Originally posted by: Mark R
On the contrary, if the treatment requires continued input in order to be effective, then its effectiveness depends on how many patients actually continue with it. For that reason, they cannot be excluded from the results.Originally posted by: Whisper
Well, if the aim of the study was to examine the effectiveness of the treatment, then dropping those patients was probably the right thing to do, as they didn't adhere to the entire treatment protocol. However, what could've been done would be to compare their improvement (if any) to the adherence group; that'd go a long way in improving external validity.
For example, lets say there's a new pill for IBS, but it has a side effect of causing severe headaches. It completely relieves the IBS symptoms as long as you take it. However, no one can continue to take it for more than a few days because of the headaches. Is this an effective treatment? No. Because, no one will take it. It's therefore pointless recommending this treatment, if there is a better alternative.
This case series is difficult to interpret, because of the lack of a control group. As such, it lacks internal validity. The external validity can't be assessed, as the abstract provides insufficient information on the selection procedures.
Not necessarily. Leaving out those individuals who didn't adhere to the treatment regimen damages generalizability of the findings, definitely. However, if you're just interested in seeing how effective the treatment is, then you only need to examine the cases in which the treatment is followed 100%; otherwise, you're examining the effects of the treatment plus its realistic ability to be completed.
Although I guess with such a study, efficacy would be a better word to go with than effectiveness, to an extent, as you're assessing the ability of the treatment to improve the condition in an ideal environment. But when the treatment is newly-developed, you have to run such "perfect world" studies to determine efficacy before you can begin to go after effectiveness. Once the ability of the treatment to alleviate the condition has been shown in a controlled setting, then you can start to examine how realistic it would be to implement.
The lack of a control group is the key shortcoming, though, that's to be sure.
