1186
Bullard et al.
d
WIRELESS TECHNOLOGY
Supporting Clinical Practice at the Bedside Using Wireless Technology Michael J. Bullard, MD, David P. Meurer, RN, BScN, Ian Colman, MSc, Brian R. Holroyd, MD, Brian H. Rowe, MD, MSc Abstract Objectives: Despite studies that show improvements in both standards of care and outcomes with the judicious application of clinical practice guidelines (CPGs), their clinical utilization remains low. This randomized controlled trial examined the use of a wirelessly networked mobile computer (MC) by physicians at the bedside with access to an emergency department information system, decision support tools (DSTs), and other software options. Methods: Each of ten volunteer emergency physicians was randomized using a matched-pair design to work five shifts in standard fashion (desktop computer [DC] access) and five shifts with a wirelessly networked MC. Work pattern issues and electronic CPG/DST use were compared using end-ofshift satisfaction questionnaires and review of a CPG/DST database. Repeated-measures analysis of variance was used to examine between-shift differences. Results: A total of 100 eight-hour shifts were evaluated; 99% compliance with postshift questionnaires was achieved. Using a seven-point Likert scale (MC values first), MCs were rated as being as
fast (5.04 vs. 4.54; p = 0.13) and convenient (5.08 vs. 4.14; p = 0.07) as DCs. Overall, physicians rated MCs to be less efficient (3.18 vs. 4.30; p = 0.02) but encouraged more frequent use of DSTs (4.10 vs. 3.47; p = 0.03) without impacting doctor–patient communication (2.78 vs. 2.96; p = 0.51). During the study period, physician use of an intranet Web application (eCPG) was more frequent during shifts assigned to the MC when compared with the DC (eCPG uses/shift, 3.6 vs. 2.0; p = 0.033). Conclusions: The MC technology permitted physicians to access information at the bedside and increased the use of CPG/DST tools. According to physicians, patients appeared to accept their use of information technology to assist in decision making. Development of improved computer technology may address the major limitation of MC portability. Key words: clinical practice guidelines; computers; decision support; emergency. ACADEMIC EMERGENCY MEDICINE 2004; 11:1186–1192.
Medical knowledge as well as diagnostic and therapeutic options continue to grow at an exponential rate. No matter how conscientious the practitioner, he or she will be faced with the daunting task of recalling new and established management strategies and clinical practice guidelines (CPGs) that form the
standards by which physicians are judged. This is especially true in a broad field such as emergency medicine, in which a wide variety of diseases first present and an equally large number of decision support tools (DSTs) are available. Medical errors are a recognized problem, and many are predictable and attributable to characteristics of our cognitive function.1 Systems that are designed to simplify or automate tasks, especially using computer technologies, hold tremendous promise in guiding clinicians with patient care and increasing the safety of this care in the frenetic emergency department (ED) setting. Initial attempts to computerize EDs have often met with only modest success, in part due to technical limitations.2 Today, the power and capability of existing technology have dramatically improved; however, surprisingly little has been incorporated into clinical medical practice,3 despite the fact that the majority of systematic reviews of decision support systems have demonstrated improved compliance with clinical recommendations but limited improvement in clinical outcomes.4–6 Moreover, one study demonstrated a decrease in adverse drug reactions, inappropriate drug prescribing, and decreased drug costs through the integration of a decision support system with electronic patient data.7
From the Division of Emergency Medicine (MJB, DPM, IC, BRH, BHR), Department of Public Health Sciences (BHR), University of Alberta, Edmonton, Alberta, Canada; and Capital Health (MJB, BRH, BHR), Edmonton, Alberta, Canada. Presented at the Canadian Association of Emergency Physicians annual meeting, Winnipeg, Manitoba, Canada, June 2003, and the SAEM annual meeting, Boston, MA, May 2003. Supported by the Alberta Medical Association Clinical Practice Guideline Program, the Alberta Medical Association/Medical Services Branch Health Innovation Fund Grant Program, and the Division of Emergency Medicine of the University of Alberta. Mr. Colman was supported in part by the Canadian Association of Emergency Physicians Research Consortium, and Dr. Rowe was supported by the Canadian Institute of Health Research as a Canada Research Chair. Dr. Bullard, Mr. Meurer, Dr. Holroyd, and Dr. Rowe are codevelopers of the eCPG and eTRIAGE programs. Address for correspondence and reprints: Michael J. Bullard, MD, Division of Emergency Medicine, University of Alberta, Room 1G1.58 WMC, 8440-112 Street, Edmonton, Alberta, Canada T6G 2B7. Fax: 780-407-3314; e-mail:
[email protected]. doi:10.1197/j.aem.2004.08.013
ACAD EMERG MED
d
November 2004, Vol. 11, No. 11
d
www.aemj.org
