BP Management System User Acceptance Testing

Nurse-subject quantitative survey response (yes/no) about any confusion or perceived inaccuracies of the system; perceived risk of patient management error; or perceived distraction that impacted care delivery

System malfunction (yes/no) observed in real-time by the Technical Observer

Completeness (yes/no) of flowsheet (data archive of the investigational System) without any lapses in input data from the bedside monitor during connectivity; and without lapses in associated BP forecasts.

Error-level and critical-level software errors in the runtime log (yes/no)

Median percentage (%) of time, per patient, during the session when inclusion criteria were otherwise met that the BP forecast was operative (i.e., time intervals when no InOp criteria detected by the System leading to the InOp state)

Accuracy of the future 20 min median MAP forecast (average error between the MAP forecast and the future 20 min median MAP; for purposes of computing future 20 min median MAP, MAP < 10 mmHg and MAP > 250 mmHg is treated as a missing number) over time intervals during the session when inclusion criteria were otherwise met; the BP forecast was operative; and the analysis window criteria were met [at least 19 minutes remaining to the end of the record from the time of the forecast; and the analysis window has no more than 9 min of non-physiological MAP data (non-physiological MAP criteria are MAP < 10mmHg or MAP > 250mmHg or 'NaN')].

Median percentage (%) of time, per patient, that the true future 20 min median MAP falls within the computed forecast cone over time intervals during the session when inclusion criteria were otherwise met; the BP forecast was operative; and the analysis window criteria were met [at least 19 minutes remaining to the end of the record from the time of the forecast; and the analysis window has no more than 9 min of non-physiological MAP data (non-physiological MAP criteria are MAP < 10mmHg or MAP > 250mmHg or 'NaN')]

SHE incidence per 24 hrs [Sustained hypotensive event "SHE" is any interval commencing upon MAP < goal & 20-min future median MAP is also < goal; terminating upon time point when MAP >= goal and & 20-min future median MAP >= goal; SHEs occurring sequentially within 10 min are merged; any remaining SHEs not persisting at least 10 min are excluded]. [Clarification: We are reporting the mean "events per patient-day" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) prior to the SHE's earliest hypotension [hypotension is defined as MAP < goal, with the exclusion of MAP < 10 mmHg, which is treated as a missing number]. Notification events are intervals that begin when there is an Index value that meets the notification criteria and terminate when there is an Index value output that does not meet the notification criteria.

AWTs (median) for SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) prior to the SHE's earliest hypotension [i.e., MAP < goal]. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) prior to the SHE's earliest hypotension [i.e., MAP < goal]

AWTs for SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) prior to the SHE's earliest hypotension [i.e., MAP < goal]. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) prior to continuous hypotension [intervals of ≥ 10 minutes of physiological MAP continuously below the hypotension threshold (physiological MAP are samples that do not meet the non-physiological criteria)]. NOTE: excluding SHEs predicted prior to earliest hypotension

AWTs for SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) prior to continuous hypotension and excluding SHEs predicted prior to earliest hypotension. NOTE: A negative AWT denotes a notification event that occurs chronologically after the onset of the SHE event. For example, an AWT of -3 would mean the first notification event occurred 3 minutes after the onset of the SHE event, although still in advance of continuous hypotension.

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) prior to continuous hypotension [intervals of ≥ 10 minutes of physiological MAP continuously below the hypotension threshold (physiological MAP are samples that do not meet the non-physiological criteria)]. NOTE: excluding SHEs predicted prior to earliest hypotension

AWTs for SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) prior to continuous hypotension and excluding SHEs predicted prior to earliest hypotension. NOTE: A negative AWT denotes a notification event that occurs chronologically after the onset of the SHE event. For example, an AWT of -3 would mean the first notification event occurred 3 minutes after the onset of the SHE event, although still in advance of continuous hypotension.

