Karteikarten
Randomised Controlled Trials - Statistics - 32
Karten in diesem Stapel (19)
- 1The purpose of randomised controlled trials:To evaluate the effects of an intervention
- 2Randomised Controlled Trials allow for the isolation of the effects of the intervention throughControlling the experimental condition, and having a control group to compare to.
- 3Samples in RCT'sAre usually not representative of a population of interest, and it is also rare to take random samples. Usually trialling a treatment so only those with a disease are selected, or are trialling a particular affect of something so pick people that are likely to complete the requirements of the trial.
- 4Control Groups provide information on:What would have happened without the experimental intervention.
- 5The intervention may become biased if:People know which group they are in (Experimental or Control) or if the experimenter knows which group a participant is in.
- 6Placebos are used:To isolate the effects that may occur if people believe they are getting a particular intervention. Also to allow blinding of the intervention and control groups, so that no bias occurs
- 7PlacebosSubstances with no biological effect on the outcome.
- 8If participant groups have the same or similar exposure to factors which effect the outcome than:Any differences between the two groups can be attributed to effects of the treatment.
- 9Historical ControlsRelates to using old data from previous experiments and studies, and comparing that with the present study data.
- 10Problem with historical controlsPrevious populations may be different from present populations, and new knowledge on the disease may effect changes in treatment over time (Better standard of care) (Therefore may not be comparable)
- 11Concurrent Non Randomised ControlsEliminates effects of changes over time. There are 2 options that patients of physicians may choose from. (Old or new treatment)
- 12Problems with Concurrent NRC'sCharacteristics of those choosing new treatment may be different to the characteristics of those who choose the old treatment. This gives unreliable information.
- 13Randomised control trials are the best way to allow us to control forConfounding. Although there is no guarantee of total control of confounding
- 14RandomisationAn uninformed mechanism to divide people in to similar groups. Makes the groups comparable, as we can control for known and unknown confounders through randomisation. Works well in large groups, with small groups the chance for confounding increases. (both groups have confounding basically - just spread it out)
- 15Concealment of AllocationDouble Blind. Nobody knows who is in what group. As long as this occur we end up with comparable groups
- 16Intention to Treat AnalysisNot all who are allocated to either the experimental group or the control group will receive the effects of the group, due to drop out, loss of follow up, death etc, and in order to maintain control of confounding one analyses them regardless.
- 17Excluding participants that dropped out of the study creates:Bias, where those who dropped out hold characteristics that those participants who stayed in do not hold. This means there will be differences between groups and the effect of the treatment is not isolated.
- 18Selection Bias in RCT'sDoes not occur because it is uninformed, computer generated groups
- 19Information Bias in RCT'sMissing follow up on people, or getting error on the measurement of outcome. Potential for bias if your measurement is generalised away from your objective e.g - those who had upper respiratory infection vs influenza.
Informationen
- Beschreibung
- Statistics Karteikarten am Randomised Controlled Trials - Statistics - 32, erstellt von Bex Chrystall am 01/10/2017.
- Erstellt von
- Bex Chrystall
- Aufrufe
- 0
- Bewertung