Research 101: Part 5

Disseminating findings

In this series, we are taking you behind the scenes of a research project at the Institute for Work & Health (IWH), from start to finish.

We met the lead researcher, IWH Scientist Dr. Peter Smith, who told us about his study comparing trends in lost-time versus no-lost-time workers’ compensation claims in Ontario (Part 1). We learned about data collection (Part 2), early results (Part 3) and the factors considered when picking a journal for potential publication (Part 4). Now, almost three years later, journal articles are finally being published.

First paper in print

In early 2012, five years since Smith submitted his grant proposal for funding, the first paper from the study is in print. Pages 84 to 91 in the January 2012 issue of the American Journal of Industrial Medicine (vol. 55, no. 1) carry an article titled “Comparing the risk factors associated with serious versus less serious work-related injuries in Ontario between 1991 and 2006.”

Interestingly, although the first paper to be submitted under this project, it was not the first to appear in print. That’s because the road to getting published sometimes takes a circuitous route.

First paper takes two tries

Smith and his team planned on writing three papers based on this research project: two based on an analysis of the administrative data from Ontario’s Workplace Safety and Insurance Board (WSIB), and one based on the more detailed “coded” data (Part 2).

The first of the papers based on the administrative data compared lost-time claims (LTCs) and no-lost-time claims (NLTCs). It was initially submitted to the journal Injury Prevention in July 2010.

We chose that journal first because it’s more general, with a high impact factor and reaching a broader audience, explains Smith (see Part 4 on the factors considered by researchers when choosing a journal). The paper was not accepted.

Nonetheless, the exercise of submitting to Injury Prevention was not in vain. The worst-case scenario is that you send your article to a journal and it says your study is just not new or interesting enough, says Smith. That was not the case here. The reviewers were interested in the question and could see the importance of comparing LTC and NLTC data.

What the reviewers were concerned about was a novel approach the team took to a final regression model. Two out of the three reviewers gave us very constructive feedback and helped us refocus the paper, Smith says. We ended up using a more traditional approach to the regression model, explaining in more detail what it meant.

Based on the feedback, the research team also tightened up the introduction to the article to focus the paper more clearly on the question that they sought to answer. When you are working with administrative data, there are going to be some questions that cannot be answered, so you need to make sure your introduction focuses your reviewers to clearly understand, ‘Here’s the question, here’s why it’s important, and now we’re going to answer it,’ says Smith.”

First paper finds a home

In the end, although the paper didn’t get accepted by Injury Prevention, the research team knew it was worthy of publication elsewhere. Sometimes a paper is rejected harshly, says Smith. Luckily, we got some good reviews back. We were rejected with enough optimism to keep going.

Keep going they did. After making changes to the article, Smith and his team submitted it to the American Journal of Industrial Medicine (AJIM) in January 2011. We decided to submit to a journal with a specific focus on occupational health he says.

As well, Smith knew the journal had recently published papers looking at questions similar to his. This meant the authors of those papers would likely be among his reviewers. It helps when someone reviewing your paper already has his or her head in the question you’re asking, he says.

Smith got comments back from AJIM in April 2011, resubmitted the paper in June and got accepted the following month. The paper then appeared online ahead of print in August and, finally, in print at the beginning of this year.

Second article gets published first

In the meantime, Smith and his co-authors wrote the second article based on the administrative data, this one looking at health-care expenditures associated with NLTCs. This was submitted to the Journal of Occupational and Environmental Medicine (JOEM) in September 2010. This journal was chosen because it, too, has a specific focus on occupational health and had recently published papers on a similar topic.

This paper turned around quickly. Smith got comments back three weeks after submitting it. He revised the paper and resubmitted it in November. Ten days later, he learned the paper was accepted. “Trends in the health care use and expenditures associated with no-lost-time claims in Ontario: 1991 to 2006” was published in the February 2011 edition of JOEM (vol. 53, no. 2, pp. 211-217). It turned out the second paper came out a lot earlier than the first, says Smith.

This wasn’t a problem for Smith’s project, because the two papers stood on their own. But sometimes a research team writes a more conceptual or theoretical paper first, and relies on this paper to support its argument in papers that follow. It can be real pain in the neck to have to hold off on submitting a second paper until you hear about the first, says Smith. We would have been holding off for over a year.

Third article is put on hold

With the two papers based on WSIB administrative data in the works, the research team was ready to tackle the third paper based on the analysis of 10,000 randomly selected NLTCs that had been coded by an expert seconded to the research team. The coder added to the NLTCs the type of detailed information usually only collected by the WSIB for LTCs; e.g. information on nature of injury, body part injured, worker’s job tenure and more (Part 2).