Clinical practice guidelines have been developed by research agencies, medical associations, and specialty groups and, when followed, have been shown to improve clinically important outcomes.7–11 Despite these facts, compliance with guidelines is frequently found to be low, with wide variation in practice.7,12–14 To be effective, many believe that DSTs and CPGs need to be accessible at the point of care and in a format that enhances workflow.15,16 Efforts to date have rarely included the emergency setting, where medication mistakes often occur due to transcription error or knowledge-based error.17 Over a two-year period, the emergency medicine research group at the University of Alberta developed a large number of electronic support tools accessible through an intranet Web application (eCPG, a copyrighted Web application). This was URL-linked to an existing ED information system (EDIS)—a ‘‘triage and tracking’’ application for easy access by clinical users. Included were DSTs that ranged in sophistication from simple patient education sheets to best-practice tools, drug dose–calculating interactive order sets, and assessment and management tools based on decision rules and CPGs that generated an editable order set based on information input. Point-of-care access was required by clinicians in order to take advantage of the sophisticated clinical assessment DSTs that replaced charting, helped risk-stratify the patient, and generated an order set based on responses to a variety of preset clinical questions. Through the use of wireless technology and mobile laptop computing, these DSTs were made available to the study participants anywhere in the ED. The primary hypothesis was that by having networked computer access at the point of care, physician utilization of CPGs, order sets, discharge instructions, and other electronic DSTs would increase. The secondary hypothesis was that physician satisfaction would be increased by having access to a computer configured to support work practice.
METHODS Study Design. This prospective, unblinded study of emergency physicians compared work using a wirelessly networked mobile computer (MC) with work using an existing central networked desktop computer (DC) on a shift-by-shift basis. Concealed, block randomization was used to allocate the work mode (MC or DC) for each physician’s ten assigned shifts. Utilizations of the various DSTs within eCPG by each study physician were compared according to shift assignment (MC vs. DC). Physician satisfaction was evaluated through the use of end-of-shift questionnaires. This study was approved by the Health Research Ethics Board of the University of Alberta. Physicians were not expected to change their practice in any way except for having the ability to make use of more
1187 interactive clinical tools at the bedside (those complex charting tools requiring responses to large numbers of detailed clinical questions to help risk-stratify patients and generate integrated rules-based orders) during their MC shifts. Each physician provided written informed consent and was free to utilize the additional functional capabilities of the MC based on his own comfort level and perception as to whether or not it increased his efficiency. All clinical information was submitted to a secure research database whenever completed DSTs were printed. The printed forms then became part of the permanent patient record. Study Setting and Population. This study was conducted in an academic tertiary care ED with 75,000 annual visits. The ED had a two-year ongoing experience with an EDIS to triage, track, assign, and discharge patients with 41 nurse/physician-accessible DCs. The study was conducted from June 24 to September 30, 2002. Volunteer emergency physicians were recruited from the full-time ED staff. Before the study, the wireless technology was tested to ensure adequate signal throughout the ED, to ensure adequate application speed, and to identify and eliminate any interference with signal transfer. Study Protocol. Patients were assigned and managed as usual in the study ED; however, emergency physicians were encouraged to use the preexisting DSTs or any additional applications they believed would assist their ED work activity. For the study day, afternoon, and evening shifts, only the noncritical ED areas were selected for randomization. Questionnaires were completed by the physicians at the end of each shift, and all study physicians then completed a final exit questionnaire. The unit of randomization and analysis was the physician. Each of the ten emergency physicians worked five randomly allocated shifts each with the MC and the DC in the noncritical ED areas. This equaled ten shifts per physician with an average of 15–25 patients seen per shift (total patient number, 1,500–2,500). This study was designed to detect a onepoint difference in satisfaction (seven-point Likert scale; SD = 1.0) between MC and DC shifts with a = 0.05 and power of 0.20 assuming that the withinphysician correlation was 0.5.18 Materials. To achieve bedside computing capabilities, an IBM personal computer (IBM Corp., White Plains, NY) equipped with the EDIS, Internet Explorer 5 (Microsoft Corp., Redmond, WA), Adobe Acrobat Reader 5 (Adobe Systems, Inc., San Jose, CA), and the Agfa Web server (electronic radiograph retrieval) (Agfa-Gevaert Group, Mortsel, Belgium) was secured onto a mobile stand that the study physicians used during their selected shifts in the University of Alberta Hospital ED (Figure 1). The stand was a rugged six-wheeled pole with an adjustable arm
1188
Bullard et al.
