CARDIOVASCULAR disease remains a common

Transcription

1 ACADEMIC EMERGENCY MEDICINE June 2000, Volume 7, Number Continuous Electrocardiographic Monitoring and Cardiac Arrest Outcomes in 8,932 Telemetry Ward Patients MICHAEL J. SCHULL, MD, MSC, DONALD A. REDELMEIER, MD, MSC Abstract. Objective: To estimate the benefit of routine electrocardiographic (ECG) telemetry monitoring on in-hospital cardiac arrest survival. Methods: In a tertiary care hospital, all telemetry ward admissions and cardiac arrests occurring over a five-year period were reviewed. Ward location and survival to discharge were determined for all patients outside of critical care areas. Results: During the study period, 8,932 patients were admitted to the telemetry ward, and 20 suffered cardiac arrest (0.2%; 95% CI = 0.1 to 0.3). Telemetry monitors signaled the onset of cardiac arrest in only 56% (95% CI = 30 to 80) of monitored arrests. Three patients survived to discharge, and in two of these three patients the arrest onset was signaled by the monitor. This yields a monitor-signaled survival rate among telemetry ward patients of 0.02% (95% CI = 0 to 0.05). All survivors suffered significant arrhythmias prior to their cardiac arrests. Conclusions: Cardiac arrest is an uncommon event among telemetry ward patients, and monitor-signaled survivors are extremely rare. Routine telemetry offers little cardiac arrest survival benefit to most monitored patients, and a more selective policy for telemetry use might safely avoid ECG monitoring for many patients. Key words: telemetry; cardiac arrest; continuous ECG. ACADEMIC EMERGENCY MEDI- CINE 2000; 7: CARDIOVASCULAR disease remains a common killer, with ischemic heart disease and acute myocardial infarctions (AMIs) accounting for a large part of these deaths. 1 Mortality rates for AMI began to fall in the 1960s, 2 4 following the recognition of arrhythmias as an important cause of death, 5 7 and the development of external defibrillation as an effective form of therapy. 8,9 These two factors led to the creation of coronary care units (CCUs), with continuous electrocardiographic (ECG) monitoring and specially trained staff. 2,3 The advent of CCUs was credited with a 30% drop in mortality following AMIs. 2 4 These successes prompted great enthusiasm for ECG monitoring technology. Monitoring expanded outside the CCU into intermediate coronary care units (ICCUs), operating rooms, emergency departments (EDs), and the out-of-hospital setting. 19 From the Clinical Epidemiology Unit, Sunnybrook and Women s College Health Sciences Centre (MJS, DAR), Division of Emergency Medicine (MJS), Department of Medicine (MJS, DAR), University of Toronto, Toronto, Ontario, Canada. Received September 23, 1999; revision received November 17, 1999; accepted November 29, Presented at the SAEM annual meeting, San Francisco, CA, May Supported by the Clinical Epidemiology Unit, Sunnybrook and Women s College Health Sciences Centre, University of Toronto. Address for correspondence and reprints: Michael Schull, MD, MSc, G-140, Clinical Epidemiology Unit, Sunnybrook and Women s Health Sciences Centre, 2075 Bayview Avenue, Toronto, Ontario, Canada, M4N 3M5. Fax: ; e- mail: mjs@ices.on.ca A related commentary appears on page 687. However, there exists no definitive evidence of benefit outside the CCU, 8,10 12,14 17 and the technology is costly. 17,20 Moreover, the routine use of monitors may cause significant delays in patient care, and has been blamed for ED overcrowding and ambulance diversion. 21,22 The primary objective of this study was to assess the benefit of routine continuous ECG monitoring in the cardiac telemetry ward of a tertiary care hospital over a five-year period. The primary objective was to determine the benefit of telemetry in cardiac arrest, by estimating the proportion of telemetry patients who were discharged alive after a cardiac arrest, and whose arrest onset was signaled by the ECG monitor (monitor-signaled survivors). The secondary objectives were to estimate the rate of cardiac arrest among telemetry ward patients and to estimate the rate at which telemetry monitors signal the onset of cardiac arrest. METHODS Study Design. All telemetry ward admissions and cardiac arrests occurring over a five-year period were reviewed to estimate the benefit of routine ECG telemetry on in-hospital cardiac arrest survival. Approval was obtained from the hospital s research ethics board for the study. Study Setting and Population. The study setting was the Sunnybrook and Women s College

