Research 101: Part 2

Dealing with the unexpected 

In this series, Research 101, we are taking you behind the scenes of a research project at the Institute for Work & Health (IWH), from start to finish.

In Part 1, we introduced you to the project’s lead researcher, IWH Scientist Dr. Peter Smith. We explored the formulation of the research question: Why, over a 14-year-period, have lost-time claims in Ontario decreased by more than 40 per cent while no-lost-time claims have only decreased by four per cent? And we followed Smith and his IWH research team through to the good news that his $200,000-plus grant proposal had been awarded by the Workplace Safety and Insurance Board (WSIB)’s Research Advisory Council.

Roadblocks threaten timelines

In January 2008, the research begins. As is often the case in research, unanticipated roadblocks crop up. Smith and his team come up against three challenges in particular that threaten to derail the project’s timeline.

First, in order to explore trends in no-lost-time WSIB claims over time, 10,000 such claims must be “coded.” What Smith and his colleagues soon find out is that “coders” are in high demand, and it can take up to nine months to train a person to do this specialized work correctly. That kind of time just isn’t built into the two-year research proposal.

Second, the research team needs a tailor-made database to code and store the no-lost-time claim information. One will have to be designed.

Third, in order to compare lost-time and no-lost time claims, the research team needs to manage information from a large number of claims. This takes special expertise.

When roadblocks happen, you have to think of alternatives — fast, says Smith. You have a fixed budget and a fairly inflexible timeline, so you don’t have the luxury of waiting around for things to fix themselves.

A WSIB expert is found

To answer the research question, Smith and his team need to understand changes in the types of no-lost-time claims submitted over the study period of 1991 to 2005. To do this, the researchers propose to extract and store information on 10,000 no-lost-time WSIB claims.

Both lost-time and no-lost-time claim forms are stored at the WSIB. However, detailed injury and claimant information is routinely coded and electronically stored by the WSIB for lost-time claims only. That is, only lost-time claims have numerical codes assigned to indicate information such as the nature of the injury, the body part injured, the event leading to injury, the worker’s job tenure and more.

This more in-depth information is not coded for no-lost-time claims. That means it’s up to Smith and his team to ensure 10,000 randomly selected WSIB no-lost-time claims are coded. We thought we’d simply hire a project manager to train four coders, says Smith. We thought it might take about a month to train a coder — if the person had the right background in nursing or physiotherapy, for example.

It turns out Smith’s estimate is way off. Linda Kacur, an expert coder, is seconded from the WSIB for a 15-month term as project manager. She informs Smith that training people to the specific standards set by the WSIB in order to produce information that is both reliable and valid can take up to nine months. Furthermore, already-trained coders are scarce. An alternate route is needed.

Smith and Kacur put their heads together and come up with a solution: Kacur volunteers to take on the mammoth task of coding all 10,000 claims herself. She estimates she can code 50 claims a day and, in the end, it will only take two months longer to code the 10,000 claims than originally anticipated. That’s far better than the extra eight months the research team faced had they been forced to train coders from scratch.

Internal IWH experts solve problems

While this roadblock is being cleared, Smith also faces another challenge: the project needs a customized database so that the no-lost-time claims can be easily coded and securely stored. This time, Smith turns to an internal resource to help him out: Institute Analyst Michael Swift. He constructs a tailor-made “access” database system that is user-friendly and can pull the data from the claims needed for the project. The database is ready to go in two months.

Another key part of the research involves analyzing and comparing trends during the study period, 1991 to 2005, for no-lost-time and lost-time claims across age, gender and industry group — the only information routinely collected from no-lost-time clams. Again, IWH expertise is available: Data and Information Systems Research Associate Marjan Vidmar has expert skills in the management of large datasets.

If it wasn’t for the expertise here at IWH, I don’t know if we would have overcome these obstacles, says Smith.

* * *

With Kacur coding the files (by the end of November 2008, she is more than halfway through), the customized database up and running, and the claims from the WSIB extracted and securely stored at the IWH, the research is back on track. When all is said and done, the roadblocks caused a two-month delay, says Smith. Not bad, when you consider the delay could have been half a year longer.

The team now hopes to have enough information to begin examining trends in no-lost-time claims by the spring of 2009.

In Part 3: Will the research team have preliminary results ready to share in time for already-scheduled presentations?

Source: At Work, Issue 55, Winter 2009: Institute for Work & Health, Toronto.