The Ability of Chronic Pain Patients to Self-diagnose Their Chief Source of Low Back Pain

The investigative team will provide 27 slides with bullet points and images of pain referral patterns for different causes (e.g., herniated disc, facet joint arthritis) for patients seen for a new visit with a chief complaint of chronic low back pain. This material is all publicly available but has been compiled in presentation form. This will have educational benefit for patients, discussing clinical signs and symptoms, risk factors and basic treatments. There will also be a smaller control group of that gets a condensed 4-slide presentation. After review of the slides, an independent observer will ask the patient what they think are the 2 most likely causes of their LBP (in order of likelihood) and match that with the attending physician and trainee, who will independently do the same. The investigative team will then determine how concordant the patient's answers are with the physicians and also record outcomes.

Artificial intelligence (AI), the growth of the internet and internet access, direct-to-patient advertising, and more recently the COVID-19 pandemic with a proliferation of telehealth visits has transformed medicine. Patients come in with a wealth of information, some accurate but some inaccurate, about their condition, often with preconceived notions about what condition they have and how they want to be treated. For conditions such as chronic pain with a high prevalence rate of abnormal imaging findings in asymptomatic individuals, the absence of biomarkers for clear-cut diagnoses, and subjective outcome measures, this has led to unnecessary tests and treatment, doctor shopping, high rates of burnout among providers and low success rates. The COVID-19 pandemic shed light on accuracy of diagnoses via telehealth, with studies finding a high concordance rate between telehealth visits without the benefit of a physical exam, and in-person pain management consultant, which is similar to that found in other specialties. The proliferation of AI in electronic medical record systems that confer diagnoses based on patient and physician input of symptoms and signs suggests that in the future, patients with access to the information will be able to self-diagnose their chronic pain and other conditions. Many guidelines also recommend education and self-care in their back pain treatment algorithms, though the effect of education on outcomes is not well-known. With this in mind, the purpose of this study is to determine how accurate patient diagnoses are when they are furnished with readily available information on the different etiologies for chronic low back pain (LBP), the leading cause of disability worldwide. The plan is to enroll 269 patients in a 3:1 allocation ratio to either the 27-slide educational group or a condensed 4-slide control group. The patient will have the opportunity to ask questions, after which they will rate their top 2 diagnoses, in order. A trainee (resident or fellow) and the attending will do then do the same. Outcomes will be recorded at 4 weeks (e.g., for simple injections such as epidural steroids and sacroiliac joint injections, medications, physical therapy) or at 12 weeks for more invasive procedures such as spinal cord stimulation, radiofrequency ablation, or vertebral augmentation. Analyses will be performed to: 1) Determine whether the educational program improves the likelihood that the patient correctly self-diagnoses the cause of their back pain using the attending physician as the reference standard compared to the control group; and 2) Whether the educational program improves treatment outcomes.