2 648 TELEMETRY Schull, Redelmeier TELEMETRY MONITORING AND CARDIAC ARREST OUTCOMES Health Sciences Centre (Sunnybrook site) in Toronto, Ontario, Canada, during a five-year period from April 1994 to March The hospital is a 1,100-bed tertiary care facility, with a nine-bed CCU and a 21-bed cardiac telemetry ward. With the exception of a single bed on a medical ward, the only other ECG monitoring units are located in intensive care units, the post-cardiac surgery ward, the surgical recovery room, and the ED. Cardiology patients at the study hospital were initially admitted to the CCU or the telemetry ward. Most patients admitted to the CCU were subsequently transferred to the telemetry ward prior to discharge home. The telemetry ward is equipped with a Hewlett-Packard M2350A Central Monitor system (Andover, MA). A nurse-to-patient ratio of approximately 1:5 is maintained during daytime hours, but no staff person is specifically assigned to watch the central monitor system (Tupis A, personal communication, May 1999). Physician staffing consists of attending cardiologists, cardiology fellows, and residents. All resuscitations occurring on the telemetry ward are conducted by a hospital arrest team, which is made up of internal medicine residents and CCU nurses [all with Advanced Cardiac Life Support (ACLS) training], and with consultant backup. The team is equipped with a standard crash cart, including a manual defibrillator, advanced airway equipment, and ACLS medications. Study Protocol. All patients admitted to the telemetry ward over the study period were included. Information regarding patient demographics and final diagnosis was obtained from the medical records department. The decision to use telemetry was at the discretion of the admitting physician, and there were no prespecified criteria or policies in use to determine which patients should be monitored. We defined a cardiac arrest according to the in-hospital Utstein recommendations: 23 patients were included if a cardiac arrest team was called and they received airway interventions, defibrillation, cardiac pacing, cardiopulmonary resuscitation (CPR), or intravenous (IV) medications. Arrests were excluded if they occurred outside of the telemetry ward, if the arrest record was illegible, or if survival could not be determined. Data describing cardiac arrests were obtained from arrest records made contemporaneously during resuscitations on a standard chart by designated nurses. Copies of all arrest records were obtained from the cardiology department, where they are stored in a central file. The data were abstracted by a single trained abstractor. Accuracy was maximized by entering data directly from arrest records onto computer using an electronic form designed for the study, by ensuring that data fields could not be left blank, and by setting data field limits to clinically plausible ranges. The following information was collected from each arrest record: baseline characteristics (hospital file number, age, sex, date and time of arrest, location of arrest, past history of cardiac disease, lung disease, or other disease), arrest characteristics (witnessed, bystander CPR, pulse at arrest onset, breathing at arrest onset, initial cardiac rhythm), interventions (intubation, time of intubation, doses of epinephrine and atropine, number of defibrillation attempts, time to first defibrillation, duration of resuscitation), and outcome (arrest survival). Survival to discharge was determined from data provided by the medical records department or, where necessary, by chart review. In addition, the charts of all patients who had cardiac arrests while on the telemetry ward were reviewed to determine prior hospital course and ECG monitor use. A patient was deemed to be on telemetry at the time of cardiac arrest if there was a physician s order for telemetry (and no order discontinuing it) prior to arrest, or nursing notes indicating a telemetry monitor was in use in the period immediately prior to the arrest. For monitored patients, the cardiac arrest onset was assumed to be monitor signaled unless nursing notes clearly indicated that the arrest was recognized as a result of an event other than a monitor abnormality (e.g., a fall, patient calling for nurse, roommate calling for nurse). The primary outcome was the proportion of telemetry patients who were discharged alive after a cardiac arrest, and whose arrest onsets were signaled by their ECG monitors (monitor-signaled survivors). Our required sample size was based on the formula for the desired confidence interval width for a proportion. 24 To achieve a 95% confidence interval of width 0.5%, we used a conservative estimate of 1% for the rate of monitor-signaled survivors; thus, a sample of 1,521 monitored patients was required. A larger sample of telemetry patients provides still more precision around the primary estimate, and allows for more meaningful secondary and subgroup analyses. Data Analysis. The 2 or Fisher s exact test was used to test for differences of proportions. Microsoft Access 97 (Redmond, WA) was used for data entry, and SAS 6.12 (Cary, NC) was used for data analysis. RESULTS During the five-year interval, there were 8,932 patients admitted to the cardiac telemetry ward. Demographic, admission, length of stay, final diagnosis, and mortality data are provided in Table

