Assessing the performance of AI chatbots in answering patients' common questions about low back pain.

OBJECTIVES: The aim of this study was to assess the accuracy and readability of the answers generated by large language model (LLM)-chatbots to common patient questions about low back pain (LBP). METHODS: This cross-sectional study analysed responses to 30 LBP-related questions, covering self-management, risk factors and treatment. The questions were developed by experienced clinicians and researchers and were piloted with a group of consumer representatives with lived experience of LBP. The inquiries were inputted in prompt form into ChatGPT 3.5, Bing, Bard (Gemini) and ChatGPT 4.0. Responses were evaluated in relation to their accuracy, readability and presence of disclaimers about health advice. The accuracy was assessed by comparing the recommendations generated with the main guidelines for LBP. The responses were analysed by two independent reviewers and classified as accurate, inaccurate or unclear. Readability was measured with the Flesch Reading Ease Score (FRES). RESULTS: Out of 120 responses yielding 1069 recommendations, 55.8% were accurate, 42.1% inaccurate and 1.9% unclear. Treatment and self-management domains showed the highest accuracy while risk factors had the most inaccuracies. Overall, LLM-chatbots provided answers that were 'reasonably difficult' to read, with a mean (SD) FRES score of 50.94 (3.06). Disclaimer about health advice was present around 70%-100% of the responses produced. CONCLUSIONS: The use of LLM-chatbots as tools for patient education and counselling in LBP shows promising but variable results. These chatbots generally provide moderately accurate recommendations. However, the accuracy may vary depending on the topic of each question. The reliability level of the answers was inadequate, potentially affecting the patient's ability to comprehend the information.

Secular trends in gabapentinoid dispensing by compensated workers with low back pain: a retrospective cohort study.

OBJECTIVES: The increase in gabapentinoid prescribing is paralleling the increase in serious harms. To describe the low back pain workers compensation population whose management included a gabapentinoid between 2010 and 2017, and determine secular trends in, and factors associated with gabapentinoid use. METHODS: We analysed claim-level and service-level data from the Victorian workers' compensation programme between 1 January 2010 and 31 December 2017 for workers with an accepted claim for a low back pain injury and who had programme-funded gabapentinoid dispensing. Secular trends were calculated as a proportion of gabapentinoid dispensings per year. Poisson, negative binomial and Cox hazards models were used to examine changes over time in incidence and time to first dispensing. RESULTS: Of the 17 689 low back pain claimants, one in seven (14.7%) were dispensed at least one gabapentinoid during the first 2 years (n=2608). The proportion of workers who were dispensed a gabapentinoid significantly increased over time (7.9% in 2010 to 18.7% in 2017), despite a reduction in the number of claimants dispensed pain-related medicines. Gabapentinoid dispensing was significantly associated with an opioid analgesic or anti-depressant dispensing claim, but not claimant-level characteristics. The time to first gabapentinoid dispensing significantly decreased over time from 311.9 days (SD 200.7) in 2010 to 148.2 days (SD 183.1) in 2017. CONCLUSIONS: The proportion of claimants dispensed a gabapentinoid more than doubled in the period 2010-2017; and the time to first dispensing halved during this period.

Clinical care standards for the management of low back pain: a scoping review.

The objective of this study is to compare and contrast the quality statements and quality indicators across clinical care standards for low back pain. Searches were performed in Medline, guideline databases, and Google searches to identify clinical care standards for the management of low back pain targeting a multidisciplinary audience. Two independent reviewers reviewed the search results and extracted relevant information from the clinical care standards. We compared the quality statements and indicators of the clinical care standards to identify the consistent messages and the discrepancies between them. Three national clinical care standards from Australia, Canada, and the United Kingdom were included. They provided from 6 to 8 quality statements and from 12 to 18 quality indicators. The three standards provide consistent recommendations in the quality statements related to imaging, and patient education/advice and self-management. In addition, the Canadian and Australian standards also provide consistent recommendations regarding comprehensive assessment, psychological support, and review and patient referral. However, the three clinical care standards differ in the statements related to psychological assessment, opioid analgesics, non-opioid analgesics, and non-pharmacological therapies. The three national clinical care standards provide consistent recommendations on imaging and patient education/advice, self-management of the condition, and two standards (Canadian and Australian) agree on recommendations regarding comprehensive assessment, psychological support, and review and patient referral. The standards differ in the quality statements related to psychological assessment, opioid prescription, non-opioid analgesics, and non-pharmacological therapies.

