CORR Insights®: Surgeons Receiving Information About Patient Language Reflecting Unhelpful Thoughts or Distress About Their Symptoms Identify Such Language More Often Than Those Who Do Not Receive This Information

Where Are We Now?

Resisting a longstanding tradition of a mind-body dichotomy, recognition of the intersection of mental and musculoskeletal health continues to expand. For example, the patient-reported outcome-based performance measure for THA and TKA from the Centers for Medicare and Medicaid Services risk adjusts for preoperative global mental health [4]. For individual patient care, there is increased attention to beliefs (for example, unhelpful thoughts such as catastrophizing or positive perceptions such as self-efficacy) and emotional distress (for example, symptoms of depression or anxiety) [10]. These factors can affect health status and engagement in recovery-related behaviors, such as exercise or symptom management. There is also growing appreciation of the related role of language, including clinicians’ language affecting patients’ beliefs, emotions, and behaviors [5], as well as patients’ language reflecting their mental health [2].

Disregarding patients’ mental health can fundamentally limit the benefits of musculoskeletal interventions. However, there are missed opportunities to identify and address these concerns because musculoskeletal specialists such as orthopaedic surgeons often have limited training in these areas. Detecting psychosocial signals in practice seems intuitive for some clinicians but not others. Can this recognition be improved?

In this month’s Clinical Orthopaedics and Related Research®, Brinkman et al. [3] completed a survey-based randomized trial enrolling international orthopaedic, plastic, or general trauma surgeons. They examined whether receiving “supportive information” about unhelpful thoughts or distress was associated with increased identification of patient language as reflecting those mental health concepts. Excerpts were taken from transcripts from English-language new musculoskeletal specialty visits. The authors found that surgeons who received definitions and examples identified slightly more instances of unhelpful thoughts or distress in the excerpts. Although the effect size was small, the intervention was simple and inexpensive, highlighting a potentially pragmatic approach to improve detection of patients’ mental health support needs. Based on these findings, clinicians can familiarize themselves with examples of patient language reflecting unhelpful thoughts or emotional distress to identify opportunities to intervene in helpful ways.

Where Do We Need To Go?

The authors [3] describe the supportive information as “priming,” a psychologic phenomenon in which “recent experience of a stimulus facilitates or inhibits later processing of the same or a similar stimulus” [1]. “Priming” may imply a subconscious, immediate effect. Practically, should surgeons or other clinicians review examples of language reflecting unhelpful thoughts or distress at the beginning of each day so they are primed to have increased awareness of related concerns during the clinical encounters that immediately follow? Or should they attend a training workshop so they are educated for more long-lasting and potentially actionable awareness? Clarifying the proposed mechanism of interventions relative to the desired impact would guide next steps in the development and evaluation of an intervention. Furthermore, although this is a valuable proof-of-concept study [3], would similar effects of related interventions be seen in busy real-world clinical environments? Reading cleaned transcript segments with the explicit goal of categorizing statements is a fundamentally different task than engaging in a dynamic conversation with a patient while integrating complex information for clinical decision-making.

The present focus on patient language to detect beliefs and distress also raises questions about intersections with formal psychosocial assessments (such as the Pain Catastrophizing Scale or Patient-Reported Outcomes Measurement Information System-Anxiety). To minimize response burden, should formal assessments only be administered to patients whose language indicates potential concerns? However, can emphasizing language-based detection introduce bias, particularly in combination with cues related to factors such as gender presentation, race or ethnicity, or cultural dress? Without thoughtful attention, this approach may exacerbate longstanding disparities in musculoskeletal care, particularly because beliefs or distress often intersect with pain in this clinical population [7]. Universal depression screening is recommended in primary care, given well-known disparities in detection and treatment [6]. Similarly, is uniform administration of brief, standardized, culturally validated psychosocial assessments a means to more systematically and equitably detect opportunities to intervene in mental health in musculoskeletal care? Of course, standardized assessments pose their own challenges (for example, literacy concerns, cultural stigma about mental health questions, and implementation costs for low-resource settings). Therefore, the best balance between formal assessment and informal language-based detection remains unclear. Guidelines such as the “Appropriate Use Criteria for the Early Screening for Psychosocial Risk and Protective Factors” [8] for musculoskeletal injury may help identify those for whom a standardized assessment should be prioritized.

Additionally, what actions result from increased identification of unhelpful thoughts or distress? A study of orthopaedic clinicians highlighted challenges in addressing psychosocial factors, including perceived lack of resources [9]. The present study [3] focused on volume versus accuracy of detection, given that there is no ideal way to evaluate accuracy and because there is a lack of harm in overidentification. Indeed, overidentification is preferred over underidentification and may be entirely unproblematic if the intended response is to adapt one’s communication. However, if identification is intended to trigger supportive resources that are scarce or perceived as scarce, “accurate” identification may be essential for resource optimization in value-based care models. More precise identification of support needs based on beliefs or distress can guide tiered care pathways, such as self-management versus referral to supervised rehabilitation versus referral to supervised rehabilitation plus counseling.

How Do We Get There?

Future research should extend this important work [3] by developing and piloting more-comprehensive interventions with the potential for larger, more long-lasting improvement in musculoskeletal care team members’ identification of unhelpful thoughts and distress. As the authors note [3], such interventions can be evaluated in more methodologically complex studies using observations or video recordings of clinical encounters. Standardized patients may be one option to simulate real-time detection in a clinical context while allowing two clinicians who are randomized to different conditions to interact independently with the same patient. Additionally, language-based detection can be triangulated with scores from standardized assessments, potentially completed by patients post-visit to avoid priming their language. Points of convergence and divergence may highlight overlap or complementary roles of formal and informal mental health assessment for personalized care planning. Impact evaluation of interventions to improve clinician recognition of mental health concerns should also encompass any actions triggered by detection. If detection does not translate to action, decision supports should be considered, such as electronic reminders to intervene or a flowchart suggesting care pathways based on combined musculoskeletal and mental health presentation. Furthermore, to avoid introducing or exacerbating disparities, research should intentionally examine whether detection or related action is disproportionately limited in vulnerable populations. Variation can be examined based on factors like patient gender, race or ethnicity, and health literacy or clinician-patient concordance on sociodemographic characteristics, which may affect patient expression or clinician perception.

Broadly, research to improve biopsychosocial health in musculoskeletal care would continue to benefit from innovative perspectives and methods. In the context of interdisciplinary collaboration, teams may benefit from specific nonmusculoskeletal expertise such as medical sociology or linguistics. Additionally, qualitative and mixed-methods approaches are essential to contextualize the “accuracy” of language-based detection and understand dynamic barriers and facilitators to assess and address mental health from patient and clinician perspectives. As we aim to relieve the burden of musculoskeletal conditions by embracing a biopsychosocial approach, this study by Brinkman et al. [3] injects optimism by highlighting ways we can improve our practice.