This case report illustrates the use of remote continuous postoperative electrocardiography (ECG) monitoring, which is routed through a smartphone to provide detection and initiate clinical management of myocardial ischemia. It was part of a feasibility observational trial, PROSE 3 (Clinicaltrials.gov NCT01423136), in which 89 patients were enrolled and 69 completed the study. Electrocardiogram ST segment analysis during Holter monitoring has been studied prior to, during, and after surgery,1-3 but only the postoperative ST segment changes were associated with postoperative ischemic cardiac events in a multivariate analysis model.2 Postoperative myocardial ischemia can be treated by decreasing heart rate reductions, thereby interfering with the myocardial ischemic cascade.4,5 Nevertheless, as postoperative ECG analysis is typically performed offline, days after the associated myocardial ischemic event, timely treatment is not possible. To address this shortcoming, PROSE 3 was conducted to examine the feasibility of real-time ST segment analysis as demonstrated with this case.

Clinical features

A 42-yr-old male (weight, 83 kg; height, 163 cm) was scheduled for open elective radical prostatectomy with bilateral pelvic lymphadenectomy. Prior to surgery, the patient had consented to remote real-time ECG monitoring as part of the PROSE 3 trial. His medical history included a myocardial infarction in 2002, coronary angioplasty in 2006, hypertension, and class II angina. His daily medications prior to surgery included 75 mg venlafaxine, 20 mg rosuvastatin, 25 mg hydrochlorothiazide, and 100 mg metoprolol bid. Cardiology was consulted prior to surgery and no further optimization was deemed necessary. Preoperative hemoglobin was 139 g·L−1 and the 12-lead ECG was unremarkable. Spinal anesthesia and analgesia were achieved with 0.5% bupivacaine 10 mg with intrathecal preservative-free morphine 200 µg as per institutional protocol. Standard monitoring included noninvasive blood pressure, ECG (leads II and V5), and pulse oximetry.6 General anesthesia was induced with propofol 90 mg, fentanyl 150 µg, and rocuronium 45 mg and maintained with desflurane, oxygen, air, and repeated doses of fentanyl to a maximum of 225 µg. Surgery was uneventful and lasted two hours 33min. The estimated blood loss was 385 mL, and normal saline 2,073 mL iv was given intraoperatively. Reversal of muscle paralysis was achieved with neostigmine 1.5 mg and glycopyrrolate 0.2 mg. In the postanesthesia care unit (PACU), blood loss from surgical drains was 50 mL; intravenous administration included normal saline 390 mL iv and Voluven® 500 mL iv. The patient’s stay in the PACU was uneventful, and he was discharged to a standard surgical ward at 19:37, three hours and 45min postoperatively. Remote real-time heart rate and ST segment monitoring was then initiated via antenna pickup connected to a Spacelabs Ultraview monitor (Spacelabs Healthcare Canada, Mississauga, ON, Canada) with full disclosure settings. Leads II and V5 were monitored, with ST depression measured at 0.06 sec after the J point. Changes of at least 1 mm in amplitude lasting longer than ten minutes were considered significant. Heart rate alarm limits were set to < 50 and > 120 beats·min−1. Electrocardiogram heart rate and ST alarms were routed to the study clinician’s smartphone via secured hospital e-mail (Fig. 1).

Fig. 1
figure 1

Electrocardiogram monitoring received remotely via smartphone at 19:43 after the patient arrived on the surgical ward. Leads II and V5 (upper and lower traces, respectively) were monitored. No alarm condition

Full size image

Within a five-hour 56min interval (19:38-01:34) 22 false heart rate alarms were received, with an upsloping ST depression < 1 mm. These were considered to be motion artifacts after the study clinician reviewed them in real time. During this interval, the patient was regularly checked and assessed to be doing well. His vital signs recorded at midnight were blood pressure 135/77 mmHg, heart rate 98 beats·min−1 and regular, and pulse oximetry SpO2 98%. During 01:54-02:09 AM, however, two ST alarms were received on the clinician’s smartphone, showing horizontal ST segment depressions that gradually increased from -1.12 to -1.84 mm (Fig. 2). At 1:54 AM, a heart rate of 127 beats·min−1 also triggered a heart rate alarm. Upon request by the study clinician, the nurse checked the patient at 02:15 AM, who assessed the patient to be sleeping “comfortably” and easily aroused. There was no complaint of chest pain and the nurse did not notice any cyanosis. It was noticed, however, that the patient did not receive his evening dose of metoprolol. The study clinician notified the on-call surgical and anesthesia personnel at 02:20. Subsequently, a 12-lead ECG, complete blood count (CBC), blood chemistry, and troponin I tests were ordered. Oxygen via nasal prongs was administered. The 12-lead ECG showed a heart rate of 88 beats·min−1 with new T wave inversions and Q waves inferiorly. Metoprolol 10 mg iv was given in 1-mg increments titrated to maintain the patient’s heart rate at 60-80 beats·min−1. The CBC showed a drop in hemoglobin from 102 g·L−1 (postoperatively) to 85 g·L−1, and one unit of packed red blood cells was transfused (hemoglobin 88 g·L−1 following the transfusion). At 04:19, ST segments normalized and heart rate was maintained at 60-80 beats·min−1. Cardiology service was consulted. Although they were unavailable to see the patient until later in the morning, no further actions were deemed necessary and they supported our assessment of supply-demand ischemia. The patient was discharged three days postoperatively and has been without cardiac-related problems for over four years.

