Objectives To determine emergency department (ED) physicians’ perceptions regarding hospital companions being prohibited from accompanying the patient during COVID-19. Design Two qualitative datasets were combined. Data collected included voice recordings, narrative interviewing and semistructured interviews. A reflexive thematic analysis was conducted and guided by the Normalisation Process Theory. Setting Six hospital EDs in the Western Cape, South Africa. Participants Convenience sampling was used to recruit a total of eight physicians working full time in the ED during COVID-19. Results The lack of physical companions provided an opportunity for physicians to assess and reflect on a companion’s role in efficient patient care. Physicians perceived that the COVID-19 restrictions illuminated that patient companions engaged in the ED as providers contributing to patient care by providing collateral information and patient support, while simultaneously engaging as consumers detracting physicians from their priorities and patient care. These restrictions prompted the physicians to consider how they understand their patients largely through the companions. When companions became virtual, the physicians were forced to shift how they perceive their patient, which included increased empathy. Conclusion The reflections of providers can feed into discussions about values within the healthcare system and can help explore the balance between medical and social safety, especially with companion restrictions still being practised in some hospitals. These perceptions illuminate various tradeoffs physicians had to consider throughout the pandemic and may be used to improve companion policies when planning for the continuation of the COVID-19 pandemic and future disease outbreaks.

Background/Objectives: Unruptured intracranial aneurysms (UIAs) affect 2.3–6.6% of the population. Traditional complication classification systems focus solely on therapeutic interventions, underweighting the burden of neurological deficits in neurovascular patients. This study applied the multidimensional Therapy-Disability-Neurology (TDN) grade to a consecutive cohort of patients treated for UIA in order to characterize the distribution, severity and clinical correlates of adverse events (AEs); the comparison between treatment modalities was a secondary, exploratory objective. Methods: We retrospectively analyzed 296 patients treated for UIA between 2016 and 2025 (microsurgical treatment n = 135, 45.6%; endovascular treatment n = 161, 54.4%). The TDN grade, a three-domain assessment incorporating the therapeutic resources required to treat an AE, functional disability and neurological deficits, was assigned for all AEs occurring during the index hospitalization. AEs were defined as any deviation from the expected post-treatment course, including minor events not usually reported as complications. The primary endpoint was the TDN grade; secondary endpoints included individual AE categories, severe AEs (TDN ≥ 4), functional outcome and length of stay. Results: AEs occurred in 89 patients (30.1%); 37 patients (12.5%) were graded TDN 1 or 2 and 52 (17.6%) TDN 3 to 5, the range corresponding to morbidity as conventionally reported. Four patients (1.4%) died during the index hospitalization or within 30 days of treatment. Median TDN grades were 0 in both groups (microsurgical IQR 0–2; endovascular IQR 0–1; and p = 0.575). The spectrum of AE categories differed between treatment modalities. The TDN grade correlated with GOSE and KPS at discharge and with length of hospital stay (all p < 0.001); length of stay was explained better by the three subgrades together than by the therapy subgrade alone (R² 0.18 vs. 0.13). Treatment modality was not associated with the TDN grade in unadjusted analysis; exploratory adjusted associations with any AE depended on model specification, and no association was found for severe AEs. Conclusions: In this cohort treated for UIA, the TDN grade captured the clinical severity of AEs and may serve as a component of multidimensional outcome assessment in future prospective studies, provided its domains are defined a priori.