Table 2

Recommendations for national patient safety incident reporting systems

No.Recommendation% Consensus (n=26)
Role of reporting systems
1.Reporting systems should be used to identify the types of safety problems that exist96.2
2.Reporting systems should be used to detect rare events not picked up by other methods92.3
3.Reporting systems should be used to share learning between hospitals92.3
4.Reporting system data should be used as indicators of the safety culture of a hospital80.8
5.Mandatory system reporting systems should be used to measure the rate of specific types of reported harm (eg, wrong site surgery) in a hospital73.1
Roles reporting systems cannot fulfil
6.Reporting systems are a not a valid and reliable measure of how safe a hospital is80.8
7.Reporting system data should not be used to measure the national incidence of harm (eg, within the national health service)76.9
 Reporting systems should not be used to identify unsafe hospitals69.0
Voluntary system reporting systems should not be used to measure the rate of harm in a hospital65.4
Reporting systems should not be used to identify unsafe healthcare professionals (eg, doctors and nurses)65.4
Methods to maximise learning from reporting systems
8.Near misses or no harm events should be reported96.2
9.Anonymous reporting data should be readily available to research groups for analysis92.3
10.Incident reports should be used in educational programmes for trainees to promote quality improvement92.3
11.Incident classification systems should be standardised to/allow comparisons84.6
12.Minimum data set should include hospital number, patient age, time/date and location of incident and specialty caring for patients84.6
13.Staff should be encouraged to propose solutions for events at the time of the report84.6
14.Reports should contain patient identifiers so they can be linked to other data sets73.1
15.There should be national priorities for reporting certain events77.0
16.The quality of incident reports submitted to a reporting system is more important for learning than the quantity of reports77.0
17.All reporters should have the option to keep their report anonymous so that they are not identified73.1
Minimum data set should include national identifying number (eg, NHS number or social security number)53.8
Fewer and better described incidents should be encouraged compared to more and less well described42.3
Minimum data set should include the consultant or attending involved in patient care or attending in charge of care42.3
Where different types of incidents should be reported
18.Hospitals should submit their solutions to safety problems nationally for shared learning88.0
19.Device incidents should be reported both nationally and locally88.0
20.Never events should be reported both nationally and locally88.0
21.Hospital-acquired infections should be reported both nationally and locally80.8
22.Medication incidents should be reported both nationally and locally76.9
23.Staff shortages should be reported locally only72.0
24.Initiatives to prevent harm should be generated at a hospital level72.0
25.Incidents that lead to harm should be reported both nationally and locally77.0
Reports should objectively classify what harm was caused and not the potential for harm65.3
There should be specific criteria for what to report65.4
Reports of misconduct by other healthcare professionals should be reported locally only61.5
Events from morbidity and mortality meetings should be reported to a national system46.2
Complaints about other members of staff or staffing issues should NOT be reported to a national system46.2
Near misses should be reported locally only (other options: nationally or nationally and locally)42.3
Professional misconduct incidents should be reported locally only (other options: nationally or nationally and locally)42.3
Voluntary and mandatory data capture
26.There should be mandatory reporting of never events or serious events such as wrong site surgery92.0
27.Near misses should be captured by a voluntary system88.0
28.Medication incidents should be captured by a voluntary system80.8
29.Device incidents should be captured by a mandatory system80.8
30.Hospital-acquired infections should be captured by a mandatory system77.0
Staff shortages or risk assessments should be captured by a voluntary system53.8
Investigation of incidents and accountability
31.Hospitals should have an executive board member responsible for patient safety100
32.Hospitals should be accountable for investigating their own reports84.6
33.Hospitals should not determine their own reporting priorities84.6
34.Reporting systems should give individual feedback to reporters84.6
35.Never events and incidents leading to death and severe harm should be prioritised for investigation80.8
36.Reporters who report deaths should receive specific feedback after analysis80.8
37.Reporters who report never events should receive specific feedback after analysis76.9
38.Analysis of all incident types is desirable; however, near miss incidents are less of a priority for investigation than never events and incident leading to death and severe harm73.1
Reporters who report device incidents should receive specific feedback after analysis69.2
Clinical teams external to the hospital should investigate reports of severe patient harm61.5
Who should provide feedback for harm or death incidents (multiple-choice options):
 National patient safety experts34.6
 External investigating clinician30.8
 Consultant/attending involved in patient care26.9
 Local risk manager7.7
Staff training
39.Greater emphasis needs to be placed on training staff to identify and report safety incidents80.8
40.Senior nurses, doctors, managerial staff and other healthcare professionals should be trained to routinely investigate patient/safety incidents73.1
 Junior nurses, administration staff and medical students should be trained routinely to investigate patient safety incidents65.3

  • Italicised statements indicate areas not reaching consensus.