d
WIRELESS TECHNOLOGY
when available only on a DC. Before the start of the study, all participating emergency physicians were briefed on the use of the MC in a one-on-one training session with the project coordinator (DPM). The signon password was common for all study physicians, as was the postshift questionnaire, which was completed online and submitted into a secure database. Measurements. All computer entry information was documented in real time and sent to the EDIS database or the forms database (captured and maintained without unique patient identifiers). Prestudy and intrastudy electronic form utilization by all emergency physicians within the department was used to determine whether the study had any overall impact on electronic tool utilization. Postshift questionnaire responses were entered electronically by the physician using a unique identity code. An exit questionnaire assessed study physician satisfaction with the EDIS in general and enhanced functionalities within the various electronic tools to determine whether these enhancements were perceived as beneficial. A researcher, blinded to physician identity, entered the exit questionnaire data into the computer using each study participant’s coded unique identifier. Figure 1. A wirelessly networked personal computer secured on an adjustable arm on a stable wheeled pole with an external fuel cell in one basket and operated by the principal investigator (MJB).
and computer platform, which allowed the computer to be raised or lowered and tilted to the optimal angle for each user. Through the use of two Cisco 350 series air ports (Cisco Systems, Inc., San Jose, CA), the area was wirelessly networked. A 12- to 16-hour external electric fuel cell (Electrovaya, Toronto, Ontario) provided the necessary additional energy to power the MC for a complete eight-hour shift. The fuel cell is lightweight (2.4 lb), 8.75 by 11.75 inches in size, and 3/ 8 inch thick. The electronic DSTs were all available on an intranet Web site URL linked into the EDIS application for user convenience (accessible to both MC and DC users). The electronic order sets, CPGs, charting templates, and discharge forms were created using Adobe Acrobat and underwent a series of functional modifications designed to enhance user friendliness and optimize utility while minimizing the amount of computer interaction required based on user feedback. Physicians were asked to rate some of the enhanced functionalities of the Adobe Acrobat forms. Training and Quality Control. All participating physicians had accumulated more than two years of experience with the entire library of electronic CPGs, order sets, discharge instructions, and other PDF tools available on the DCs. Some of these tools were new to the clinicians because they were designed to be interactive at the bedside and were not as effective
Data Analysis Statistical Analysis of eCPG/DST Use. Utilizations of DSTs (e.g., charting tools, CPGs, order sets, discharge instructions, and outpatient referral forms) for each individual were compared between shifts (while assigned to the MC vs. the DC only). Statistical Analysis of End-of-shift Questionnaires. Each physician completed an end-of-shift questionnaire for each of the ten shifts (five with the MC and five with the DC). To account for both the variation between physicians and the variation within the answers for each physician, a repeated-measures analysis of variance was used. The repeated-measures analysis of variance calculated an F statistic, which used df of 1 (2 computer types 2 1) and 9 (10 physicians 2 1), and compared the mean responses between the MC shifts and the DC shifts. Differences were considered significant if the associated p-value was less than 0.05. Adjusted marginal means (and 95% confidence intervals), which account for the between-physician variability, were calculated. These analyses were conducted using SPSS version 11.5 (SPSS Inc., Chicago, IL) using the generalized linear model (GLM) procedure programmed to calculate effects for the work mode, the shift number, and the interaction between these factors. Statistical Analysis of End-of-study Questionnaire. Each physician completed one end-of-study questionnaire. For questions in which physicians rated a
ACAD EMERG MED
d
November 2004, Vol. 11, No. 11
d
1189
www.aemj.org
response on a scale, means were calculated and reported with a 95% confidence interval around the mean. For questions in which physicians noted a yesor-no response, counts and percentages of yes responses are reported.
