Model Naked: Some Thoughts on Transparency in Health Economics Writing

By Caitlin Rothermel, MA, MPH

Caitlin Rothermel

Caitlin Rothermel

Recently, Patti Peeples at HealthEconomics.Com interviewed me for a podcast on health economics writing. You can listen to it here (and I hope you will), but after it was over I realized I had a bit more to say about modeling, writing, and transparency.

First off, I really like writing health economics research. When studies are well designed, it is a pleasure to explain researchers’ assumptions, model design, and study outcomes. It is like putting together a very attractive puzzle, one that’s just complex enough to hold your interest all the way through.

That said, some of us are good at building research (that’s probably you, reader), and some of us are good at organizing and presenting it (that’s me). I think that one of the biggest concerns today in health economics communications is creating and presenting genuinely transparent cost models. And of course, developing a transparent model is only the first step in this process – it has to maintain transparency and make sense to a range of readers when it’s summarized in print.

Recently, I read an interesting presentation from a 2011 Drug Information Association workshop. Medina and colleagues surveyed payers (n=12, primarily pharmacy directors) and pharmaceutical manufacturers (n=13) to evaluate the relative usefulness and importance of dossier submissions’ clinical and economic sections. This research focused on Academy of Managed Care Pharmacy dossiers, but its take-home is key for any publication type: “Payers felt that [economic] models were only somewhat transparent, whereas manufacturers believed their models were transparent to most, if not all, health systems.” In addition, payers thought most models were not clear and/or easy enough to use.

It might be fair to argue that part of the problem is that pharmacy directors are not versed in health economics. There may be a genuine need for more education in this area. But this justification isn’t much consolation when your hard work is thrown in the trash because the target audience just doesn’t get it.

I’ve picked up some good tips from listening to experts and payers. Here’s the first: As much as possible, make sure your model design reflects the real world. Transparency doesn’t mean very much if a model’s steps or transitions don’t match the actual course of disease progression or what would be considered standard practice in clinical care. For example, when modeling a chronic disease, think twice about using an approach that assumes patients will remain on the same treatment for decades or even a lifetime. A hallmark of chronic disease is its progressive nature and the fact that treatment is constantly updated.

Don’t torture the data just because you want to present outcomes in a certain way. The harder you have to work to make your point, the more confusing your explanations are likely to become. You’re digging a hole for yourself, and readers don’t necessarily want to follow you down that hole. Think of it this way – most of us are exposed regularly to tweaked budgets of some sort, fudged or failed projections, or estimates once considered accurate and now recognized as horribly off. Educated readers are increasingly savvy to potential distortions, and are likely to overlook monetary details or calculations that don’t make reasonable sense.

Another way to look at your writing is to consider what turns readers off, and do what you can to avoid it. When people become confused by a document’s content, they tend to put it down and not pick it up again. I’ve known people who refused to continue reading a document with the inappropriate use of hyphenation. Here’s something I’ve seen in health economics – not providing recommended information because it is considered intuitively apparent. For example, what is the point in explaining why a model covers only a specific time frame, when this is readily apparent from context? In this case, imagine an in-hospital treatment where all side effects will show up within one year or less. It’s generally best to insert a sentence stating the reason for the decision. Yes, it may seem simplistic, and your intention may be evident to 90% of readers. But what about the other 10% who simply stop in their tracks and don’t go further when they reach a point of non-understanding?

Of course, a consistent barrier to writing up health economics research is having too much to say. It is always going to be challenging to communicate health economics data in existing journal length format. In general, the word count for original research in a medical journal (whether it’s a traditional journal or one focused on health economics) is 3,500 or 4,000 words. A few publications, like Health Economics, allow up to 5,000 words, but they are the exception and not the rule.

For the writer, this means you need to be concise and to the point. If you can’t achieve all of your goals in the space allotted, take advantage of the online appendices that many journals offer (I would especially encourage using appendices to provide information on study assumptions). If this is not possible, you’ll have to work harder to streamline the content you are able to present. Don’t make non-intuitive leaps or leave gaps in your model explanation just because space is tight. If you find yourself in a situation where you run out of room to summarize your model methods or results and are thinking of cutting back, ask yourself honestly if you’ve padded your Introduction and Discussion sections. Remember, good opening and closing arguments for your study’s relevance will only go so far if the reader is not convinced that your research strategy or results are valid.

One way to make sure your document is concise and complete from the first draft on is to outline it first, based on the sections of the document. To get an idea of the word count you are headed towards, take a look at the word count of your outline; mentally add on the amount of additional content that will be needed to fully flesh it out. Stay within the limits you set.

This gives me a chance to make a plug for having a health economics-oriented medical writer help you out, either at the beginning of your project, or at least before final submission. Medical writers are all about attention to detail, looking for gaps, and making sure the content we’ve been given matches the publishers’ specifications. Even if you can’t afford a professional writer, accept that there will typically be gaps in your documents that you will not be able to see. Prior to submitting your document, have someone who you trust review it closely. Choose someone quantitatively oriented who is willing to look at your model from the roots up, and who will let you know if there are points where their understanding falls off.

The work health economists do is complicated and meticulous. If your writing is clear, even non-professionals should be able to review and basically grasp your studies’ goals, implications, and key findings. Bottom line – strive for transparency, clarity, and simplicity. When writing achieves these goals, it not only facilitates the accurate communication of information, it builds trust between stakeholders.

Caitlin Rothermel, MA, MPH is a medical and health economics writer. She lives in Vashon, WA with her family. You can learn more about Caitlin by visiting http://www.MedLitera.com.

Caitlin Rothermel

Caitlin Rothermel, MA, MPH is a medical and health economics writer. She lives in Seattle, WA with her family. Thanks to genetic sequencing, her husband now knows that he is of Irish extraction (this made him very happy). You can learn more about Caitlin by visiting http://www.MedLitera.com.

Leave a Reply

Your email address will not be published. Required fields are marked *