LAUSR

Pain researchers and scientists acknowledge that pain is dynamic and can vary substantially. Yet, much of what we know about pain comes from studies that assess pain at single time points. The study by Mun et al. is important for several reasons. First, it focuses attention on the advantages of assessing pain as a dynamic process. Second, it provides a comprehensive analysis of the state of the science with regard to intensive longitudinal data methods and related statistical approaches. Along these lines, they highlight the utility of certain statistical indices of variability (eg, the probability of acute change) that can provide new insights into pain phenomena (eg, acute pain flares that may occur in patients with sickle cell disease, rheumatoid arthritis, or at the end of life). Finally, they offer clear and practical guidance on how to choose measurement approaches, analyze data, and interpret variability in pain in a more valid and meaningful fashion. The study by Mun et al. represents a paradigm shift for most pain scientists. Much of pain science relies on either single or a limited set of pain assessments. Unfortunately, in clinical samples, these are often retrospective, self-report measures subject to recall and other reporting biases. This traditional approach does a poor job of meeting the ultimate goal of pain science, ie, to understand, predict, and influence pain as dynamic phenomenon. Technology provides important and new opportunities to study variations in pain. Smartphone apps are available or can be programmed in ways that enable researchers to easily sample pain at regular intervals (eg, time contingent sampling) or much closer in time to events of importance (eg, event contingent sampling). No longer do we need to rely on paper and pencil methods that are more burdensome and further removed from momentary experience, both of which may increase bias. The benefits of repeated assessments of pain for clinicians and patients with pain are potentially enormous. First, these approaches can help the clinician better understand and validate patients’ daily experiences. Second, repeated pain assessments are a form of self-monitoring that can provide patients with perspective on meaningful patterns and the insights gained may serve tomotivate behavior change. Third, as pointed out by Mun et al., intensive daily recordings of painmay reveal both expected and unexpected treatment effects. Finally, these approaches fit well with the growing emphasis on precision medicine by improving the prediction of treatment outcomes. With any paradigm shift, it takes the field a while to embrace and refine the use of new methods. Several concerns are commonly raised regarding intensive daily pain assessments. First, some argue that tracking pain on a frequent basis may impact the pain experience (ie, reactivity) and draw more attention to it, when the goal of pain treatment is often to move attention away from pain and towards resumption of valued activities. Reactivity is a feature of any form of pain assessment and it potentially can work to either benefit a patient (eg, by decreasing pain) or work against a patient (eg, by increasing pain). Interestingly, intensive longitudinal studies with patients with chronic pain have found little statistical evidence of reactivity. Indeed, research shows that the reactive effects of recording may actually be reduced when sampling is continued over a prolonged period. A second concern is that daily recordingmay be burdensome to patients leading to high rates of missing data. However, many daily diary studies of patientswith chronic pain have shownmissing data rates that are lower than those reported by Mun et al. (eg, completion rates often over 90%). New technologies (eg, entering one’s data on a smartphone) can reduce the burden of recording. In addition, careful preparation of the patient before recording and provision of ongoing incentives (eg, simple graphs and periodic summaries or, for research studies, financial incentives) can increase the likelihood of obtaining high rates of diary completion. In clinical settings, having the clinician review data with the patient is perhaps the best and most reinforcing method of enhancing completion of daily records. A third concern with intensive longitudinal designs is that they can yield a great deal of data. How is one to deal with this? Mun, et al. highlight the fact that sophisticated statistical approaches are being developed to manage these data (and data missingness when that occurs) and provide concrete examples of how these approaches can be used to provide meaningful and userfriendly indices of change. Simple approaches to summarizing longitudinal data also can be quite useful (eg, graphing ofmultiday averages over recording periods of weeks) and are increasingly being built into apps that track pain and pain-related variables. When clinicians and patients review and discuss the resulting patterns, it can be very therapeutic and inform intervention decisions. The study by Mun et al. primarily focuses on longitudinal reports of pain. Pain is one of the most important outcomes to patients and their families. However, intensive longitudinal methods need not and should not be limited to pain reports. Research shows that other pain-related outcomes also vary considerably over time (eg, reports of emotional distress, coping efforts, and catastrophizing). An intensive analysis of how variations in these outcomes relate to important domains of functioning can be quite revealing. A particularly important direction for the future is identifying valued and meaningful personal goals (eg, time with family, relationships with coworkers, Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.