Measures of performance and proficiency in robotic assisted surgery: a systematic review.

Robotic assisted surgery (RAS) has seen a global rise in adoption. Despite this, there is not a standardised training curricula nor a standardised measure of performance. We performed a systematic review across the surgical specialties in RAS and evaluated tools used to assess surgeons' technical performance. Using the PRISMA 2020 guidelines, Pubmed, Embase and the Cochrane Library were searched systematically for full texts published on or after January 2020-January 2022. Observational studies and RCTs were included; review articles and systematic reviews were excluded. The papers' quality and bias score were assessed using the Newcastle Ottawa Score for the observational studies and Cochrane Risk Tool for the RCTs. The initial search yielded 1189 papers of which 72 fit the eligibility criteria. 27 unique performance metrics were identified. Global assessments were the most common tool of assessment (n = 13); the most used was GEARS (Global Evaluative Assessment of Robotic Skills). 11 metrics (42%) were objective tools of performance. Automated performance metrics (APMs) were the most widely used objective metrics whilst the remaining (n = 15, 58%) were subjective. The results demonstrate variation in tools used to assess technical performance in RAS. A large proportion of the metrics are subjective measures which increases the risk of bias amongst users. A standardised objective metric which measures all domains of technical performance from global to cognitive is required. The metric should be applicable to all RAS procedures and easily implementable. Automated performance metrics (APMs) have demonstrated promise in their wide use of accurate measures.

Performance of large language models at the MRCS Part A: a tool for medical education?

INTRODUCTION: The Intercollegiate Membership of the Royal College of Surgeons examination (MRCS) Part A assesses generic surgical sciences and applied knowledge using 300 multiple-choice Single Best Answer items. Large Language Models (LLMs) are trained on vast amounts of text to generate natural language outputs, and applications in healthcare and medical education are rising. METHODS: Two LLMs, ChatGPT (OpenAI) and Bard (Google AI), were tested using 300 questions from a popular MRCS Part A question bank without/with need for justification (NJ/J). LLM outputs were scored according to accuracy, concordance and insight. RESULTS: ChatGPT achieved 85.7%/84.3% accuracy for NJ/J encodings. Bard achieved 64%/64.3% accuracy for NJ/J encodings. ChatGPT and Bard displayed high levels of concordance for NJ (95.3%; 81.7%) and J (93.7%; 79.7%) encodings, respectively. ChatGPT and Bard provided an insightful statement in >98% and >86% outputs, respectively. DISCUSSION: This study demonstrates that ChatGPT achieves passing-level accuracy at MRCS Part A, and both LLMs achieve high concordance and provide insightful responses to test questions. Instances of clinically inappropriate or inaccurate decision-making, incomplete appreciation of nuanced clinical scenarios and utilisation of out-of-date guidance was, however, noted. LLMs are accessible and time-efficient tools, access vast clinical knowledge, and may reduce the emphasis on factual recall in medical education and assessment. CONCLUSION: ChatGPT achieves passing-level accuracy for MRCS Part A with concordant and insightful outputs. Future applications of LLMs in healthcare must be cautious of hallucinations and incorrect reasoning but have the potential to develop AI-supported clinicians.

Inference under unequal probability sampling with the Bayesian exponentially tilted empirical likelihood.

Fully Bayesian inference in the presence of unequal probability sampling requires stronger structural assumptions on the data-generating distribution than frequentist semiparametric methods, but offers the potential for improved small-sample inference and convenient evidence synthesis. We demonstrate that the Bayesian exponentially tilted empirical likelihood can be used to combine the practical benefits of Bayesian inference with the robustness and attractive large-sample properties of frequentist approaches. Estimators defined as the solutions to unbiased estimating equations can be used to define a semiparametric model through the set of corresponding moment constraints. We prove Bernstein-von Mises theorems which show that the posterior constructed from the resulting exponentially tilted empirical likelihood becomes approximately normal, centred at the chosen estimator with matching asymptotic variance; thus, the posterior has properties analogous to those of the estimator, such as double robustness, and the frequentist coverage of any credible set will be approximately equal to its credibility. The proposed method can be used to obtain modified versions of existing estimators with improved properties, such as guarantees that the estimator lies within the parameter space. Unlike existing Bayesian proposals, our method does not prescribe a particular choice of prior or require posterior variance correction, and simulations suggest that it provides superior performance in terms of frequentist criteria.

Novel tumour suppressive protein encoded by circular RNA, circ-SHPRH, in glioblastomas.

The previously unchartered gene expression territory governed by circular RNAs is becoming clearer with the onset of deeper sequencing technologies. The translation of circular RNAs remained a controversial theory until earlier this year, when two studies [1, 2] showed endogenous circular RNA translation in vitro and in vivo, and have further provided mechanistical evidence. In this edition of Oncogene, Zhang et al., provide evidence for the first circular RNA translated with relevance to cancer; a novel tumour suppressor protein, SHPRH-146aa, produced by circ-SHPRH driven by IRES elements. The novel tumour suppressor protein produced by the circular RNA was found to work in synergy with the full-length protein, behaving as a protective decoy molecule to decrease degradation, and thus increasing the tumour suppressive functionality of the gene. An extended patient survival time was seen in glioblastoma patients with elevated levels of SHPRH-146aa. This study also marks the discovery of the first circular RNA with an overlapping initiation and termination codon, resulting in the translation of the full circRNA, exploring a mechanism not previously found or seen.