Clinical indicators to monitor health care in low back pain: a scoping review.

Clinical care indicators for low back pain can be used to monitor healthcare practices and consequently be used to evaluate success of strategies to improve care quality. The aim of this study was to identify the clinical care indicators that have been used to measure appropriateness of health care for patients with low back pain. We conducted a systematic search of five electronic databases and Google to identify clinical care indicators that have been used to measure any aspect of care for people with low back pain. Care indicators were narratively described according to their type (i.e. structure, process, or outcomes) and categorized by their purpose (e.g. to measure aspects related to assessment, imaging requests, treatment/prevention, and outcomes). A total of 3562 and 2180 records were retrieved from electronic databases and Google searches, respectively. We identified 280 indicators related to low back pain care from 40 documents and publications. Most quality indicators were process indicators (n = 213, 76%), followed by structure (n = 41, 15%) and outcome indicators (n = 26, 9%). The most common indicators were related to imaging requests (n = 41, 15%), referral to healthcare providers (n = 30, 11%), and shared decision-making (n = 21, 7%). Our review identified a range of clinical care indicators that have been used to measure the quality of health care for people with low back pain. Our findings will support a Delphi study to reach international consensus on what would be the most important and feasible indicators for a minimum dataset to be collected globally.

Prioritisation of clinical trial learning needs of musculoskeletal researchers: an inter-disciplinary modified Delphi study by the Australia & New Zealand musculoskeletal clinical trials network.

BACKGROUND: There is a need to increase the capacity and capability of musculoskeletal researchers to design, conduct, and report high-quality clinical trials. The objective of this study was to identify and prioritise clinical trial learning needs of musculoskeletal researchers in Australia and Aotearoa New Zealand. Findings will be used to inform development of an e-learning musculoskeletal clinical trials course. METHODS: A two-round online modified Delphi study was conducted with an inter-disciplinary panel of musculoskeletal researchers from Australia and Aotearoa New Zealand, representing various career stages and roles, including clinician researchers and consumers with lived experience of musculoskeletal conditions. Round 1 involved panellists nominating 3-10 topics about musculoskeletal trial design and conduct that they believe would be important to include in an e-learning course about musculoskeletal clinical trials. Topics were synthesised and refined. Round 2 asked panellists to rate the importance of all topics (very important, important, not important), as well as select and rank their top 10 most important topics. A rank score was calculated whereby higher scores reflect higher rankings by panellists. RESULTS: Round 1 was completed by 121 panellists and generated 555 individual topics describing their musculoskeletal trial learning needs. These statements were grouped into 37 unique topics for Round 2, which was completed by 104 panellists. The topics ranked as most important were: (1) defining a meaningful research question (rank score 560, 74% of panellists rated topic as very important); (2) choosing the most appropriate trial design (rank score 410, 73% rated as very important); (3) involving consumers in trial design through to dissemination (rank score 302, 62% rated as very important); (4) bias in musculoskeletal trials and how to minimise it (rank score 299, 70% rated as very important); and (5) choosing the most appropriate control/comparator group (rank score 265, 65% rated as very important). CONCLUSIONS: This modified Delphi study generated a ranked list of clinical trial learning needs of musculoskeletal researchers. Findings can inform training courses and professional development to improve researcher capabilities and enhance the quality and conduct of musculoskeletal clinical trials.

1,4-Diaryl-1,2,3-triazole neolignan-celecoxib hybrids inhibit experimental arthritis induced by zymosan.

