Methods of data collection
| Accreditation surveys | Performed during accreditation, not always current and checked for accuracy. Best for structural elements |
| Computer tracking systems | Collected as part of routine care—essential for tracking timeliness of care but often not collected concurrently with care and open to data manipulation |
| Clinical notes | Often poor and inconsistent documentation, labour intensive to collect |
| Administration systems | Collected for billing and epidemiological purposes, not always checked by clinicians, often not sufficient data for adequate risk adjustment |
| Surveys | Patient surveys useful to gain qualitative data regarding service, numerical data of limited value |
| Audits of clinical topics | Labour intensive but useful to drive process changes associated with evidence-based practice |
| Random chart audits | Labour intensive but useful if sample a percentage. Drives good documentation |
| Registries | Most useful for high-risk/high cost procedures and illness. Essential for examining clinical outcomes over time. Also linking key processes with outcome |
| Video | Limited use in certain scenarios, such as resuscitation, team training. Immediate feedback to staff, can drive behavioural change rapidly |
| Incident/sentinel event reporting | Useful to alert ED to developing issues and engage staff. Should not be used quantitatively |
| Mortality plus morbidity meetings | Important to engage staff and discuss issues. Must be performed in non-punitive manner |
Commonly available data sources are listed. A major impediment to developing a comprehensive framework for measuring quality is the lack of adequate data systems in emergency departments (ED).