Introduction

Pain is subjective and multidimensional, making it a complex phenomenon to measure for research purposes. Musculoskeletal pain refers to pain that is primarily localized to muscles, tendons, ligaments, joints, and bones, and may or may not include referred (secondary) pain in anatomically-predictable areas [1]. Musculoskeletal conditions are typically characterized by persistent pain that limits mobility and functionality and inhibits people’s ability to work and participate in society [2]. Musculoskeletal pain is a global epidemic, and a major determinant of years lived with disability [3]. The burden from musculoskeletal pain is predicted to increase globally especially in low- and middle-income countries due to population growth, aging, and increasing risk factors for other musculoskeletal conditions, such as obesity, injury, and a sedentary lifestyle [4]. Although the impact of musculoskeletal pain is likely underestimated, existing evidence supports an urgency in addressing conditions of musculoskeletal pain holistically [4]. For the overall understanding of pain mechanisms and the evaluation of methods to control pain, accurate and purposeful measurement of pain characteristics is crucial [5]. However, beyond the ubiquitous Numeric Pain Rating Scale, there is little consistency or standardization in the way pain is measured and understood in pain research [6].

Pain (or body) maps are tools used for assessing pain/symptom characteristics including pain extent, location, or distribution, and can be used for clinical or research purposes [7]. They require a transformational representation of pain experiences into a graphical, numerical, or descriptive form that typically requires patients to indicate their region of pain and may include additional information such as qualitative or intensity aspects of the pain. The mapping ‘method’ refers to the specific procedures or techniques used to assess and capture the spatial characteristics of pain (e.g., shading, marking, grid, etc.), while the ‘mode’ of assessment refers to the mediums or platforms through which the patient interacts with the pain mapping tool (e.g., paper and pen, digital, etc.) [8]. Pain mapping has gained more recognition in recent times and the methods and modes of implementation have also evolved from pen-on-paper to computer-generated images, web-based applications, and mobile applications [9]. Different methods and modes of implementation of pain maps have been described, from relatively simple checklists or two-dimensional line drawings to interactive three-dimensional digital maps [9]. The methods and modes of implementing pain mapping, such as the resolution of the selectable area (e.g., individual pixels vs. anatomic regions) have a considerable impact on the resulting spatial estimate of pain characteristics [10]. There are many potentially important inconsistencies in how pain maps are built, administered, and interpreted [11]. For example, the Collaborative Health Outcome Information Registry (CHOIR) body map contains 36 anterior and 38 posterior coded body segments highlighting anatomical or pain conditions with sex demarcation [12], which is quite different from the Pain-QuiLT body map which has over 100 codified body locations [13]. Others such as the Corlett and Bishop body map, Michigan body map, and Nordic body map each have their unique and separate features.

Most pain maps present a body image that represents an atypical well-defined healthy-appearing, fully intact, binary male or female, with body morphometric characteristics most closely aligned with a pseudo-idealized white European body structure. We are unaware of any pain map validated and published specifically for use in people with severe obesity [14], people of Black African or Pacific Islander ancestral background, or people with amputations, as examples. As the transformational representation required to complete a pain map requires that the respondents can ‘see themselves’ in the body image being presented to them, these are non-trivial considerations. The optimal methods of collecting spatial pain estimates for use in musculoskeletal pain research has yet to be adequately synthesized [15], making comparisons of different study results often complicated or impossible [16].

Machine learning (ML) and artificial intelligence (AI) are transforming pain mapping by enabling more accurate and comprehensive analysis of pain patterns beyond a simple summation of ‘areas selected’ or ‘pixels shaded’. These technologies can process large datasets of pain maps to identify subtle trends in pain location, extent or distribution, intensity, and changes over time, helping to create personalized treatment plans. Goldstein and colleagues [17] have shown that ML/AI models have the potential to integrate patient-specific factors like demographics and medical history to reduce trial-and-error in pain management and predict the progression of musculoskeletal conditions. Furthermore, ML can contribute to the development of standardized and objective pain maps, improving consistency and accessibility for diverse populations. Sankaran and colleagues [18] argued that the integration of AI/ML with electronic health records (EHRs) promises to streamline data collection, advancing both clinical practice and research in pain management. In essence, ML and AI are driving a paradigm shift in pain mapping by offering more accurate, comprehensive, and personalized insights into pain patterns [18]. These advancements not only improve clinical outcomes but also promise to make pain management more efficient, effective, and accessible across a variety of patient populations.

