HospitalInspections.org

Based on observation, interview, and document review, the hospital failed to ensure an effective governing body legally responsible for the conduct of the hospital when:

A. The hospital failed to set up and maintain a safe environment of care when a clinical alarm monitoring system utilizing equipment and devices was set up without validating the system was safe and reliable. Refer to A 724.

B. The hospital did not provide adequate nursing services during a major change in the clinical alarm notification system (CANS) on two of two telemetry units. Refer to A 385.

C. The hospital failed to develop, implement and maintain an effective, ongoing, hospital-wide, data-driven quality assessment and performance improvement program that reflected the complexity of the hospital's organization and services. Refer to A 263.

These failures put patients at risk for adverse events, delays in rescue for life-threatening conditions, and death.

The cumulative effect of these systemic problems resulted in the inability of the Hospital to comply with the statutorily-mandated Condition of Participation for Governing Body.

Findings:

Review of the 12/4/15 governing body (GB) bylaws Section F-3 read, "It is recognized that the [GB] has ultimate legal and moral responsibility for health care and services rendered in the hospitals ..."

In an 11/30/18 at 12:50 p.m. interview with the Assistant Secretary to the Board (ASB), who also served as a corporate Senior Director of Quality and reported to the Executive Vice President for Quality/leader of Quality and Health Improvement Committee, ASB acknowledged that the governing body was ultimately responsible for all aspects of operations and patient safety at each of its 44 hospitals. ASB also indicated that governing body minutes captured only an "overview" of information received and discussed at each meeting, and did not show the content of specific information provided that was relevant to each hospital. ASB acknowledged that the minutes did not clearly display the governing body's obligation to oversee each hospital's performance, and ensure patient safety and quality of care.

Based on interview and document review the hospital failed to demonstrate that the governing body was responsible for all services furnished under contract/agreement as the overseer of such services to ensure that each vendor complied with all hospital conditions of participation and standards relevant to the vendor services for hundreds to thousands of vendors (hospital was unable to determine how many).

This failure put patients at risk for harm or adverse events due to potentially unsafe practices by a vendor's product or services.

Findings:

In an interview on 11/29/18 at 1:25 p.m., Administrator 1 (who managed the daily operations of the hospital) stated that a single governing body (GB) oversaw several large corporate entities including a health plan, dozens of licensed hospitals, clinics, hospices, and home health agencies. Administrator 1 indicated that the governing body delegated the authority to arrange for outside vendors to perform services in the hospital to several categories of leaders: National leaders who acquired and managed corporate contracts, Regional leaders who acquired and managed other vendor services across geographic regions of the country, leaders who managed California-specific contracts, and some contracts arranged and managed by a hospital site administrator like Administrator 1. For example, Vendor 1 provided middleware (software platform) to connect a mobile telemetry heart monitoring system that was utilized on 2 patient units of the hospital to a wireless messaging system to notify nurses taking care of the monitored patients when serious or life-threatening heart rhythms occurred. A national corporate leader acquired and managed that contract. Evaluations of the vendor's performance were completed by a shared information technology group outside the hospital. The national leader was accountable to the governing body, but Administrator 1 was not aware of how the national leader showed the governing body how that vendor performed and whether they complied with all applicable conditions of participation and standards for that patient care service.

Review of GB minutes between January 2017 and November 2018 showed no entries related to performance and compliance of Vendor 1, or other outside vendors providing patient care services in the hospital.

In an 11/30/18 at 12:50 p.m. interview with the Assistant Secretary to the Board (ASB), who also served as a corporate Senior Director of Quality and reported to the Executive Vice President for Quality/leader of Quality and Health Improvement Committee, ASB indicated that a corporate level of leaders handled national contracts like that of Vendor 1, and management was not delegated to the local hospital site Administrator for such national and regional contracts. ASB indicated that there might be hundreds to thousands of such vendor contracts involving patient care services in this hospital. ASB acknowledged that the governing body meeting minutes provided a generally broad overview of activities for all entities reporting to this single governing body, not just hospitals. ASB confirmed that the GB did not routinely hear presentations or reports of compliance for every vendor performing services in each site hospital. Therefore, the GB minutes did not show that the GB validated such vendor compliance.

Review of Policy SSC-1300 titled "Contract Oversight - Providers and Services (NCAL)," documented under item 1.0, "Contracts to support patient needs must assure that services are provided safely and effectively ... and oversight must be provided to assure compliance [with regulatory requirements]." The policy described the procedure for managing locally obtained contracts and local oversight for those. However, the policy made reference to oversight by a committee outside the hospital for sub-regional, regional, and national contracts. These latter categories were not incorporated into the local hospital's obligation to review, validate, or integrate into its own quality program metrics or reporting to local hospital and medical staff leaders.

Review of the Hospital Quality and Patient Safety Program Description Annual Work Plan and Evaluation, approved 1/9/18, read, "At least annually (more often as necessary), the hospital assesses the quality monitoring of the agencies, organizations and individuals with which it contracts for the provision of care, treatment and services provided to the hospital's patients. The hospital's clinical contract list will be approved by the Medical Executive Committee annually based on review of quality and performance data." "The individual assigned responsibility for each contract is accountable to review the contract expectations and establish appropriate quality and operational indicators and monitoring frequencies, and to report performance through the established quality structure." The Quality Council was responsible to prepare reports related to patient safety activities to be submitted to the GB QHIC, through the Medical Executive Committee on an ongoing basis.

Review of the 12/4/15 governing body bylaws documented under Quality and Health Improvement Committee (QHIC) Section E-10-B that the QHIC's duty was to provide oversight of systems designed to monitor on behalf of the [GB] that quality care and services were provided at a comparable level to all patients throughout the Program [entities], and provide oversight of the [organization that owns multiple entities] Program's quality assurance and improvement systems. Other duties included identifying and addressing deficiencies in quality, reviewing and approving standards for patient safety and risk management, and evaluating results of quality improvement activities. The Duties did not specify the evaluation of contracted services.

Further in the 12/4/15 GB bylaws Section H-3 read, "The [GB] may authorize any officer or officers and any agent or agents to enter into any contract or execute any instrument in the name of, and on behalf of, the corporation." No further entries documented the delegation for contract management to specific corporate leaders, and the hospital presented no policy that framed that responsibility, for contracts handled at various corporate levels.

Based on interview and document review the hospital failed to show a system to ensure that services performed under contract were evaluated to be safe and effective, evaluations were reported to the governing body, and an evaluation system was incorporated into the hospital's quality assessment and performance improvement (QAPI) plan for hundreds to thousands of vendors who provided patient care services to the hospital (hospital was unable to determine how many).

This failure put patients at risk for harm or adverse events due to potentially unsafe practices by a vendor's product or services.

Findings:

In an 11/29/18 at 1:25 p.m. interview, Administrator 1 (who managed the daily operations of the hospital and was accountable to the governing body) stated that the contracted services provided by outside vendors were evaluated either by staff at the hospital, or by regional or national corporate departments and leaders above the hospital. For local contracts, the hospital department managers were responsible to identify what to evaluate, create reports, and present them to Administrator 1. However, the hospital QAPI program did not collect performance data and outcomes for the hundreds to thousands of national and regional vendors not managed by the hospital.

In an 11/30/18 at 12:50 p.m. interview, the Assistant Secretary to the Board (ASB), who also served as a corporate Senior Director of Quality, stated that the hospital's quality and risk directors did not incorporate quality performance metrics from national/regional vendors (including Vendor 1 who provided medical device services for continuous physiologic monitoring patient care) who did provide services to patients at the local hospital.

Review of the Hospital Quality and Patient Safety Program Description Annual Work Plan and Evaluation, approved 1/9/18, read, "At least annually (more often as necessary), the hospital assesses the quality monitoring of the agencies, organizations and individuals with which it contracts for the provision of care, treatment and services provided to the hospital's patients. The hospital's clinical contract list will be approved by the Medical Executive Committee annually based on review of quality and performance data."

In a 12/5/18 at 9:30 a.m. interview, the Director of Risk Management Director (DRM) acknowledged that the local hospital quality department did not routinely collect the results of contract evaluations for vendors managed by various corporate level managers (including Vendor 1) and did not submit such results to the hospital's administration and medical staff as part of their responsibility to review safety and effectiveness of contracted vendors.

Based on interview and document review, the hospital failed to present a complete list of all outside vendors who performed patient care services.

This organizational failure posed confusion about who was accountable to evaluate and validate the services of outside vendors.

Findings:

Review of a document titled "Clinical Contracted Providers and Services" dated 2/28/17 listed 44 vendors who provided patient care-related services. The list did not include vendors who serviced various medical devices, including some related to intestinal surgery visualization and transmission of heart monitoring device alarms.

In an 11/29/18 at 1:15 p.m. review of a list of vendor contracts with Administrator 1, who managed daily operations of the hospital, Administrator 1 stated that the list contained only the local contracts that Administrator 1 directly managed. Hundreds to thousands of other contracts were managed by corporate leaders at higher levels of the organization who owned the hospital.

Review of Policy SSC-1300 titled "Contract Oversight - Providers and Services (NCAL)," required each hospital to have a list of patient care-related contracts currently used by the hospital, and approved by the hospital's medical staff leadership. The same policy distinguished authority to acquire contracts among local, sub-regional, regional, and national leaders, while not clearly setting an expectation for the contracts list to incorporate all relevant contracts from all levels of acquisition.

In an 11/30/18 at 3:30 p.m. interview, the Director of Risk Management (DRM) stated that the hospital had no complete list of vendors that included all patient care services provided by outside entities.

Review of the Hospital Quality and Patient Safety Program Description Annual Work Plan and Evaluation, approved 1/9/18, read, "At least annually (more often as necessary), the hospital assesses the quality monitoring of the agencies, organizations and individuals with which it contracts for the provision of care, treatment and services provided to the hospital's patients. The hospital's clinical contract list will be approved by the Medical Executive Committee annually based on review of quality and performance data."

Based on observation, interview and document review, the hospital failed to develop, implement and maintain an effective, ongoing, hospital-wide, data-driven quality assessment and performance improvement (QAPI) program that reflected the complexity of the hospital's organization and services, as evidenced by:

A. The hospital failed to show a system to ensure that services performed under contract were evaluated to be safe and effective, evaluations were reported to the governing body, and an evaluation system was incorporated into the hospital's QAPI plan for hundreds to thousands of vendors who provided patient care services to the hospital (hospital was unable to determine how many). Refer to A 084.

B. The hospital failed to set priorities for its performance improvement activities that focused on the high volume, high-risk and problem prone area of continuous heart monitoring alarm notification and timely response, determine the extent and severity of problems in that area, and assess/determine the impact of such problems on patient safety and outcomes. Refer to A 283.

C. The hospital leaders failed to ensure that the QAPI program prioritized patient safety concerns for the reliability of a clinical alarm notification system (CANS) that was used for continuous monitoring of potential heart function instability. Hospital leaders failed to ensure the CANS concerns were formally and effectively evaluated, in order to improve quality and safety for all patients affected by the CANS system. Refer to A 309.

D. The hospital leaders failed to ensure that adequate resources were allocated for assessing, improving and sustaining performance of a clinical alarm notification system (CANS) that was used for continuous monitoring of potential heart function instability, and for reducing risk to patients. Refer to A 315.

These failures put patients at risk for adverse events, delays in rescue for life-threatening conditions, and death.

The cumulative effect of these systemic problems resulted in the inability of the Hospital to comply with the statutorily-mandated Condition of Participation for QAPI.

Based on observation, interview and document review, the hospital failed to set priorities for its performance improvement activities that focused on the high volume, high-risk and problem prone area of continuous heart monitoring alarm notification and timely response, determine the extent and severity of problems in that area, and assess/determine the impact of such problems on patient safety and outcomes when:

Problems were identified with delayed and unreliable receipts of text messages from a telemetry patient's heart monitor recording center to the mobile phone upon which nurses relied to be informed of a patient's life-threatening heart activity. Phone reliability and functionality problems were repeatedly reported by mobile phone users from the start of a change in the telemetry system equipment and design in August 2017, continued through focused analysis following a patient death (Patient 1) potentially related to the telemetry system weaknesses, and persisted until the present time.

This failure put up to 48 of 48 telemetry patients on any given day at risk for delays in rescue from life-threatening conditions.

This failure precluded the hospital from taking a formal, systematic approach to problem-solving for a significant issue in a vulnerable patient population, to reduce risks to patients for delays in rescue for life-threatening conditions, and to improve care quality.

Findings:

Review of the Hospital Quality and Patient Safety Program Description (HQPSPD), last approved 1/9/18, under Section 1 Purpose documented that the governing body's Quality and Health Improvement Committee (QHIC), Senior Leaders, Medical Staff and Hospital Staff would demonstrate a consistent and collaborative approach to deliver safe, effective, patient centered and timely care within a quality assurance and performance improvement (QAPI) framework. The framework included "Proactive and prioritized performance improvements to prevent failure, mitigate hazards, and improve systems and process reliability." The hospital listed its values to include, "Ensuring that every patient has the right to care that is reliable, effective and safe."

The HQPSPD further described Safe Care as to ensure that the actual and potential hazards associated with high risk procedures, processes, and patient care populations were identified, assessed, and controlled in a way that demonstrated continuous improvement. Safe Place meant that the design, construction and operation of the hospital would maintain a safe environment of care, and ensured the evaluation, purchase and utilization of equipment and products in a way that promoted the effectiveness with which safe healthcare was provided. Safe Systems meant support systems would be identified, implemented and maintained to provide the right information, to the right people, at the right time, and would include meaningful measures of risk and safety.

During a tour of the hospital's third floor telemetry unit on 11/27/18 at 9:30 a.m., the Telemetry Manager (TM) indicated that the types of patients routinely admitted to the telemetry units included patients with heart attacks, recovery from invasive imaging studies of the heart, strokes, chest surgery, head and neck surgery. All were high risk for a life-threatening event and required continuous monitoring of heart function, and for some, blood oxygen levels. The hospital had two telemetry units, one on the third floor and one on the fourth floor, with capacity for a total of 48 telemetry beds.

