Dear Mr. Hodson, Thank you for bringing to our attention the incident involving the above cited product. The complaint details, subsequent investigation and resultant findings are described below. Incident as reported to Philips Healthcare: On August 27, 2021, Philips received the Coroner’s report titled ‘Regulation 28 Report to Prevent Future Deaths ’ issued by you, Adam Hodson, HM Assistant Coroner. The following was reported: “On 1 April 2021 I commenced an investigation into the death of [redacted]. The investigation concluded at the end of the inquest. The conclusion of the inquest was: Death was a consequence of natural causes. The deceased was admitted to Good Hope Hospital on 26 February 2021 where [redacted] was treated for heart failure. [Redacted] was commenced on continual cardiac monitoring from 6th to 8th March, where two periods of ventricular standstill were recorded but were missed due to a combination of policy, staffing, workplace and equipment issues. [Redacted] suffered a cardiac arrest on 8 March 2021 and was treated on ICU for 22 days in total. Despite treatment, [redacted] deteriorated rapidly on [redacted] final day, and [redacted] died at 13:40, 30 March 2021. Had the periods of ventricular standstill been detected, [redacted] would have been admitted to CCU for monitoring, but [redacted] subsequent cardiac arrests could not have been prevented.” The report summarized concerns of a risk that future deaths will occur unless action is taken. Specifically, the report discussed a retrospective review of the telemetry monitoring carried out on this patient during the period of March 6-8, 2021. Two periods of ventricular standstill were noted: one at 12:43 on March 6, 2021 where the telemetry recorded a period of 4 seconds of ventricular standstill; and one at 13:09 on March 8, 2021 where the telemetry recorded a period of 10 seconds of ventricular standstill. The report further indicated that Medical Engineers were asked to analyse the telemetry and noted that on March 8, 2021 the monitor's alarm had triggered and that it was a self-terminating alarm as the heart rhythm had corrected itself. Medical Engineers and staff indicated there was no evidence that staff muted the alarm or that there was any fault with the equipment. Thus, it was concluded per the report that this self-correcting function is an intended function of the monitor.

Philips Healthcare

Investigation:

Strip Review As the Device Manufacturer, Philips’ first contact from the Hospital in regard to this incident was on June 29, 2021. This was an informal request to investigate to a Philips Clinical Application Specialist (CAS) whilst visiting Heartlands Hospital on another matter. The question from the Matron was whether a Ventricular Standstill would cause an alarm and, if so, how long it would last for. A general ECG waveform strip and an Asystole Alarm strip, printed out around the time of the incident, was emailed to the CAS on July 18, 2021. The Asystole Alarm strip showed this alarm generated at the time of the Ventricular Standstill, but without the Clinical Audit Logs, the duration of this alarm and any other actions taken around the time of this alarm are unknown.

The configuration of the PIC iX was reviewed by the CAS at the time of this first site visit and found the systems functioning as designed. Red alarms were enabled and cannot be terminated without end user intervention.

Further review of the available information found that there was not a strip for the incident on March 6, 2021 to review. A strip was provided for March 8, 2021 13:09:13 that revealed an ECG waveform with beat labels. A (artifact), M (missed), ? (questionable), and one beat labeled N (normal) along with a period of ventricular standstill. The strips do not include the response taken by the user nor the length of time the alarm sounded. Philips was not notified of the March 2021 incidents until greater than three months following the events. As a result, Clinical Audit Logs required for a complete investigation were no longer available.

Self-Terminating Alarms Per the report, hospital Medical Engineers were asked to review the Telemetry and the hospital Medical Engineers noted that on March 8, 2021, the monitors alarm had triggered and that it was a self-terminating alarm. Philips is interpreting this description as the alarm stopped on its own accord without user interaction. Any ventricular standstill of 4-10 seconds of duration would meet the definition of Asystole which is a red arrhythmia alarm. There is not a configuration to change Asystole or any other red arrhythmia alarm to self-terminate for a Telemetry / Central Station setup. In summary, the product is not designed to self-terminate red alarms nor is Philips able to reproduce the described behavior.

Good Hope Hospital Site Visit Philips support personnel performed a site visit to Good Hope Hospital on September 24, 2021.

An ECG simulator was utilized to test the customer’s system using the same equipment that was used on the patient in March. An Asystole alarm was generated, and the simulator was reset to normal sinus rhythm and the alarm continued even though the asystole condition had ended. Testing revealed no self-terminating of alarms. The devices were found to perform as expected.

The configuration was further reviewed by the Business Unit Clinical Specialist from the data collected during this site visit and the initial assessment of red alarms cannot be terminated without intervention was verified. While the original description of the issue was reported as the Information Center iX; this issue is with alarm detection and communication from the Telemetry to the PIC iX. When Telemetry is used with PIC iX, the alarm settings are sent from the PIC. Red alarms cannot be set to self terminate in the PIC configuration.

Philips Healthcare Complaint Review Complaints were analyzed for the last 3 years world-wide and found one other similar complaint. This complaint was a non-adverse event, it was found that the alarm had been silenced by an end user initially and was giving an alarm reminder with brief audible alerts since alarm reminders were enabled. The equipment was operating as designed.

