Observational studies and routine data Flashcards
test accuracy of diagnostic test?
cross-sectional design
test prognosis of disease?
cohort study
test aetiology and risk factors of disease?
various non-randomised designs
test population healthcare needs?
various, ecological studies
test treatment efficacy?
RCT
What is a descriptive study ?
describe distribution of factors or disease in relation to
- person (age, sex, race, marital status, lifestyle, job)
- place (variation between/within countries)
- time (variation over time and season)
What types of data can be used for a descriptive study?
routine (births, deaths)
survey
performance management (Quality and Outcomes Framework for GPs)
What is routine data?
= data routinely collected and recorded in ongoing systematic way without research question in mind at time of collection
What are types of routine data?
- health outcome data (deaths, hospital admissions, primary care consultations, prescriptions, well being levels from national surveys)
- exposures and health determinant data (smoking, air pollution, crime stats)
- disease prevention data (screening and immunisation uptake)
- demographic data (census population counts)
- geographical data (health authority boundaries, location of GP practices)
- health service provision data (bed staff counts)
- births
Advantages of routine data
cheap already collected/available standardised collection procedures comprehensive (population coverage, large numbers) wide range of recorded items
Disadvantages of routine data
- may not answer the question
- not every case captured
- variable quality due to variable diagnostic fields
- validity may be variable
- disease labelling may vary over time/by area
- coding changes may create artificial increase/decreases in rates
- need careful interpretation
Examples of heath outcome data
mortality cancer notification of infectious diseases termination of pregnancy congenital abnormalities road traffic accidents
Why are cross sectional studies important?
e. g. census every 10 years
- useful for care providers to allocate resources efficiently and plan effective prevention
- provide clues leading to hypotheses which can be used in analytical studies
- describe status of individuals re absence/presence of exposure and disease
However cannot distinguish between whether exposure preceded disease
How is mortality recorded?
death certificates
- local registrars of births and deaths
- Office of National Statistics for coding/processing
- produced as routinely published tables (mortality stats)
How is cancer registered?
- voluntary notification to local cancer registry
- also from death certificates
- used for incidence and survival information
- linked to hospital admissions and national clinical audits
Info allows identification of most common cancer types
How are infectious diseases notified?
reported by doctors
incidence of disease
What is the Quality and outcomes framework?
- component of General Medical Services contract for GPs from April 2004
- rewards practices for provision of quality care and helps to fund further improvements in delivery of clinical care
- collected in national database system called Quality Management Analysis System
What is Administrative hospital admissions data?
UK’s way of classifying treatment
What is a finished consultant episode?
the time spent under the cont care of a specific consultant
What is admission?
comprises of 1+ episodes and/or transfers between hospitals
What are the benefits of case control studies?
cheap
quick to conduct
investigate many exposures simultaneously
good for rare diseases
What are the disadvantages of CCs?
need cases and controls to compare those who were exposed and unexposed to the risk factors suspected
selection bias of controls
subject to recall bias
uncertainty of exposure (disease time relationship)
poor for rare exposures
cannot calculate incidence rates directly as fixed number of controls in sample
What is the principle of CCs and general procedure (3)?
comparing odds of exposure between cases and controls
- select cases with disease, controls without
- obtain information on past exposures, other factors
- compare proportions of people exposed in cases and controls
How are controls selected in case control studies?
control must be free of disease (outcome of interest) during same period of time in which cases were identified
representative of population of individuals identified as cases
Sources of controls?
general population
- neighbourhood
- friends/relatives
- hospital
What does the choice of control vary in?
Amount of recall bias
response rates
selection bias
cost
What is a cohort?
group of people with something in common
represents outcome-free population from which cases will arise
What are the types of cohort study ?
prospective - compare disease rates in exposed and unexposed groups after time
retrospective - with routine data to look at relationship between exposure and outcome
Advantages of cohort?
look at multiple outcomes
follow natural disease history
good design to look at risks related to rare exposures
calculate incidence
minimise bias in estimating exposure if prospective
Disadvantages of cohort?
inefficient for rare disease expensive time consuming if prospective loss to follow up introduces bias healthy worker/volunteer effect affects generalisability
What is the STANDARDISED MORTALITY RATIO?
= rate ratio adjusted for age that shows the number of observed deaths (cases of disease), O, in a population to the expected number if the population has the same morbidity/mortality as a standard population corrected for differences in age/sex
What does SMR allow?
compare population death rates as it is adjusted for age/sex
Formula for SMR?
number of observed deaths/number of expected deaths if same age specific rates as standard population
