Home  Publications Essays Ecology and the Internet |
|
|
Page 1 of 2
Ecology, the Internet and the Sampling Population Health
John E. Ware Jr., PhD
|
|
To survey the health of populations, at least three distinct sampling tasks are involved. First, and most obvious, such a survey requires a sample of people. Second, it is easy to forget, that we must also draw a sample of descriptions of functional status and well-being to represent the content of health in our survey and for that we have the domain sampling theory that underlies psychometrics. Third, we must sample time to represent the peaks and valleys of life. Kerr White and his collaborators made profound comments about all of these sampling tasks in a classic article about the ecology of medical care published nearly half a century ago (White et al., 1961). For the domains of health, they emphasized individuals’ “perceptions of their health and well-being” as opposed to rates of morbidity and mortality. With respect to time, they argued that one month is “a more realistic” unit of time or, in today’s terms, a month is the right length of survey recall for many purposes.
Focusing on the first sampling task, White and his colleagues reported startling statistics about who does and does not seek medical care when they need it leading to doubts about the representativeness of samples drawn from academic medical centers. Specifically, they observed that out of 1,000 adults about 750 will experience illness in a month and of those only one will be referred to a university medical center. A more recent replication of the original study yielded virtually the same results (Green et al, 2001). These findings combined with the additional considerations of the costs of sampling and surveying patients under care at academic medical centers are good reasons to consider whether the Internet might be a better place to fulfill Kerr White’s search for an understanding of population health and its determinants and its consequences.
With advances in electronic data capture and growing access to the Internet – approaching 80% in US cities – there have been many opportunities to test the use of the Internet in population health research. As we reported in the Newsletter in 2004, in one of our first tests we were able to quickly and inexpensively sample and complete surveys of the general US population using electronic data capture and processing on the Internet.
<< Start < Prev 1 2 Next > End >>
|
|
