Audibility of Patient Clinical Alarms to Hospital Nursing Personnel

Transcription

1 MILITARY MEDICINE, 171, 4:306, 2006 Audibility of Patient Clinical Alarms to Hospital Nursing Personnel Guarantor: Josef Sobieraj, PhD Contributors: Josef Sobieraj, PhD*; Claudia Ortega, BS ; COL Iris West, ANC ; Leo Voepel, MBA ; MAJ Steven Battle, MC FS #; SGT DaShawn Robinson, MC USA The audibility of patient clinical alarms to nursing personnel was tested during the first shift on a medical/surgical in-patient ward at William Beaumont Army Medical Center. This study was conducted during normal hospital operations, and we tested the Joint Commission on Accreditation of Healthcare Organizations 2004 National Patient Safety Goal, Section 6b to determine whether patient alarms could sufficiently compete against environmental background noises. Patient clinical alarms were audible at distances of 95 feet with room doors open but were not sufficiently audible to hospital staff members when room doors were closed or during floor-buffing activities. This study suggests that, under these circumstances, hospitals may not meet Joint Commission on Accreditation of Healthcare Organizations 2004 National Patient Safety Goal, Section 6b. Because the audibility of patient clinical alarms directly affects patient safety, satisfaction, and quality of care, we provide recommendations for engineering controls and modifications to work routines. *Industrial Hygiene Service, William Beaumont Army Medical Center, El Paso, TX Department of Environmental Science, University of Texas at El Paso, El Paso, TX Clinical Quality Management, William Beaumont Army Medical Center, El Paso, TX Current address: U.S. Army Medical Department Activity, Fort Drum, NY Patient Safety, William Beaumont Army Medical Center, El Paso, TX Preventive Medicine Service, William Beaumont Army Medical Center, El Paso, TX #Current address: Department of Emergency Medicine, York Memorial Hospital, York, PA Presented at the 7th Annual Conference on Force Health Protection, August 9 12, 2004, Albuquerque, NM. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Army, the Department of Defense, or the U.S. government. This manuscript was received for review in November 2004 and was accepted for publication in April Reprint & Copyright by Association of Military Surgeons of U.S., Introduction he audibility of patient clinical alarms directly affects patient safety, quality of care, and patient satisfaction, as pro- T mulgated in the Joint Commission on Accreditation of Healthcare Organizations 2004 National Patient Safety Goal, Section 6b, Assure that alarms are activated with appropriate settings and are sufficiently audible with respect to distances and competing noise within the unit. 1 Manufacturers are required to design alarms that are sufficiently audible but do not pose a noise hazard, in accordance with the Association for the Advancement of Medical Instrumentation guidelines. 2 With the exception of studies on alarms during anesthetic management 3,4 and nurses responses to alarms in a neonatal intensive care unit, 5 there have been few detailed investigations of the audibility of clinical alarms in a functioning hospital environment. Those previous investigations were not designed to evaluate the 2004 National Patient Safety Goal, Section 6b. The purpose of this study was to evaluate the audibility of patient clinical alarms to nursing personnel in a medical/surgical patient ward during a typical work shift. Experimental Site The research site is located on a wing of the ninth floor medical/surgical patient ward at William Beaumont Army Medical Center (WBAMC) in El Paso, Texas. Patient rooms 9014, 9016, and 9018 were selected for this study because of their close, intermediate, and far distances, respectively, from the nurse station, and trials were conducted with patient room doors both open and closed (Figs. 1 3). High-risk patients are located in room 9014, which is visible to personnel from the nurse station, and lower-risk patients are located down the hall in rooms 9016 and Staffing of a normal shift for this combined patient ward (both wings, with a total of 52 beds) includes 7 to 10 registered nurses, 2 licensed vocational nurses, and 7 or 8 certified nursing assistants. Audibility of alarms is a function of an individual s hearing acuity, which in turn is a function of age and other factors. The ages of 35 nursing volunteers working on the ninth floor yielded mean and median ages of 38 years, SD of 10 years, minimal age of 19 years, and maximal age of 54 years. Environmental background noise levels were evaluated by using noise dosimeters to continuously record data during the 10-hour work shift from 7:00 a.m. to 5:00 p.m. Sampling was conducted for 5 consecutive days, with dosimeters located behind the nurse station and in the hallway in front of rooms 9014, 