What’s in a Number? Rethinking a Patient’s Pain Rating
Let’s do an experiment: write down a number between 0 and 10 that indicates how happy you are: 0 means not happy at all, and 10 means extremely happy.
Now that the number is written down, you may not change it. Why did you choose that number? What information did you need to choose a number? What information did you use to choose a number? What additional information did you want?
What else did you use? What information did I give you in order for you to return a number to me? What information did I not give you?
One key component missing from this exercise is the social interaction piece (you can’t see me and I can’t see you). That aside, if you’re like most people, the information you used to choose a number is far more complex and diverse than you realize.
The Goalposts
Let’s start with the scale itself. The scale uses 11 potential response options, 0 – 10. Why 11? The answers to this question are both conceptual and empirical but are far from definitive.
Why use a 0 – 10 scale?
From a conceptual standpoint, 0 – 10 makes sense to most adults. It’s a scale encountered in our lives, at athletic events like diving or gymnastics, in common use like the Richter scale, and it’s easily converted to a percentage. The familiarity means people can easily understand what you’re asking.
It seems to work empirically too. On balance, the 0-10 Numeric Rating Scale (NRS) works at least as well as others with more complex options or different ranges.1,2 Even though concept and empiricism appear to hold, we still have not answered, “Why 0-10?”
What do the anchors represent?
Consider the anchors for 0 and 10 that I gave – no happiness to extreme happiness.
The 0 anchor is usually straightforward on 0-10 scales, representing a complete absence of the construct, in this case a complete absence of happiness. If 0 is a complete absence, then what’s the opposite of nothing?
[Note: We cannot assume 0 to mean extreme sadness. These are two separate constructs and the only way to capture sadness is to ask a dedicated question. Opinion-based scales – those that tend to range from strong disagreement to strong agreement – use a true ‘0’ (a complete absence of opinion) in the middle of scale, where the neutral response lies.]
Mathematicians would tell us the answer is infinite (1/0). However, we need to place an anchor on it so the question can be answered and interpreted.
What anchors do you choose for pain ratings?
The pain rating transaction between asker and rater depends on the anchor you place on the extreme limits:
- 10 is the worst pain you’ve ever experienced.
- 10 is the worst pain you can possibly imagine.
These are two very different anchors. Most adults have never experienced the worst imaginable pain. For most patients, the worst pain they’ve experienced is lower than the worst imaginable, meaning the scale range – the goalposts – are different, and affect how people understand the given options.
In a clinical setting, we must recognize that by shifting these goalposts, we can influence a patient’s score and alter the understanding of their condition or treatment.
Aversion to Extremes
The second consideration requires a quick story:
In 2013 I wrote a chapter on pain assessment for the 3rd edition of the American Society of Hand Therapists book ‘Clinical Assessment Recommendations’. As part of that process, I reported some normative values for different clinical populations. I reviewed clinical trials or epidemiology papers for many conditions, such as complex regional pain syndrome, diabetic neuropathy, osteoarthritis, and repetitive strain problems. Upon completion, I found remarkable consistency across conditions – the mean of 0 – 10 NRS scores neared 5, with a standard deviation of around 2. This pattern was so persistent that when I see a mean pain rating dramatically different from 5 or SD 2, I question if the sample was representative.
Why does this pattern exist? Let’s consider two possibilities:
- The single-item 0 – 10 NRS is exceptionally psychometrically sound.
- Patients are deliberately avoiding the extremes.
I favor the latter and see it as another example of interpersonal transactions. Think back to your happiness score – while happiness is less stigmatized than pain, you still probably didn’t choose 10.
As a patient being asked about their pain, you consider both the intensity of the current experience (which itself is difficult to quantify), but also how you perceive the clinician will respond. If you appear strong and rate low, you risk not receiving the appropriate care. If you try to be honest and rate high, you fear the clinician won’t believe you or think you are weak. When we are boxed in, we tend to run straight up the middle. The aversion to rating extremes is a well-recognized bias, and nothing to do with the actual intensity.3
The Man in the Mirror
The rater influences the patient’s response in this transaction as well. Abundant research suggests that we can use operant conditioning to influence a patient’s rating depending on our responses (either reinforcing or extinguishing feedback).4 Interestingly, it seems the influence of the clinician (‘rater’) on responses of patients is different between men and women.5 Additional influences on pain assessment and ratings include:6,7
- Clinician’s demeanor and warmth
- Clinician’s tone of voice
- Patient’s feelings on being heard and supported
The ubiquitous 0 – 10 Numeric Rating Scale is not simple or universal. You can tap into the other influences on a patient’s experience of pain (including psychological, other social, and psychophysical influences) but the interpersonal transaction is unavoidable. We’re asking the patient a question, and there’s no way to separate the experience of pain from the interpersonal transaction – the patient considers not only what they’re feeling but also what they believe you need to hear to properly treat them.
We are social animals, and the 0 – 10 NRS is not immune to these effects. Consider this the next time you ask a patient to rate their pain, especially if you think their answer may be disproportionate to their experience. Perhaps the reason for their strange response is staring back at you in the mirror.
- Jensen MP, Turner LR, Turner JA, Romano JM. The use of multiple-item scales for pain intensity measurement in chronic pain patients. Pain. 1996;67(1):35-40.
- Jensen MP, Miller L, Fisher LD. Assessment of pain during medical procedures: a comparison of three scales. Clin J Pain. 1998;14(4):343-349.
- Neumann N, Bockenholt U, Sinha A. A meta-analysis of extremeness aversion. J Consum Psychol. 2016;26(2):193-212.
- Jolliffe CD, Nicholas MK. Verbally reinforcing pain reports: an experimental test of the operant model of chronic pain. Pain. 2004;107(1-2):167-175. doi:S0304395903004329 [pii].
- Gijsbers K, Nicholson F. Experimental pain thresholds influenced by sex of experimenter. Percept Mot Skills. 2005;101(3):803-807.
- Jackson T, Iezzi T, Chen H, Ebnet S, Eglitis K. Gender, interpersonal transactions, and the perception of pain: an experimental analysis. J Pain. 2005;6(4):228-236. doi:10.1016/j.jpain.2004.12.004.
- Shields CG, Finley MA, Elias CM, et al. Pain assessment: the roles of physician certainty and curiosity. Health Commun. 2013;28(7):740-746. doi:10.1080/10410236.2012.715380.