RESULTS Physician Sample. More than 75% of the study-site emergency physicians expressed an interest in being involved in the study. All volunteer physicians were full-time clinicians, and all were male; six (60%) had certification in emergency medicine from the College of Family Physicians of Canada, and four (40%) had specialty (American Board of Emergency Medicine or Fellow of the Royal College of Physicians of Canada) training (more than four years of program training). The median age of the group involved was 35 years (interquartile range, 32–37). This breakdown is representative of the makeup of the ED group at the study site. Physician Satisfaction. There was a 99% completion rate of the end-of-shift questionnaires. Table 1 outlines the comparative results of these questionnaires. Despite apparent satisfaction with computing speed, configurability, and availability of the MC, when asked about overall efficiency, physicians believed that use of the MC slowed their work (mean 3.2 [SD 0.3] vs. 4.3 [SD 0.1]; p = 0.02). Patients did not make negative comments about the MC, and it did not prove an impediment to patient or staff communication. The other question that generated a significant difference related to eCPG use. Physicians believed that they used eCPGs more when assigned the MC (mean 4.1 [SD 0.2] vs. 3.5 [SD 0.2]; p = 0.03). The database confirmed this impression by demonstrating
more frequent eCPG use during the MC shifts than the DC shifts (mean 3.6 [SD 1.7] vs. 2.0 [SD 1.9]; p = 0.033). Computer Speed and Sensibility. Responses to this question demonstrated that there was no statistically significant difference between MCs and DCs in perceived computing speed (p = 0.13). The physicians were asked whether the computer(s) were conveniently configured to their work practice. Analysis of these data demonstrated that although there was a trend toward physicians’ finding the MC more conveniently configured than the DC, this result was not statistically significant (p = 0.07). Shift Computer Use. A series of questions asked the users to identify which other applications they used during MC shifts. In addition to the EDIS and the eCPG Web application, 80% used the Agfa Web viewer to access electronic radiographs and to show patients their radiographs, 80% used a Web browser for knowledge access and information searches, 20% used e-mail to keep up with their correspondence, and 20% kept shift logs using Microsoft Word (Microsoft Corp.) to give a report to the study coordinator. End-of-study Satisfaction. The exit questionnaire focused on user satisfaction with the EDIS in general, which electronic tool enhancements were most helpful, and which further enhancements are desired by current users. All physicians completed these questionnaires. While there was strong agreement that the size of the cart had a negative impact on the use of the MC, there was an equally strong feeling that the EDIS should be maintained and enhanced (Table 2). All respondents wished to be able to both order and retrieve their laboratory results and radiographs
TABLE 1. Questions and Responses Regarding the Impact of the Mobile Computer on Clinical Shift Performance Mobile Computer Question Computer speed was fast Computer configuration was convenient Computer availability Impact on efficiency Increased use of clinical practice guidelines Computer saved time Computer reduced communication with staff Computer reduced communication with patients Patients made negative comments about computer Accessibility in three locations
Desktop Computer
Adjusted Mean
Adjusted 95% CI
Adjusted Mean
Adjusted 95% CI
p-value
1 = unacceptable; 7 = excellent
5.0
4.3, 5.8
4.5
4.0, 5.1
0.13
1 = unacceptable; 7 = excellent 1 = unacceptable; 7 = excellent 1 = much slower; 7 = much faster
5.1 5.2 3.2
4.4, 5.8 4.5, 6.0 2.6, 3.8
4.1 4.5 4.3
3.5, 4.8 3.9, 5.1 4.0, 4.6
0.07 0.13 0.015
1 = strongly disagree; 7 = strongly agree 1 = strongly disagree; 7 = strongly agree
4.1 3.1
3.6, 4.6 2.3, 3.9
3.5 4.2
2.9, 4.0 3.6, 4.7
0.03 0.05
1 = strongly disagree; 7 = strongly agree
2.9
2.1, 3.7
3.1
2.4, 3.7
0.57
1 = strongly disagree; 7 = strongly agree
2.8
2.0, 3.6
3.0
2.4, 3.6
0.51
1 = strongly disagree; 7 = strongly agree 1 = unsatisfied; 7 = extremely satisfied
2.4 4.4
1.6, 3.2 3.7, 5.0
2.6 3.9
1.9, 3.3 3.2, 4.7
0.42 0.20
Scale
1190
Bullard et al.