9016, and Average decibel levels varied for these spatial locations, with the nurse station and room 9014 having the highest levels (64.1 and 65.1dBA, respectively) and rooms 9016 and 9018 having lower levels (62.0 and 59.4 dba, respectively). During evaluation of the environmental background noises in space and time, only floor-buffing activities (Cougar 2000 model burnisher, Unisource, Norcross, Georgia) by the housekeeping staff generated noise levels (77 82 dba) that exceeded the environmental background levels for significant periods of time (in this case, minutes). The acoustical properties of building materials influence the audibility of alarms and environmental background noises in this experimental setting. The building materials include vinyl compound floor tiles overlying 4 to 6 inches of concrete, Structo Lite premixed perlite gypsum plaster (United States Gypsum Company, Chicago, Illinois); supported by lath plastic material on the walls of the corridors and patient rooms, wooden patient room doors with a 20-minute fire rating, and Cortega acoustical ceiling tiles (American Micro Industries, Inc., Chambersburg, Pennsylvania; noise reduction coefficient, 0.55; ceiling attenu- 306

2 Audibility of Patient Clinical Alarms 307 Fig. 1. Alarm trials in room 9014 with open and closed doors. ation class, 0.33) composed of wet-formed mineral fiber with latex paint finish. The acoustical properties of WBAMC, like most hospitals, are such that there are reverberant rooms and hallways that allow for easy cleaning but generally have low sound absorption coefficients. 6 Methods Audibility testing was conducted with an ALARIS Medical Systems volumetric infusion pump (ALARIS Medical Systems, Inc., San Diego, California), model 7130, 7,8 which is the most common patient alarm used at WBAMC. For consistency, the same ALARIS pump (serial no ) was used throughout this study; it operated at the maximal volume setting of 79.1 dba, with the alarm signal represented by a beeping with 4-second intervals indicating an occlusion downstream. Trials were conducted for 8 weeks, Monday through Friday, from 7:00 a.m. to 5:00 p.m., to test the audibility of alarms during the most active and loudest periods on this ward. For each of the six experimental variables (3 rooms with open and closed doors), we attempted at least two trials for each 1-hour interval of time within the 10-hour work shift, thus making our sample size 20 samples per variable. Because patients and personnel were constantly entering and leaving the experimental area, we assumed that the 8-week sampling period was adequate in capturing the variability within this environment. For a trial event, the dedicated ALARIS pump was brought into room 9014, 9016, or Before alarm activation, patients were briefed on the study and gave written consent for alarm activation in their room. The audibility of alarms to nursing personnel was based on the assumption that, if an alarm was audible, then nurses would enter a patient room in response to the alarm, as required by their training. To determine whether patient alarms were sufficiently audible against competing background noises in a functioning work environment (i.e., Joint Commission on Accreditation of Healthcare Organizations 2004 National Patient Safety Goal, Section 6b), we made every effort to eliminate bias in our experimental design so that nurses assumed it was an actual patient alarm. The following measures were taken for each trial. (1) Personnel activating the alarms had normal duties on the patient ward; therefore, their presence did not create suspicion among the nurses. (2) The alarm was transported to a patient room when nursing personnel were unaware of this activity (e.g., no personnel in the hallway at the time of transport). (3) Nurses responding to an alarm were asked whether their responses were biased (i.e., whether they were aware that this was a trial), and biased responses were discarded. To avoid disrupting the activities of on-duty personnel, we attempted a maximum of four trials per day, with no more than two trials in a room per day. A 3-minute response time was used in this study, on the basis of clinical management and patient environmental considerations. Although alarms in a clinical setting are frequently attributable to benign causes, such as alerting personnel to replenish the infusing fluid or to adjust a catheter kinked because of patient positioning, more serious