Smith and his team were getting ready to submit this third paper when they started to raise concerns about the interpretation of some of their findings based on the data available. They decided they needed more information before being able to complete the research and the final paper. They wanted to know how changes in the coded NLTCs compared to changes to matched LTCs over a similar time period. We couldn’t look at all LTCs because our sampling strategy involved a very particular subset of claims, Smith explains.

The team decided to apply for some top-up funding that would allow it to match LTCs to the coded NLTCs. It applied to the WSIB’s Research Advisory Council during its 2011 round of funding. A more comprehensive story could be told with some additional analyses, says Smith. So we decided to hold off on publishing that paper until we find out if we get the extra funding. More analysis will allow us to answer the ‘so what next’ questions and, ultimately, make the paper more compelling.

Unexpected “side paper” gets published

Although the research team only planned on three papers related to this project grant, a fourth did get published—a short report in the Occupational and Environmental Medicine that was first published online in September 2010 (vol. 67, no. 12, pp. 878-880).

Led by research team member Sara Morassaei, the research operations coordinator at IWH, the study examined whether switching to and from Daylight Savings Time (DST) is associated with an increase in work injuries. We had the data, it was an interesting question, we thought we could answer it, so we went ahead and tackled it, says Smith, the paper’s co-author. Sometimes an expected paper emerges from the results collected by a study.

The study showed that switching to and from DST does not make a difference to claim rates. This finding differed from previous studies that showed switching did made a difference in certain industry groups. A couple of papers published since ours also show it doesn’t make a difference, says Smith.

Findings reported to funder, stakeholders

Although the third paper is not yet published, the results of the analysis of the coded NLTCs have been presented publicly. Smith shared the results in October 2010 at a seminar held at IWH that was open to the public. (He had already presented the results based on the administrative data at an earlier seminar in March 2009.)

The most important finding emerging from the analysis of the coded claims is that the distribution of claims submitted as NLTCs did not change a lot over the 1991-2006 period, at least not among the firms sampled in this study. The hypothesis that we would see an increase in more traumatic injuries like factures and concussions among NLTCs didn’t happen, he says. They have remained consistent, accounting for about four per cent of NLTCs throughout the time period.

That suggests an increase in claims management is not the driver behind the increasing rate of claims submitted as NLTCs. But as Smith puts it, we need to know what is happening to LTCs among this particular group over the same time period so we get both sides of the picture—thus the request for extra funding.

Following the plenary, the study’s results were presented to WSIB policy-makers in November and to an injured workers’ group in December. A number of IWH publications also covered the study. It was the subject of an August 2011 Issue Briefing (http://www.iwh.on.ca/briefings/no-lost-time-claims) and a page-one story in Fall 2011 issue of At Work (http://www.iwh.on.ca/at-work/66/understanding-no-lost-time-claims-in-Ont...). It will also be covered in a future Research Highlight to be posted on the Institute’s website.

It’s always hard to measure the impact of any one research study, but Smith believes the project’s findings have added to the important discussion of how best to track what is happening within the workers’ compensation and injury prevention system. It’s fairly well accepted in the stakeholder community now that the seriousness of an injury cannot be determined only by whether time is lost or not from work, says Smith. I think we’ve added to that discussion. During the course of the project, WSIB moved away from focusing on LTCs only to focusing more on all types of claims.

Thinking back, looking ahead

Smith learned more from this project than the answers to his questions about LTCs and NLTCs. We learned a lot about how to study these claims, about the strengths and weaknesses of the WSIB administrative data and about the claim coding process, he says.

We also learned what information needs to be collected in order to truly understand the implications of changing trends in LTCs and NLTCS, he adds. In fact, the 90-page report to the funder, the WSIB’s Research Advisory Council, included a recommendation that NLTCs should note if the person is back at work at full capacity or on modified duties.

Smith found out during the last couple of weeks of 2011 that his grant application for top-up funding to further analyze LTCs had been approved. But the logistics have changed. Smith moved back to his native Australia in January 2012 to take up the position of senior research fellow with the Centre of Occupational and Environmental Health at Monash University.

Smith nevertheless remains heavily involved in a number of important IWH studies, including this one analyzing NLTCs and LTCs. This project isn’t over yet …

Source: At Work, Issue 68, Spring 2012: Institute for Work & Health, Toronto