TABLE 2. Questions and Responses Regarding the Mobile Computer and the Current Emergency Department Information System
d
d
Question I found the mobile computer made me more efficient.* Number of shifts with mobile computer before proficient. The computer had a negative impact on the physician–patient relationship.* The size of cart was a major impediment to using the mobile computer.* What is your opinion/expectation regarding the future of an EDIS?y I would get rid of the EDIS in our department altogether.* Maintain the existing EDIS functionality, do not worry about enhancements.* Want the ability to do my charting and ordering electronically.*
Mean (95% CI) d
3.30 (2.33, 4.27) 2.80 (2.04, 3.56)
2.60 (1.87, 3.33) d
6.30 (5.13, 7.00) d
5.67 (5.20, 6.13) 1.80 (1.31, 2.29) 2.30 (1.79, 2.81)
online. They all wanted to be able to send their diagnostic imaging interpretation electronically to the radiologist, fax or transmit referrals or visit summaries directly to consultants and primary care physicians electronically, and e-mail their patient list at the end of each shift to themselves using a secure intranet. There was also 100% approval for the development of files of special-needs patients by category with their individualized treatment plans (patients with hemophilia, patients undergoing transplantation, and so on). Finally, 80% wanted full bedside charting capability combining a templated style with voice recognition–integrated additional notes. Satisfaction with eCPG Application Enhancements. The responses to questions regarding the utility of specific enhancements built into the DSTs using a seven-point Likert scale are shown in Table 3. The majority of innovative enhancements within the electronic DSTs were considered helpful or extremely helpful by the participants. These included the following:
d
d
WIRELESS TECHNOLOGY
The ability to clinically override a decision rule with the opportunity to comment on the rationale for the override Automatic dosing calculations for weight-based medications Preset common drug doses and intravenous rates for common problems Order sets created automatically based on predetermined response patterns in the clinical assessment tools (physician able to change orders before signing them off) Reset buttons for specific sections to protect other data entry Specific print buttons for initial visit and repeat visit for the same problem (e.g., cellulitis) to prevent unnecessary reprinting of patient instructions and the ability to select optional reassessment times from buttons on the order sheet
5.50 (4.52, 6.48)
EDIS = emergency department information system. *1 = strongly disagree, 7 = strongly agree. y1 = doomed to failure; 7 = will be paperless in five years.
d
d
d
Automatic default buttons, which check common groups of orders or clinical information with a single button A ‘‘not applicable’’ button to allow the physician to avoid reviewing unnecessary items quickly Interactive decision rules that help calculate patient pretest probability (Well’s criteria for deep venous thrombosis, Pneumonia Severity Index for community-acquired pneumonia, and so on) and risk scores
DISCUSSION This study represents a prospective and detailed attempt to evaluate both the functional capability and the suitability of an MC to support the use of electronic DSTs at the bedside in an ED setting. This study clearly indicated that current wireless technology is capable of supporting networked computer applications at the point of care in this setting. In addition, this goal was accomplished without any appreciable loss of computing speed when compared TABLE 3. Questions Regarding the Application Enhancements That Are Useful/Not Useful Question The ability to default items/orders at the push of a button Single-button skip function through ‘‘not applicable’’ sections The interactive decision rules that help stratify risk The ability to clinically ‘‘override’’ decision support suggestions The automatic dosing calculations for weight-based medications The preset intravenous rates and drug doses The gender- and age-driven automatic template modifications The pop-up messages The allergies box with the top 20 allergies reported The reset buttons for isolated sections of the forms The templates that build your orders based on the patterned responses The ability to select a reassessment time convenient for the physician and/or the patient Different print button options on the forms The page-up/page-down button options 1 = extremely unhelpful; 7 = extremely helpful.