3 ACADEMIC EMERGENCY MEDICINE June 2000, Volume 7, Number During the same interval, 367 cardiac arrests occurred in 357 patients throughout the hospital. Three hundred forty-seven were excluded: 339 occurred outside of the telemetry ward, the charts of four were illegible, and survival could not be determined for four others (the arrest record included no identifying information such as name, hospital file number, date, or location with which the patients could be identified). None of the excluded arrests occurred on the telemetry ward. The characteristics of the remaining 20 cardiac arrests that occurred on the telemetry ward are listed in Table 2. These 20 cardiac arrests occurred among 8,932 telemetry ward patients, yielding a rate of 0.2% (95% CI = 0.1 to 0.3). This equals about one arrest per 500 telemetry ward patients. Nineteen of the 20 patients were on telemetry monitors at some point while on the telemetry ward (95%; 95% CI = 75 to 100), but in only 16 of the 20 was telemetry in use at the onset of cardiac arrest (80%; 95% CI = 56 to 94) (Fig. 1). Among these 16, the arrest onset was recognized as a result of a monitor abnormality in nine cases (56%; 95% CI = 30 to 80). Three telemetry ward patients survived to discharge following cardiac arrest, all of whom had ventricular tachycardia or fibrillation as initial rhythm, and all were on ECG monitors at the time of the arrest. However, in only one case was the arrest onset clearly monitor-signaled; the second arrest was triggered by the noise made as the patient fell to the floor, and in the third arrest the onset signal could not be determined. Assuming that the third survivor s arrest was monitor-signaled, the estimated rate of patients with monitorsignaled arrests who survived to discharge was two per 8,932 telemetry ward patients (0.02%; 95% CI = 0.0 to 0.05). Among cardiac arrest patients on the telemetry ward, there was no significant difference in survival between monitor-signaled and non-monitorsignaled arrests (two of nine vs one of seven; survival difference 7.9%; 95% CI = 30 to 46). Survival rates were also not significantly different between monitored and unmonitored patients (three of 16 vs zero of four; survival difference 18.8%; 95% CI = 0.4 to 38). Of the four unmonitored patients, two survived their cardiac arrests but died prior to discharge; one was in cardiogenic shock with a normal sinus rhythm when the arrest team was called and one died of a suspected ruptured thoracic aneurysm. Age, gender, and time to first defibrillation were not significant predictors of survival among telemetry ward patients in univariate analyses. All secondary analyses were of limited statistical power. A subgroup analysis of patients with chest pain NYD (not yet diagnosed), angina, unstable TABLE 1. Characteristics of the Telemetry Ward Patients (n = 8,932) Demographics Sex female 39% Age mean 66 yr Admitted from: Emergency department 52% Other acute care hospital 15% Home or clinic 32% Urgent or emergent admission Average length of stay 64% 4.5 days Final diagnosis Number (%) Number of Deaths Acute myocardial infarction 993 (11%) 23 Ischemic syndromes 4,304 (48%) 16 Valvular heart disease 85 (0.9%) 2 Congestive heart failure 638 (7%) 31 Other heart disease 1,208 (14%) 12 Vascular 82 (0.9%) 4 Lung disease 980 (11%) 4 Other disease 642 (7%) 8 Total 8,932 (100%) 100 TABLE 2. Characteristics of the Telemetry Ward Cardiac Arrest Patients (n = 20) Patient characteristics Gender female 11 (55%) Age mean SD yr Known heart disease 11 (55%) Known lung disease 2 (10%) Known other disease 3 (15%) Arrest characteristics Witnessed arrest 11 (55%) Pulse present at onset 6 (30%) Respiration present at onset 2 (10%) Initial rhythm Ventricular fibrillation (VF) 5 (25%) Ventricular tachycardia (VT) 3 (15%) Asystole 3 (15%) Pulseless electrical activity 0 (0%) Bradycardia/block 2 (10%) Supraventricular 2 (10%) Respiratory arrest 1 (5%) Other or unknown 4 (20%) Arrest interventions Bystander CPR 7 (35%) Mean SD time to 1st defibrillation min Mean SD time to 1st defibrillation for VT or VF min Mean SD number defibrillation attempts Mean SD number defibrillation attempts for VT or VF Mean SD doses epinephrine Mean SD doses atropine Intubated 14 (70%) Mean SD resuscitation duration min Survival to discharge 3 (15%)

