Can you please help me decide?

#1
Hello everyone,

I am a healthcare professional looking to publish literature on a study that I have done. I am so confused about some things and I would really appreciate your insight and help. I put a lot of effort into my study. The sample size was huge, think between 100 and 1000 and there were several follow-ups. It was extremely strenuous. I initially thought of putting the study aside and not publishing it because of the many drawbacks it suffered. However, many seniors in my field have gone through it and said the study is too extensive to go to waste and encouraged me to publish it. I want to be as honest as possible in publishing it. So, I will list out the drawbacks and concerns down below. Could you please let me know whether there are methods to overcome/ deal with these concerns during analysis and publication allowing me to be as transparent as possible? For example, I know to there are methods to handle missing data, such as imputation. I am not good at statistical analysis, so I will be looking out for a biostatistician to help me, but I need to know your thoughts. Thank you in advance!

The list:
There was a maximum of 15 percent attrition in some of the follow-ups. The attrition varied at various follow-ups. I don't have a way to pinpoint how much attrition occurred in each of the two groups at each of the follow-ups.

I also don't have the data on which participants were lost to follow-up at the visits. Say, I don't know whether participant 'a' was present in all the visits. Is there any way to kind of take out a certain number of participants out of the analysis to make up for this even though I won't be able to say whether I am taking out the participants that were present at all the visits?

I was able to give the intervention to some of the participants during some visits but not during all visits. I don't remember. I don't if I should then consider this as a repeated intervention or a single-time intervention given at the start of the study.

I would really appreciate your help. Please do let me know if you know any bio/statisticians that can work with me through every step. Thanks.
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
You need to describe the basics. Exactly how many people did you start with? Did everyone get an intervention or was it randomized? It is unclear why there are repeated visits and what was suppose to be collected. Also, you never describe the outcome.

How can you have values across time but not know who they belong to? Did people just complete anonymous surveys. If so, this may be a major issue that cannot be overcome with analytic methods.
 

katxt

Well-Known Member
#3
Please do let me know if you know any bio/statisticians that can work with me through every step.
Try your local university. It's a complex problem and you want to do it right. The advice we can give here can be useful background, but there is a limit to what we can do.
You could invite a biostatistician to be a co-author, or if it's an important project, get your department to hire one for a few hours to get you started.