People who have pain are faced with the challenge of finding a way to express what is a subjective and multidimensional experience, usually by choosing a number on a pain rating scale or by verbal pain report. But how do individuals make such decisions about their pain? It’s an important question for researchers and clinicians, who rely on pain scales and verbal reports to conduct experiments and provide treatment for patients.
While many cognitive psychology studies have examined the effect of metacognition, or “thinking about thinking,” on a variety of decision-making tasks, no research had examined what role this higher-level cognitive process may play in subjective pain ratings.
Now, with a study of healthy volunteers subjected to experimental heat pain, researchers at the National Institutes of Health’s National Center for Complementary and Integrative Health, Bethesda, US, show that reaction time – how quickly subjects rated their pain – predicted their confidence in their subjective pain rating. This association shows that people can make metacognitive judgments about pain, namely that they are not always confident in their judgments and are aware of that.
The findings are consistent with studies of other decision-making tasks. The authors say future pain studies should take metacognitive factors like confidence into account, which could ultimately lead to more tailored treatments for patients.
“Measuring pain has always been a challenge in the absence of an external ‘gold standard’ metric. Pain is very subjective, similar to other emotional experiences,” said Johan Vlaeyen, University of Leuven, Belgium.
“This study offers a sensible, observational approach to better understand the parameters that influence decision making during a pain intensity rating task, with a focus on metacognition,” according to Vlaeyen, who was not involved with the new research.
The study was published December 7, 2020, in Scientific Reports.
An experimental setup to assess metacognition about pain
Lauren Atlas, lead author on the study, said her prior work testing expectancy effects on pain naturally led her to wonder whether there are decision-making biases that contribute to pain reduction in response to a placebo.
“There’s been some debate whether placebos affect nociception or are just influencing decision making in some way,” Atlas explained. “We, as well as other placebo and expectancy researchers, have looked at computational models to look at the decision-making process itself. And we thought we could easily build on standard pain assessments by measuring confidence.”
“There’s already a huge body of work on confidence and metacognition in decision-making research that we could potentially extend to pain,” Atlas continued. “This kind of approach might give us some insight on how people are making decisions, which we can later apply to studies of placebo or other types of cognitive factors that may influence pain.”
Focusing on the relationship between confidence and pain ratings, Atlas, first author Troy Dildine, and Elizabeth Necka recruited 80 healthy volunteers without a history of chronic pain for their study. Subjects underwent 24 trials of noxious thermal stimulation of varying intensity on their non-dominant forearm and then rated their pain on a visual analogue scale ranging from 0-10 (0 meaning no sensation and 10 the most pain imaginable).
The participants were also fitted with an eye tracking device. This was so the researchers could assess where the subjects’ gaze went on a computer screen on which the subjects saw the pain scale so they could make their pain rating decisions. Previous research had shown that less confident decisions are associated with an increased number of eye fixations on visual alternatives.
For each trial, to help calibrate the eye tracking device, participants were asked to fixate for 500 milliseconds on a black box that appeared at the center of the computer screen. Then, they experienced eight seconds of a heat stimulus, which was followed by the appearance of the pain rating scale on the screen. The scale stayed on the screen for three seconds and during that time participants were asked to think about their rating as eye movements were recorded.
Further, once an arrow appeared on the screen, participants used the arrow to select the number that matched their pain level and also verified that rating verbally. Finally, study participants then rated their certainty of their pain rating on a scale of 0 (completely certain) to 100 (completely uncertain).
With the above setup in place, the researchers could now measure reaction time (how long it took participants to enter their pain rating on the computer), as well as eye movements and confidence scores.
Less confidence means slower reaction times and more eye fixations
The first finding from the study was that there was some variation in how confident participants felt in their pain ratings. Further, participants took longer to rate their pain when they were more uncertain about their rating, but uncertainty also went down with later trials. When looking only at trials in which participants reported pain, the lower the temperature of the heat stimulus, the greater the uncertainty.
To delve further, the researchers next used a multilevel statistical model that considers not only within-subject factors upon which the initial findings were based, but also between-subject factors. Here too, they found an association between uncertainty and slower reaction times; the more time participants took to make a rating, the more uncertain they were about their choice.
The researchers also found that more eye fixations were associated with a greater likelihood of participants’ being uncertain of their ratings, in those who were more reliable in associating temperature and pain. For less reliable participants, the more eye fixations they had, the greater the likelihood they were certain of their pain ratings. Finally, the researchers saw once again that, as the trials went on, participants showed less uncertainty.
Together, the results show that, at least in a lab setting with healthy volunteers subjected to experimental pain, people make metacognitive judgments about their pain and have awareness of their confidence levels. In addition, their judgments are linked to how sensitive they are to changes in temperature.
Reconsidering patient encounters and pain research
Troy Dildine, first author of the study, says the results argue for a more critical approach to pain scales.
“There’s a lot of consensus in the field that subjective pain reports are very important to making decisions about treatment, so I don’t want to minimize these scales in any way,” he said. “But I think it is important to consider that we may need other resources or different ways of getting information from the patient to really understand their experience. We should be critical of how we are using these scales. And we should think about how patients understand the scale and how to use it so we can be sure we are getting the information from them that we really need.”
Atlas added that an understanding of metacognition about pain could help the doctor-patient interaction.
“If you find out a patient is less confident about a pain rating, you might ask about different features of the pain or the emotional burden of the pain to help the patient communicate what their experience is like to the physician,” she said. “Having that kind of information could be quite useful.”
Vlaeyen said it is likely too early to derive clinical implications from these results. He also would like to see the authors apply a drift diffusion model, a type of statistical model that can describe behavioral performance across multiple factors for decision-making tasks. Such a model, he argued, could help reveal how prior knowledge about the expected stimulation intensity may influence ratings. That said, he thinks the current work offers some relevant insights to pain researchers and may inspire them to think more about how they design future studies using pain rating scales.
“The focus on metacognition and the influence of uncertainty on pain ratings, with the relationships with reaction times and eye fixation variability as implicit markers of such confidence, is admirable,” he said. “There are some interesting implications for researchers, such as the inclusion of practice trials to have participants learn to better judge the intensities of the painful stimuli.”
Both Atlas and Dildine wish to continue to use reaction times and eye fixations to study other contextual influences on pain. She also agreed with Vlaeyen that using a drift diffusion model to further investigate the results would be of value. The current study, she said, provides a baseline for better understanding of pain decision-making in future work.
“We didn’t even know if people would be able to rate their confidence in their subjective rating in a meaningful way because pain is ultimately defined as a person’s subjective experience,” Atlas said. “This study really was a necessary first step to see if people could introspect on their pain experience, and we’ve shown that they can.”
Kayt Sukel is a freelance writer based outside Houston, Texas.