also recommend that you think of the kinds of variables you're collecting to answer the question: are they quantitative or qualitative, or will you have both? How can they best be measured?

This doesn't cover everything, but I think a big issue is the lack of research question(s). I see this frequently with medical students and physician researchers; they have a general idea for a study that they think has direction rather than a specific question to investigate. The latter is frequently the correct starting point, from which you can design the study and come up with an analysis plan.

 

[memadridista]

I skimmed it very briefly and I read the feedback you posted, and I believe it's congruent with my initial statement. It seems you have a general idea for a project without a specific question or set of questions you want to answer. I notice the proposal says you want to determine the role of...my question for you is can you more clearly define what that means?
I also see that most objectives are "to determine..." which are pretty general; this isn't necessarily bad if you are trying to get some descriptive information about your patient population and the cases you're seeing, but "to determine the correlation between..." isn't really a defined question. That goal is literally saying you want to calculate a correlation coefficient or some measure of association between two variables. The issue with that is that there is no stated goal or reason for calculating the correlation. If the study is exploratory, it should be stated as such, but if it is specifically testing a hypothesis, it should be clearly stated. Once that question is posed, then you can determine what you need to answer that question including the appropriate statistical approach. In short, what would you do with the calculate correlation, for example? How what are you trying to say from it? This would help you better formulate a specific set of questions and then you can decide on the appropriate way to answer them. I would also recommend that you think of the kinds of variables you're collecting to answer the question: are they quantitative or qualitative, or will you have both? How can they best be measured?

This doesn't cover everything, but I think a big issue is the lack of research question(s). I see this frequently with medical students and physician researchers; they have a general idea for a study that they think has direction rather than a specific question to investigate. The latter is frequently the correct starting point, from which you can design the study and come up with an analysis plan.

Hey Thanks for the feedback. I have started getting the point but I am still kind of confused.
Let me try to clarify some stuff

1. The role of abdominal x ray ; basically this study will be aimed at looking at the radiological features that patients with acute abdomen present with .. That was meant by the role ...

2. Regarding the last objective of determining correlation , i really dont want to calculate any of the above you mentioned I just need to look at wether the findings that are pre op based on x ray, related to those found in the OR ; for e.g. If they found out it might be an acute abdomen , was it really what they found out in the OR. So i wanted to see that kind of relation basically qualitative data really.


So upto where I am am really confused on the way forward, maybe you can advice me on how I can come up with those research question(s) which would make sense. Also if there is anyway I can modify the proposal.

Please I would be really thankful if you can help me out as this is my first time doing a study by myself.

Regards,
Mohammad Omar

 

[751866]

Hey Thanks for the feedback. I have started getting the point but I am still kind of confused.
Let me try to clarify some stuff

1. The role of abdominal x ray ; basically this study will be aimed at looking at the radiological features that patients with acute abdomen present with .. That was meant by the role ...

So this part seems more descriptive. Are you looking to learn the features of any patient presenting or only the patients who, by some gold standard, were determined to have a definite pathology?

2. Regarding the last objective of determining correlation , i really dont want to calculate any of the above you mentioned I just need to look at wether the findings that are pre op based on x ray, related to those found in the OR ; for e.g. If they found out it might be an acute abdomen , was it really what they found out in the OR. So i wanted to see that kind of relation basically qualitative data really.

Is it fair to say you want to know if patients had a real pathology based on what the x-ray suggested? In other words, if an x-ray indicates a pathology, and we get to the OR but find there is no pathology, is this something you're interested in knowing? You could still calculate a measure of association like an odds ratio, or look at the sensitivity and specificity of the x-ray, assuming you're trying to see if it has diagnostic value--just be sure to compare it to the gold standard in your location.

I hope I've understood what you were going for with this.

 

[memadridista]

So this part seems more descriptive. Are you looking to learn the features of any patient presenting or only the patients who, by some gold standard, were determined to have a definite pathology?

Anyone who presents with acute abdominal pain and who did an abdominal xray +/- chest xray

Is it fair to say you want to know if patients had a real pathology based on what the x-ray suggested? In other words, if an x-ray indicates a pathology, and we get to the OR but find there is no pathology, is this something you're interested in knowing? You could still calculate a measure of association like an odds ratio, or look at the sensitivity and specificity of the x-ray, assuming you're trying to see if it has diagnostic value--just be sure to compare it to the gold standard in your location.

I hope I've understood what you were going for with this.

Aha I see now what you meant i really needed that kind of a measure in that part, so does that part change that particular objective ? , and do i need to add this in the data analysis ; in terms of mentioning that Odds ratio will be used ?