3 308 Audibility of Patient Clinical Alarms Fig. 2. Alarm trials in room 9016 with open and closed doors. complications can result from intravenous infusions, particularly during administration of pharmaceutical agents that may result in necrosis (e.g., calcium formulations, dopamine, vancomycin, acyclovir sodium, and multiple oncology medications 9 ). A final justification for a 3-minute limit was to avoid generating anxiety among patients and their visitors in the event of a lack of response by nursing personnel to an activated alarm. For all trial events, responders were asked to show the spatial locations on the ward where they first heard the alarm, thus providing distances in which the alarms are audible over competing environmental noises. We assumed that the 8-week period, with two trial events within each 1-hour period of the 10-hour workday, was sufficient in accounting for the variability of the spatial locations of nursing personnel on the ward at the time of their responses. In addition to measuring distances in which nursing personnel could hear alarms over competing background noises, we also used the following Likert scale to determine the relative audibility during each trial: 1, I did not hear the alarm (no responder); 2, alarm was faintly audible or I was not sure I heard the alarm but I responded anyway; 3, alarm was somewhat audible; 4, alarm was clearly audible; 5, alarm was definitely audible and too loud. Results Our trials show that activated alarms are sufficiently audible and can compete against environmental background noises when patient room doors are open. For rooms 9014, 9016, and 9018 with open doors, maximal audibility distances were 89 feet, 53 feet, and 95 feet, respectively. Deception used for each trial in this study did not allow for simultaneous determination of an alarm s audibility to all nurses on the ward (this would prime nurses to an alarm and generate bias in their responses), so it is possible that alarms were audible beyond distances of 95 feet. Alarm audibility was significantly reduced when patient room doors were closed. For rooms 9014, 9016, and 9018 with closed doors, maximal audibility distances were 5 feet, 22 feet, and 45 feet, respectively. Spatial locations of responding nurses are extremely useful in evaluating the audibility of alarms. Figure 1 shows the locations where nurses first heard an alarm in response to a trial event in room 9014, which houses relatively high-risk patients in proximity to the nurse station. In the 20 trials in room 9014 with an open door, nurses responded within 3 minutes for 16 of those trials. Figure 1 shows that, with open doors, nurses responded primarily from the nurse station but also from considerable distances in all directions. Furthermore, these alarms were generally somewhat audible to clearly audible on our Likert scale for assessing relative audibility. Only six unbiased trials were conducted in room 9014 with a closed door, because personnel became immediately suspicious. In these six trials, there were only three responses by nurses within 3 minutes, and these all occurred 5 feet from the closed door, with the alarms being only faintly audible on our Likert scale. Figure 2 shows the locations where nurses first heard an

4 Audibility of Patient Clinical Alarms 309 Fig. 3. Alarm trials in room 9018 with open and closed doors. alarm in response to a trial event in room 9016, which houses relatively lower-risk patients 30 feet down the hallway from the nurse station. Nurses responded within 3 minutes in 11 of the 21 trials. Figure 2 shows that nurses responded to alarms from considerable distances in all directions, with alarms being clearly audible and somewhat audible even at the nurse station. In the 21 trials in room 9016 with a closed door, nurses responded within 3 minutes for 5 trials. Responses occurred directly outside the door or in an adjacent patient room, with varying degrees of relative audibility. Floor-buffing activities occurred during one trial with an open door and one with a closed door, and nurses did not respond during those trials. Figure 3 shows the locations where nurses first heard an alarm in response to a trial in room 9018, which houses relatively lower-risk patients 75 feet down the hallway from the nurse station. Nurses responded within 3 minutes in 10 of the 21 trials. Figure 3 shows that nurses responded to alarms from considerable distances in all directions, with alarms being clearly audible and somewhat audible even as far away as the nurse station. Floor-buffing activities occurred during three trials with an open door, and nurses did not respond during those trials. In the 20 trials in room 9018 with a closed door, nurses responded within 3 minutes for 6 trials. Responses occurred directly outside the door (with a single outlier) for room 9016, and relative audibility was generally faintly audible. Discussion Our results indicate that the maximal volume setting of the ALARIS alarm in rooms 9014, 9016, and 9018 with open doors was sufficiently audible to nursing personnel at considerable distances ( 95 feet) along this ward, but the alarm in any room was not sufficiently audible to nursing personnel when patient doors were closed or there were floor-buffing activities. Although the