Kaiser Permanente has released a study from its EMR database looking at use of vaccines in its Colorado region. KP in Colorado has data on about 480,000 members dating back to the mid-1990s from when they started implementing the first EMR. After that system was retired and they moved to Epic the old data is in PDF format for current records, but is also in a database for research use. I spoke to the researchers Jason Glanz & Ted Palen from the KP Colorado Institute for Health Research late last week.
Essentially the problem is that several studies have shown vaccines to be safe but some parents are really concerned, prompted in large part by certain celebrities (with former Playboy model Jenny McCarthy being among the most vociferous claiming that vaccines cause autism), and partly because they don't believe the diseases the vaccine prevents are serious.
The prevalence of whooping cough is not huge. In 2007 CDC estimated there were about 10,400 cases in US, with an average of 800 per year in Colorado. The school exemption rate for vaccination is 1–2% which is a proxy for refusals. The rate of refusal for KP members is less than 1% (which may be just missing one dose of one vaccine from the EMR data). The rough estimate of whooping cough is 20 cases per 100,000 per year. KP’s rate in Colorado is estimated at 40 per year—but the real rate is much more as only an estimated 10% of actual cases are reported.
Whooping cough, defined as a cough lasting more than 14 days (including incessant coughing & vomiting) can be very serious and can be cause of hospitalizations and death in young children. In 1930’s & 40s it had a 10% mortality rate, although now it’s only about 1%. So for now it’s not such a big deal but the concern is that numbers will go up overall, and increasingly more kids won’t be protected. The vaccination rates need to be well above 90% plus to be effective for the community as a whole.
The KP researchers identified the cases of kids with whooping cough, selected a parallel group who didn't have the condition, and looked into the record to see who had vaccinations. There were 156 cases over 12 years. In that population, 18 had refused the vaccine, which means that the unvaccinated were 23 times more likely than those fully vaccinated. 1 per 20 ‘refusers’ got the disease versus 1 per 500 vaccinated kids.
As a side-bar, this type of study is one coming out not just from the electronic records at KP, but also from an alliance of 16 HMOs are in a virtual data warehouse which has records on 15 million people (in US and in Israel). That data can be queried for different projects. A series of other studies are now being done including looking at cancer, heart disease, drug safety, vaccine safety, plus consortium studies for looking at rare diseases. Glanz said that the trend is to collaborate over these and to combine this clinical data with genetics research.
So expect much more information on the impact of real time health care using these massive databases, and let’s hope that these data drive decisions. Of course in a world in which there’s massive disagreement over the science from people like Jenny McCarthy, it may not make all the difference we could hope for.