Rheumatoid arthritis (RA) is a chronic autoimmune disease that causes cartilage damage. Anti-inflammatories are widely used in the management of RA, but they can have side effects such as gastrointestinal and/or cardiovascular disorders. Studies published by our group showed that the synthesis of hybrid triazole analogs neolignan-celecoxib containing the substituent groups sulfonamide (L15) or carboxylic acid (L18) exhibited anti-inflammatory activity in an acute model of inflammation, inhibited expression of P-selectin related to platelet activation and did not induce gastric ulcer, minimizing the related side effects. In continuation, the present study evaluated the anti-inflammatory effects of these analogs in an experimental model of arthritis and on the functions of one of the important cells in this process, macrophages. Mechanical hyperalgesia, joint edema, leukocyte recruitment to the joint and damage to cartilage in experimental arthritis and cytotoxicity, spread of disease, phagocytic activity and nitric oxide (NO) and hydrogen peroxide production by macrophages were evaluated. Pre-treatment with L15 and L18 reduced mechanical hyperalgesia, joint edema and the influx of leukocytes into the joint cavity after different periods of the stimulus. The histological evaluation of the joint showed that L15 and L18 reduced cartilage damage and there was no formation of rheumatoid pannus. Furthermore, L15 and L18 were non-cytotoxic. The analogs inhibited the spreading, the production of NO and hydrogen peroxide. L15 decreased the phagocytosis. Therefore, L15 and L18 may be potential therapeutic prototypes to treat chronic inflammatory diseases such as RA.

Efficacy, safety, and dose-dependence of the analgesic effects of opioid therapy for people with osteoarthritis: systematic review and meta-analysis.

OBJECTIVE: To evaluate the efficacy and safety of opioids for analgesic therapy for people with osteoarthritis. STUDY DESIGN: Systematic review and meta-analysis of randomised, placebo-controlled trials of opioid therapies for treating the pain of osteoarthritis. The primary outcome was medium term pain relief (six weeks to less than 12 months). Quality of evidence was assessed with GRADE criteria. DATA SOURCES: MEDLINE, EMBASE, the Cochrane Database of Systematic Reviews and Central Register of Controlled Trials, CINAHL, PsycINFO, AMED, and the WHO International Clinical Trials Registry; trials published to 31 October 2020. DATA SYNTHESIS: We extracted pain, disability, health-related quality of life, and adverse events data for 36 eligible trials (overall dose range: 10-210 oral morphine milligram equivalents [MME] per day). Continuous pain and disability outcomes were converted to common 0-100-point scales; changes of less than ten points were deemed to be very small effects. Differences in dichotomous outcomes were expressed as risk ratios. Data were pooled for meta-analysis in random effects models. The evidence from 19 trials (8965 participants; dose range, 10-126 MME/day) for very small medium term pain relief (mean difference [MD], -4.59 points; 95% CI, -7.17 to -2.02 points) was low quality, as was that from 16 trials (6882 participants; dose range, 10-126 MME/day) for a very small effect on disability (MD, -4.15 points; 95% CI, -6.94 to -1.35 points). Opioid dose was not statistically significantly associated with either degree of pain relief or incidence of adverse events in a meta-regression analysis. Evidence that opioid therapy increased the risk of adverse events (risk ratio, 1.43; 95% CI, 1.29-1.59) was of very low quality. CONCLUSIONS: Opioid medications may provide very small pain and disability benefits for people with osteoarthritis, but may also increase the risk of adverse events. PROSPERO REGISTRATION: CRD42019142813 (prospective).

Public and patient perceptions of diagnostic labels for non-specific low back pain: a content analysis.

PURPOSE: An online randomised experiment found that the labels lumbar sprain, non-specific low back pain (LBP), and episode of back pain reduced perceived need for imaging, surgery and second opinions compared to disc bulge, degeneration, and arthritis among 1447 participants with and without LBP. They also reduced perceived seriousness of LBP and increased recovery expectations. METHODS: In this study we report the results of a content analysis of free-text data collected in our experiment. We used two questions: 1. When you hear the term [one of the six labels], what words or feelings does this make you think of? and 2. What treatment (s) (if any) do you think a person with [one of the six labels] needs? Two independent reviewers analysed 2546 responses. RESULTS: Ten themes emerged for Question1. Poor prognosis emerged for disc bulge, degeneration, and arthritis, while good prognosis emerged for lumbar sprain, non-specific LBP, and episode of back pain. Thoughts of tissue damage were less common for non-specific LBP and episode of back pain. Feelings of uncertainty frequently emerged for non-specific LBP. Twenty-eight treatments emerged for Question2. Surgery emerged for disc bulge, degeneration, and arthritis compared to lumbar sprain, non-specific LBP, and episode of back pain. Surgery did not emerge for non-specific LBP and episode of back pain. CONCLUSION: Our results suggest that clinicians should consider avoiding the labels disc bulge, degeneration and arthritis and opt for labels that are associated with positive beliefs and less preference for surgery, when communicating with patients with LBP.