focus of our study involved testing the audibility of alarms according to 2004 National Patient Safety Goal, Section 6b, an outcome of this study was that nurses can hear but do not necessarily respond in a timely manner ( 3 minutes) to patient alarms (i.e., common lack of response by nurses to alarms for rooms 9016 and 9018 with open doors). On the basis of observations during trials, the lack of response within 3 minutes can be explained, at least in part, by the following reasons: (1) simultaneous activation of multiple patient alarms, keeping nurses occupied; (2) acoustical properties of hallway construction materials, making it difficult to identify the spatial location of the alarm; (3) floor-buffing machines generating a background noise level that impairs the audibility of alarms; and (4) nursing personnel perceiving that the activated alarm is either a false alarm or a nonurgent matter and disregarding it. On the basis of this study, we offer engineering solutions and work routine adjustments for improving both the audibility of patient alarms and the response time of nurses. Perhaps the

5 310 Audibility of Patient Clinical Alarms most significant recommendation involves proposing guidelines establishing when it is safe to close the door to a patient room. Hospitals are noisy environments, with patients wanting privacy and rest, and recommendations are made for nurses to routinely close doors to reduce noise for patients. 10 We show that alarms at the maximal volume setting are relatively inaudible to nurses when patient room doors are closed; therefore, these doors should be closed only when there is little to no risk to the safety of the patient. Nurses must carefully consider the benefits and risks for patient privacy and comfort vs. patient safety. An engineering solution to the problem is the installation of a nurse call system that incorporates clinical patient alarms. Hospitals without such a system should conduct a cost-benefit analysis to determine whether this engineering solution best serves their needs. We recommend that nurses increase the frequency of patient room visitations during hospital floor-buffing activities. Patient alarms are not sufficiently audible and cannot compete with background noises during floor buffing, thus violating 2004 National Patient Safety Goal, Section 6b, for relatively short periods of time (10 20 minutes). Nurses should be informed of this finding and should consider adjusting their routines accordingly. In addition, we recommend that nurses use the maximal volume setting on patient clinical alarms during the relatively noisy first shift. Finally, we recommend that nurses assume that all activated patient alarms are real (as opposed to false) events and vigilantly respond to activated alarms. We recognize that this may be the most challenging recommendation to implement, because of the vast number of alarm types and noises, the large number of false alarms, and the numerous background environmental noises that constantly compete for attention. Acknowledgments This study was possible only through the participation and cooperation of the ninth floor patients and staff members, with special thanks to MAJ Gollasch, SFC Waltenburg, and SGT Taylor. This study was approved and funded by the institutional review board at WBAMC (protocol 04/13). References 1. Anonymous: Clinical alarms goal remains intact: Joint Commission expands 2004 goals. Biomed Instrum Technol 2003; 37: Association for the Advancement of Medical Instrumentation: ID26, Section Standard, Medical Electrical Equipment, Part 2: Particular Requirements for the Safety of Infusion Pumps and Controllers, Ed 2. Arlington, VA, Association for the Advancement of Medical Instrumentation, Kestin IG, Miller BR, Lockhart CH: Auditory alarms during anesthesia monitoring. Anesthesiology 1988; 69: Morris RW, Montano SR: Response times to visual and auditory alarms during anaesthesia. Anaesth Intensive Care 1997; 24: Sabar R, Zmora E: Nurses response to alarms from monitoring systems in NICU. Pediatr Res 1997; 41: Grumet GW: Pandemonium in the modern hospital. N Engl J Med 1993; 328: ALARIS Medical Systems: IVAC Vital Check 4400 Series: Directions for Use. San Diego, CA, ALARIS Medical Systems, ALARIS Medical Systems: IVAC Signature Edition Technical Service Manual. San Diego, CA, ALARIS Medical Systems, Brown S: Complications with the use of venous access devices. US Pharm 1998; 23(8). Available at newlook/ files/feat/acf2ff9.cfm&pub_id 8&article id Cmiel CA, Karr DM, Gasser DM, Oliphant LM, Neveasu AJ: Noise control: a nursing team s approach to sleep promotion: respecting the silence creates a healthier environment for your patients. Am J Nurs 2004; 104: The Federal Medical Chiefs honoring Retiring AMSUS Executive Director RADM Fred Sanford, along with Board of Managers Chairman, Col Ben Daughtry, (far right)