|
 |
|
|
Ending a Study Early
An assisted living facility has increasingly high levels of infection among residents admitted to its two intensive care nursing wards. The facility obtains funding for a 1-year study of techniques that might reduce the risk of infection in such an environment. Researchers enroll every patient in one of the wards, obtain informed consent from them, and begin applying the study protocols in that ward. A few months later, they see a thirty per cent decrease in infections on the study ward. Infections in the other, control ward continue at a high level, meanwhile, and several patients from that ward must be hospitalized. At the end of the year, comparison of rates of infection in the study and control wards supports use of the clinical interventions. Should the study have been ended early? Should patients in the control ward have been asked for informed consent?
J1G says: Control subjects are subjects. They should've been consented. As for ending the study early -- hard to say, since the study wasn't conducted correctly anyway.
R3H says: Without knowing the details of the study it is difficult to have a definitive answer. Most studies of this type have a data review panel which can suggest early termination. This was the case, for example, with the laser treatment of diabetic retinopathy. If there is no interim analysis, it is hard to really assess if there is an actual difference. In this example, however, one must conclude that there are interim analyses else how would one know there was a difference. If a clear medically and statistically significant difference is demonstrated, as it appears to be in this case, it is ethically required that the study be stopped and intervention done in the control group. The study at that point has been concluded, a benefit or increased risk has been demonstrated and there is no longer any justification for increased exposure to risk.
|
