Quality of Large Language Model Responses to Radiation Oncology Patient Care Questions.

Importance: Artificial intelligence (AI) large language models (LLMs) demonstrate potential in simulating human-like dialogue. Their efficacy in accurate patient-clinician communication within radiation oncology has yet to be explored. Objective: To determine an LLM's quality of responses to radiation oncology patient care questions using both domain-specific expertise and domain-agnostic metrics. Design, Setting, and Participants: This cross-sectional study retrieved questions and answers from websites (accessed February 1 to March 20, 2023) affiliated with the National Cancer Institute and the Radiological Society of North America. These questions were used as queries for an AI LLM, ChatGPT version 3.5 (accessed February 20 to April 20, 2023), to prompt LLM-generated responses. Three radiation oncologists and 3 radiation physicists ranked the LLM-generated responses for relative factual correctness, relative completeness, and relative conciseness compared with online expert answers. Statistical analysis was performed from July to October 2023. Main Outcomes and Measures: The LLM's responses were ranked by experts using domain-specific metrics such as relative correctness, conciseness, completeness, and potential harm compared with online expert answers on a 5-point Likert scale. Domain-agnostic metrics encompassing cosine similarity scores, readability scores, word count, lexicon, and syllable counts were computed as independent quality checks for LLM-generated responses. Results: Of the 115 radiation oncology questions retrieved from 4 professional society websites, the LLM performed the same or better in 108 responses (94%) for relative correctness, 89 responses (77%) for completeness, and 105 responses (91%) for conciseness compared with expert answers. Only 2 LLM responses were ranked as having potential harm. The mean (SD) readability consensus score for expert answers was 10.63 (3.17) vs 13.64 (2.22) for LLM answers (P < .001), indicating 10th grade and college reading levels, respectively. The mean (SD) number of syllables was 327.35 (277.15) for expert vs 376.21 (107.89) for LLM answers (P = .07), the mean (SD) word count was 226.33 (191.92) for expert vs 246.26 (69.36) for LLM answers (P = .27), and the mean (SD) lexicon score was 200.15 (171.28) for expert vs 219.10 (61.59) for LLM answers (P = .24). Conclusions and Relevance: In this cross-sectional study, the LLM generated accurate, comprehensive, and concise responses with minimal risk of harm, using language similar to human experts but at a higher reading level. These findings suggest the LLM's potential, with some retraining, as a valuable resource for patient queries in radiation oncology and other medical fields.

Alignment prior to total knee arthroplasty in high tibial osteotomy patients has no effect on subsequent functional outcomes.

INTRODUCTION: The influence of prior high tibial osteotomy (HTO) on total knee arthroplasty (TKA) functional outcomes remains widely debated. Alignment of failed HTO can pose technical challenges with subsequent TKA. The primary aim of this study was to evaluate the influence of HTO alignment on the clinical outcomes of subsequent TKA. The secondary aim was to compare the time to TKA for each HTO alignment type. METHODS: Patients who underwent TKA post lateral closing-wedge HTO for symptomatic medial compartment osteoarthritis between 2001 and 2014 were prospectively followed up for 2 years. A total of 159 patients were assigned to three groups based on their pre-TKA femora tibia angles using long lower limb radiographs: varus alignment (VrA) ≤ 3o valgus, neutral alignment (NA) 3-9o valgus alignment, valgus alignment (VlA) ≥ 9o valgus. Functional outcomes were quantified using Knee Society Function Score and Knee Scores (KSFS and KSKS respectively), modified Oxford Knee Score (OKS), Short Form 36 Physical Component Score (SF-36 PCS), and SF-36 Mental Component Score (SF-36 MCS). Pre-operative and post-operative knee range of motion were also measured. RESULTS: Mean pre-TKA KSKS in VrA patients (35 ± 18) was significantly lower than both NA (51 ± 19) and VlA (40 ± 21) patients (p < .05). Otherwise, there was no significant difference in functional outcome scores (KSFS, KSKS, OKS, SF-36 PCS and SF-36 MCS) or range of motion at 6 months and 2 years post-TKA. The mean duration from HTO to TKA was 12 ± 7 years with no significant differences between VrA, NA, and VlA HTO to TKA (13 ± 7 years, 13 ± 6 years and12 ± 8 years respectively, p > .05). CONCLUSION: HTO alignment did not influence time to subsequent TKA. HTO alignment did not influence early outcomes as well as radiological outcomes of subsequent TKA. LEVEL OF EVIDENCE: III.

Traumatic intradural disc herniation following a cervical facet dislocation: a case report.

We report a unique case of a patient who sustained an intradural disc herniation from a left C5-6 unilateral facet dislocation after a fall. This was not easily identified on pre-operative imaging. We explain the details of our surgical approach in this case report. A 65-year-old male fell into a 2 m drain and sustained a left C5/6 unilateral facet dislocation. He then sustained an American Spinal Injury Association (ASIA) B cord injury. His power was 0/5 from C8 downwards bilaterally but sensation was intact throughout. Magnetic resonance imaging (MRI) showed severe compression at C5/6 but no overt intradural disc herniation. This patient subsequently underwent a closed reduction in the operating theatre followed by a combined anterior and posterior approach for the disc herniation. Cerebral spinal fluid (CSF) leakage was noted upon completion of the C5/6 discectomy and it was discovered that there was a traumatic dural tear from the traumatic disc herniation. The decision was made not to repair the dural tear due to the friable nature of the dura and the potential for adhesive glue to propagate through the spinal cord. An anterior drain was placed for 3 days and then removed, he subsequently underwent rehabilitation and was able to regain power in the affected myotomes. Intradural disc herniations can be easily missed on MRI in the setting of cervical spinal trauma. Hence, the anterior approach is an increasingly acceptable approach to tackle disc herniations in unilateral cervical facet dislocations (CFD) surgery.