Association of serum electrolytes and smoking with salivary gland stone formation.

To further define potential factors that may contribute to stone formation in salivary glands (sialolithiasis), a retrospective chart review was performed of patients diagnosed with sialolithiasis between March 1, 1998 and February 29, 2012. Information on salivary gland stone number, location and size, medical history, medications, and serum electrolyte levels were collected. Associations between electrolyte levels and stone characteristics (such as stone number and size) were examined. Fifty-nine patients were identified; their median age was 58 years (range 25-89 years) and most were male (95%). Salivary stones were most commonly located in the submandibular glands (83%). Thirty-five patients (59%) had a smoking history, with 16 (27%) reported as current smokers. There was a significant association between current smoker status and stone size (mean largest stone size 12.4±8.8mm vs. 7.5±4.8mm in current smokers vs. non-smokers; P=0.03). Serum sodium levels (r=0.32, P=0.014) and serum potassium levels (r=0.31, P=0.017) showed significant positive correlations with stone size. While the aetiology of sialolithiasis remains unclear, smoking (which can contribute to reduced saliva flow) and higher serum sodium levels (which can reflect volume depletion) are associated with larger salivary stones.

Prevalence of levator ani muscle injury in Chinese women after first delivery.

OBJECTIVES: To assess the prevalence of levator ani muscle injury in Chinese women after their first delivery and investigate associated factors. METHODS: A prospective observational study was conducted involving Chinese nulliparous women recruited in the first trimester of pregnancy. Translabial ultrasound was performed at 35-38 weeks' gestation and 8 weeks postpartum and three-dimensional volume datasets were obtained. Offline analysis to detect levator ani muscle injury was performed by investigators blinded to the delivery details. RESULTS: 339 women, with a mean age of 30.6 ± 3.9 years, completed the study. Overall, 201 (59.3%) had a spontaneous vaginal delivery, 62 (18.3%) an operative vaginal delivery (48 ventouse extraction and 14 forceps delivery), 14 (4.1%) an elective Cesarean section and 62 (18.3%) an emergency Cesarean section. No levator ani muscle injury was detected in any woman antenatally. After vaginal delivery, 57 (21.7% (95% CI, 16.7-26.7%)) women had levator ani muscle injury. The rates of injury for spontaneous vaginal delivery, ventouse extraction and forceps delivery were 15.4%, 33.3% and 71.4%, respectively. There was no levator ani muscle injury in the Cesarean section groups. Logistic regression analysis showed that only operative vaginal delivery (odds ratio, 3.09) was associated with an independent increase in the likelihood of levator ani muscle injury. Intrapartum epidural analgesics, duration of second stage of labor and infant birth weight were not independently associated with levator ani muscle injury. CONCLUSIONS: The prevalence of levator ani muscle injury in Chinese women after their first vaginal delivery was 21.7% (95% CI, 16.7-26.7%). Operative vaginal delivery was found to increase the likelihood of women suffering such injury. A longer follow-up of these women and future studies on the effects of episiotomy are proposed.

Characterization of a functional AT1A angiotensin receptor in pancreatoma AR4-2J cells.

Functional angiotensin receptors were characterized in the rat pancreatic acinar cell line AR4-2J. Angiotensin II stimulated a dose-dependent release of amylase and production of inositol phosphates. Results of high-performance liquid chromatography separation of inositol phosphates indicated that angiotensin stimulated the rapid accumulation of inositol 1,3,4-trisphosphate. Angiotensin II and angiotensin III were at least an order of magnitude more potent than angiotensin I in the stimulation of amylase release. The angiotensin II-stimulated amylase release was blocked by losartan, a selective AT1 angiotensin antagonist. The selective AT2 angiotensin receptor ligands CGP42112 did not alter angiotensin II-stimulated amylase released. However, CGP42112 stimulated amylase release at micromolar concentrations with a potency similar to angiotensin I. Analysis of mRNA expression by reverse transcription polymerase chain reaction suggested that AT1A was the predominant type-I angiotensin receptor expressed in the AR4-2J cells.

Immunohistochemical localization of type-II (AT2) angiotensin receptors with a polyclonal antibody against a peptide from the C-terminal tail.

A polyclonal antibody has been prepared against a synthetic peptide derived from the C-terminal tail of the cloned rat AT2 angiotensin receptor, corresponding to amino acid residue 341-351. The antibody was of high titer and displayed monospecific activity toward the synthetic peptide in the ELISA assay. Western blot analysis indicated that the antiserum recognised only a single protein band with a mean apparent molecular mass of 75.4 kDa in the rat adrenals. Immunohistochemical studies with affinity purified antibody localised immunoreactive AT2 angiotensin receptor in medulla cells of the adrenals. Immunoreactivity was also observed in pyramidal tract, but no specific immunoreactivity can be detected in regions of rat brain that are known to express AT2 angiotensin receptors, including inferior olive, locus coeruleus and cerebellum.