The TM explained that telemetry patients were monitored for potential unstable heart conditions by way of mobile devices attached to the patient (boxes) that transmitted heart activity signals to a remote central station display, which in turn transmitted alarm signals to a middleware software server, which in turn transmitted alarm text messages to end user mobile phones carried (in their pockets) by nurses caring for the telemetry patients. The alarm/alerts were supposed to transmit to the phones both audible and visual signals. The alarm messaging system was known as the Clinical Alarm Notification System (CANS).

Patient 21 Observation

During observations of care in the 3rd floor telemetry unit with the TM, the Clinical Quality Director (CQD), and the Medical Director for Quality and Risk (MDQR) on 11/27/18 at 10 a.m., Registered Nurse (RN) 5 sat temporarily in front of electronic display screens (central station) that showed the waveform heart rhythms for 12 or more telemetry patients. Each patient in the unit was attached to a wireless box that transmitted heart monitoring data to the display's central computer. RN 5 explained that for Patient 21 within the past hour (at 9:08 a.m.) the heart monitoring system detected a very slow heart rate, should have triggered an audio and visual alarm signal to the central station, and from there [through a middleware software medical device operating on a different connectivity sphere and back to separate internal wireless network] should have transmitted a text message red alert signal to RN 5's mobile phone. But RN 5 did not see or hear the phone alarm. At the time of the incident, RN 5 was providing care to another patient in a room distant from the central station where audio and visual alarms registered.

When RN 5 completed the patient care task at 9:20 a.m., RN 5 "woke the mobile phone" and discovered the red alert text message was waiting to be "accepted." RN 5 immediately checked to see that Patient 1 did not need rescue intervention. Then RN 5 reported the missed red alert to Assistant Nurse Manager (ANM) 2. From the central station display, RN 5 was able to show the waveform for Patient 1 that triggered the 9:08 a.m. red alert. It showed that Patient 21's heart rate was 39 beats per minute (normal is 60-100) for 6 seconds beginning at 9:07 a.m.

Charge Nurse (CN) 1 joined the discussion. CN 1 indicated that her duties included providing relief and support to the unit RNs, and often (but not continuously) she might sit close to the central station display where alarms were visible and audible. However, for Patient 21's 9:08 a.m. alarm, CN 1 did not hear or see such an alarm from where she was located at that moment. CN 1 indicated that no staff was specifically assigned to continuously stay at the central display to recognize critical heart activity and alarms, or to recruit others to respond when a concerning waveform persisted or alarms were unanswered.

In an 11:30 a.m., 11/27/18 interview, the TM and the Director of Clinical Adult Services (DCAS) verified the facility no longer routinely used monitor technicians or other qualified staff on their telemetry units to watch the central monitors. In addition, there was no other area within the hospital used for remote continuous monitoring of primary physiological alarm notification.

CN 1 also carried a mobile phone programmed to receive the same red alert text messages that the 3rd floor telemetry RNs received when the RN did not respond to a red alert message within 30 seconds. ANM 2 similarly carried a mobile phone programmed to receive the RN red alerts for up to 48 patients in two telemetry units, and was expected to respond if the responsible RN did not respond within 30 seconds. This was considered the escalation pathway to ensure timely response to clinical alarms. CN 1 indicated that neither CN 1 nor ANM 2 heard or saw the 9:08 a.m. red alarm for Patient 21.

In an interview on 11/27/18 at 10:40 a.m., MDQR, who routinely admitted and cared for patients in telemetry units and elsewhere) stated that when MDQR ordered continuous cardiac monitoring, he expected care staff to respond to alarms and notifications of serious heart rhythm conditions timely, from immediately to several minutes depending on the specific condition. A delay of 13 minutes to respond to a rhythm such as Patient 21 displayed was too long. MDQR expected nurses to reliably hear and recognize alarms in real time.

During observation and interview of Patient 21 on 11/27/18 at 10:40 a.m., Patient 21 appeared tired and serious, sitting in a chair beside her patient bed, with one arm stationary in her lap and not used throughout the meeting. Patient 21's son explained that she was 79 years old and had had a kidney transplant 6 years ago. But she was admitted to the hospital this time on 11/26/18 for a stroke, for which various rehabilitation therapies were being provided.

Review of Patient 21's medical record with the TM on 11/29/18 at 10:10 a.m. confirmed the son's account. Also identified was an abnormal heart rhythm condition at the time of admission for which continuous cardiac monitoring was ordered. The records showed that Patient 21's rhythm was still not stabilized and shifted back and forth from a normal rate to various high risk states both fast and slow.

In a collaborative review of the Clinical Alarm Notification System (CANS) report (generated from the middleware software medical device that transmitted telemetry patient alarms to the various nurse-held mobile phone devices) on 11/27/18 at 4:15 p.m., DCAS showed that Patient 21's 9:08 a.m. slow heart rate alarm text signal was "delivered" within 10 to 30 seconds to 4 mobile phones. One phone was assigned to RN 5, one phone was assigned to CN 1, one phone was assigned to ANM 2, and one phone was not in use at the time. The report showed that only RN 5's phone "accepted" and "removed" the message at 10:38 a.m. (after showing it to ANM 2 as explained above).

Review with TM of additional CANS reports for Patient 21 eight hours prior to the 11/27/18 at 9:08 a.m. event revealed that a red alarm for a temporary period where Patient 21's heart stopped beating entirely was triggered at 1:28 a.m. That alarm was delivered and accepted by an RN phone, but was also delivered to an ANM and unassigned phone, neither of which registered as responding to the alarm.

The Joint Commission, an agency which accredits healthcare facilities including this hospital, established the following National Patient Safety Goal on Alarm Management: "As of January 1, 2016, educate staff ...about the purpose and proper operation of alarms systems for which they are responsible."

Telemetry system safety problems were identified and communicated to hospital leaders.

In a 12/4/18 at 6 p.m. interview, Registered Nurse (RN) 8 stated that although she was trained as a "super-user" prior to the CANS conversion, and trained many of the nurse end users, the training did not permit each nurse to "play with the phone" in order to master the new procedures to receive and process telemetry alarm alerts. "Right away we had problems. Then we had more training." Problems continued. CANS-related concerns began to be discussed with the CNE in the Professional Performance Committee on 8/9/17. RN 8 stated that she repeatedly raised the phone function concerns at staffing acuity (GRASP) meetings and Nursing Quality Forum meetings since January 2018. RN 8 did file some formal complaints, but often at the time of an incident, the communication was informal to the assistant nurse manager on duty at the time.

In an 11/29/18 at 11:30 a.m. interview, the Clinical Education, Practice and Informatics Manager (CEPIM, who served as the local CANS Administrator and helped to organize and implement conversion from a prior telemetry alarm notification system to the current CANS) stated that she and her leadership team expected minimal changes in how the new system would work or how it would be used by nurses caring for telemetry patients. CEPIM indicated that the previous notification system did not require "return to home screen" by a user in order to hear and see alerts as they were delivered (a significant change in phone functionality). The new software placed cardiac alarms and alerts in phone queues until the devices were manually returned to a home screen which allowed transmission. Nurses experienced back-ups of previously-transmitted alarms and alerts. The configuration of the phones was also changed to deliver additional alerts from a bed-exit sensor (for when patients left their beds) and alerts from a patient-pushed call system (for patients asking for help from staff).That added "a lot more alarms than expected."

In an 11/30/18 at 4:15 p.m. interview, the DCAS indicated that she began serving in the current position to oversee the telemetry services in April 2018. Between August 2017 and April 2018, CEPIM was responsible to respond to CANS concerns, and that was the basis for revising the education tool to "CANS 2.0" soon after the CANS conversion in 2017.

In a 1:22 p.m., 12/4/18 meeting, the Area Quality Leader (AQL) indicated that CANS "wasn't working" as anticipated and hospital leadership met with the union representing RNs to problem-solve.

In a 10:43 a.m., 11/28/18 interview, the CEPIM stated that after a discussion with union labor partners, it was agreed that additional and formalized CANS education ("CANS 2.0") would be offered to staff.

In an 11:17 a.m., 11/30/18 interview, the CEPIM noted that the hospital continued to receive staff complaints about the CANS system after the conclusion of the CANS 2.0.

In an 11/29/18 at 9:38 a.m. interview, ANM 2 stated that incoming alerts could not be seen or heard unless the user had the phone on the home screen. ANM 2 also indicated that there had been times when on the home screen, the alerts were still not heard. In response to Patient 21's slow heart rate alert on 11/27/18 at 9:08 a.m., ANM 2 did not recall hearing, toggling the volume or clearing, the alert. [The CANS report showed no response by ANM 2.] ANM 2 wondered if it automatically "toggled" or cleared. ANM 2 was carrying the phone in her coat pocket. Phone functionality and alarm escalation design "was a big change" when the new system was adopted.

In an 11/29/18 at 10 a.m. interview, RN 6 stated that it is easy to bump the phone when carried in a user's pocket, easy to unknowingly press a key, and any key takes the phone off the home screen.

In a group interview on 11/27/18 at 1:50 p.m., attended by the Regional Clinical Informatics Director (RCID), local Area Information Officer (AIO responsible for electronic device management and communications systems), CEPIM, the local Support Services Manager (SSM), and DCAS, the representatives stated that the expectation was that the nurse users of the mobile phones issuing alarm alerts would hear the alarms and respond promptly. However, they listed several obstacles to achieving that including activities that frequently removed the user from the home screen (active telephone call or text, active alarm being delivered, and not clearing prior messages). The phones had limited capability to deliver audible alerts while the user was engaged with other tasks. An attempt to try the only mode available from the phone produced a distracted background sound which doctors disliked, so it was rejected. Another attempt resulted in losing prioritization of alerts so that the least important last message would show first, and viewing more important messages was delayed. "Neither are very good options." The leaders claimed that no other means to drive prioritized alerts through to be promptly recognized when off the home screen was possible from this particular phone model.

In a 12/5/18 at 9:13 a.m. interview, ANM 1 (who served as interim TM between January 2018 and August 2018 during early stages of CANS implementation) stated that staff was "pretty vocal, pretty much came up to me every day," about the phone not being safe. ANM 1 reported the concerns (informally) to the CEPIM and previous DCAS, and stated they were brought them up at the Nurse Quality Forum (NQF) meetings attended by the Chief Nursing Executive (CNE). At times, incidents were investigated by simulation testing by the hospital's biomedical engineering staff, which always showed "the phones functioned as intended." Pending results of an incident investigation, a temporary monitor technician was employed for short periods (1-2 shifts). There were some instances of concluding the phone in fact did not get an alert. Concerns were "escalated" to senior leaders many times. CEPIM "reached out to" the Regional Clinical Informatics Director (RCID) and an information technology director. But no changes from the original design were implemented until more recently (in September when some bed exit alarms were removed from alerting on the phone).

In an interview with the DCAS on 12/4/18 at 1:30 p.m., she stated there had been another recent instance of a nurse "not getting an audible alert". DCAS stated that as a result, additional CANS functionality testing was performed on 12/3/18. The Department had entered the hospital six days before.

In a review of the recorded results of the 12/3/18 CANS testing, the following was validated, "Extreme Brady (very slow heart rate) configured to escalate to Charge RN Phone and ANM both, Upon initial alert no sound or vibration to CN device ... Alert was present in the log on the phone. Tried multiple times to replicate with same device and other devices, however unable to recreate."

No meaningful data collection to analyze extent of problems.

In an 11/27/18 at 4:15 p.m. interview, DCAS indicated that telemetry unit nursing staff had been reporting similar stories of not hearing/seeing mobile phone text alerts, regardless of which screen the phone was idling from. When investigated, the nurse managers and biomedical engineer staff were not able to replicate the user experience. CANS reports generated from the middleware program kept showing that the messages were delivered electronically to the phones. Investigators were not able to determine why the phones did not create audible alarms consistently. They didn't know what more to investigate and had no means to correct it. They also were not able to retrieve precise phone functionality activity logs in order to better understand why users reported that they didn't hear alarms even when on the home screen. DCAS believed such instances did occur, but the limitations of the equipment to analyze potential failure points led to a "dead end." Assumptions were made that the user failed to use the phone correctly (return to home screen) even though no data measures were produced to prove that. Investigators couldn't actually determine why all delayed/missed alarm message incidents occurred, couldn't fix it, and therefore couldn't prevent it from occurring again. DCAS viewed this unresolved problem as a threat to patient safety.

In the 11/27/18 at 1:50 p.m. group interview, CEPIM stated that when complaints surfaced, a monitor technician was assigned to the telemetry central display station temporarily, to serve as an additional layer to ensure end users of the phones responded quickly to the alerts. Assumptions were made that the end users were not using the phones correctly and needed more education/training. During those intervention periods (which CEPIM was not able to produce documentation of the complaints, discussions with senior leaders, corrective action plan, duration of temporary intervention, or data capture of impact of interventions), complaints were reduced. But that improvement was not sustained. Complaints of unsafe incidents continued to resurface, temporary monitor technicians were put in place again, education given again, complaints lessened, but improvement was not sustained.

In the 11/28/18 at 10:10 a.m. group interview, attended by AIO, CEPIM, DCAS, the Regional Director of Patient Care Services (RDPCS), RCID, the Regional Director of Nursing Informatics (RDNI), and the Senior Director of Operations for Southern California (SDOSoCal) the following was presented:

It was noticed very early in CANS implementation that recognition of alerts by users of the end phone device was inconsistent. Attempts to understand what users were describing included the collection of data on frequency of alarm delivery per user per shift (alarm volumes,) but did not include frequency of failed alarm recognition. No time-motion studies or "follow the nurse" shadowing observations over multiple shifts and days were recorded to generate meaningful data. Attempts to bring in Vendor 1 and experts in the phone functionality were not documented. When isolated phone failure incidents were identified, the end-to-end testing performed by the hospital's biomedical engineer did not simulate real-world system conditions and user workflow activities. It was always concluded that the system "worked as intended." Therefore, no effective validation testing of the safety and effectiveness of the new continuous cardiac monitoring alarm notification system was documented.

