A just released report says that surveyed injured workers are mostly satisfied with their medical care. When you look closely at the report, you find some disturbing problems which throw into question the summary conclusion.  

The authors can be thanked for attempting the hard task of measuring claimant experience. Maybe their report is useful, in part, as a case study of pitfalls in conducting a worker survey.

The Pennsylvania Department of Labor and Industries periodically surveys injured workers who file and are awarded workers’ comp benefits. Its most recent study was reported in the January 2016 issue of the Journal of Occupational and Environmental Medicine (“Factors That Predict Satisfaction With Medical Care: Data From 27,212 Injured Workers Surveyed for 14 Years”).

From 2000 to 2013, between 80% and 88% of survey respondents said they were satisfied with their medical care.

The article says, “Patient perception of a correct initial diagnosis and the explanation of WC rights were strong predictors of treatment satisfaction.” Among the 64% who said their initial diagnosis was correct, there was a 92% satisfaction rate, compared to a 69% rate among those whose initial diagnosis was incorrect. Roughly the same levels of and contrasts in satisfaction levels are found when the doctor explained the diagnosis and explained treatment options.

I would like to take these results at face value. They point to how injured worker experience can improve by inducing doctors to practice better medicine, including patient counseling. However, there are several flaws in the surveys used.

First is a low response rate: about 20%. The response rate varied a lot among sub-state regions.

Second, the article cites some unsettling findings that did not make their way into the sunny impression found in the article’s abstract. (Abstracts in research articles are often as far as readers get.) The adverse findings are at least as interesting as the positive ones.

For example, about 40% of injured workers did not have their workers’ comp rights explained to them by the insurer.  And 80% of the time, the insurer did not explain the worker’s right to see another provider if dissatisfied with care.

Third, the raw survey data might show big differences in response by severity of disability, but these were not reported. They put all respondents into one group. You have to go to the state’s own May 2015 report to learn that, historically, about 10% of respondents had medical-only injuries; that rose to 30% in the most recent survey. And neither the state’s report nor the research article segmented lost-time compensable injuries by duration of disability. One would expect that workers who returned to work within a month have a much different medical care experience than those who are disabled for an extended time.

And there is another issue, an admittedly hard one to address: Most work injuries are self-healing. The body does the work by itself. The doctor’s role is relegated to the role of a coach, or a placebo. This role should not be discounted, but it is dramatically different from where the severity of the injury really tests the doctor’s diagnostic and treatment acumen. What does worker satisfaction with care mean when the body heals itself?

There’s much to admire in the Pennsylvania report. The state’s report especially includes a lot of interesting findings on workers and on medical providers. For instance, the survey asked workers about what they felt about employer choice of provider. Depending on how the question was phrased, the responses differed. Medical treatment within provider networks set up by the insurers was associated with a much lower duration of disability than treatment without a network.

Dig into these details, but don’t pay much attention to the summary conclusions.