Patients assigned to the intervention arm received every element of conventional emergency neurosurgical care plus a structured, prospective safety overlay derived from Healthcare Failure Mode and Effect Analysis. The add-on programme was delivered by a nine-member, hospital-authorised HFMEA team (one chief neurosurgeon, two attending neurosurgeons, one head nurse, four bedside nurses, one quality-manager) who had completed an 8-hour training course on HFMEA principles, risk-priority-number (RPN) scoring and decision-tree analysis. The programme ran from the minute the participant arrived in the emergency department until 30-days post-operation and consisted of six sequential steps that were re-applied every month.

The control group received conventional nursing risk management, including:

1. Admission assessment: vital signs monitoring, neurological function assessment, Hunt-Hess grading;
2. Preoperative nursing: condition observation, psychological care, preoperative preparation;
3. Postoperative nursing: close monitoring of vital signs, neurological function observation, complication prevention;
4. Health education: disease knowledge, precautions, rehabilitation guidance;
5. Discharge guidance: medication instructions, follow-up arrangements, lifestyle recommendations.

Alongside standard care, these patients were managed with an HFMEA-based safety bundle. A trained nine-member team had pre-identified 12 highest-risk failure points (delayed ICP checks, missed re-bleeding signs, vasospasm, seizures, infection, etc.). From admission to day-30, nurses followed printed checklists and electronic order-sets: neuro-vitals every 15-30 min, BP target 140-160 mmHg, daily TCDs, automatic "red-flag" escalation for sudden headache/GCS drop, Triple-H protocol for velocities >120 cm/s, prophylactic levetiracetam for severe grades, 30° head positioning, chlorhexidine/line bundles, pain-delirium scale, bed-alarm, SBAR hand-over and a 5-minute family video with teach-back. Compliance was tracked in real time and reviewed monthly; measures were updated if failure rates did not fall within six weeks.

Incidence of perioperative nursing adverse events, including:

• Rebleeding
• Cerebral vasospasm
• Increased intracranial pressure
• Seizures
• Pulmonary infection
• Pressure ulcers
• Deep vein thrombosis
• Catheter-related infections
• Falls/bed exits
• Medication errors, etc.

Nursing Quality Score: Newcastle Satisfaction with Nursing Scale (NSNS), including four dimensions: condition observation, basic nursing, specialized nursing, and health education, each scored 0-25 points, total score 0-100 points, with higher scores indicating better nursing quality.