Tuesday, June 19, 2007

Online health survey proven valid

Medical knowledge is changing so fast these days that it is impossible to follow without using the World Wide Web. Its resources may be used in two ways: as a database for retrieving what is known and as a research tool for gaining new knowledge.

1. The WWW as data base

yahoo healthA group of Danish prostate cancer specialists, after having tested several search strategies, has put up this top ten list of online cancer resources:

1. Yahoo Health,
2. Doctor's Guide,
3. The National Electronic Library for Medicines,
4. CancerPage,
5. Medscape,
6. MedlinePlus,
7. HealthAndAge,
8. CancerConsultants,
9. EurekAlert,
10. Oncolink.

Yahoo at top and Google nowhere is quite a surprise to me. My forecast: In ten years or so we may find the first cancer science blog in the top ten.

2. The WWW as research tool

Online surveys are fast, but are they reliable enough for medical research? Most studies on risk factors, nutrition, stress, drug use and many more public health issues are done by classical, slow and very expensive surveys such as the 2004 Behavioral Risk Factor Surveillance System (BRFSS), the 2003 National Health Interview Survey (NHIS), and 2001-2002 National Health and Nutrition Examination Survey (NHANES).

All these surveys have gathered the national smoking prevalence (20.9%, 21.5%, 24.5%) and, in a new study, serve as a gold standard. Against this, the current smoking prevalence (24.0%) from Harris Poll Online (HPOL) have been compared. The online value is well in the range of the classical surveys. The authors conclude: «Online surveys may be useful for public health surveillance of the US population.» (One of the authors is affiliated with Harris Online.)

And poor old Mousetrapper concludes that soon he will have to watch even more public health survey studies.

Wikipedia: Behavioral Risk Factor Surveillance System

No comments: