What’s the score in pain assessment?

Self-report measures of pain intensity provide important primary outcomes in randomised trials of pain-relieving interventions. Scores from numerical, visual analogue, and other scales serve well in comparing group outcomes.1 But in the context of individual clinical treatment, certain catchphrases are commonly employed. Self-report of pain is “the gold standard”. Pain intensity scores are “the fifth vital sign”. Analgesic decisions are sometimes based on an “algorithm” that includes self-report of pain. Although such concepts help to draw attention to pain management, each one contributes to oversimplification of the process of assessing pain.

Studies show only modest concordance of self-report measures of pain intensity with observational measures and with judgements based on context, such as knowledge of the patient’s medical condition or of factors influencing his or her self-report. Self-report scores are often higher or lower than estimates based on these other sources, leading to a dilemma for clinicians. There is no logical way to combine and balance estimates of pain intensity from different sources in order to arrive at a single “true” pain score. Yet studies have shown that health care providers often do so: they alter patients’ self-reported pain scores when recording them in medical charts, or substitute their own informal ratings for patients’ self-report. Partly, this may be because flow sheets (paper and electronic) typically contain only a box for a single pain score at each time point, as for the true vital signs. There is little room and little time to record conflicting or complex information.

Pain is defined primarily in subjective terms as an “unpleasant sensory and emotional experience . . .”.2 Thus, self-report is the primary source of information about pain. But it is an overstatement to claim that self-report represents a “gold standard” — a score which, in the presence of conflicting estimates, must be taken as the true score. Contrary examples abound. Consider a 6-year-old who is sweating, rigid, groaning, and displaying a facial expression of pain a few hours after a major operation. Perioperative anaesthesia is assumed to be beyond its half-life and no other analgesia has been provided. The child gives a pain intensity score of 0 out of 10, meaning no pain. Is this a gold standard score? Clearly not: the child might misunderstand the scale, or might believe that reporting no pain will lead to going home sooner, or might not wish to worry his parents, or might fear that reporting pain will lead to another needle poke. A decision about providing analgesia must take into account these other possibilities.

In a concise statement of the issue, Berde and McGrath wrote,

Craig and colleagues have proposed a theoretical model of pain behaviour (including self-report) and pain assessment as a social communication.4 In this view, a painful stimulus leads to basic biological and cognitive processes that are encoded in verbal and non-verbal behaviour (intentional or involuntary). The external signals are perceived by other people, including caregivers, who decode them as signs of the presence or absence of pain, and then interpret them as the basis for actions, such as provision of analgesia. At each step, numerous social and cultural factors — for example, the level of confidence that the clinician and patient have in each other — impinge on the process.

Thus, the path from the experience of pain, through its expression by the patient, interpretation by the clinician and final arrival at a decision on management, is not straight, but meandering. Let’s discard the simplistic catchphrases. We need to accept that recording a score for pain intensity is not a true measurement task like measuring temperature or heart rate, but a multidimensional judgement based on many factors.

Self-reported pain intensity scores are a key starting point in this process, and should be obtained whenever possible. But adequate pain assessment often also includes the following steps,5 not necessarily in order: a comparison of pain scores with patients’ goals for their own comfort and function; consideration of probable sources of pain; observation of the patient’s behaviour; obtaining others’ report of the patient’s pain behaviours; and assessing the effects of trials of pharmacological, physical and psychological interventions to prevent and relieve pain.

The challenge is to make those steps simple enough to carry out routinely, yet rich enough to reflect the multifaceted reality of pain.