Oscar is a two year old cat who lives and works in a hospice for elderly people. It seems he is not generally a sociable cat. Every morning, however, he makes the rounds with the nurses. He only curls up on the beds of those people who are going to die within the next few hours. If Oscar is kicked out of a dying person’s room, he paces and meows outside the door. Apparently he has predicted deaths with startling accuracy. He never curls up with anyone else. As soon as the death occurs, Oscar leaves the room. Check it out at:
http://content.nejm.org/cgi/content/short/357/4/328
What are some possible explanations?
1) the cat is causing the people to die (I sure hope not!)
2) the cat is picking up behavior from the nursing staff who are experienced with predicting death and may be giving off overt cues
3) the cat is picking up cues from the visiting family
4) the dying person is often the centre of attention – maybe the cat is merely curious or enjoys being at the centre of the attention
5) one news report speculated that heating pads are placed on the dying and the cat is only enjoying the extra warmth. The original article does not discuss if heating pads are used this way
6) the cat really does sense that the person is going to die, and is drawn to either the person or the event in some way. Remind me not let that cat curl up on my bed!
Please indulge me as I use Oscar to stretch my mind in the area of designing research.
This sounds a lot like an urban myth and I would dismiss it except that the report was published in the New England Journal of Medicine. So it might be a reliable source. What do I believe?
I believe that the staff at the hospice believe that Oscar can predict impending death. It has happened often enough that the staff have noticed a connection. Or at least they have created a connection in their minds. Coincidence? Maybe. Or maybe by expecting it, they are influencing the cat’s behavior. Or maybe it is simply a case of formulating a notion (which may or may not be true) and then when events prove it, they notice it. When events disprove it, they don’t even notice it. They might be applying a filter to how they see events. It is like when a guy thinks that all women drivers are bad. He only notices it when a woman makes a driving mistake. If it is a man who makes a mistake, or a woman who is driving well, he/she does not even notice. In this way, his opinion is strengthened and “proven”, even though the data are flawed.
How would I research this one?
First I would have to decide if the cat’s behavior is correlated to impending deaths. I would need to follow the cat and see what he does on a regular basis. I would need to follow the cat in a manner that did not influence him or impede him in any way. A kitty cam? I would have to observe every time he curls up on a person’s bed. I would have to decide what “curl up” means. How long would the cat have to lie there for it to count? I would have to see how often the curling up behavior actually was accompanied by a death. What if the cat curls up but the resident, inconveniently, doesn’t die? Or doesn’t die within a few hours? How long is a few hours? What percentage of people die without the cat curling up on their bed?
I suspect the study would end there with no correlation between the cat’s behavior and deaths. But what if I found one? What then? How could I design a study to figure out why there is a correlation?
I would have to identify the variables and try to control them.
What if the cat is picking up cues from the nurses? I would have to allow the cat access to patients without the nurses around. This would be difficult and unethical because the nurses need to be there for patient care. The RRU ethics board would never approve it! So maybe we have to leave the nurses in the equation and accept it as a flaw in the study. There could be steps we could take to minimize the presence of the nurses while the cat is doing “rounds”.
Are heating pads placed on the dying? Only the dying or anyone who is sick? Is the cat lying on or near the heat? Another ethical problem – we can’t deny heat to the dying. How about a heating pad for every patient in the hospice?
Does the cat accurately predict deaths in cases where there is no family attending? If not, it may indicate that the cat is indeed picking up cues from the other people in the room or that he enjoys being in the middle of the attention.
Finally, if I can control all the variables and I find the cat able to predict death with accuracy, how does he do it? Do cats have an ability to sense something humans cannot? How could I measure it? Are there certain electromagnetic signals that we could detect? Changes in brain activity? One thing is for sure, I would not have a lot of elderly people lining up to be subjects in this study!
Most of these measures are quantitative. Qualitative would be pretty hard. The cat can’t talk and the people who are dying are not going to want to be interviewed. Not a lot of help there. I could make anecdotal notes as to how things seemed at the time, remembering that such notes would be filtered through my biases and beliefs.
Lastly, in order for the results of a study to be considered accurate, the results need to be replicable. That would not be possible in this case because I only know of one Oscar and he only works in one hospice. To take Oscar to another hospice would certainly disrupt his routine and be hard on him. Again, there are ethics issues.
Please don’t think that I am seriously suggesting that a study be done. This was a brain exercise only. It reminds me of Dr. Muirhead’s comment on gravity… “it fell because the earth gods made it fall”. An impossible statement to prove, and equally impossible to shake the convictions of those people who believe it.
So I hope Oscar keeps up the good work. If the cat helps at the hospice, I’m all for it. But that is one cat scan that I would rather not have. Just in case. J