Technology Despite Objection Forms ("TDOs")

Review of TDOs (RN union communication tools reporting device-related concerns) received by the hospital revealed the first report of a CANS-related issue in January, 2018. The hospital was notified that a red alarm was not received on the phones of the primary nurse or those in the escalation pathway. 43 additional CANS-related TDOs were received by the hospital between February - August, 2018.

Telephone testing summary logs reviewed did not reflect any correlation with receipt of these TDOs.

In a 1:50 p.m., 11/27/18 interview, the DCAS estimated she'd received "50 - 75 TDOs over a three month period" reporting alarms that staff believed should have been sent to phones but weren't received. She stated she could recall three incidents in which alarms were known not to have been received on nurses' phones The DCAS indicated the hospital had been "mostly reactive" in its approach to possible CANS malfunctions and stated there was "no data we collect," including staff response time to critical heart rhythm ("red") alarms.

In a 2:14 p.m., 12/4/18 quality leadership group interview, the CNE indicated "there were quite a few TDOs" written regarding the CANS. She stated TDOs were responded to by unit managers and were not sent to hospital quality or risk departments for tracking. If equipment failure was identified as a possible patient safety incident cause, the manager would be expected to complete an Electronic Responsible Reporting Form (eRRF), notifying the Risk Management Department of the event.

In the same 2:14 p.m., 12/4/18 interview, the Director of Risk Management (DRM) indicated the Risk Management Department had "no idea" about the volume of CANS-related TDOs being received by the telemetry department manager. The DRM stated that Risk Management had received 10 eRRF reports of CANS-related issues in the previous 18 months. However, because that number was so small, "it didn't trend" as a patient safety issue.

In an 11/30/18 at 10:34 a.m. interview, DRM stated that from the inception of the new telemetry CANS system, staff formally generated complaints related to end users not effectively receiving telemetry critical alarms on the mobile phones through two pathways that did not speak to each other, and were not captured together to show trending volumes of similar incidents. That is, the TDO and eRRF incident reports were not integrated into a single quality indicator data file. Informal complaints were not captured into any quality measurement pathway either. Various committee chairs and unit managers who were hearing about phone functionality concerns were not successful in elevating those concerns to the Quality and Risk leaders, so the system could be formally analyzed. The DRM acknowledged that Quality and Risk leaders did not elevate known information to initiate a formal analysis of CANS.

Hospital Expectations to capture problems through data-driven process.

In an 8:24 a.m., 11/29/18 interview, the CEPIM stated that CANS system reference materials included the reference "CONTACT for CANS Extension Engage Support" which indicated. "Technical Issues (after Go Live): Local CANS Administrator...Tracks medical center issues and improvement requests ...."

Review of related telemetry unit job descriptions reflected the responsibility for patient safety to be widespread and assigned to every level of nursing, risk/patient safety, quality, and leadership:
- The CNE job description indicated she "provides strategic leadership as the patient care executive position responsible for all nursing ...Principal Responsibilities ...Generates and encourages innovation and creative solutions to issues or problems ...Achieves quality outcomes by utilizing problem identification, process improvement, RCA [root cause analysis, a problem-solving method] ...Implements the organization's performance improvement program, including planning, setting priorities, conducting systematic performance assessments, implementing improvements based on such assessments ...."
- The DCAS job description reflected, "In collaboration with medical staff and facility leadership ensures ...a safe environment with [sic] patients ...Manages and resolves ...department safety and risk management issues ...."
- Review of the telemetry units' 2018 "Plan for Provision of Patient Care" reflected, "Medical care is under the direction of the Medical Director who is ...responsible for assuring the overall quality and safety of care ...."
- The Manager job description directed, "Designs and evaluates processes to improve systems and patient care results ...Manages and resolves ...department safety and risk management issues ...Oversees and develops ...quality improvement activities ...."
- The ANM job description read, "Manages ...quality improvement. Ensures the highest quality of care is provided ...Supervises ...risk management ...Works with health care providers outside the unit to achieve optimal patient care ...."
- The Quality Nurse Consultant job description stated, "Acts as a consultant for quality ...risk, patient safety ...Reports all occurrences which may lead to medical center liability and follows up as necessary ...Analyzes risk management/patient safety principles and provides recommendations ...."
- The Staff Nurse - Telemetry job description stated, "Participates in unit and Department Performance Improvement activities as directed...Identifies and solves problems effectively ...."

Some levels failed to meet their quality and patient safety responsibilities.

Review of hospital patient safety and quality-related committee charters reflected that several groups were also charged with responsibility for ensuring a safe patient environment yet did not address CANS-related issues in a timely fashion:
- The nursing unit "Department Quality Council Charter" indicated the group's purpose was to "Improve the quality and safety of patient care in the department." The group "addresses quality care issues ...to develop interventions, design new processes, and adapt ideas from others that improve care...." Its scope includes "safety issues ...problem-solving ...performance improvement ...."
- The Quality Council's 2018 charter indicated its responsibilities included, "Set priorities for data collection ...Monitoring of hospital programs ...Improving the performance of systems ...required to deliver the highest quality care ...Provide necessary leadership to ensure implementation of quality improvement programs ...Identifying problem areas in health care or clinical performance ...Monitoring and evaluating the effectiveness of corrective actions taken ...."
- There were no references to CANS issues in any Quality Council meeting minutes from 9/1/17 to present, however.
- The Risk Management Patient Safety Committee charter for 2018 read, "Provides oversight to the proactive, comprehensive Risk Management Patient Safety Program ...the committee is expected to embrace, promote and actively support ...patient safety ...Seeks to identify, assess, prioritize, and implement strategies to eliminate/mitigate hazards for identified high risk ...processes and patient populations ...."
- Risk Management/Patient Safety Committee minutes reflected no references to CANS-related problems until 9/14/18.

Patient 1 - Missed opportunity to reduce patient risk and improve quality and safety.

In interviews and document reviews on 11/28/18 at 1:45 p.m. and on 12/4/18 at 1:20 p.m., DRM indicated that the risk management department handled the review of a potential adverse event (Patient 1) related to delay in care from staff relying on the CANS alerts. DRM confirmed the following:

Patient 1 was 68 years old, admitted on 6/23/18 after a week of non-typical heart symptoms that progressed to shortness of breath. His electrocardiogram (EKG, a waveform image of the electrical activity of the heart) showed a non-typical pattern for lack of oxygen to the heart muscle but was consistent with a heart attack. He was treated with blood thinner and other medications for heart failure, and placed on telemetry unit 3 South for continuous heart monitoring. He was at risk for further oxygen deprivation to the heart muscle. He was awake and able to eat, and scheduled for an ultrasound study of the heart on 6/24/18. According to the CANS report, at 6:05 a.m. on 6/24/18 a critical alert for high heart rate was transmitted to a phone held by RN 7. 19 seconds later, a second alert for a more dangerous high heart rate was transmitted to RN 7's phone. A third alert for a fatal heart rhythm was sent 56 seconds after the second alert, at 6:06 a.m. A hospital-wide Code Blue alarm was signaled at 6:09 a.m. The Code Blue team arrived at 6:10 a.m. Rescue measures were begun prior to the Code Blue team arriving by telemetry nurses including RN 7. There was concern and discrepancies regarding the timely response to the critical alarms starting at 6:05 a.m. by RN 7 and others in the vicinity, which prompted a deep dive to review the event as a potential preventable delay in rescue. Patient 1 did not survive the rescue attempts.

The event investigation and analysis was conducted by Quality Nurse Consultant (QNC) 2, DRM, CNE, AQL, MDQR with input from the CEPIM, DACS, Regional Risk leader, RCID, a clinical technology representative, interim TM, and telemetry unit assistant managers. Several potential system failures were identified. Some corresponded to what staff had reported previously:

a. End users of the phones struggled with the limitations and complexity of the phone's functionality and reliability.

b. Telemetry staff relied on the mobile phones as the primary means of alerting to critical heart conditions.

c. Redundancy backup notifications through the escalation of alerts to others in the pathway was not reliable.

d. Excessive generations of alerts to a single user device produced "alarm fatigue" which can motivate users to ignore and fail to respond to the alerts.

Of these system flaws that jeopardized patient safety, the hospital had no meaningful baseline prevalence data or validation to assess and determine how significant any of these identified risks actually were. Of the corrective actions proposed, only alarm fatigue resulted in the collection of baseline data on current frequency of alarms, and not on frequency of failed alarm recognition. An intervention to remove one source of alerts was to generate comparison data on frequency of alarms, but again not on alarm recognition. Other proposed actions that related to purchase of additional primary notification alarm displays throughout the telemetry units or in each patient room to allow nurses to more readily view the waveforms with the alerts in real time, or the purchase of improved mobile phones, were stalled for months awaiting approvals by leaders outside the hospital. There was no solution to the problem of workflow.

What did NOT result from the patient event analysis and proposed actions were key opportunities to improve the telemetry system. Opportunities that leaders indicated were not taken included:

a. Conduct a survey of end users to identify specific obstacles related to phone reliability and functionality. Seek solutions from end users.
b. Time-motion studies to record the effect of alerts on workflow through each shift, over multiple days, and to validate the effectiveness of user re-education/retraining over time.
c. Further consultation with mobile phone experts to explore a means to capture phone screen activity and functionality retrospectively, validate limitations of the phone, and present data to budget decision makers.
d. Assign additional staff to ensure immediate response to all critical alerts until safety measurements could validate that the system was reliable and safe.

By November 2018 most of the proposed corrective actions that were related to equipment

Based on interview and document review, the hospital leaders failed to ensure that the Quality Assurance and Performance Improvement (QAPI) program prioritized (in accordance with hospital governance documents) patient safety concerns for the reliability of a clinical alarm notification system (CANS) that was used on 2 of 2 telemetry units, where up to 48 patients were routinely admitted for continuous monitoring of potential heart function instability. Hospital leaders failed to ensure the CANS concerns were formally and effectively evaluated, in order to improve quality and safety for all patients affected by the CANS system.

These failures put all telemetry patients at risk for delayed rescue for life-threatening heart conditions, preventable organ damage, and death.

Findings:

Telemetry system safety problems were identified and communicated with hospital leaders.

During a tour on 11/27/18 at 9:30 a.m. of the hospital third floor telemetry unit, the Telemetry Manager (TM) explained that telemetry patients were monitored for potential unstable heart conditions by way of mobile devices attached to the patient (boxes) that transmitted heart activity signals to a remote central station display, which in turn transmitted alarm signals to a middleware software server, which in turn transmitted alarm text messages to end user mobile phones carried (in their pockets) by nurses caring for the telemetry patients. The alarm/alerts were supposed to transmit to the phones both audible and visual signals. The hospital had two telemetry units, one on the third floor and one on the fourth floor, with capacity for a total of 48 telemetry beds.

In a 12/4/18 at 6 p.m. interview, Registered Nurse (RN) 8 stated that although she was trained as a "super-user" prior to the CANS conversion, and trained many of the nurse end users, the training did not permit each nurse to "play with the phone" in order to master the new procedures to receive and process telemetry alarm alerts. "Right away we had problems. Then we had more training." In January 2018, RN 8, along with another RN and 2 nurse managers were sent to an off site simulation facility. RN 8 stated that she was expecting to train with a different phone. The same phones were used and there were failures in validating recognition of alerts even in the simulations. Problems continued. Since January 2018 RN 8 repeatedly raised the phone function concerns at Professional Practice Committee meetings, staffing acuity (GRASP) meetings, and Nursing Quality Forum meetings. RN 8 did file some formal complaints, but often at the time of an incident, the communication was informal to the assistant nurse manager on duty at the time.

In a 11/27/18 at 1:50 p.m. group interview with Regional Clinical Informatics Director (RCID), local Area Information Officer (AIO, who was responsible for electronic device management and communications systems), the local Director of Clinical Education Practice and Informatics (CEPIM), the local Support Services Manager (SSM), and the local Director of Clinical Adult Services (DCAS), the hospital leaders indicated that nurse users of the telemetry alarm notification end device (mobile phone) "started complaining about a year ago."

In a 11/29/18 at 11:30 a.m. interview, CEPIM (who served as the local CANS Administrator and helped to organize and implement conversion from a prior telemetry alarm notification system to the current CANS) stated that she and her leadership team expected minimal changes in how the new system would work or how it would be used by nurses caring for telemetry patients. CEPIM indicated that the previous notification system did not require "return to home screen" by a user in order to hear and see alerts as they were delivered (a significant change in phone functionality). The configuration of the phones was also changed to deliver additional alerts from a bed-exit sensor (for when patients left their beds) and alerts from a patient-pushed call system (for patients asking for help from staff).

In a 11/30/18 at 4:15 p.m. interview, DCAS indicated that she began serving in the current position to oversee the telemetry services in April 2018. Between August 2017 and April 2018, CEPIM was responsible to respond to CANS concerns, and that was the basis for revising the education tool to "CANS 2.0" shortly after initiating the CANS conversion in 2017.

No Effective Data Collection-Analysis

In a 1:50 p.m. 11/27/18 interview, the DCAS estimated she'd received "50-75 TDOs over a three month period" reporting alarms that staff believed should have been sent to phones but weren't received. (A TDO is a Technology Despite Objection tool, a written report, utilized by union nurses to voice device-related complaints and concerns.) The DCAS indicated the hospital had been "mostly reactive," in its approach to possible CANS malfunctions and stated there was "no data we collect," including staff response time to red (critical, urgent) alarms.

In the 11/27/18 at 1:50 p.m. group interview, CEPIM stated that when complaints surfaced, a temporary monitor technician was placed at the central display station for 1-2 weeks to serve as an additional layer to ensure end users of the phones responded quickly to the alerts. Assumptions were made that the end users were not using the phones correctly and needed more education/training. During those intervention periods (which CEPIM was not able to produce documentation of the complaints, discussions with senior leaders, corrective action plan, duration of temporary intervention, or data capture of impact of interventions), complaints were reduced. But that improvement was not sustained. Complaints of unsafe incidents continued to resurface, temporary monitor technicians were put in place again, education given again, complaints lessened, but improvement was not sustained.

