Subtitle: "Or how you can associate almost anything with a weak study design."
One of my favorite epidemiological study designs is the temporal association "ecological" study that attempts to infer causation by showing one exposure increasing and one outcome increasing and then implying that the exposure is causing the outcome. You know, "Hey, they are both going up so one thing causes another." Vaccine use and autism rates anyone?
So, just for fun I've produced the graph and as you can see, through the efforts of CDC, WHO, VA and many individually hard-working IPs, hospital epidemiologists and clinicians, hand hygiene compliance has increased. And as you can also see obesity is also increasing, ergo hand hygiene causes obesity! Just try to disprove it!
Now, why am I wasting time with such an exercise? Because there is a paper in this June's ICHE that uses a similar study design and comes to an equally incorrect and perhaps dangerous conclusion. The study used 2008-2011 data from Ontario to compare yearly hand hygiene compliance rates to quarterly MRSA rates and monthly CDI rates. The study found that despite increases in hand hygiene compliance there was little change in MRSA and CDI rates over this period. The author then concluded: "This study supports the emerging evidence that once a threshold level of hand hygiene compliance is achieved, there is very little if any benefit to attempting to achieve higher rates of hand hygiene compliance among health care providers."
Well, except that you can't conclude that from such a study design. For one, the author didn't have exposure and outcome from the same time periods. Why would we think average hand hygiene compliance over an entire year would correlate with monthly CDI rates and quarterly MRSA rates? And how can we not consider other factors at play like the emergence of NAP1 or CA-MRSA during this period? Maybe there's even a Simpson's Paradox here, but that's a topic for another day. Oh, and keep washing your hands. I doubt we've reached a "threshold" of compliance!
Addendum: Probably the biggest flaw in this study is the accuracy of the reported hand hygiene compliance rates. No doubt the rates are lower than reported.
Eli N. Perencevich, MD, ACP Member, is an infectious disease physician and epidemiologist in Iowa City, Iowa, who studies methods to halt the spread of resistant bacteria in our hospitals (including novel ways to get everyone to wash their hands). This post originally appeared at the blog Controversies in Hospital Infection Prevention.