Mean (95% CI) 6.00 (5.59, 6.41) 5.88 (5.19, 6.56) 5.80 (5.31, 6.29) 5.33 (4.68, 5.99) 6.43 (5.71, 7.00) 6.00 (5.35, 6.65) 5.43 (4.85, 6.01) 4.75 (4.13, 5.36) 4.00 (2.86, 5.13) 6.25 (5.31, 7.00) 5.38 (4.55, 6.20)
5.86 (5.35, 6.37) 5.25 (4.63, 5.86) 4.33 (3.68, 4.99)
ACAD EMERG MED
d
November 2004, Vol. 11, No. 11
d
1191
www.aemj.org
with hardwired DCs. With a 100 Base-T Ethernet line linked into the wireless transmitter and signals sent out at 28 Mb/s, prestudy testing showed no performance drop-off on the MC in terms of any application performance. Moreover, the system was well received by emergency physicians and their patients. A key component of the successful deployment was the availability of a portable external power source. While advertised to provide 12–16 hours of continuous power, variable user computing patterns affected the usage rate of the external power supply (range, 7– 11 hours). For users continuously using four applications (e.g., EDIS, Agfa, Internet, and one other application) during their MC shift, a backup fuel cell (two external fuel cells were with the computer at all times) was occasionally required; however, most of the physicians were able to complete a shift using a single external fuel cell. Recharging a discharged fuel cell took approximately eight hours, so if multiple MC users were planned from shift to shift, an adequate supply of fuel cells or an equally effective alternate power supply would be required. Study participants had two major concerns throughout the study that need to be addressed before broader use of mobile computing in the ED can be realized. First, the size of the MC and the additional effort required to maneuver it for an eight-hour shift were universally expressed. The size was particularly limiting in the fast-track area of the ED, where the space between beds is limited and congestion is common. Second, due to the limited number of DSTs deployed on the intranet at that time, most patient charting remained paper-based. This dilemma limited the need for the computer at the bedside and will change as template charting and increased DSTs are developed. Presently, the completed clinical assessment tools are applicable to less than 10% of the ED patients, and the electronic order sets, patient discharge instructions, and referral forms are all equally convenient on the DCs. While additional eCPG tools are in the development stage, until electronic charting becomes the primary record, it will be hard to encourage broad compliance. An unexpected finding in this study was the very strong support shown by the study participants for the present EDIS and their desire and expectations that enhancements will materialize. Overall, the volunteer physicians appear to believe that a full range of clinical support tools should eventually be successfully introduced within the EDIS. All of the eCPG tool functions that were designed to simplify data input requirements by the physician, while at the same time supporting best-evidence standards of care, were viewed as helpful or extremely helpful. The three enhancements that received only moderate endorsement were pop-up messages (providing teaching information or alerts), a drop-box list of common allergies, and separate page-up and page-down but-
tons. These options appeared to be less popular, largely because they did not provide tangible clinical efficiencies to experienced clinicians. Parenthetically, many junior learners find the pop-up messages valuable teaching tools in some of the DSTs.