In a 12/5/18 at 9:13 a.m. interview, ANM 1 (who served as interim TM between January 2018 and August 2018 during early stages of CANS implementation) stated that staff was "pretty vocal, pretty much came up to me every day," about the phone not being safe. ANM 1 reported the concerns (informally) to the CEPIM and previous DCAS, and brought them up at the Nursing Quality Forum (NQF) meetings attended by the Chief Nursing Executive (CNE). At times, incidents were investigated by simulation testing by the hospital's biomedical engineering staff, which always showed "the phones functioned as intended." Pending results of an incident investigation, a temporary monitor technician was employed for short periods (1-2 shifts, not 1-2 weeks). There were some instances of concluding the phone in fact did not get an alert. Concerns were "escalated" to senior leaders many times. CEPIM "reached out to" RCID and an information technology director. But no changes from the original design were implemented until more recently (in September when some of the bed exit alarms were removed from alerting on the phone).

In a 11/28/18 at 10:10 a.m. group interview attended by AIO, CEPIM, DCAS, RCID, the Regional Director of Patient Care Services (RDPCS), the Regional Director of Nursing Informatics (RDNI), and the Senior Director of Operations for Southern California (SDOSoCal, who served as the national clinical lead for CANS), SDOSoCal stated that the telemetry CANS conversion was designed and directed by a core group of regional leaders who provided support to the hospital. Both the hospital and the regional leaders were governed by the same national corporation. The core group determined the new system was safe to implement for 39 hospitals, including this one. SDOSoCal described a very different design for workflow in Southern California hospitals, most of which did not rely on the mobile phone as a primary notification device. Half the Southern California hospitals had upgraded to different phones or a voice-activated wearable receiver, and were not using the same model phones used at this hospital. SDOSoCal provided on site support during the hospital's conversion rollout and met with the Chief Nursing Executive. SDOSoCal did not recall concerns voiced by users then, or any "huge problems" since then.

CNE and Quality/Risk Leaders to Region Responsibility

In a 12/4/18 at 1:20 p.m. group interview, attended by the Director of Risk Management (DRM), CNE, the Area Quality Leader (AQL), and the Clinical Quality Director (CQD), CNE acknowledged that some unexplained incidents of nurse users of the telemetry mobile phones were identified and investigated. CNE agreed that assumptions were drawn that nurse users simply did not use the phone correctly. On two or three occasions during periods of heightened concern, telemetry monitoring technicians were assigned to the central display stations, temporarily, until it was determined to be safe to remove them because alarm failure incidents were not repeated during the brief intervention. CNE stated that the hospital had relied on the phone alerts (via a secondary alarm notification CANS system) as a primary notification means "for years." CNE indicated that the phone problems may have come up at NQF meetings. CNE did at some point "elevate" the concerns to Regional leaders (aka Region). "It was our failure to resolve. We kept pushing." Conversations between Regional Risk Management and Regional Patient Care Services occurred (no documentation presented). CNE submitted a request to Regional leaders to place heart monitor displays in every patient room. That request was rejected "for now." DRM considered the problems a threat to patient safety. Each time an incident or TDO surfaced, the feedback from Regional Patient Care Services was, "the phone functioned as intended, get staff to use them correctly."

CNE further indicated that she asked for new phones, but was told 'no' by Region . DRM indicated that Region often responded that equipment must be standardized, so if your hospital changes phones, all hospitals change phones. CNE acknowledged that end devices among the 39 CANS hospitals were quite variable (per SDOSoCal's claim at the 11/28/18 at 10:10 a.m. group interview), and indeed not standardized.

The CNE job description indicated she "provides strategic leadership as the patient care executive position responsible for all nursing ...Has oversight for the professional practice of nursing and the provision of nursing care ...Oversight is defined as having the responsibility for regulatory, quality ...nursing staff competency and evaluation of the overall delivery of nursing care ...Principal Responsibilities ...Generates and encourages innovation and creative solutions to issues or problems ...Achieves quality outcomes by utilizing problem identification, process improvement, RCA [root cause analysis, a problem-solving method] ...Implements the organization's performance improvement program, including planning, setting priorities, conducting systematic performance assessments, implementing improvements based on such assessments ...."

In interviews and document reviews on 11/28/18 at 1:45 p.m., and on 12/4/18 at 1:20 p.m., DRM indicated that the risk management department handled the review of a potential adverse event (Patient 1) related to delay in care from staff relying on the CANS alerts. Three sequential alarms for high heart rate progressing to a fatal heart rhythm triggered over 1.5 minutes on 6/24/18 beginning at 6:05 a.m. The timeliness of response to the alarms from nurses who were holding the phones that were to notify them of the dangerous rhythms raised questions for how reliable the notification system performed in the existing workflow design. The event analysis identified several contributing factors that warranted corrective actions. The proposed corrective actions included:

a. install display monitors in every patient room that showed waveforms and alarms for all patients in the same telemetry unit for RNs to view from nearly all parts of the unit, and

b. convert to end device that did not have messages held "in queue" when off home screen, using phone for other functions, etc.

DRM indicated that a preliminary analysis and development of solutions for the 6/24/18 event was prepared by late August 2018 for reporting to the medical staff. Review of the event summary updated in October 2018 showed that item (a) above could not be implemented until March 2019, depending on funding decisions by Regional leaders. On 10/1/18 the limitations of the existing telemetry alarm notification system was discussed with Regional Risk Management leaders by telephone. A request to have an end device capable of delivering a wave form was also dependent on Regional approval, and "ruled out at this time." For item (b) above, the results of "escalation to Region" were also framed as "ruled out at this time." To date neither had been achieved. There was no documented evidence that either request was submitted to the hospital Administrator and similarly rejected.

Administrator Responsibility

In a 11/29/18 at 1:25 p.m. interview, Administrator 1 stated that she was responsible for the operations of the hospital to ensure patient safety. In that capacity, Administrator 1 had authority to "halt the line" when unsafe systems were identified. In a subsequent 12/5/18 at 10:15 a.m. interview, Administrator 1 indicated she was aware of the unresolved telemetry alarm notification system concerns, and that she had the authority to implement requests for products and services when deemed necessary by local hospital leaders.

Review of the 12/4/15 governing body bylaws, under Section D-12 documented that the Hospital Administrator was responsible for the management and administration of the hospital, including quality of care, provision of information and support systems, and physical and financial assets.

Review of the job description for Administrator 1 documented duties to provide primary accountability for overseeing all aspects of hospital and Area operations across the region, ensure Hospital oversight responsibilities were appropriately exercised, and ensure operational compliance with regulatory requirements for safety.

DQC and Medical Staff/Governing Body Responsibility

In an 11:17 a.m., 11/28/18 interview, the DCAS stated that CANS issues had been discussed in the telemetry units' Department Quality Councils (DQCs). In a 9:37 a.m., 12/5/18 interview, ANM 1 stated that one of the staff co-leads for the group "read a list of staff concerns about CANS" during one DQC meeting and requested a "town hall meeting with upper management to vent their concerns." The ANM stated the DCAS and the CNE were aware of the staff's request.

In a 4:31 p.m., 12/5/18 interview, the Quality Nurse Consultant for the telemetry units (QNC 1) stated that when the topic came up in the DQC, he and the Interim Telemetry Manager (ANM 1) "refocused" the group on other performance improvement projects, indicating CANS-related issues were being worked on at the leadership level. The DQC stated ANM 1 told the group she would attempt to have a senior leader come to DQC to discuss CANS issues. The CANS issue was not documented in unit DQC minutes and was not reflected in hospital Quality Council minutes as having been carried forward to the latter group.

Physician involvement in resolution of CANS-related issues was minimal. In a 3:38 p.m., 12/4/18 interview, the DRM confirmed there had been no formal communication sent to physicians regarding potential CANS reliability issues. In a 10 a.m., 12/5/18 interview, Medical Doctor (MD) 1, who served as the Medical Director of the telemetry units, denied having any "specific" information related to CANS issues. Review of the telemetry units' 2018 "Plan for Provision of Patient Care" reflected, "Medical care is under the direction of the Medical Director who is ...responsible for assuring the overall quality and safety of care ...."

Review of the Quality Council 12/12/17 minutes stated, "The purpose of [the Quality and Patient Safety Plan] is to provide the mechanism for improving hospital quality and safety ...Staff demonstrate a consistent and collaborative approach to deliver safe ...patient-centered care within a quality assurance and performance improvement framework. This plan informs the improvement processes for ...applying remediation strategies in response to system or process failure ...Foundational elements ...proactive and prioritized performance improvements to prevent failure, mitigate hazards and improve systems and process reliability."
There were no references to CANS issues documented in any Quality Council meeting minutes from 9/1/17 to present.

Review of the 2018 Professional Staff Bylaws (for the hospital's medical staff) documented under Preamble that the medical staff was responsible in matters involving the quality of medical care. Section E-10 directed, "The Quality Council shall develop and implement a hospital wide Quality Improvement Program ... shall evaluate the quality and safety of patient care through ongoing monitoring and analysis of data and performance improvement activities ... functions shall include identifying problem areas in health care and clinical performance, monitoring and evaluating the effectiveness of corrective actions taken, and identifying opportunities to improve health care."

Review of meeting minutes for the governing body from January 2017 to November 2018 showed no entries related to the hospital's continued problems related to an unreliable heart monitoring alarm system. Review of medical staff leadership meeting minutes for the same time period showed reporting at the 7/3/18 meeting of a potential delay in care (Patient 1's 6/24/18 event) that would be investigated. No subsequent discussion was documented until the 11/6/18 meeting that announced the initiation of an analysis with action plan to reduce failure points for Patient 1's 6/24/18 delay in care review. The leadership group closed the case, made no recommendations, took no actions, and determined "no further action required."

Review of the 12/4/15 governing body bylaws, under Section E-10 item B documented, "The Quality and Health Improvement Committee (QHIC) is created to ... provide oversight of systems designed to monitor on behalf of the [governing body] that quality care and services are provided at a comparable level to all members and patients." The QHIC was directed to provide direction in oversight of quality systems and identification/addressing deficiencies in quality, and review of standards for patient safety and risk management.

Telemetry safety was not elevated to formal quality improvement activity/project.

In a 12/4/18 at 1:20 p.m. group interview, DRM, AQL and CQD indicated that formal analysis of the safety risks for the CANS system had not become a priority for a quality activity or performance improvement project until this week. AQL stated there is "no proactive monitoring from Quality or Risk" of CANS-related issues. DRM stated that the number of formal complaints and incident reports related to unreliable telemetry phone performance "did not trend," and therefore did not qualify to be tracked, monitored, and analyzed. The CQD expected telemetry unit managers to enter complaints including TDOs into the hospital's incident reporting system (eRRF). If they were not entered, the hospital had no means to capture the informal complaints. There was no proactive effort to monitor the telemetry phone complaints, or to monitor the reliability of the telemetry end user phones as integral to the overall telemetry alarm monitoring system.

Review of the Hospital Quality and Patient Safety Program Description (HQPSPD), last approved 1/9/18, under Section 1 Purpose documented that the governing body's Quality and Health Improvement Committee (QHIC), Senior Leaders, Medical Staff and Hospital Staff would demonstrate a consistent and collaborative approach to deliver safe, effective, patient centered and timely care within a quality assurance and performance improvement (QAPI) framework. The framework included "Proactive and prioritized performance improvements to prevent failure, mitigate hazards, and improve systems and process reliability." The hospital listed its values to include, "Ensuring that every patient has the right to care that is reliable, effective and safe."

Based on interview and document review, the hospital leaders failed to ensure (in accordance with hospital governance documents) that adequate resources were allocated for assessing, improving and sustaining performance of a clinical alarm notification system (CANS) that was used on 2 of 2 telemetry units, where up to 48 patients were routinely admitted for continuous monitoring of potential heart function instability, and for reducing risk to patients.

These failures put all telemetry patients at risk for delayed rescue for life-threatening heart conditions, preventable organ damage, and death.

Findings:

During a tour on 11/27/18 at 9:30 a.m. of the hospital third floor telemetry unit, the Telemetry Manager (TM) explained that telemetry patients were monitored for potential unstable heart conditions by way of mobile devices attached to the patient (boxes) that transmitted heart activity signals to a remote central station display, which in turn transmitted alarm signals to a middleware software server, which in turn transmitted alarm text messages to end user mobile phones carried (in their pockets) by nurses caring for the telemetry patients. The alarm/alerts were supposed to transmit to the phones both audible and visual signals. The hospital had two telemetry units, one on the third floor and one on the fourth floor, with capacity for a total of 48 telemetry beds.

In a 11/29/18 at 11:30 a.m. interview, the Director of Clinical Education, Practice and Informatics (CEPIM, who served as the local CANS Administrator and helped to organize and implement conversion from a prior telemetry alarm notification system to the current CANS) stated that she and her leadership team expected minimal changes in how the new system would work or how it would be used by nurses caring for telemetry patients. CEPIM indicated that the previous notification system did not require "return to home screen" by a user in order to hear and see alerts as they were delivered (a significant change in phone functionality). The configuration of the phones was also changed to deliver additional alerts from a bed-exit sensor (for when patients left their beds) and alerts from a patient-pushed call system (for patients asking for help from staff).

Resources for additional display monitors, better phones, dedicated staff at display station, and technical expertise were not provided.

In a 12/4/18 at 1:20 p.m. group interview, the Chief Nursing Executive (CNE) acknowledged that some unexplained incidents of nurse users of the telemetry mobile phones were identified and investigated. CNE agreed that assumptions were drawn that nurse users simply did not use the phone correctly. On two or three occasions during periods of heightened concern, telemetry monitoring technicians were assigned to the central display stations, temporarily, until it was determined to be safe to remove them, because alarm failure incidents were not repeated during the brief intervention. CNE stated that the hospital had relied on the phone alerts (via a secondary alarm notification CANS system) as a primary notification means for years. CNE indicated that the phone problems may have come up at Nurse Quality Forum (NQF) meetings. CNE did at some point "elevate" the concerns to Regional leaders. "It was our failure to resolve. We kept pushing." Conversations between Regional Risk Management and Regional Patient Care Services occurred (no documentation presented). CNE submitted a request to Regional leaders to place heart monitor displays in every patient room. That request was rejected "for now."

