Prognosis Flashcards
What is prognosis?
- Risk of future health outcomes in people with a given disease or health condition
- Applies to a particular clinical context
- Defined by current clinical approaches in diagnosing, characterising & managing patients with a symptom or disease
What are the streams of prognosis research?
- The clinical course of a condition
- Factors associated with prognosis
- Statistical models to predict individual risk
- Tailored treatment based on prognostic factors
What are the different types of prognosis study designs?
- Clinical observation
- Prospective & retrospective cohort studies
- Systematic reviews of cohort studies
- RCTs
What are the sources of bias in clinical observation studies?
- Patients are not seen systematically in
long-term follow-up
- Follow-up is only on a subset of patients
- Large numbers of patients are needed to get accurate estimates
What is the aim of a prospective cohort study?
- Identify exposure first (e.g. knee injury)
- Collect outcome data second (e.g. future OA)
What is the aim of a retrospective cohort study?
- Collect outcome data first (e.g. signs of knee OA)
- Identify exposure second (e.g. years since ACL rupture)
What are the limitations of prospective & retrospective cohort studies?
- Retrospective studies make it difficult to determine the portion lost to follow-up
- Quality & completeness of data collection may be an issue
What is the aim of systematic review of cohort studies?
- Provide transparent & minimally biased overview of the best data available regarding prognosis
- Difficult to find on the topic of interest
How do RCTs provide hidden prognostic information?
- Control group = natural history
- Intervention group = clinical course/treated history
- Factors such as sampling & length of follow-up should be considered
How are prognosis & RCTs linked?
- RCTs can contribute to interpretation in terms of generalisability of clinical trial results
- Average prognosis of patients in the trial without treatment can be compared with average prognosis in particular populations
What are the two groups of people considered when asking prognostic questions?
- Those who receive no treatment (natural history of a condition)
- Those who are receiving an intervention (clinical course of a condition)
What are the important components of a prognostic question?
PIOT
- Patient or problem
- Intervention or issue
- Outcome
- Time
What is the best way to avoid the survivor bias in studies about prognosis?
Recruit an inception cohort
What is the most fundamental error with this statement about prognosis;
People treated with early ACL reconstruction after ACL rupture have a 16% risk of developing osteoarthritis.
The follow-up time period is not specified
True or false:
Survival curves are only used in studies of mortality
False