Conclusion:

Philips Healthcare investigated the reported incident and concluded that the device operates per specification.
• A full investigation into the alleged self-terminating alarm incident was not possible as the required logs were not obtained when the incident occurred.
• There is not a configuration available to enable asystole or any other red arrhythmia alarm to self- terminate for Telemetry / Central Station monitoring.
• Termination of asystole or other red arrythmia alarm with the current configuration requires end user intervention, as shown with the simulation test with NHS personnel present.
• Based on the current configurations of the system, if a red alarm is silenced by the user but persists, the expected behavior is for the audible alarm to end, the visual alarm to remain and a brief series of audible tones to be provided every 3 minutes, which stop after several seconds on their own as long as the condition remains.
• An on-site visit found the equipment to be operating as designed.

A full explanation of the configuration options is available in the Configuration Guide. Philips’ clinical support continues to be available to assist with any changes to desired workflow.

Please be assured that Philips Healthcare is committed to providing quality products and solutions. If you have any questions on the investigation of this incident, please contact us.

Sincerely,

Philips Healthcare

Quality Manager Complaint Investigator

Philips Electronics UK Limited Quality and Regulatory

Philips Centre, Ascent 1 One Deerfield Centre

Aerospace Blvd 13560 Morris Road

Farnborough, GU14 6XW Alpharetta, GA 30004

Cc: Good Hope Hospital

Dear Mr Hodson Inquest touching the death of Anne Geraghty Inquest – 25 August 2021 I write further to the Inquest touching the death of Anne Geraghty in which you issued a Regulation 28 Report to Prevent Future Deaths to Philips in relation to their central monitoring stations. Following receipt of the Report, a meeting took place with our clinical leads and representatives from Philips to discuss the equipment and the issues that had arisen in this case. Philips have also subsequently attended and undertaken an inspection of the specific machine used in monitoring Ms Geraghty. During the Inquest, the evidence presented to you in both our internal investigation and through witness evidence, was that the alarms on the monitoring equipment had self-terminated. This was based on evidence obtained from nursing staff and our medical engineering team that the alarm had not been silenced and therefore it was assumed that it had self-terminated on Ms Geraghty’s heart rhythm returning to normal. During our discussions with Philips subsequent to the Inquest Hearing we have become aware that the machines are not configured in a way that allows alarms to self-terminate upon the heart rhythm returning to normal. The alarm must therefore have been silenced and we believe this is likely to have been a user error based on the fact the trace would have shown a normal rhythm within seconds of the alarm sounding. In light of this information, we would like to apologise for our honest and mistaken belief that the alarms on the equipment self-terminated. From our subsequent enquiries it appears that the confusion as to the functionality of the equipment is likely to have arisen during conversations with our ward staff and medical engineers, and subsequently with Philips’s support personnel. There is no evidence of any deliberate provision of incorrect information. As stated above, it is considered most likely that a member of staff must have silenced the alarm, noted that the heart was in normal rhythm and took no further action. Once an alarm has been silenced, the configuration of the equipment does not automatically highlight the abnormal section of

the ECG and therefore no further action was taken at that point. It was only on a later examination of the log that the episode became apparent. Our clinical evidence remains that this episode, and the failure to identify the same at the time, would not have affected the outcome as regards Ms Geraghty’s unfortunate death.

Action Plan

As soon as we became aware of the machine configuration, following the inspection undertaken by Philips, we took a number of immediate steps to raise awareness of the alarm system in place which included an email to all cardiology staff on each of our sites detailing how the alarm system functions. This information was also discussed in ward safety huddles.

In addition to the above, we have also put in place a number of other actions as set out below.

Education and training

In addition to internal education, the Trust is working with Philips to provide additional refresher training and education to our staff around the monitoring equipment and relevant human factors in the management of the same, including efforts to avoid/minimise any risk of ‘alarm fatigue’. This will include an Alarm Management Programme across our cardiology service at all our sites.

Both our divisional and medical device team educators will be involved with the refresher training, and we are currently in discussion with Philips to agree dates. Our aim is to have completed the education and training within the next 3 months.

Machine configuration

As set out above, once an alarm is silenced, there is no facility which allows for the immediate review of the abnormal ECG if no longer on view. In partnership with Philips, we will be exploring whether it is possible to alter the configuration of the software and to explore with them the possibility of putting in place a facility which allows for review of the abnormal section of the ECG at the point it is silenced.

Retention of traces

Currently trace logs are only retained for a period of 50 days before they are removed from the system. In this case, the traces had already been removed at the point where Philips were asked to review them and so this could not assist the authors of the SI report relied upon at the Inquest Hearing. In recognition of this issue the Trust is exploring the retention of trace logs locally for an extended period which would allow for retrospective review.

I would like to assure you that we have reflected on this case and have taken immediate steps to ensure our teams are fully aware of how the alarm systems function on the monitoring equipment. We will continue to work with Philips and implement the actions set above which we hope provides you with assurance that steps are being taken to prevent any similar occurrences in the future.

Once again, we apologise to both the Court and the parties for the inadvertent incorrect evidence provided. This has been reflected on within both the Service and the Trust’s legal department and we can assure you that reflection on this incident will ensure increased scrutiny of evidence before it is provided to your office.