As technology advances, we can expect new designs, methods, and modes of implementing pain maps. Eliav and Grace [19] proposed that for the measurement of pain experiences with pain/symptom maps to be accurate, the methods should be standardized and used consistently with each patient. We note the rapid development in electronic health management that has created a strong need for a standardized body map to efficiently collect self-reported pain data to enhance practice, research, and data integration [20]. To contribute towards standards for pain/symptom mapping it is important to understand the current state of the field, including an appreciation for the types and uses of pain maps previously published, seeking similarities and differences in their evolution. The purpose of this study is to scope the existing literature on pain maps to better understand the current state and conduct important groundwork for further work to come.

Methods

Selection of sources of evidence

All identified citations will be uploaded to Covidence, and the screening process will be documented in a PRISMA flowchart [26] (Fig 1), including reasons for exclusion. Duplicate citations will be removed through automatic text screening. Two independent reviewers will screen titles and abstracts to identify relevant studies, followed by full-text screening to assess eligibility. Disagreements will be resolved through discussion between the reviewers; if consensus is not reached, a third experienced reviewer will be consulted to make the final decision. However, the nature of the review occasioned by broadness of the topic, diversity of study designs, less focus on quality assessment, and the potential inconsistencies in framework for deploying intervention, data charting priorities, and iterative and reflective process [25], complicates defining a kappa threshold and makes it less important to report inter-rater reliability.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flow diagram of study selection processes.

https://doi.org/10.1371/journal.pone.0319498.g001

Discussion

Pain maps have historically been viewed as adjunctive tools for clinicians and researchers to better understand a patient’s pain through largely qualitative means and have accordingly often been designed without rigorous quantification in mind. Hägg and colleagues [39] found that the predictive value of pain drawing in predicting outcomes for surgical and non-surgical interventions is not properly documented. Conversely, some recent studies have shown that computational quantification of pain maps has the potential to enhance comprehensive pain assessment, facilitating research and personalized care over time for patients with various musculoskeletal pain conditions [40,41]. Similarly, pain mapping has significantly improved pain reporting [42], and tracking the effectiveness of treatment pathways [23].

Applying the methodology of a scoping review in collating historical and current ideas in the type, methods, and modes of implementing pain maps will potentially mitigate the inconsistencies observed in musculoskeletal pain drawing. However, in keeping with the broad scope of this review, there is a limitation in language inclusivity. This review focuses on identifying the characteristic features of common pain maps and scoring strategies, which are typically reported in multiple publications, although the highest-impact journals in the field require, at a minimum, an English-language translation. Additionally, included studies will not be typically assessed for quality as there are no standards for operationalizing a pain map which is probably deployed differently in different study designs [42].

As pain mapping evolves from a qualitative tool to a more quantifiable and data-driven resource, it holds considerable promise in both personalized patient care and research outcomes. In clinical settings, pain maps can enhance personalized patient care as a novel communication channel between patient and clinician, as a means of quantifying spatial pain measures to optimize treatment targeting, and enabling clinicians to track progress and evaluate treatment effectiveness beyond the historic reliance on only pain intensity ratings. A more structured approach to pain mapping can complement patient narratives, fostering consistent, patient-centric and evidence-based recommendations. The potential for personalized care is amplified by advancements in artificial intelligence (AI) and machine learning (ML), which when combined with standardized pain maps can detect subtle pain patterns, predict outcomes, and suggest personalized treatment plans [17]. Integrating pain maps into electronic health records (EHRs) enhances the holistic view of a patient’s health [18]. From a research perspective, the scoping review will synthesize existing pain mapping methodologies and identify gaps, guiding future studies. Standardizing pain map protocols across studies would enhance comparisons and collaboration, advancing pain management research.

This scoping review is intended to provide a clear picture of the prior use and current state of pain/symptom maps in the musculoskeletal pain field. As technology is rapidly evolving and core data sets are becoming increasingly important for making decisions at the individual and population health level, there is a clear window of opportunity for the development, implementation, and interpretation of pain/symptom maps to also evolve in new ways that contribute unique and relevant data across health conditions. Notably, the rise of increasingly accessible artificial intelligence and machine learning platforms appears to signal that the information gleaned from pain/symptom maps could go much further than a simple summation of pixels or body regions. As a first step towards this level of deep rigor and an ‘ideal’ pain/symptom map, we find it prudent to first present the scope and summarize the historic and recent trends that have led the field to where it is today.

Supporting information