Today’s collection links bedside physiology, diagnostic testing, emerging therapeutics and the systems needed to turn clinical data into safer care. Hypokalaemia and sleep apnoea require clinicians to match investigation and treatment to mechanism. Liquid biopsy and troriluzole show why a technically impressive test or promising signal is not the same as proven clinical utility. Digital surgery asks the same question at service level: what decision will the data improve?

The acute-care material begins with hypokalaemia. A low potassium result may reflect true loss through diuretics, vomiting or diarrhoea, or a temporary intracellular shift after continuous albuterol. That distinction matters because a mild shift without ECG change or other risk features may not require the same response as substantial body potassium depletion. The ECG adds information beyond the laboratory number. U waves can appear, while fusion of the T and U waves into a TU complex indicates prolonged repolarisation and greater risk of ventricular arrhythmia.

Replacement should therefore account for symptoms, ECG findings, cause and severity. Stable moderate hypokalaemia is often managed orally, whereas values below 2.5 usually prompt intravenous replacement and consideration of admission. Magnesium should be assessed and replaced when appropriate because hypomagnesaemia commonly accompanies low potassium. Episodic marked weakness without an obvious history of potassium loss should also raise the possibility of hypokalaemic periodic paralysis.

The respiratory material moves from recognition to test selection and treatment follow-up. Obstructive sleep apnoea involves recurrent upper-airway collapse despite continued respiratory effort; central sleep apnoea reflects reduced drive to initiate breathing. Loud snoring, witnessed apnoeas or gasping, daytime sleepiness, difficult-to-control hypertension, obesity and upper-airway anatomy can increase suspicion, while reduced ejection fraction heart failure, chronic opioid use and neuromuscular disease make central events or hypoventilation more relevant.

Home sleep apnoea testing works best for uncomplicated patients with a high pre-test probability of obstructive disease. It is better at ruling disease in than ruling it out, so a technically negative home study should not end assessment when the history remains strongly suggestive. In-laboratory polysomnography is more appropriate when important cardiopulmonary or neuromuscular comorbidity is present, central sleep apnoea is suspected or uncertainty remains after home testing.

Report terminology also matters. The apnoea-hypopnoea index uses measured sleep time, whereas the respiratory event index uses recording or monitoring time when sleep is not directly measured. Once positive airway pressure treatment begins, early review of mask fit, leak, ramp, expiratory pressure relief, humidity, residual events and pressure requirements may prevent discomfort from becoming treatment failure. Oral appliances, positional approaches and weight management can form part of a tailored plan when appropriate.

The oncology episode examines circulating tumour DNA, or ctDNA, among a much larger background of normal cell-free DNA. In metastatic disease, a liquid biopsy may identify an actionable mutation and guide targeted treatment. After curative-intent treatment, tumour-informed assays can track mutations identified in the original cancer and look for molecular residual disease.

Their high specificity does not remove important limitations. A negative result may miss a low-shedding tumour, does not exclude a new primary cancer and should not be treated as proof of cure. Clinical validation also does not automatically establish that changing treatment improves outcomes. De-escalating standard adjuvant therapy solely because ctDNA is negative remains unsupported outside a monitored trial, and low-level positivity should be interpreted alongside imaging, treatment history and specialist review. Pre-test counselling should include the exact clinical question, false-negative risk, use of finite tumour tissue and how each possible result might alter management.

The neurological evidence update applies similar discipline to troriluzole in spinocerebellar ataxia. The biological rationale centres on glutamate modulation and possible reduction of excitotoxic injury, but the 48-week randomised trial misses its primary endpoint despite a numerical direction favouring treatment. Slower-than-expected progression in the placebo group may have reduced the ability to detect a difference.

Longer externally controlled follow-up suggests slower progression, yet propensity matching can balance only measured characteristics and cannot remove the bias of a non-randomised comparator. The appropriate clinical message is neither dismissal nor premature reassurance: troriluzole remains an advanced candidate, but approval has no clear timeline and current counselling should separate biological plausibility, randomised evidence, observational follow-up and regulatory decisions.

Digital surgery broadens the discussion from individual tests to learning systems. Its value lies not in passively storing operative video, but in combining visual data, sensors, analytics and connectivity to answer defined questions about technique, workflow, education or safety. Short, step-specific clips can support deliberate practice; whole-room review can reveal communication and team behaviours; and objective data can expose avoidable inefficiency such as repeated camera cleaning or unused instruments.

These benefits depend on psychologically safe multidisciplinary review rather than blame. Telepresence may extend specialist input to remote settings, but it remains clinical care and requires agreed consent, credentialing, licensure, liability, compensation, reliable connectivity and clear responsibility. A digital platform without curriculum, feedback processes and governance risks becoming another unused repository rather than a tool that changes care.