4 650 TELEMETRY Schull, Redelmeier TELEMETRY MONITORING AND CARDIAC ARREST OUTCOMES angina, or rule-out MI admission diagnoses was carried out for all telemetry ward cardiac arrests. Only one such patient had a cardiac arrest (one day after admission the patient was diagnosed as having AMI; five days later the patient experienced reinfarction, had a cardiac arrest, and died). The denominator for this subgroup can be estimated by adding this patient to the 4,304 patients discharged with noninfarction ischemic syndromes (confirmed or rule-out ); therefore, only one of 4,305 (0.02%) admitted with chest pain NYD, angina, unstable angina, or rule-out MI suffered a cardiac arrest. Of the three patients who suffered cardiac arrests but survived to discharge, two had admitting and final diagnoses of AMI with ventricular arrhythmias (four days and seven hours prior to arrest, respectively), while the third was digoxintoxic (5.0 nm/l) with arrhythmias and confusion on admission. DISCUSSION Figure 1. Cardiac arrests and electrocardiographic (ECG) monitoring among 8,932 patients on the telemetry ward. *A combination of airway interventions, cardiopulmonary resuscitation, defibrillation, and intravenous medications was used in 19 of 20 resuscitations; one involved IV medications only. MS = monitor-signaled; NMS = monitored, not monitor-signaled; UM = unmonitored. Includes one patient for whom arrest onset signal could not be determined. VT/VF = ventricular tachycardia/ventricular fibrillation. Our results suggest that among telemetry patients outside of critical care wards, cardiac arrest is a rare but survivable event. Among all 8,932 telemetry ward patients, only about one in 5,000 (0.02%) were monitor-signaled survivors of cardiac arrest, suggesting very little benefit on cardiac arrest survival from the routine use of ECG monitoring in these patients. Nonetheless, the three survivors were on monitors, and all had ventricular arrhythmias prior to their arrests; ECG monitoring may have conferred a benefit to these three patients. For the other 8,929 telemetry ward patients, we observed no benefit on cardiac arrest survival. We wonder, therefore, whether a policy of selective rather than routine ECG monitoring would have conferred the same benefit to these patients. Two of the three were known to have AMI and had suffered ventricular arrhythmias prior to arrest, while the third had arrhythmias and confusion due to digoxin toxicity, a well-known cause of arrhythmias. It is very likely that a policy of selective ECG monitoring would have included these patients. Indeed, based on a previously published protocol for selective ECG monitoring, telemetry would have been indicated in these three patients, 25 but might have been avoided for a great many other patients in our cohort. Studies that have suggested a survival benefit from the routine use of ECG monitoring outside the CCU have lacked control groups 10,11 or relied only on proxy outcomes, 12 and were conducted in an era when arrhythmia prophylaxis was common. 26 Numerous other studies have either found no survival benefit 13,15,16,27 or found it limited to patients who had required prior antiarrhythmic therapy, 14 suggesting that selective rather than routine ECG monitoring post-ccu discharge might be appropriate. Evidence of other benefits is also lacking. One study prospectively followed 467 telemetry patients and found that in 98.9% the monitoring added no significant information, and that patients who deteriorated were identified clinically without appreciable contribution from the monitors. 26 In a second study of 2,240 telemetry patients, monitor abnormalities resulted in transfer to the intensive care unit for only 0.8% of patients, and for only 7% did such abnormalities lead to any modification in treatment. 27 In a third prospective

5 ACADEMIC EMERGENCY MEDICINE June 2000, Volume 7, Number study, telemetry led to changes in management for only 11% of patients, and previously published selective telemetry guidelines would have appropriately selected all of them. 28 Among telemetry ward patients, we observed a nonsignificant trend toward improved survival of cardiac arrest when on telemetry monitors. However, we believe this represents a chance finding and it is unlikely that additional telemetry would have conferred any benefit in our cohort. Of the four unmonitored patients who had cardiac arrests, two survived their arrests, one died of a suspected thoracic aneurysm rupture, and one was in cardiogenic shock with normal sinus rhythm when the cardiac arrest team arrived. For these four patients, either telemetry monitors could not have altered the arrest outcome or telemetry alarms would not have been triggered with