![]() |
![]() |
![]() |
front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |review |
Scenario 1:
You are asked to find out the prevalence of accidental falls in men and women aged 65 and
over in a particular geographic area. What would be your data sources and the best study
type? The most obvious source to explore would be the accident and emergency database and injury surveillance database at hospitals or health centres. This obviously contains information on severe cases who require hospital care in the area of interest, but the time period to analyse could go back to several years. Some sentinel medical practices might hold limited data on patients who have experienced a fall but do not require other care. These may be the majority. A population survey using random digit dialling or the electoral roll or a combination, may detect additional cases of falls in the elderly who did not consult any health service at all. This would only be suitable if you are exploring recent episodes, as recall bias may be an issue. Pensioner’s lists may provide a neat register of all elderly entitled to these benefits (and potentially those at greatest risk of falling) but it excludes all other elderly not entitled or registered elsewhere. GIVEN THE LIMITATIONS OF DATA SOURCES, WHATEVER YOU CHOOSE, YOU WILL BE LIMITED IN THE REPRESENTATIVENESS OF YOUR SAMPLE. SO A COMBINATION OF POPULATION SURVEY AND EXISTING INJURY REGISTER COULD BE THE BEST OPTION. |