Fun with data

I've recently discovered that Google has embedded a fun Flash data visualization tool in Google Docs. I've tested it out using data on the amount of buprenorphine (a drug used to treat opioid addiction) distributed from 2005-2009, by state. Grams of buprenorphine are shown both overall and adjusted for the number of opioid users in each state. The other variable is the supply of physicians who are licensed to prescribe buprenorphine adjusted for the number of opioid users in each state. Check it out.

Happiness and focus

I've been taking part in the Track Your Happiness project for the past few weeks, and as of today I finished the 50th sample and have access to my happiness report.

Overall, I am slightly happier at work than at home, cooking and doing homework make me happy, and I'm happiest when doing something I want to do that I don't have to do.

I found this chart especially interesting (happiness on a scale of 0 to 100 is on the vertical axis, and focus is on the horizontal axis).

Focus

Drug costs

I'm working on a project involving a health economic evaluation of a clinical trial. The trial collected data on drugs prescribed for each patient over the course of, let's say, a year. My job is to assign a cost for each row in a very large Excel spreadsheet (each of which represents one prescription). The data I have are sometimes incomplete, but in general, I have: Drug Dose (a number), Drug Unit (eg, milligrams), Drug Frequency (free text), Drug Route (oral, topical, intramuscular, etc.), and Drug Name (free text).

The process involves a cumbersome program called Red Book for Windows, which allows me to look up the drugs by name or by National Drug Code (NDC) -- unfortunately, I don't have the NDC, so I have to look them up by name. The first problem is that many of the drug names don't match. Let's take, for example, ciprofloxacin, a generic antibiotic often referred to as cipro.  Cipro is sold in dozens of forms (pills, eyedrops, eardrops, etc.) in dozens of strengths by dozens of manufacturers. It's listed in Red Book in lots of different ways, including: CIPRO, CIPROFLOXACIN, ciprofloxacin, ciprofloxacin/ciprofloxacin hydrochloride, ciprofloxacin HCL, CIPROFLOXACIN HYDROCHLORIDE, ciprofloxacin hydrochloride, ciprofloxacin hydrochloride/dexamethasone, ciprofloxacin hydrochloride/hydrocortisone, CIPROFLOXACIN IN DEXTROSE, CIPRO HC, CIPRO IV, and CIPRO XR.

The cost of a 500mg tablet may be listed under any of these, and I have to locate the lowest cost available (because we're using a conservative approach to costing - rather than use an average, we use the lowest cost in order to underestimate the potential burden or cost of the illness). Then I can apply that cost to all the instances where it shows up in my spreadsheet. Then I have to go find the lowest cost for 200mg delivered intravenously, and so on.

So this fun project has been eating up lots of my time lately. It's not exactly brain surgery, but it can't be automated and can't be outsourced - and it has to be done very carefully. It would go a lot faster if the Red Book program were easier to query, but, in my opinion, it's not a user-friendly application at all!

Health reform: preventive care for all

I honestly didn't pay that much attention to the measures in the health reform bill. I was supportive, in theory, while being cynical about its chance of passing (especially since January) and about how much good it might actually do. But now that it's passed, I've been taking a closer look. Guess what? There are some really, really great things in the bill.

As a public health person at heart, I am especially thrilled about the measures in the bill that will encourage prevention of disease with first-dollar coverage of important screening procedures. Here are just a few of the recommended services which insurers (private and public) will have to offer and provide with no deductibles, co-payment, or co-insurance within the next 6 months:

• mammography and genetic risk assessment for breast and ovarian cancer;
• cervical cancer screening (Pap smears);
• colorectal cancer screening;
• screening adults for depression;
• intensive behavioral dietary counseling for adult patients with known risk factors for cardiovascular and diet-related chronic disease;
• oral fluoride supplementation to preschool children older than 6 months of age;
• screening for high blood pressure in adults aged 18 and older;
• screening and behavioral counseling interventions to reduce alcohol misuse by adults, including pregnant women, in primary care settings.

For the whole list, see this AHRQ page.
And for more, see this Health Reform Talk post.

Comparative effectiveness

Comparative effectiveness research (CER) is the latest buzzword in my field, in large part because Obama has promoted it as an important component of healthcare reform and included $1 billion of funding for it in the stimulus bill. This Reuters article discusses a new study by Hochman  and McCormick finding that only 32% of drug studies published in top medical journals compare the effectiveness of existing treatments. Most of the time, studies compare the new drug to placebo because that's what the FDA requires. In other countries, such as the UK, health technology assessments are required that compare the new drug to the standard of care.

The study is published in this week's issue of JAMA - here's a link to the abstract.

Couple of thoughts:

• They only looked at articles in the six top-ranked general medicine and internal medicine journals. If I were going to try to publish a comparative effectiveness study, I would think it probably wouldn't get in to a top-ranked general journal (JAMA, NEJM, etc.), so I'd probably target it to a specialty journal. 
• Pharma company-sponsored research often seems to be less favorably reviewed than academia-sponsored research, plus companies are often in a hurry and don't want to go through the excruciatingly long review processes at the top journals.
• They point out the lack of cost-effectiveness (CE) studies. Again, if I were going to try to publish a CE study, JAMA and NEJM would be my last choice for a target journal unless I thought I had a real blockbuster... they just don't publish many CE studies.

In short, I think this study is majorly confounded by the conventions of the medical publishing world.

Lit review methods at the abstract stage

With the volume of medical literature that's published, even a relatively narrow search may get thousands of hits in PubMed.  A lot of people have published on the topic of rating full text articles - see this, for example - but when you're doing a large-scale lit review, it's often necessary to do some paring down at the abstract stage. 

What I usually do is work through my pile of abstracts (easiest for me to do this on hard copy) and make clear notes on each one about whether or not it fits my search criteria. This is usually not that hard.

For example, say I'm doing a lit review of cost-effectiveness of treatments for multiple sclerosis. Up front, I have a list of reasons for excluding articles altogether:

• Not an economic analysis
• Review/opinion article with no meta-analysis
• Fewer than 30 subjects per study arm
• Trial not conducted in a relevant country (often I'm looking specifically for US studies)
• Study concerns a treatment that is no longer in use in the US
• etc.

Once I have narrowed it down to just the abstracts for articles that might contain something relevant, I usually pull the full texts of all of them. Even if a particular article itself isn't useful, I still need to go through each bibliography to identify articles that I may have missed in my searches of lit databases.

Mistakes Were Made

Database analyses are the hardest part of our work, in my opinion.  I've mostly managed analyses of large government databases so far - like Medicare claims data and data from the Hospital Cost and Utilization Project (HCUP). The difficulty is really that it's hard to know when you get the wrong answer.  If you get something that looks right, you have to pretty much trust it once you've checked all the SAS code and any formulas used to get results into Excel or wherever the results are being displayed. 

It's different from clinical trial data analysis because in a clinical trial, you always have a source document to go back to.  The source data in a large government database are often so numerous that it's really hard to go back and verify that the analysis is correct.  I guess it's similar to modeling in some respects, but models can often be recreated in another language to verify the results.  Databases could conceivably be analyzed in multiple programs (like SAS and STATA, for example), but that would probably add 10-25% to the timeline and budget.  In my small company, at least, we don't have the luxury.

At least the error I caught today was detected before we had already published the results!