the onset of the cardiac arrest. In our cohort of more than 8,900 telemetry ward patients, only about one in 5,000 was a monitor-signaled survivor of cardiac arrest. This is explained by our three other observations among telemetry ward patients: cardiac arrest is rare (one in 500), it is frequently not signaled by the cardiac monitor (44%), and it has a high mortality (85%). The performance of telemetry might be improved if it were used more selectively for higher-risk patients, since, as in other studies, 17,18,27 we found a broad variety of discharge diagnoses among telemetry ward patients. The largest group of telemetry ward patients (48%) were those with chest pain syndromes other than AMI, among whom we estimate a cardiac arrest rate of only 0.02%. Indeed, this likely represents an overestimate, since this subgroup s denominator was defined by discharge diagnosis, and some patients initially admitted with angina or rule-out MI would ultimately have other final diagnoses, thereby reducing the denominator and artificially inflating the rate. More selective use of telemetry based on admitting diagnosis is supported by prospective studies as well. Arrhythmias were rarely identified in a study of telemetry patients admitted with chest pain and normal or nonspecific cardiograms, 29 and another study suggested that telemetry is less helpful for chest pain than it is for arrhythmia or syncope. 27 Reducing telemetry use among patients with noninfarction chest pain syndromes might offer the greatest potential for safe and selective monitoring. LIMITATIONS AND FUTURE QUESTIONS Given the retrospective study design, some missing data were encountered. Arrest records made in haste during a cardiac arrest are not always complete, and double data entry was not done. Data regarding all telemetry ward patients were obtained from computerized medical records, which may contain coding errors. However, the impact of such errors was minimized by basing the primary analysis on routinely collected data such as survival to discharge and ward location. Furthermore, the charts of all telemetry ward cardiac arrest patients were reviewed to confirm ward location, interventions during the arrest, and survival. Since cardiac arrests were identified by means of arrest records supplied by the cardiology department, it is possible that some record sheets were never completed or lost. However, we believe such losses are likely to be very few, since each arrest team includes a nurse specifically tasked with record keeping, arrest forms are reviewed in the cardiology department, and mandatory reports are made to hospital committees. A second limitation is that we did not assess the possibility of benefits beyond improved cardiac arrest survival. However, the proven benefits of ECG monitoring lie in its ability to detect cardiac arrest. 2 4 Third, since not all patients on the telemetry ward are always being monitored, the potential benefit of ECG monitoring might be underestimated. However, we found that telemetry monitors were used in 95% of cardiac arrest patients on the telemetry ward, and it is estimated that, on average, 90% of all telemetry ward patients are being monitored at any given time (Tupis A, personal communication, May 1999). Finally, the absence of a dedicated monitor watcher in our telemetry unit may have impacted the detection of arrhythmias. However, only about half of ICCUs in the United States use dedicated monitor watchers, 30,31 and their use has not been shown to improve patient outcomes. 30,31 Nonetheless, the generalizability of our results will be limited to hospitals that use similar forms of monitoring in similar clinical settings. Our results and those of previous studies suggest that the routine use of ECG monitoring in telemetry units should perhaps be scrutinized. Future studies may better delineate those groups at high risk of cardiac arrest and may allow for the safe and selective use of telemetry. Future studies should also examine the routine use of ECG monitoring technology elsewhere, such as EDs and the out-of-hospital setting. CONCLUSIONS To the best of our knowledge, our study represents the largest published series of telemetry ward patients and cardiac arrest, and is the first to examine the frequency with which telemetry monitoring signals cardiac arrest onset. We do not call into question all telemetry monitoring; rather, we