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) after onset of continuous hypotension

AWTs for SHEs predicted by an SHE-contiguous notification event (i.e., Index > 35%) after onset of continuous hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) after onset of continuous hypotension

AWTs for SHEs predicted by an SHE-contiguous notification event (i.e., Index > user's Index threshold setting) after onset of continuous hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage of total SHEs with false resolutions [i.e., MAP >= goal transiently during SHE]

Percentage (%) of total SHE false resolutions with a contiguous notification event where {Index > 35%}

Percentage of total SHE false resolutions that were not detected by {Index > 35%} and had InOp state occurring during the index false resolution episode

Percentage (%) of total SHE false resolutions that were not detected by {Index > 35%} and did not have InOp state occurring during the index false resolution episode

Percentage (%) of total SHE false resolutions with a contiguous notification event where {Index > user's Index threshold setting}

Percentage (%) of total SHE false resolutions that were not detected by {Index > user's Index threshold setting} and had InOp state occurring during the index false resolution episode

Percentage (%) of total SHE false resolutions that were not detected by {Index > user's Index threshold setting} and did not have InOp state occurring during the index false resolution episode

Incidence of notification events {Index > 35%} that are contiguous with SHEs, excluding notification events with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} that are contiguous with SHEs, excluding notification events with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} that are sentinel notifications [sentinel notification is a notification event that occurs within 60 min prior to an SHE; recorded dose increase; or during stuttering hypotension (defined in 3.7.1)], excluding notification events with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} that are sentinel notifications, excluding notification events with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} that are false notifications (defined as notification events not contiguous with an SHE and not a sentinel notification, excluding notification events with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} that are false notifications. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} contiguous with an SHE false resolution. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} contiguous with an SHE false resolution. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} contiguous with an SHE true resolution. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} contiguous with an SHE true resolution. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} contiguous with an SHE true resolution after excluding all notification event interval that were within 8 min of a documented clinical intervention. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} contiguous with an SHE true resolution after excluding all notification event interval that were within 8 min of a documented clinical intervention. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > 35%} with onset during an SHE. [Clarification: We are reporting the mean "notification events / patient-day (24h)" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Incidence of notification events {Index > user's Index threshold setting} with onset during an SHE. [Clarification: We are using "notification events / patient-day (24h)" as the Unit of Measure for several reasons. First, different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Median duration of false notification events {Index > 35%}

Median duration of false notification events {Index > user's Index threshold setting}

Stuttering hypotension incidence [Stuttering hypotension event commences upon hypotension and with >= 10 min of cumulative hypotension within subsequent 60 min; terminates upon next time point when onset condition is no longer met AND when MAP >= goal; intervals that qualify as an SHE are excluded from stuttering hypotension]. [Clarification: We are reporting the mean "events per patient-day" as the Unit of Measure. Different subjects were observed for different intervals of time, with a maximum 8 hours specified in the protocol. Because subjects were observed for different time intervals, it is helpful to standardize findings as events per unit of time. Standardizing the events to units of time allows for valid inter-subject comparison when subjects have differing observation intervals. Next, events per DAY is a conventional Unit of Measurement that is both valid and easy to interpret, even though subjects were not observed for an entire day.]

Percentage (%) of total stuttering hypotension episodes detected by {Index > 35%} prior to the stuttering hypotension's first hypotension

AWTs for stuttering hypotension episodes detected by {Index > 35%} prior to the stuttering hypotension's first hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total stuttering hypotension episodes detected by {Index > user's Index threshold setting} prior to the stuttering hypotension's first hypotension

AWTs for stuttering hypotension episodes detected by {Index > user's Index threshold setting} prior to the stuttering hypotension's first hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total stuttering hypotension episodes detected by {Index > 35%} prior to the occurrence of 10 cumulative minutes of hypotension within that stuttering hypotension episode (excluding episodes detected prior to first hypotension)

AWTs for stuttering hypotension episodes detected by {Index > 35%} prior to 10 cumulative min of hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total stuttering hypotension episodes detected by {Index > user's Index threshold setting} prior to the occurrence of 10 cumulative minutes of hypotension within that stuttering hypotension episode (excluding episodes detected prior to first hypotension)

AWTs for stuttering hypotension episodes detected by {Index > user's Index threshold setting} prior to 10 cumulative min of hypotension. NOTE: A positive AWT denotes a notification event that occurs chronologically in advance of the onset of the SHE event. For example, an AWT of +3 would mean the first notification event occurred 3 minutes before the onset of the SHE event.

Percentage (%) of total stuttering hypotension episodes not detected by {Index > 35%} prior to the occurrence of 10 cumulative minutes of hypotension within that stuttering hypotension episode

Percentage (%) of total stuttering hypotension episodes not detected by {Index > user's Index threshold setting} prior to the occurrence of 10 cumulative minutes of hypotension within that stuttering hypotension episode