In the same 12/4/18 at 1:20 p.m. interview, the Director of Risk Management (DRM) stated that she considered the CANS problems a threat to patient safety. Each time an incident or complaint surfaced, the feedback from Regional Patient Care Services was "the phone functioned as intended, get staff to use them correctly."

CNE further indicated that she asked for new phones, but was told 'no' by Region (Regional leaders). DRM indicated that Region often responded that equipment must be standardized, so if your hospital changes phones, all hospitals change phones. CNE acknowledged that end devices among the 39 CANS hospitals were quite variable, and indeed not standardized. CNE also asked Region for permission to assign a staff member to continuously view the telemetry central displays, and was told "you have successfully been able to work without one, so no." CNE indicated that she had the authority to assign such staff in previous situations for patient safety, though those were only temporary.

In a 11/27/18 at 1:50 p.m. interview, the Director of Clinical Adult Services (DCAS) indicated that some ideas to ensure reliable responses to CANS alarms included (1) placing a waveform and alarm monitor display in every patient room, (2) obtaining a cost projection from a monitor vendor for a better configuration of the primary waveform and alarm central display, (3) providing/staffing dedicated telemetry technicians at the central display, and (4) providing rolling display monitors in the halls of each telemetry unit for nurses to access waveforms close to where they were actually working.

In interviews on 11/28/18 at 1:45 p.m., and on 12/4/18 at 1:20 p.m., DRM indicated that the risk management department handled the review of a potential adverse event related to delay in care from staff relying on the CANS alerts. The event analysis identified several contributing factors that warranted corrective actions. The proposed corrective actions included:

a. install display monitors in every patient room that showed waveforms and alarms for all patients in the same telemetry unit for RNs to view from nearly all parts of the unit; and

b. convert to an end device that did not have messages held "in queue" when off home screen, when using the phone for other functions, etc.

DRM indicated that a preliminary analysis and development of solutions for the 6/24/18 event was prepared by late August 2018 for reporting to the medical staff. Review of the event summary updated in October 2018 showed that item (a) above could not be implemented until March 2019, depending on funding decisions by Regional leaders. On 10/1/18 the limitations of the existing telemetry alarm notification system were discussed with Regional Risk leaders by telephone. A request to have an end device capable of delivering a wave form was also dependent on Regional approval, and "ruled out at this time." For item (b) above, the results of "escalation to Region" were also framed as "ruled out at this time."

In an interview on 12/5/18 at 1:20 p.m., CEPIM reviewed a vendor cost quote dated 11/16/18 to reconfigure a central monitoring station to serve both telemetry units, up to 48 beds, from a single location manned by a dedicated telemetry technician. That idea was not listed on the 6/24/18 adverse event analysis proposals, nor did it appear in any committee, unit, department, medical staff, or governing board minutes.

In a 12/5/18 at 9:40 a.m. interview, the Area Information Officer (AIO) stated that the acceptance and timeline to implement the work project had not been finalized or started, but he hoped it could be completed within a few weeks. Such a design change could serve as a temporary fix to ensure primary monitoring of telemetry patients until a more expensive undertaking to get monitor displays in every patient room was achieved (not yet approved).

In a 12/5/18 at 10:15 a.m. interview, Administrator 1 stated that she had access to funding that she was authorized to use for products and services deemed necessary by the local hospital leaders. When asked if that funding was sufficient to purchase 32 to 40 mobile phone devices with superior functionality to those currently in use for the telemetry alarm notification system, Administrator 1 said "yes," if that was determined to be an effective solution. Administrator 1 indicated that she had ongoing knowledge of the CANS staff complaints over the past year. Administrator 1 was aware of the system problems identified from the event analysis, and its proposed solutions. Yet, funding for the proposals that were rejected by Regional leaders was not provided to urgently resolve the ongoing threats to patient safety.

Review of the 12/4/15 governing body bylaws, under Section D-12 documented that the Hospital Administrator was responsible for the management and administration of the hospital, including quality of care, provision of information and support systems, and physical and financial assets.

Review of the job description for Administrator 1 documented duties to provide primary accountability for overseeing all aspects of hospital and Area operations across the region, ensure Hospital oversight responsibilities were appropriately exercised, and ensure operational compliance with regulatory requirements for safety.

In a 12/4/18 at 1:20 p.m. interview, CNE indicated that the hospital utilized the mobile phone alerts as a primary monitoring means, when the CANS system with end user phones was intended to be a secondary means of notifying nurses to respond to critical patient conditions. Even as this problem formally emerged from the 6/24/18 event analysis as a system design flaw with high risk for patient safety, permanent staffing of a dedicated telemetry technician was not immediately put in place by CNE and other nurse managers. Ensuring that primary alarm notifications were promptly responded to by layering additional staff to verify receipt/recognition of alarms was not achieved, when that resource was readily available.

The CNE job description indicated she "provides strategic leadership as the patient care executive position responsible for all nursing ...Has oversight for the professional practice of nursing and the provision of nursing care ...Oversight is defined as having the responsibility for regulatory, quality ...nursing staff competency and evaluation of the overall delivery of nursing care ...Principal Responsibilities ...Generates and encourages innovation and creative solutions to issues or problems ...Achieves quality outcomes by utilizing problem identification, process improvement, RCA [root cause analysis, a problem-solving method] ...Implements the organization's performance improvement program, including planning, setting priorities, conducting systematic performance assessments, implementing improvements based on such assessments ...."

In a 11/30/18 at 11 a.m. interview, Biomedical Engineer (BE) 1 indicated that he was not an expert in use of the CANS end user mobile phones, although he performed most of the end-to-end testing of the CANS system in response to reported incidents. The corporate Regional structure had technology staff off site who had expertise with various aspects of the technology involved in the CANS system, but no one employed locally had comprehensive expertise.

In a 11/30/18 at 4:30 p.m. telephone interview, the national Director of Integrated Clinical Mobility (DICM, who managed the Facilities Portfolio for Information Technology services) indicated that expertise for the CANS software was provided by a national group known as Applications Services Group. There was another Senior Manager with expertise in phone functionality who was considered "in-house." And a "Network Team" handled reports for monitoring and testing of various wireless networks. The Network Team could audit lost wireless calls. DICM indicated he was familiar with the phones. "My team deals with the alarm platform" when it rolled out in August 2017. DICM stated that the phone had limited memory, but could produce logs to show certain features, including delays in functionality. The capability to access logs was usually turned off (except for testing) in order to not slow the phone's performance. DICM stated he could offer advisory services, was not informed of phone user problems, and was not asked to troubleshoot.

On 11/30/18 at 2:30 p.m., the AIO presented a list of Personnel for the national Application Support Group. Review of the job descriptions for them noted:

Lead Systems Administrator - incident and problem management process, resolving complex issues (e.g., root cause analysis, coordinate prevention plan execution), addressing escalations, escalates unresolved or critical issues, risks or problems to system engineers, implementation of hardware and software changes into environments and one or more additional IT (information technology) production environments by performing impact analyses of system changes

Senior Systems Administrator - troubleshooting, and resolving complex incidents to minimize system issues, plans and performs complex system configuration

Systems Administrator - Assists in the implementation of hardware and software changes into environments

In a 11/30/18 at 9:30 a.m. interview, AIO indicated that although he oversaw the information technology related operations at the hospital, he relied on the project managers and regional consultants for expertise. He had avenues to seek expertise and consultation when needed, including the manufacturer of a product. AIO acknowledged that hospital leaders had not engaged the regional and national experts and phone manufacturer to reconfigure, reprogram, or otherwise fix the mobile phones in a way that offered the needed functionality identified by users from the start of the CANS conversion.

Based on observations, interviews and document review, the facility did not provide adequate nursing services during a major change in the clinical alarm notification system (CANS) on two of two telemetry units (inpatient hospital units that provided continuous cardiac monitoring) as evidenced by:

1. The absence of reliable continuous cardiac monitoring by staff for safety-net redundancy (A-392);
2. The absence of reliable and timely supervisory response to unacknowledged escalated critical alerts (physiological alarms) (A-392);
3. Nursing management did not provide adequate training for nursing personnel for the initial CANS implementation or subsequently thereafter (A-397) and;
4.The above failures were identified by nursing management but not integrated with the QAPI program (A-263).

As a result, registered nurses (RNs) relied on the alerts from their unit mobile phones (end-user, secondary notification device), via the new CANS extension service, as a primary source of notification for life-threatening cardiac rhythms. The lack of staff knowledge and understanding about the new CANS extension service and the use of the mobile phone compromised the effectiveness of the intended use of the phone to assist in patient care.

These failures had the potential to impede the immediately availability of a RN for bedside care of the telemetry patient with a cardiac emergency.

In addition, based on staff interview and clinical record review, the facility failed to ensure pain was assessed and medication was administered as prescribed and according to facility policy for 2 of 24 patients (A-405).

The cumulative effect of these systemic problems resulted in the inability of the Hospital to comply with the statutorily-mandated Condition of Participation for Nursing Services.

Based on observations, interviews and document review, the facility failed to provide an adequate number of supervisory and staff personnel for two of two telemetry units (a total of 48 beds; an inpatient hospital unit that provides continuous cardiac monitoring; continuous observation of the heart's electrical activity in patients with conductive disturbances and in those at risk for life-threatening cardiac rhythms) to ensure the immediate availability of a registered nurse (RN) for bedside care of any patient as evidenced by:

1. The absence of reliable continuous cardiac monitoring by staff for safety-net redundancy and;

2. The absence of reliable and timely supervisory response to unacknowledged escalated critical alerts (alarms).

As a result, RNs relied on the alerts from their unit Vendor 2 mobile phones (end-user, secondary notification device), via the middleware application, as a primary source of notification for life-threatening cardiac rhythms of patients on both telemetry units.

Findings:

Background

The Staff Nurses (RNs), the Charge Nurse (CN) (when scheduled) and the Assistant Nurse Manager (ANM) on telemetry units 3 South and 4 South are assigned a Vendor 2 mobile phone at the start of every shift. According to the September 2018 facility policy titled, "Clinical Alarm Notification System (CANS) NCAL Regional Policy", the wireless phones receive routed physiological alarms, originating from the patient ...based on patient assignments and escalation rules of the middleware application. These alarms are defined according to severity: a red alert, or lethal arrhythmia (life-threatening cardiac rhythm disturbance) requires an immediate response and validation; a yellow alert, or non-lethal arrhythmia, is a top priority and prompt response is required. A blue, or inoperative alert, such as "no signal" or "replace/change battery" should be answered promptly. These contextual notifications facilitate situational awareness in which RNs can better determine a treatment plan for their patients. Unacknowledged red alerts are escalated exclusively to the Assistant Nurse Manager/CN phone; unacknowledged yellow and blue alerts are escalated to a Buddy RN's phone (a partner to the primary RN that responds to alarms when the nurse is occupied and cannot respond to the alarm).

1.a. Absence of continuous cardiac monitoring

A tour of the telemetry unit 3 South was performed 11/27/18 at 9:55 a.m. with the Telemetry Manager (TM). TM explained 3 South and 4 South each consisted of 24 beds. 4 South was one floor up from 3 South and had the identical structural "foot-print". Patients were admitted to both units with various diagnoses and required continuous cardiac monitoring. During the tour, it was noted that no one was assigned to watch the central cardiac monitors for primary alarm notification.

During a separate visit to the Telemetry Unit 3 South on 11/27/18 at 10:40 a.m., no one was observed watching the two central cardiac monitors at the main nurses' station.

On 11/27/18 at 11:30 a.m., Charge Nurse 1 (CN 1) acknowledged the absence of continuous central cardiac monitoring; "That's correct", CN 1 said, "several years ago", monitor techs were removed from both telemetry units.

In a concurrent interview with TM and the Director of Clinical Adult Services (DCAS) on 11/27/18 at 11:35 a.m., they verified the facility no longer routinely used monitor technicians or other qualified staff on their telemetry units to watch the central monitors. In addition, there was no other area within the hospital used for remote continuous cardiac monitoring for primary physiological alarm notification.

The 3 South and 4 South Telemetry Nursing Assignment sheets, dated 6/22/18 through 6/30/18, 7/1/18 through 7/12/18, 7/18/18, 11/3/18 to 11/5/18, 11/10/18 to 11/12/18, 11/14/18 and 11/26/18 through 11/27/18 were reviewed. Documented evidence that a monitor technician, or other qualified staff member, was assigned to watch the centralized monitors for continuous cardiac monitoring did not exist.

1.b. Change of shift

RN 2 was interviewed on 11/29/18 at 11:20 a.m. RN 2 explained that at each change of shift the off-going primary nurse gives a verbal report on each patient to the on-coming nurse. During this period of time, the break nurse collects the wireless phones (up to 5- 6) from each staff nurse. "At the end of my (day) shift," RN 2 explained, "the break nurse collects the phones between 2:50 p.m. and 2:55 p.m. She'll have the phones until 3:08-3:10 p.m., when the RNs come back from their huddle." RN 2 stated, "The Break RN should be watching the central monitor" when the phones are being cleaned and reprogrammed. This does not always happen, RN 2 said.

During an interview with RN 1 on 11/27/18 at 11:45 a.m., RN 1 discussed the lack of staff coverage for adequate alarm notification and response during the hand-off of phones at shift change. "At shift change", RN 1 reported, "the break nurse has all the phones for up to 20 minutes" and "there's a lack of sufficient staff to respond to the alarms. It's a busy time for nurses."