6 652 TELEMETRY Schull, Redelmeier TELEMETRY MONITORING AND CARDIAC ARREST OUTCOMES suggest that a more selective approach to continuous ECG monitoring might be safe. However, prospective studies are required to better elucidate which patient groups would benefit most from continuous ECG monitoring. Such a policy could result in a lessening of patient backlogs in units that now routinely admit to telemetry beds, such as EDs and CCUs. A selective approach to continuous ECG monitoring could be as beneficial to patients, and significantly more efficient for hospitals, thereby maximizing the utility of an important technology. The authors thank Dr. C. Joyner, Department of Cardiology, Sunnybrook and Women s College Health Sciences Centre, for his help in obtaining data necessary for the completion of this work. References 1. Statistics Canada. The Leading Causes of Death at Different Ages Canada Ottawa: Statistics Canada, Jul Lown B, Fakhro AM, Hood WB, Thorn GW. The coronary care unit. New perspectives and directions. JAMA. 1967; 199: Killip T, Kimball JT. A survey of the coronary care unit: concepts and results. Prog Cardiovasc Dis. 1968; 11: Day HW. Effectiveness of an intensive coronary care area. Am J Cardiol. 1965; 15: Spann JF, Moellering RC, Haber E, Wheeler EO. Arrhythmias in acute myocardial infarction. A study utilizing an electrocardiographic monitor for automatic detection and recording of arrhythmias. N Engl J Med. 1964; 271: Julian DG, Valentine PA, Miller GG. Disturbances of rate, rhythm and conduction in acute myocardial infarction. Am J Med. 1964; 37: Stock E, Goble A, Sloman G. Assessment of arrhythmias in myocardial infarction. BMJ. 1967; 2: Weinberg SL. The current status of instrumentation systems for the coronary care unit. Prog Cardiovasc Dis. 1968; 11: Klassen GA, Broadhurst C, Peretz DI, Johnson AL. Cardiac resuscitation in 126 medical patients using external cardiac massage. Lancet. 1963; 1: Grace WJ. Acute myocardial infarction: the course of the illness following discharge from the coronary care unit. A description of the intermediate coronary care unit. Chest. 1971; 59: Frieden J, Cooper JA. The role of the intermediate cardiac care unit. JAMA. 1976; 235: Gorfinkel HJ, Kercher L, Lindsay J. Electrocardiographic radiotelemetry in the early recuperative period of acute myocardial infarction. Its value in detecting arrhythmias in the setting of a progressive coronary care unit. Chest 1976; 69: Hubner PJ, Goldberg MJ, Lawson CW. Value of routine cardiac monitoring in the management of acute myocardial infarction outside a coronary care unit. BMJ. 1969; 1: Graboys TB. In-hospital sudden death after coronary care unit discharge. Arch Intern Med. 1975; 135: Weinberg SL. Intermediate coronary care. Observations on the validity of the concept. Chest. 1978; 73: Leak D, Eydt JN. An assessment of intermediate coronary care. Arch Intern Med. 1978; 138: Weinberg SL. The intermediate coronary care unit. A promise unfulfilled [editorial]. Arch Intern Med. 1982; 142: Lipskis DJ, Dannehl KN, Silverman ME. Value of radiotelemetry in a community hospital. Am J Cardiol. 1984; 53: Grace WJ. The mobile coronary care unit and the intermediate coronary care unit in the total systems approach to coronary care. Chest. 1970; 58: Turkie W, Brown AK. The use of radiotelemetry after discharge from the coronary care unit. J R Coll Physicians Surg Lond. 1990; 24: Kollek D. Overwhelmed in emergency: examining ER status at Hamilton Civil Hospitals. Ont Med Rev. 1990; 57: Ontario Hospital Association. Region 3 [Metropolitan Toronto] Emergency Services Working Group: Final Report. Toronto: Ontario Hospital Association, Apr 1998, p Cummins RO, Chamberlain D, Hazinski MF, et al. Recommended guidelines for reviewing, reporting, and conducting research on in-hospital resuscitation: the in-hospital Utstein style. Acad Emerg Med. 1997; 4: Moore DS, McCabe GP (eds). Introduction to the Practice of Statistics, Second Edition. New York: Freeman, 1993, pp Emergency Cardiac Care Committee. Recommended guidelines for in-hospital cardiac monitoring of adults for detection of arrhythmia. J Am Coll Cardiol. 1991; 18: Estrada CA, Prasad NK, Rosman HS, Young MJ. Outcomes of patients hospitalized to a telemetry unit. Am J Cardiol. 1994; 74: Estrada CA, Rosman HS, Prasad NK, et al. Role of telemetry monitoring in the non-intensive care unit. Am J Cardiol. 1995; 76: Sivaram CA, Summers JH, Ahmed N. Telemetry outside critical care units: patterns of utilization and influence on management decisions. Clin Cardiol. 1998; 21: Hollander JE, Valentine SM, McCuskey CF, et al. Are monitored telemetry beds necessary for patients with non-traumatic chest pain and normal or non-specific electrocardiograms? Am J Cardiol. 1997; 79: Funk M, Parkosewich JA, Johnson CR, Stukshis I. Effect of a dedicated monitor watcher on patients outcomes. Am J Crit Care. 1997; 6: Stukshis I, Funk M, Johnson CR, Parkosewich JA. Accuracy of detection of clinically important dysrhythmias with and without a dedicated monitor watcher. Am J Crit Care. 1997; 6: