| | From the study I hyperlinked to above:
Unlike prior studies of caregiving, which typically compare caregivers (e.g., people with a spouse in poor health) with noncaregivers (e.g., people with a spouse in good health), this study assessed caregiving (hours of care provided to a spouse) for every individual. Let me get this straight. You want to find out if giving care to the sick or dying shortens or extends your life and, in your data sample, you include the sick and dying (as examples of non-caregivers)?! That's preposterous! What in the world is going on here?! Of course care giving is going to be seen to be life extending -- if half of your freaking control group is already sick or dying!
We computed survival time for respondents from the day of the 1993 interview until death or the date of the last interview.
Now, let's say you wanted to lie with statistics. What you do then is find everyone in the study who wasn't a caregiver -- and you perform your "last interview" with them real early on. Now, if you assume those people die the day after the inteview -- as it looks like these researchers might have done -- then, Voila!, you get the entirely-fabricated result that non-caregivers die before caregivers.
ˇ°On the days your spouse helped you, about how many hours per day was that?ˇ± Responses to these questions were then used to calculate the number of care hours per week. The resulting variable was nonnormally distributed, so ... we created a dummy variable to indicate whether the care recipient had received 0, from 1 to 14, or 14 or more hours of care a week from the respondent. Maybe Merlin can chime-in here, but is it really all that smart to create dummy variables? Is it honest? Can you lie with a dummy variable? Enquiring minds want to know.
Among individuals with one or more impairments in ADLs (n = 673), for example, nearly half (n = 333) reported receiving no help from their partner. For perspective, about 3370 folks were in the study and 909 (~27%) of them died. If the above is an indication of how many folks were not caregiving (n = 333) and of how many folks were caregiving (n = 340), then we've got a death total that exceeds the sample size. Must be that folks who got care weren't limited to folks with one or more impairments in Activities of Daily Living (ADLs). Besides ADLs ("eating, transferring, toileting, dressing, bathing, walking across a room"), the study also mentions "instrumental" ADLs (IADL), such as "preparing meals, grocery shopping, managing money" and basic cognitive impairment. It's a good guess that many (most?) who got care where in these last 2 groups.
The test of the unadjusted association of care hours and mortality demonstrated that the highest level of caregiving (ˇÝ 14 hr per week) was associated with a reduced risk of mortality (p = .012), but that a lower level of caregiving (1¨C14 hr per week) was unrelated to mortality risk (n.s.). When hours of care and spousal-need variables were considered simultaneously (Model 1), the hazard ratio for high care hours (i.e., ˇÝ 14 per week) was significant (p < .0001). The problem with factoring in spousal-need (someone who is really bad off) is that an assumption is attached to it and that assumption is this: if your spouse is in serious medical need, then you are more likely to die early. So, when they say they "considered" hours of care and spousal-need simultaneously, then they assume caregivers should have this astronomical death rate and -- if they don't -- well then that means that caregiving extends life. Let's say they assume that high-need spouses take 7 years off of your life. Let's say they found a normal death rate. What they do then, is proclaim that caregiving adds 7 years to your life (because of their prior assumption that high-need spouses take that much away).
Ed
(Edited by Ed Thompson on 12/05, 2:55am)
|
|