On 11/29/18 at 1:25 p.m. RN 3 was interviewed. RN 3 stated her shift starts with huddle in the break room with the ANM and CN. Before shift change, the Break RN keeps her phone and collects the other 6 bedside RN phones. While the Break RN has the phones, he/she changes the batteries in each phone as needed, clears all the previous alert messages from each phone, and cleans each phone. Then, the Unit Assistant programs the phones with new patient assignments and sends a test alert message to each phone. This process, RN 3 explained could take a total of 10 minutes and should happen near the central monitor. "We are all responsible for the alarms," RN 3 said, "but no one consistently watches the central cardiac monitors. There is no monitor tech."

Further review of the September 2018 facility policy titled, "Clinical Alarm Notification System (CANS) NCAL Regional Policy", "At the start of each shift, at least every 8 hours, phone memory must be cleared and test pages measured to ensure delivery within 10 seconds. Each phone ...will be cleaned with germicidal disposable wipes after each use ...at shift change, the off going shift must retain the telemetry assigned phone until end of shift and must hand phone off to the assigned qualified RN on the oncoming shift."

The actual change of shift practice on telemetry did not comply with hospital policy or ensure continuous cardiac monitoring.

During a group interview on 11/27/18 at 1:20 p.m., the Regional Director of Clinical Informatics (RCID), stated that shift change is "riskier" due to phone programming, the transfer of phones and alarm management.

2. Absence of reliable and timely supervisory response to escalated alerts

In a concurrent interview with TM and DCAS on 3 South on 11/27/18 at 11:15 a.m., they explained that escalated alerts go to Buddy RNs and ANMs, consistent with the facility policy and procedure. TM mentioned there were four remaining charge nurses for both telemetry units that have been "grandfathered in" that may be assigned to receive escalated alerts; the facility was no longer hiring charge nurses.

On 11/29/18 at 9:30 a.m., ANM 2 was interviewed. ANM 2 stated there was a total of 8 ANMs assigned to work telemetry. One ANM is generally scheduled on each unit every day and one ANM is scheduled for both units at night. There were no other ANM positions available on 3 or 4 South; all ANM positions were "filled". ANM 2 stated her job entailed "bedding patients, patient rounds, making the staffing assignments, giving breaks, service recovery, and the occasional transport of a patient, if "transport" is not available." When asked if she had in the past been the only ANM on duty for both floors, 3 South and 4 South, ANM 2 answered, "yes." ANM 2 remarked that when she is the only ANM scheduled for both floors, she potentially had 48 patients to account for; "I split my time between the two units [and see] what needs are present and visible." ANM 2 stated there generally is a CN scheduled in place of an ANM, but only a few CN remain; so there are times there is no replacement for either.

With further interview of ANM 2, she reported ANMs received an influx of escalated alerts when assigned to both units; "a lot of alerts in the midst of doing other stuff". ANM 2 acknowledged staff may not receive an immediate response from her. ANM 2 acknowledged there were times when she did not hear an escalated alert from her mobile phone and there were times when she heard it, but could not respond to it. ANM 2 stated she may be carrying more than one phone while on duty. ANM 2 continued saying, besides the escalated physiological alerts, she also received phone calls from the House Supervisor, staff RNs, transport, the Staffing Office and Radiology. "This takes me off the floors to transport patients", ANM 2 said. ANM 2 acknowledged she took the phone that received the escalated alerts with her.

ANM 1 was interviewed on 11/29/18 at 10:30 a.m. ANM 1 explained that between the rules of the CANS and the end user device (phone), an ANM receives escalated red alerts, code blue alerts, "OT" alerts, such a bathroom assist alarm lasting a certain duration, phone calls and unit assistant text messages. ANM 1 said if only one ANM is scheduled to cover both units and a patient needs to be transferred to or from one of the telemetry units, the ANM may transport the patient, taking her "off both units at times." ANM 1 reported she took her phone with her when she goes off the unit, unless there was another ANM scheduled. There were other occasions ANM 1 spoke of that required she routinely leave the telemetry units. At 8:30 a.m. she attended a discharge through-put huddle in the Social Services Department and at 9 a.m., all managers attended a meeting with Chief Nurse Executive. ANM 1 acknowledged "time delays" on telemetry in response to alerts. In addition, ANM 1 mentioned there is no routine monitor tech despite "multiple CANS upgrades".

During a group interview with the Regional Director Clinical Informatics (RCID), Clinical Education Practice & Informatics Manager (CEPIM), Area Informatics Officer (AIO), Support Services Manager (SSM) and DCAS on 11/27/18 at 2 p.m., it was acknowledged that the telemetry ANM(s) received escalated red alerts but were not always available, or present on the unit to safely accept or acknowledge the alerts. One of the group stated, the ANM should not accept escalations when off the unit.

According to the September 2018 facility policy titled, "Clinical Alarm Notification System (CANS) NCAL Regional Policy", "when leaving the unit, hand off phone to another relieving qualified RN, who has been assigned."

In an interview with RN 2 on 11/29/18 at 11:30 a.m., RN 2 stated, "we don't always have an ANM or CN. The ANMs go to meetings off the unit ...No one is watching the cardiac monitor continuously."

In an interview with DCAS and the CEPIM on 11/29/18 at 1 p.m., they acknowledged the facility did not verify the escalation of alerts to the ANMs, since neither monitoring nor audits of ANM escalations were performed.

The 3 South and 4 South Telemetry Nursing Assignment sheets, dated 6/22/18 through 6/30/18, 7/1/18 through 7/12/18, 7/18/18, 11/3/18 to 11/5/18, 11/10/18 to 11/12/18, 11/14/18 and 11/26/18 through 11/29/18 were reviewed. Staffing data from all three nursing shifts (days, evenings and nights) for each day was reviewed. During these periods, 22 of the 34 days included a shift where one ANM was assigned to both units. 24 out of the 34 days of assignment sheets, including all three shifts, were absent a "Transport/Rapid Response Team (RRT)" nurse assignment; and 19 of the 34 days included a shift without a charge nurse scheduled.

More specifically, the night shift "Nursing Assignment Sheet-3 South", dated 6/22/18, read, "No ANM". On the 6/23/18 day shift assignment sheets, the ANM (assigned to both units) was also assigned as the "Transport/RRT" nurse for both units.

Further review of the Nursing Assignment sheets revealed that the night shift assignments for both telemetry units, dated 6/24/18, 7/2/18-7/10/18, 11/3/18-11/5/18, 11/10/18 and 11/26/18, were missing a monitor technician, a charge nurse and a transport/RRT nurse, which potentially left the one scheduled ANM to do all three jobs for both units.

A final group interview was conducted on 12/4/18 at 3:55 p.m., including the Chief Nurse Executive (CNE). The CNE acknowledged the telemetry units had a "few charge nurses." When asked about sustained staffing solutions to ensure continuous cardiac monitoring, safety net redundancy and reliable response to escalated alarms, the CNE offered none. The CNE was asked if she had proposed a corrective action for the lack of reliable ANM response to unacknowledged red critical alerts. The CNE acknowledged ANMs were not on the units at times and this issue had been "challenging ever since the event [on 6/24/18]."

The Joint Commission, an organization of which the hospital is accredited, has published, effective January 1, 2014, National Patient Safety Goal NPSG.06.01.01 which specified, "identify the most important alarm signals to manage based on the following: ...risk to patients if the alarm signal is not attended to..."

According to the September 2018 facility policy titled, "Clinical Alarm Notification System (CANS) NCAL Regional Policy", the primary alert for physiological alarms, any alarm that alerts staff based on the patient's immediate physiological process or health status, such as cardiac monitors, pulse oximeters, etc., remains the monitor or central station, not the phones.

Based on staff interviews and document review, the hospital failed to ensure that nursing personnel received adequate training for a major change in the clinical alarm notification system (CANS) on two of two inpatient telemetry units that provide cardiac monitoring.

The lack of staff knowledge and understanding about the new CANS extension service and the use of the mobile phone compromised the effectiveness of the intended use of the phone to assist in patient care and put patients at increased risk for a delay in response to clinical alarms.

Findings:

In a group interview on 11/28/18 at 10:30 a.m., CEPIM stated initial training on the use of the CANS extension was limited to the CEPIM, the previous Telemetry Manager, two to three Assistant Nurse Managers, a few Clinical Educators, a Clinical Tech Engineer and "at will other staff." This small group of 5-7 individuals was trained by the vendor prior to its implementation on the telemetry units.

In an interview with Registered Nurse (RN 1) for telemetry on 11/27/18 at 11:40 a.m., RN 1 stated orientation to the phones in conjunction with the new extension/middleware was not provided until after its roll-out. Staff were not prepared; "we had to beg to get more training on the phones" and the new CANS extension/middleware.

During a group interview on 11/27/18 at 3 p.m., the Clinical Education Practice & Informatics Manager (CEPIM)/Local CANS Administrator acknowledged that about a month after the facility introduced the CANS "Go Live", a "subset of medical surgical and telemetry employees requested additional training".

In a concurrent interview with CEPIM and the Director, Clinical Adult Services (DCAS) on 11/28/18 at 11 a.m., they reported that "in a period of a few weeks" after the inception of the new CANS extension service with the phones, "we heard from our labor partners that staff were facing challenges with the new CANS."

On 11/29/18 at 9:50 a.m., Assistant Nurse Manager 2 (ANM 2) was interviewed. ANM 2 stated a few of the ANMs received training and were given pamphlets and time to practice using the phones. ANM 2 said there was no "on-the-job" or "training-in-motion" learning before the middleware and phones were "rolled-out" and in use together. "The functionality of the CANS and the escalation process was a big change for all staff," ANM 2 reported.

During an interview with ANM 1 on 11/29/18 at 10:45 a.m., ANM 1 remembered staff wanted more "hands-on" training and requested education on the use of the phone itself. Staff wanted to feel "more engaged with the use of the [Vendor 2] phone."

In an interview with RN 2 on 11/29/18 at 11:38 a.m., RN 2 attended the "Super-User" CANS training before its roll-out. RN 2 stated it consisted of a "30 minute to 1 hour" class where we "read a book and played with the phone". The CANS roll-out was a "big blur to everyone." When asked if the CANS training she received was sufficient to do her job, RN 2 said, "No."

RN 3 was interviewed on 11/29/18 at 1:25 p.m. RN 3 discussed the phone alerts transmitted by the new CANS middleware. RN 3 explained that a red alert, or lethal arrhythmia (life threatening cardiac rhythm disturbance), was transmitted to the primary nurse's phone first. If the red alert is not acknowledged by the primary RN, the red alert then escalates to "the ANM's phone in 30 seconds and the Buddy RN's phone in 60 seconds". When asked about yellow alerts, or non-lethal arrhythmias, RN 3 said she "doesn't receive yellow alerts." (According to the 2018 CANS policy, an unacknowledged red alert escalates to the ANM/CN only. The primary nurse can receive yellow alerts; the Buddy RN receives escalated yellow alerts.) When asked about her experience with any training regarding the new CANS extension and phone, RN 3 stated she attended "a class with papers". There has been a lot of discussion on alarm fatigue however, she said. RN 3 mentioned no learning or training has come out of the incidents that have occurred in relation to the CANS.

The Joint Commission, a hospital accreditation organization, published its National Patient Safety Goals on Alarm Management, which included, "As of January 1, 2016, educate staff and licensed independent practitioners about the purpose and proper operation of alarm systems for which they are responsible."

In a group interview with the Chief Nurse Executive (CNE) on 12/4/18 at 1:20 p.m., the CNE acknowledged staff did not understand the nuances of the new CANS. There were significant differences between the old telemetry alarm notification system and the new CANS system in place now. The CNE acknowledged that allowing the end users to practice using the phones in a "playground-like environment" would have been beneficial. "Then", the CNE explained, "staff could make mistakes without repercussions".

According to the 2018 facility Medical Equipment Management Plan, "Education and training for the end users of equipment (including use, reporting failures, and emergency procedures, etc.) is area/department specific and provided through the individual department manager. Educational topics include: capabilities, limitations and special applications of the equipment ...basic operations and safety precautions ...Department Managers/Administrators, in concert with the Clinical Education Department, will verify that each employee possesses the required core competencies relative to the safe and effective use and maintenance of equipment, as required."

In a final interview with the CEPIM on 12/5/18 at 8:30 a.m., she acknowledged the initial curriculum and training by the extension service vendor was provided to supervisory staff only. No site specific training was planned or scheduled. The phone vendor was not involved during the CANS roll out. "I would have done things differently," with regards to staff training before and after the CANS inception," the CEPIM said.

Based on clinical record review and staff interview the facility failed to ensure pain was assessed and medication was administered as prescribed by the physician and per the facility policy for two of 30 sampled patients (Patient 10 & 13), when prescribed pain medication was not administered to Patient 10 despite self-report of pain, and Patient 13 was administered pain medication for mild pain despite a non-verbal assessment of moderate pain.

This failure had the potential to cause pain to not be controlled for Patient 10 and 13 and to preclude staff from having accurate information in the medical record to make clinical decisions for providing adequate care.

Findings:

Review of the electronic medical record for Patient 10 indicated she was admitted to the hospital on 11/23/18 for diagnoses including pneumonia (lung inflammation caused by bacterial or viral infection).

A concurrent interview and electronic review of Patient 10's health record with Clinical Education & Practice & Informatics Manager/Local CANS (Clinical Alarm Notification System) Administrator (CEPIM) was done on 11/29/18 at 1:30 p.m., which revealed a documented pain assessment on 11/28/18 at 4:32 a.m. Patient 10's health record indicated a CNVI (Checklist of Non-Verbal Pain Indicators- used to implement pain interventions) pain score of 6. Further review of the Patient 10's electronic health record included the following physician order dated 11/24/18 at 12:12 a.m.: Acetaminophen Tab 650 mg (milligram-unit of measure) every 4 hours as needed... mild pain (1-3) which was noted as administered at 4:32 a.m. on 11/28/18. CEPIM verified there was no subsequent documentation of Patient 10's pain assessment noted in the electronic record as required by facility policy.

Review of the electronic medical record for Patient 13 indicated he was admitted to the hospital on 11/26/18 for diagnoses including sepsis (body's overwhelming and life-threatening response to infection that can lead to tissue damage, organ failure, and death).

A concurrent interview and electronic review of Patient 13's health record with CEPIM was done on 11/29/18 at 11 a.m., which revealed a documented pain assessment on 11/26/18 at 4:10 p.m. Patient 13's health record indicated he reported upper abdominal pain at a score of 5 (on a scale of 0-10 with 10 being the worst pain you can imagine). Patient 13's electronic health record also revealed the following physician order dated 11/26/18 at 5:26 p.m.: "Norco (an opiate pain medication) 5-325mg 1 tablet...Every four hours as needed...moderate pain (4-6). CEPIM verified there was no documented evidence of the administration of the pain medication as prescribed or any subsequent documentation of Patient 13's pain assessment until 11/26/18 at 7:27 p.m., over three hours after the documented patient self report of pain which exceeded the facility's policy of within one hour.

Review of the facility policy titled Pain Assessment and Management, dated 3/7/17, included the following notations: "Implement pharmacologic, including multimodal analgesic [medications with different mechanisms of action to produce pain relief], and non-pharmacologic pain management interventions as needed to achieve patient comfort...Reassess pain and level of sedation within 60 minutes or less of the intervention as needed to evaluate the effectiveness of pain management interventions...Maintain effective communication and notify the physician if...Patient's pain goals are not met."

Review of the facility policy titled Basic Unit Care Standards for Medical-Surgical and Telemetry Unit included the following notations: "Take patient's vital signs...every 8 hours or as ordered...Assess patient's pain/comfort level with routine vital signs and PRN (as needed) using the appropriate pain scale...Evaluate and document patient's response to intervention within 1 hour...Notify the physician if unable to achieve patient's comfort level."

The hospital failed to set up and maintain a safe environment of care when:

A. A Clinical Alarm Notification System (CANS) utilizing equipment and devices was set up without validating the system was safe and reliable (refer to A724);

B. Effective staff training and opportunities for trialing the CANS and associated devices was not provided (refer to A0397);

C. By not incorporating the safety of the CANS into the Quality Assurance and Performance Improvement program (refer to A263).

These failures put patients at risk for preventable injuries, adverse events, and delays to recognize and rescue from life-threatening events.

The cumulative effects of these systemic problems resulted in the inability of the Hospital to comply with the statutorily-mandated Condition of Participation for Physical Environment.

Based on observation, interview, and record review, the hospital failed to ensure a continuous cardiac monitoring alarm system was safe, reliable, and used as intended for two 24 bed telemetry units. The hospital used an unregulated end user device as a communications extension for cardiac monitoring that nursing staff relied on to notify them of critical patient conditions and was designed such that signals were not reliably and timely recognized by staff. There was a lack of local hospital or site specific testing and workload specific hazard and risk analysis of the system prior to and after implementation.

This failure resulted in unsafe equipment being used as a primary means of notification to guide clinical decision making by staff to provide rescue interventions for life-threatening conditions that may have resulted in significant patient harm including death.

Findings:

According to the Lippincott's Nursing Procedures and Skills (provided by the hospital), revised 8/17/18, "Cardiac monitoring enables continuous observation of the heart's electrical activity ...In telemetry, a small transmitter connected to the patient sends electrical signals to another location where they're displayed on a monitor screen ... monitors can display heart rate, rhythm, produce a printed record of cardiac rhythm, and sound an alarm if the patient's heart rate rises above or falls below specified limits. Monitors can also recognize and count abnormal beats as well as recognize changes in cardiac activity [i.e. - a heart attack, fatal heart rhythms, or cardiac arrest]." The above describes the type of central cardiac monitor in use at the hospital.

During an observation of the telemetry unit and concurrent interview with the Director of Clinical Adult Services (DCAS) and the Telemetry Manager (TM) on 11/27/18 at 10:40 a.m., the central cardiac monitoring center was noted to be unmanned. The DCAS and TM confirmed there was no staff visualizing the monitor continuously. The system was described as follows: in the event of a physiologic alarm, the central monitoring center would then send data of a clinical alarm to the middleware software (Vendor 1- software that acted as a bridge between two systems or applications). The middleware would then send the information to an end user device, in this case a phone (Vendor 2). The phone was a wireless mobile phone on the hospital's local or internal wireless network. The phones were utilized by direct care nurses and other staff. The phones were programmed to receive clinical alarms in the form of a text message from the central cardiac monitors. The phones were also programmed to receive other alarms/ alerts including patient bed exit alarms, and chair exit alarms and other text messages. The phones were also used for audio telephone calls.

The policy titled, Clinical Alarm Notification Systems (CANS) NCAL Regional Policy, last revised 12/2017, established, "The middleware application uses advanced rules to deliver calculated alerts and alarm notifications to intended nurses and other medical staff. Contextual notifications facilitate situational awareness in which caregivers can better determine a treatment plan for their patients. The primary alert for physiologic alarms remains the monitor or central station." The policy defined, "Alarm/alert: An audible and/or visual signal intended to get someone's attention ..." and, "Clinical Alarm: Any alarm designed to alert staff of any potentially life-threatening situation or need for intervention." The CANS policy defined the severity of the cardiac alarms and expected responses as the following: "Red- Validate/ respond immediately ... Yellow- Top priority and answer required promptly ...Inoperative- answer promptly."

During observations of care in the 3rd floor telemetry unit with the TM, the Clinical Quality Director (CQD), and the Medical Director for Quality and Risk (MDQR) on 11/27/18 at 10 a.m., Registered Nurse (RN) 5 explained that for Patient 21 within the past hour (at 9:08 a.m.) the heart monitoring system detected a very slow heart rate, should have triggered an audio and visual alarm signal to the central station, and from there should have transmitted a text message red alert signal to RN 5's mobile phone. RN 5 sat in front of electronic display screens (central station) that showed the waveform heart rhythms for 12 or more telemetry patients and explained the phone alarm was not seen or heard. At the time of the incident, RN 5 was providing care to another patient in a room distant from the central station where audio and visual alarms registered.

When RN 5 completed the patient care task at 9:20 a.m., RN 5 "woke the mobile phone" and discovered the red alert text message was waiting to be "accepted." RN 5 immediately checked to see that Patient 1 did not need rescue intervention. Then RN 5 reported the missed red alert to Assistant Nurse Manager (ANM) 2. From the central station display, RN 5 was able to show the waveform for Patient 21 that triggered the 9:08 a.m. red alert. It showed that Patient 21's heart rate was 39 beats per minute (normal is 60-100) for 6 seconds beginning at 9:07 a.m.

Charge Nurse (CN) 1 joined the discussion. CN 1 indicated that her duties included providing relief and support to the unit RNs, and often (but not continuously) she might sit close to the central station display where alarms were visible and audible. However, for Patient 21's 9:08 a.m. alarm, CN 1 did not hear or see such an alarm from where she was located at that moment. CN 1 indicated that no staff was specifically assigned to continuously stay at the central display to recognize critical heart activity and alarms, or to recruit others to respond when a concerning waveform persisted or alarms were unanswered.

CN 1 also carried a mobile phone programmed to receive the same red alert text messages that the 3rd floor telemetry RNs received when the RN did not respond to a red alert message within 30 seconds. ANM 2 similarly carried a mobile phone programmed to receive the RN red alerts for up to 48 patients in two telemetry units located on two separate floors of the hospital, and was expected to respond if the responsible RN did not respond within 30 seconds.

CN 1 indicated that neither CN 1 nor ANM 2 heard or saw the 9:08 a.m. red alarm for Patient 21.
The TM indicated that the types of patients routinely admitted to the telemetry units included patients with heart attacks, recovery from invasive imaging studies of the heart, strokes, chest surgery, head and neck surgery. All were high risk for a life-threatening event and required continuous monitoring of heart function, and for some, blood oxygen levels.

In an interview on 11/27/18 at 10:40 a.m., the Medical Director for Quality and Risk (MDQR, a physician who routinely admitted and cared for patients in telemetry units and elsewhere) stated that when MDQR ordered continuous cardiac monitoring, he expected care staff to respond to alarms and notifications of serious heart rhythm conditions timely, from immediately to several minutes depending on the specific condition. A delay of 13 minutes to respond to a rhythm such as Patient 21 displayed was too long. MDQR expected nurses to reliably hear and recognize alarms in real time.
During observation and interview of Patient 21 on 11/27/18 at 10:40 a.m., Patient 21 appeared tired and serious, sitting in a chair beside her patient bed, with one arm stationary in her lap and not used throughout the meeting. Patient 21's son explained that she was 79 years old and had had a kidney transplant 6 years ago. She was admitted to the hospital this time on 11/26/18 for a stroke, for which various rehabilitation therapies were being provided.

Review of Patient 21's medical record with the TM on 11/29/18 at 10:10 a.m. confirmed the son's account. Also identified was an abnormal heart rhythm condition at the time of admission for which continuous cardiac monitoring was ordered. The records showed that Patient 21's rhythm was still not stabilized and shifted back and forth from a normal rate to various high risk states both fast and slow.

In a collaborative review of the Clinical Alarm Notification System (CANS) report (generated from the middleware software medical device that transmitted telemetry patient alarms to the various nurse-held mobile phone devices) on 11/27/18 at 4:15 p.m., the DCAS showed that Patient 21's 9:08 a.m. slow heart rate alarm text signal was "delivered" within 10 to 30 seconds to 4 mobile phones. One phone was assigned to RN 5, one phone was assigned to CN 1, one phone was assigned to ANM 2, and one phone was not in use at the time. The report showed that only RN 5's phone "accepted" and "removed" the message at 10:38 a.m. (after showing it to ANM 2 as explained above).

Review with TM of additional CANS reports for Patient 21 eight hours prior to the 11/27/18 at 9:08 a.m. event revealed that a red alarm for a temporary period where Patient 21's heart stopped beating entirely was triggered at 1:28 a.m. That alarm was delivered and accepted by an RN phone, but was also delivered to an ANM and unassigned phone, neither of which registered as responding to the alarm.

In the 11/27/18 at 4:15 p.m. interview, DCAS indicated that telemetry unit nursing staff had been reporting similar stories of not hearing/seeing mobile phone text alerts, regardless of which screen the phone was idling from. When investigated, the nurse managers and biomedical engineer staff were not able to replicate the user experience. The CANS reports kept showing that the messages were delivered. Investigators were not able to determine why the phones did not create audible alarms consistently. It was a "dead end." They didn't know what more to investigate and had no means to correct it. Yet, they continued to rely on the end user mobile phone device as the primary means to notify nurses of critical life-threatening heart rhythm conditions.

In group interviews on 11/27/18 at 1:50 p.m., attended by the Regional Clinical Informatics Director (RCID), local Area Information Officer (AIO responsible for electronic device management and communications systems), the local Clinical Education Practice & Informatics Manager (CEPIM), the local Support Services Manager (SSM), and the DCAS, and on 11/28/18 at 10:10 a.m. attended by AIO, CEPIM, DCAS, the RCID, the Regional Director of Nursing Informatics (RDNI), Regional Director of Patient Care Services (RDPCS), and the Senior Director of Operations for Southern California (SDOSoCal) the following was presented:

1. A decision to replace portions of the hospital's continuous heart monitoring alarm system was made about 2 years ago when a software technology that connected the primary monitoring equipment to mobile end user devices (middleware) was approaching its end of life usage.

2. That decision plus decisions for selection of a middleware vendor (Vendor 1) and any other equipment needed, as well as the sequence and procedures to achieve the conversion at 39 site hospitals, were made by corporate leaders outside the hospital. Local hospital biomedical, clinical technology, and nursing managers did not participate in any site specific hazard and risk analysis targeted at the building design and workflow aspects of this hospital to ensure it was safe and not requiring mitigation of hazards prior to installation. The members of the group interviews were unable to produce results of a formal site specific hazard analysis (targeted for this change in setup of a continuous monitoring alarm system) prior to installing the equipment and using the changed system.

3. Each site hospital was instructed to follow the regionally directed instructions, with "support" from regional administrators and Vendor 1. The configuration for the setup at the local hospital did not identically match that of other hospitals. For example, some hospitals had a different brand of mobile phone end device, some had pagers, some had additional monitor displays, and some had dedicated staff to continuously watch the primary alarm display center. The simulation testing of the new software medical device transmitting to some type of mobile end user device occurred at an offsite location in another city. No simulations were conducted onsite using the planned equipment configuration in real-world conditions with trained site specific staff prior to using the new setup around August 2017 and relying on it for patient care ever since.

4. The hospital's mobile phone end devices for the setup had been used with a previous middleware software that did not function identically to the new middleware. There were some differences in the functionality of the phone. Users had only been using the phone previously with the old middleware for about one year. Experts on the end user mobile phone did not participate in setting up this system for this specific hospital and did not attend the implementation activities as it was rolled out.

5. It was noticed very early in implementation that recognition of alerts by users of the end phone device was inconsistent. Attempts to understand what users were describing included the collection of data on frequency of alarm delivery per user per shift (alarm volumes), but did not include frequency of failed alarm recognition. No time-motion studies or "follow the nurse" shadowing observations over multiple shifts and days were recorded to generate meaningful data. Attempts to bring in Vendor 1 and experts in the phone functionality were not documented. When isolated phone failure incidents were identified, the end-to-end testing performed by the hospital's biomedical engineer did not simulate real-world system conditions and user workflow activities. It was always concluded that the system "worked as intended." Therefore, no effective validation testing of the safety and effectiveness of the new continuous cardiac monitoring alarm notification system was documented.

6. The group leaders repeatedly stated that the alarm notification system that connected the heart monitoring equipment from patient to central station display center with waveform was the primary heart monitoring means to notify nursing staff of critical patient conditions. The group leaders stated that the alerts distributed and extended to the end user mobile phone devices were a secondary means of alarm notification. Yet, the group leaders confidently stated that the receipt of the phone alarm was part of the continuous cardiac monitoring system (i.e. an extension) on which nurses routinely relied to be notified promptly of potential life-threatening patient conditions, and it was often the secondary system that drove a nurse to act (to assess, analyze, and make care decisions).

