cueFlash

Glossary of Epidemiology: 3) Retrospective Cohort

Start Studying! Add Cards ↓

Synonyms for retro cohort
historical cohort
retro cohort defined
cohort identified and assempled in the past based on existing records and 'followed up' till event of interest
Sources of Data
Medical encounter DB, Employment records, vital records, data from previous studies.
Database Criteria for retro-cohort
Ecompass well defined population
exposure can be determined
Disease can be determined
complete
additional information if needed
Information from Medicaid
claims files
enrollment
pharmacy
inpatient
outpatient
nursing home
4 Factors when looking at DB for pharmacoepi study
Population
Exposure/Drugs
Outcomes
Confounders
DBS Population Strengths
Defined
Participation unnecessary
Can study providers
Timely and cheap
DBS population limits
sample size
generalizability
confidentiality
DBS Exposure strengths
Accurate, complete
detailed: drug, dose, duration
longitudinal
indpendent patient/phys reporting
DBS exposure limits
Other sources of exposure (OTC, samples)
non-covered drugs
compliance
DBS outcomes strengths
Serious disease: comprehensive
Utilization of health care
DBS outcomes limits
Diagnostic coding, record REVIEW needed
Requires health care encounter
missing function or 'soft' endpoints
DBS confounders strengths
Demographic data
Medical care use
DBS Confounders limits
Indication often absent
no behavioral risk factors
comorbidity possibly absent
3 problems of using medical provider records
Data Quality (QC)
Lack of denominator: population sending patients to hospital cannot be defined
Difficult to followup: other providers and migration
Problems with using previous study cohorts or RCT
Not adequate measurements for endpoint or exposure
Key issues of retro-cohort
Take advantage of already-collected data
How adequate are these data for study purpose
Retro-cohort power
Disease 1/100 - 1/10000 incidence

Add Cards

You must Login or Register to add cards