Fig. 2
figure 2

Electrocardiogram monitoring received remotely via smartphone at 01:55 showing ST segment depression (−1.12 mm) accompanied by alert. ECG = electrocardiogram

Full size image

Discussion

This case shows how technology can link clinicians to a changing clinical situation in real-time, enabling provision of appropriate and timely medical care. Technology helped to alert the perioperative team so that a potentially serious cardiac ischemic event could be appropriately interrupted. Importantly, during the ischemic event, the patient did not experience any untoward symptoms, and standard patient care on the surgical ward was insufficient to identify the evolving cardiac event. Such real-time ECG monitoring differs from previous efforts using Holter monitoring which provides information that is time delayed by hours or days. The remote monitoring, as described, also differs from traditional telemetry, which is available only in specialized locations, such as step-down, intensive care, or cardiac care units. Moreover, the integration of telemetry into these sites requires extensive nursing training and continuing education.

Indeed, telemetry monitoring for ST segment changes has been used after acute coronary syndromes and coronary artery interventions,7 but it remains an underutilized technology even in these circumstances.8 Postoperative ST segment monitoring via telemetry after non-cardiac surgery is even rarer, and its availability on surgical wards is unknown.7 While remote ECG monitoring technology has existed for some time, there is a lack of literature concerning its use for remote real-time off-site monitoring or on regular surgical wards during the postoperative period. Nevertheless, some studies have been published concerning its use in the intensive care unit,9-13 including during weaning from mechanical ventilation.14-17 Importantly, none of those studies involved remote off-site monitoring.

An important consideration for remote patient monitoring is ensuring that the technology provides the appropriate documentation, privacy, and firewall protection befitting a medical recording and transmission system. Our system has de-identified encrypted e-mail transmissions, e-mails stored behind the firewall, full disclosure behind the firewall, and “double-lock” data storage as per Research Ethics Board requirements. One of the drawbacks of remote cardiac monitoring is the relatively large number of false alarms, and this continues to be a challenge with this emerging technology. During the 24 hr of monitoring with the patient described in this report, there were 60 false alarms and three true alarms. Causes for false alarms include motion artifacts, electrolyte shifts, temperature changes, and tachycardia.18 In addition, there are biomedical engineering reasons for false alarm triggers, and work in progress on signal quality index and signal-to-noise ratio has shown improvements in alarm algorithms.19-21

Postoperative myocardial infarction is one of the most frequent and serious adverse events.22 In a study of vascular patients, the overall mortality was 3.4%, and seven of eight deaths were related to perioperative myocardial ischemic injury (PMII). Of 236 patients, PMII occurred in 42 (17.8%) patients: 22 myocardial infarctions, 11 congestive heart failures, and 12 new arrhythmias (three patients had two PMII events). The mean hospital length of stay was 16.8 and 10.0 days for patients with and without PMII, respectively (P < 0.001).23 In a study of patients ≥ 60 yr of age, 129/1,422 (9.1%) developed at least one major adverse cardiac event, and cardiac death occurred in 11 patients (0.8%).24

With the patient described in the present report, the combination of perioperative blood loss, a missed evening metoprolol dose, and tachycardia undoubtedly contributed to the evolving cardiac ischemic cascade. The synergistic combination of missing postoperative beta-blocker doses and bleeding on myocardial ischemia has been previously described.25 The rebound effect of increased heart rate and contractility after missing a beta-blocker dose in the postoperative period, combined with postoperative anemia, is likely a factor leading to an imbalance of myocardial oxygen supply and demand and the resulting ischemia. An acute coronary syndrome may be the result of plaque rupture (type 1) or an imbalance of supply and demand (type 2),22 and the ensuing ischemic cascade26 may be interrupted by early intervention. In this case, the troponin I rose to only 0.05 µg·L−1 and no new ECG changes persisted, so it is unlikely that the patient experienced a postoperative myocardial infarction. In view of the specificity of postoperative myocardial ischemia in predicting myocardial ischemic complications, we assume that the early intervention to interrupt the ischemic cascade prevented a likely postoperative myocardial infarct.2,10

In conclusion, this case report illustrates the usefulness of real-time remote monitoring of ECG signals recorded from hospitalized patients at risk of cardiac events. This case involved a postoperative surgical patient who experienced electrocardiographic evidence of myocardial ischemia during his recovery on a standard surgical floor. The patient was otherwise asymptomatic. The ECG monitoring, as described, allowed for timely detection and appropriate clinical intervention, otherwise that would not have occurred. Real-time remote ECG monitoring during the postoperative period is a potentially useful tool to help prevent perioperative myocardial infarction.