User Interviews of Functionality Concerns Confirmed by Management

Throughout the survey, front line nurse users of the phone and management leaders reported problems with configuration and functionality of the phone from the beginning of working with the new CANS system in the fall of 2017:

1. In an 11/29/18 at 9:38 a.m. interview ANM 2 stated that incoming alerts could not be seen or heard unless the user had the phone on the home screen. ANM 2 also indicated that there had been times when on the home screen, the alerts were still not heard. In response to Patient 21's slow heart rate alert on 11/27/18 at 9:08 a.m., ANM 2 did not recall hearing, toggling the volume control, or clearing the alert. [The CANS report showed no response by ANM 2.] ANM 2 was carrying the phone in her coat pocket. ANM 2 stated she wondered if it inadvertently "toggled" or cleared. Phone functionality and alarm escalation design "was a big change" when the new system was adopted.

2. In an interview with RN 1 on 11/27/18 at 11:45 a.m., RN 1 described the phones as "scary". RN 1 stated there is confusion regarding a "sleep mode" or "home screen", but regardless, nurses are not getting alarms. RN 1 stated if the phone user is looking through missed calls, alerts, or making a call, the phone is "off" the home screen and stated, "It's difficult to keep up with". RN 1 also voiced concerns regarding the phones being programed to escalate to ANMs or CNs. RN 1 stated many staff had left the telemetry units due to the fear of the functionality of the CANS system.

3. In an 11/30/18 at 10:30 a.m. interview, ANM 1 stated that alerts have been missed or delayed when the user was on an active phone call. If phone was placed on "sleep mode" or "screen saver" audible alerts were not heard; the user must exit and return to the home screen to receive audible alerts. Shortly after adopting the new system, front line nurse users "requested/needed hands-on education on use of the phone itself. You can't just rely on the phone."

4. In an 11/29/18 at 11:30 a.m. interview, RN 2 stated that alerts had to be cleared at the end of every shift. If an incident relating to a missed alert occurred, it was not retrievable directly from the phone for a manager to investigate and report later in the day.

5. In an 11/29/18 at 11:30 a.m. interview, RN 2 stated that an incident occurred on 9/24/18 when a red alert that a patient's heart had stopped was heard while RN 2 was performing a sterile procedure in a patient room (gowned and gloved and not able to touch the phone). RN 2 called out for help from the patient's room with no response. RN 2 "broke sterile technique" (removed the protective garments) to go check on the patient whose alarm triggered. When RN 2 asked the ANM whether they had heard the escalation alert, the ANM replied, "You heard an alert?" ANM never got the alert. A technician investigating the incident told her, "Couldn't explain the occurrence."
6. In an 11/29/18 at 11:30 a.m. interview, RN 2 also stated, "Have had bad reception more than once. Phones go in and out. In rooms 8 and 14 of 3 South dropped calls, dropped messages. We don't receive them sometimes in the corner rooms."

7. In an 11/29/18 at 10 a.m. interview, RN 6 stated that it is easy to bump the phone when carried in a user's pocket, easy to unknowingly press a key, and any key takes the phone off the home screen. RN 6 reviewed the tracking log for verifying receipt and response to alerts that was instituted on 11/28/18. Patient 30 triggered an alarm for low heart rate on 11/28/18 at 11:21 p.m. The primary RN assigned to Patient 30 did not hear an alert. The Tracking RN called her to verify, at which time the primary RN "woke the phone up" at 11:22 p.m. A subsequent low heart rate alert was noted by the Tracking RN at 11:23 p.m. which too was not heard by the same primary RN.

8. In an 11/27/18 at 10:10 a.m. interview, RN 5 stated that alarms stack up if the very first one was not cleared. Some patients generate very frequent alarms for the same condition like non-persistent high or low heart rates, and those alarms over-ride critical alarms from other patients, or delay recognition until the less significant frequent alarms were cleared.

9. In an interview on 11/29/18 at 2:40 p.m., RN 4 indicated that the previous phone system described as easier to use and did not require the multi-step accept-clear response to alerts.

10. In a group interview on 11/27/18 at 1:50 p.m., attended by the RCID, local AIO, CEPIM, SSM, and the DCAS, the representatives stated that the expectation was that the nurse users of the mobile phones issuing alarms would hear the alarms and respond promptly. However, they listed several obstacles to achieving that including activities that frequently removed the user from the home screen (active telephone call or text, active alarm being delivered, and not clearing prior messages). The phones had limited capability to deliver audible alerts while the user was engaged with other tasks. An attempt to try the only mode available from the phone produced a distracted background sound which doctors disliked, so it was rejected. Another attempt resulted in losing prioritization of alerts so that the least important last message would show first, and viewing more important messages was delayed. "Neither are very good options." They claimed there were no other means to drive prioritized alerts through to be promptly recognized when off the home screen was possible from this particular phone model.
The representatives also stated that the phone design was not capable of retaining activity logs or generating activity reports to better identify trends in aspects of functionality that corresponded to the issues described by users, and that persisted after education and re-education of users from the inception of using the system to today.

11. In an 11/27/18 at 4:15 p.m. interview, DCAS stated that those conducting investigations of delayed or missed alarm incidents were not able to retrieve precise phone functionality activity logs in order to better understand why users reported that they didn't hear alarms even when on the home screen. DCAS believed such instances did occur, but the limitations of the equipment to analyze potential failure points led to a "dead end." Investigators couldn't determine why an incident like that for Patient 21 occurred, couldn't fix it, and therefore couldn't prevent it from occurring again.

12. In an 11/28/18 at 10 a.m. interview, TM stated that in the first day (11/28/18) of auditing whether user RNs were hearing and receiving alarm messages, there had been instances where the RN's phone was on home screen but the alert was not heard.

13. In an interview with the DCAS on 12/4/18 at 1:30 p.m., DCAS stated additional CANS functionality testing had been completed on 12/3/18. The testing was done after the Department had entered the hospital 6 days prior. The DCAS stated there was an instance of "not getting an audible alert".

In a review of the recorded results of the testing performed on 12/3/18, the following was validated, "Extreme Brady (very slow heart rate) configured to escalate to Charge RN Phone and ANM both, Upon initial alert no sound or vibration to CN device ... Alert was present in the log on the phone. Tried multiple times to replicate with same device and other devices, however unable to re- create."

Review of a 2006 clinician survey report from the American College of Clinical Engineering (ACCE) Healthcare Technology Foundation titled "Impact of Clinical Alarms on Patient Safety" documented, "Alarms must be accurate, intuitive and provide alerts which are readily interpreted and acted on by clinicians ... adverse patient events continue to occur related to alarm system design and performance." The authors warned not to substitute alarms for providing sufficient trained staff. "When caregivers rely on alarms, it becomes essential that the alarms perform to their expectations." In addressing the challenges faced in general care units (in contrast to critical care units) such as the hospital's telemetry units, the authors discussed "enhancement solutions" such as technologies that route device alarms through a nurse call or paging system or enunciator (audible communication) devices. Reference was made to a Veterans Health Administration Patient Safety Alert dated 7/2/04. That report stated, "Medical alarm systems using paging technology are not designed or intended to be used as the primary method for alerting clinical staff of critical alarms conditions nor are they approved for this use by the FDA."

Review of the 510(k) premarket notification of intent letter to the United States federal Food and Drug Administration (FDA) dated 11-11-13 to market Vendor 1's Clinical Alert Notification System (CANS) as a Class II medical device noted in Vendor 1's supporting documents that the CANS device interfaced "with health care information systems and medical devices to provide a secondary means of annunciating and displaying patient alarm text and other information to mobile health care workers that is important to patient care, nursing vigilance and overall hospital operations." The CANS was limited to use by qualified medical professionals who have been trained on the use of the device. Patient monitors were listed under "types of medical devices supported." Quality assurance measures that were applied to the development of CANS included "risk analysis" and "usability and validation testing." That quality assurance applied only to the software application in isolation, and not to the performance of the end user device. The CANS was intended as a supplementary alarm system, with respect to safety, effectiveness, and performance.

Review of the manufacturer user manual for the end user mobile phone specifically stated, "This product is not a medical device." The user manual showed no documentation that the phone was intended for use in a health care setting. The user manual detailed the phone's features and functionality as that of a telephone communication device. It used an unlicensed frequency band that was vulnerable to radiofrequency interference from medical devices and equipment. The manufacturer product label showed no regulatory approval as a primary or secondary heart monitoring alarm notification device.

Review of the FDA guidance document titled "Medical Device Data Systems (MDDS), Medical Image Storage Devices, and Medical Image Communications Devices," issued 2/9/15, defined a MDDS per 21 Code of Federal Regulations 880.6310 as a hardware or software product that transfers, stores, converts formats and displays medical device data. MDDS does not alter or control the functions of any connected medical devices. MDDS is not subject to FDA registration, compliance, or enforcement. However, the definition of MDDS does not include devices intended to be used in connection with active patient monitoring. The clinical context for active patient monitoring, or the particular clinical condition (e.g. life-threatening heart rhythm), likely required a timely response and would not be exempt from regulation, as other data storage and display devices were. In this respect, reliance on the use of mobile phone text alerts to notify the nurse responsible for continuous heart monitoring of a patient's life-threatening heart rhythm may be considered part of the continuous heart monitoring telemetry system overall, and quite possibly subject to FDA oversight.

Review of other FDA guidance documents related to medical device safety for hardware and software, such as "Mobile Medical Applications" issued 2/9/15 and "Off the Shelf Software Use in Medical Devices" issued 9/9/1999, consistently focused on safety aspects of "intended use" in active patient monitoring and "intended use environment."
FDA guidance from "Radio Frequency Wireless Technology in Medical Devices," issued 8/14/13, stressed the need to "determine and understand the medical device functions that are to be wirelessly enabled and the intended use of the medical device. The medical device functions and intended uses should be appropriately matched with the wireless technology's capabilities and expected performance. In addition, issues relating to the integrity of data transmitted wirelessly... and safety-related requirements of your device should be considered. Potential risks that can affect consistent and timely wireless medical device functions include data corruption or loss and interference from simultaneous transmitters in a given location, which can increase latency (delay) and transmitted signal error rates." ... "If the wireless medical device will be part of a network, wireless Quality of Service (QoS) should be carefully considered in conjunction with the intended use of the wireless medical device ... While the QoS of cellular telephone networks might be acceptable for voice communication, it might not be sufficient for certain medical functions." Quality of Service assessments were part of a risk based verification and validation approach for safety. "FDA recommends that you continue to manage the risks associated with the use of wireless technology described above for the entire life cycle of your device. Your procedures for implementing corrective and preventive action must include, among other things, analyses for possible trends in nonconformance information and complaints, such as reports of failures, which could include erratic or unexpected behavior of the medical device. Examples of such behavior include reprogramming of stimulation devices, commands missed or misinterpreted by operating room controllers, unexplained inconsistencies of an infusion pump, and failure to activate alarm signals in alarm conditions."

Review of the FDA guidance for What Is A Medical Device, retrieved at https://www.fda.gov/medicaldevices/deviceregulationandguidance/overview/classifyyourdevice/ucm051512.htm, noted, " If a product is labeled, promoted or used in a manner that meets the following definition in section 201(h) of the Federal Food Drug & Cosmetic (FD&C) Act, it will be regulated by the Food and Drug Administration (FDA) as a medical device and is subject to premarketing and postmarketing regulatory controls ... an instrument ... machine ... or other similar or related article, including a component part, or accessory which is ... intended for use in the diagnosis of disease or other conditions ... or ... treatment, or prevention of disease." According to FDA at https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfStandards/detail.cfm?standard__identification_no=33226, transmissions of heart rate and alarm as part of a physiologic monitoring device system, network and communication, may be considered a physiologic monitoring medical device subject to regulatory oversight under 21 Code of Federal Regulations 870.2300.

Review of Vendor 1's middleware software contract, effective 2/1/17, between Vendor 1 and the national corporate owner of the hospital, indicated that Vendor 1 promised to maintain industry standard quality assurance and virus protection, and to participate in the hospital's quality assurance program. However, Vendor 1 set limits on its responsibility to patient care as such, "Software does not make clinical, medical or other decisions and is not a substitute for competent, properly trained and knowledgeable staff who bring professional judgment and analysis to the information presented by the Software and that [Hospital] is solely responsible for verifying the accuracy of all patient/schedule information and determining the data necessary to make medical and diagnostic decisions as well as for complying with all laws, regulations, and licensing requirements applicable to [Hospital's] delivery of healthcare services."

Review of the Scope of Work Exhibit to support Vendor 1's contract indicated that Vendor 1 agreed to "provide clinical staff with real time visualization of the alarms, alerts, and notifications from the existing input systems at each [hospital]." The integration would include physiologic monitors, replacement of the existing system owned by the physiologic monitor company (software connector to end device), and Vendor 1 agreed to set up connection to any one of several end point devices, including the mobile phone brand utilized by the hospital. Vendor 1 offered implementation methodology services including user design, system configuration, end-to-end testing, and user acceptance testing of a sample of end devices (not all). However, pre-install site specific hazard analysis was not listed as an offering by Vendor 1. The Scope of Work exhibit clearly understood that Vendor 1 was asked to arrange connectivity with a particular brand/model mobile phone, and required a Communications Manager programming feature to sort the messages and priorities to be delivered to a user phone. Vendor 1 claimed it could support that brand mobile device, did not require coordination with the manufacturer of the mobile device, and did not offer to analyze its functionality in relation to the needs of the user or the patients served.

Review of the hospital's policy titled, Clinical Alarm Notification Systems (CANS) NCAL Regional Policy, last revised 12/2017, documented under item 1.3.1, "The primary alert for physiologic alarms remains the monitor or central station." Item 4.4.1 noted, "Only [brand name] phones with [middleware] service will be used to receive patient care alarms and alerts."

In the 11/28/18 at 10:10 a.m. interview, the AIO indicated that user simulation tests performed in the hospital were performed in a patient room. The alarm generator/simulation device sends a signal to a patient box, which recognized an alarm-eligible pattern, which was recognized by the middleware, and eventually