Hospital variation in admissions for low back pain following an emergency department presentation: a retrospective study.

BACKGROUND: One in 6 patients with low back pain (LBP) presenting to emergency departments (EDs) are subsequently admitted to hospital each year, making LBP the ninth most common reason for hospital admission in Australia. No studies have investigated and quantified the extent of clinical variation in hospital admission following an ED presentation for LBP. METHODS: We used routinely collected ED data from public hospitals within the state of New South Wales, Australia, to identify presentations of patients aged between 18 and 111 with a discharge diagnosis of LBP. We fitted a series of random effects multilevel logistic regression models adjusted by case-mix and hospital variables. The main outcome was the hospital-adjusted admission rate (HAAR). Data were presented as funnel plots with 95% and 99.8% confidence limits. Hospitals with a HAAR outside the 95% confidence limit were considered to have a HAAR significantly different to the state average. RESULTS: We identified 176,729 LBP presentations across 177 public hospital EDs and 44,549 hospital admissions (25.2%). The mean (SD) age was 51.8 (19.5) and 52% were female. Hospital factors explained 10% of the variation (ICC = 0.10), and the median odds ratio (MOR) was 2.03. We identified marked variation across hospitals, with HAAR ranging from 6.9 to 65.9%. After adjusting for hospital variables, there was still marked variation between hospitals with similar characteristics. CONCLUSION: We found substantial variation in hospital admissions following a presentation to the ED due to LBP even after controlling by case-mix and hospital characteristics. Given the substantial costs associated with these admissions, our findings indicate the need to investigate sources of variation and to determine instances where the observed variation is warranted or unwarranted.

Second opinions for spinal surgery: a scoping review.

BACKGROUND: Second opinions have the goal of clarifying uncertainties around diagnosis or management, particularly when healthcare decisions are complex, unpleasant, and carry considerable risks. Second opinions might be particularly useful for people recommended surgery for their back pain as surgery has at best a limited role in the management of back pain. METHODS: We conducted a scoping review. Two independent researchers screened PubMed, EMBASE, Cochrane CENTRAL and CINAHL from inception to May 6th, 2021. Studies of any design published in any language were eligible provided they described a second opinion intervention for people with spinal pain (low back or neck pain with or without radicular pain) either considering surgery or to whom surgery had been recommended. We assessed the methodological quality with the Downs & Black scale. Outcomes were: i) characteristics of second opinion services for people considering or who have been recommended spinal surgery, ii) agreement between first and second opinions in terms of diagnoses, need for surgery and type of surgery, iii) whether they reduce surgery and improve patient outcomes; and iv) the costs and healthcare use associated with these services. Outcomes were presented descriptively. RESULTS: We screened 6341 records, read 27 full-texts, and included 12 studies (all observational; 11 had poor methodological quality; one had fair). Studies described patient, doctor, and insurance-initiated second opinion services. Diagnostic agreement between first and second opinions varied from 53 to 96%. Agreement for need for surgery between first and second opinions ranged from 0 to 83%. Second opinion services may reduce surgery rates in the short-term, but it is unclear whether these reductions are sustained in the long-term or if patients only delay surgery. Second opinion services may reduce costs and healthcare use (e.g. imaging), but might increase others (e.g. injections). CONCLUSIONS: Second opinion services typically recommend less surgical treatments compared to first opinions and may reduce surgery rates in the short-term, but it is unclear whether these reductions are sustained in the long-term or if patients only delay surgery. There is a need for high-quality randomised trials to determine the value of second opinion services for reducing spinal surgery.