Symptomatic Construct Failure after Metastatic Spine Tumor Surgery.

STUDY DESIGN: Retrospective cohort study. PURPOSE: To evaluate the incidence and presentation of symptomatic failures (SFs) after metastatic spine tumor surgery (MSTS). To identify the associated risk factors. To categorize SFs based on the management in these patients. OVERVIEW OF LITERATURE: Few studies have reported on the incidence (1.9%-16%) and risk factors of SF after MSTS. It is unclear whether all SFs, occurring in MSTS-patients, result in revision surgery. METHODS: We conducted a retrospective analysis on 288 patients (246 for final analysis) who underwent MSTS between 2005-2015. Data collected were demographics and peri/postoperative clinical and radiological features. Early and late radiological SF were defined as presentation before and after 3 months from index surgery, respectively. Univariate and multivariate models of competing risk regression analysis were designed to determine the risk factors for SF with death as a competing event. RESULTS: We observed 14 SFs (5.7%) in 246 patients; 10 (4.1%) underwent revision surgery. Median survival was 13.4 months. The mean age was 58.8 years (range, 21-87 years); 48.4% were women. The median time to failure was 5 months (range, 1-60 months). Patients with SF were categorized into three groups: (1) SF when the primary implant was revised (n=5, 35.7%); (2) peri-construct progression of disease requiring extension (n=5, 35.7%); and (3) SFs that did not warrant revision (n=4, 28.5%). Four patients (28.5%) presented with early failure. SF commonly occurred at the implant-bone interface (9/14) and all patients had a spinal instability neoplastic score (SINS) >7. Thirteen patients (92.8%) who developed failure had fixation spanning junctional regions. Multivariate competing risk regression showed that preoperative Eastern Cooperative Oncology Group score was a significant risk factor for implant failure (adjusted sub-hazard ratio, 7.0; 95% confidence interval, 1.63-30.07; p<0.0009). CONCLUSIONS: The incidence of SF (5.7%) was low in patients undergoing MSTS although these patients did not undergo spinal fusion. Preoperative ambulators involved a 7 times higher risk of failure than non-ambulators. Preoperative SINS >7 and fixations spanning junctional regions were associated with SF. Majority of construct failures occurred at the implant-bone interface.

Process evaluation of the Living Green, Healthy and Thrifty (LiGHT) web-based child obesity management program: combining health promotion with ecology and economy.

OBJECTIVE: To conduct a process evaluation of the Living Green, Healthy and Thrifty (LiGHT) program, a novel virtual child obesity management program that combines health promotion with ecology and economy (Phase 1). METHODS: We carried out a mixed methods process evaluation involving qualitative and quantitative data collection in 3 phases: among 3 child-parent units, (group 1) that informed program development; 9 child-parent units (group 2) that tested the draft program and further aided program refinement; and 17 child-parent units (group 3) for a 4-week pilot of the program. In the program pilot, we assessed participants' knowledge and readiness to change pre- and postintervention and explored perceptions of the program. RESULTS: Participants generally felt that the online format for program delivery was convenient and accessible, the content was practical, and the integration of health-environment-economy was well received. Many parents also appreciated the involvement of the family. However, the lack of visual appeal and overabundance of text was identified as a challenge, and children/youth in particular requested assurance that their personal information (e.g. weight) was not seen by their parents. The online method of program delivery holds the unique challenge of requiring special efforts to create a sense of personal connection and community. The presence of a "Way-finder" to assist participants and discussion boards/forums are potential solutions. CONCLUSION: The LiGHT online weight management program offers an accessible, convenient weight management resource that children and families appreciate for its availability, broader educational scope, and practicality. Outcome evaluation of LiGHT will be carried out in Phase 2 of the project.

Making sanitation count: developing and testing a device for assessing latrine use in low-income settings.

While efforts are underway to expand latrine coverage to an estimated 2.6 billion people who lack access to improved sanitation, there is evidence that actual use of latrines is suboptimal, limiting the potential health and environmental gains from containment of human excreta. We developed a passive latrine use monitor (PLUM) and compared its ability to measure latrine activity with structured observation. Each PLUM consisted of a passive infrared motion detector, microcontroller, data storage card, and batteries mounted in a small plastic housing that was positioned inside the latrine. During a field trial in Orissa, India, with ∼115 households, the number of latrine events measured by the PLUMs was in good agreement with that measured by trained observers during 5 h of structured observation per device per week. A significant finding was that the presence of a human observer was associated with a statistically significant increase in the number of latrine events, i.e., the users modified their behavior in response to the observer. Another advantage of the PLUM was the ability to measure activity continuously for an entire week. A shortcoming of the PLUM was the inability to separate latrine events that occurred in immediate succession, leading to possible undercounting during high-traffic periods. The PLUM is a promising technology that can provide detailed measures of latrine use to improve the understanding of sanitation behaviors and how to modify them and for assessing the intended health, livelihood, and environmental benefits of improved sanitation.