LIMITATIONS There are several potential limitations of this study. First, and most importantly, this was a single-site study involving physicians who had two years of experience in an ED utilizing a computerized triage, tracking, and disposition system. While the results documented here are encouraging, they need to be replicated in a variety of settings, including with a less computer-literate physician group. Second, we were unable to demonstrate that the use of computer technology improved patient outcomes; however, a study to accomplish this goal would require larger samples of both physicians and patients. Our goal here was to explore the acceptance of the MC concept in a busy ED environment. Finally, as with all computer technology, hardware and software advancements tend to outstrip the pace of research. As such, the computing and battery-power capabilities may already be lighter, smaller, faster, and more robust. Newer tablet PCs are lighter and more portable and may be better designed to meet the needs in the ED. While physically more appealing, they pose different yet real concerns (e.g., theft, damage from dropping, and memory limitations). When to take the plunge in moving to this kind of technology is always a calculated gamble. Knowing that technology is now robust enough to meet the needs of frontline caregivers in an ED environment, the need to address the size and portability issues expressed by our study subjects becomes more urgent. While some respondents suggested palm technology as a portable option, there are problems with this alternative. For example, the technology is not robust enough for continuous networking and the screen size is too small to support the needs of many users. The new wave of notepad technology may be able to fill the niche, and plans are under way to pilot models that fit our specifications in this center. The major technology hurdles to overcome will be battery power (recognizing we needed a 2.4-lb external fuel cell to support the MC for this study) and security for a small portable device to prevent ‘‘computer self-mobility’’ (theft) from the department if the user sets it down. An even more promising option would appear to be thin client technology, where the hard drive resides on a desktop or central server while the user carries and interacts with a wirelessly linked interactive monitor. This should help to decrease weight, decrease battery needs, and prevent the loss of private information (no data stored in the device).
1192
Bullard et al.
From the standpoint of DST development, work is continuing locally, while at the same time efforts are under way to forge partnerships and collaborations nationally and internationally. The goal is to build consensus and seek assistance in developing more evidence-based clinical support tools.
CONCLUSIONS In this emergency setting, wireless technology effectively supported mobile computing and was well received by the physician group. The system permitted emergency physicians to rapidly access information at the bedside and use DSTs more frequently. Patients generally accepted the physicians’ use of information technology to assist in decision making. The major limitations were the size and inconvenience of maneuvering an MC around for an entire shift. Ongoing research into best-fit technology is required to strike the optimal balance between size, portability, and functionality. The authors thank the following emergency medicine volunteers at the University of Alberta Hospital for their cooperation and input in this project (listed alphabetically): Barry Diner, Jeffrey Franc-Law, Tim Graham, David Hoshizaki, Richard Ibach, Shawn Janes, Malcolm Long, Curtis Rabuka, and Roger Yao. The authors also thank Sandra Blitz for reviewing the statistical analysis and the Capital Health Authority for providing the server space and technical support necessary to launch this project.
References 1. Nolan TW. System changes to improve patient safety. BMJ. 2000; 320:771–3. 2. Cordell WH. Information technologies for emergency medicine. Acad Emerg Med. 1994; 1:194–7. 3. Smith MS, Feied CF. The next generation emergency department. Ann Emerg Med. 1998; 32:65–74. 4. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computerbased clinical decision support systems on physician performance and patient outcomes: a systematic review. JAMA. 1998; 280:1339–46.
d
WIRELESS TECHNOLOGY
5. Shiffman RN, Liaw Y, Brandt CA, Corb GJ. Computer-based guideline implementation systems: a systematic review of functionality and effectiveness. J Am Med Inform Assoc. 1999; 6:104–14. 6. Balas EA, Austin SM, Mitchell JA, Ewigman BG, Bopp KD, Brown GD. The clinical value of computerized information services. A review of 98 randomized clinical trials. Arch Fam Med. 1996; 5:271–8. 7. Evans RS, Prestotnik SL, Classen DC, et al. A computerassisted management program for antibiotics and other antiinfective agents. N Engl J Med. 1998; 338:232–8. 8. Tierney WM, Overhage JM, Takesue BY, Harris LE, Murray MD, Vargo DL. Computerizing guidelines to improve care and patient outcomes: the example of heart failure. J Am Med Inform Assoc. 1995; 2:316–22. 9. McDonald CJ, Wilson GA, McCabe GP Jr. Physician response to computer reminders. JAMA. 1980; 244:1579–81. 10. East TD, Bohm SH, Wallace CJ, et al. A successful computerized protocol for clinical management of pressure control inverse ratio ventilation in ARDS patients. Chest. 1992; 101:697–710. 11. Zielstorff RD. Online practice guidelines: issues, obstacles, and future prospects. J Am Med Inform Assoc. 1998; 5:227–36. 12. Nelson EC, Splaine ME, Batalden PB, Plume SK. Building measurement and data collection into medical practice. Ann Intern Med. 1998; 19:248–53. 13. Lobach DF, Hammond WE. Computerized decision support based on clinical practice guideline improves compliance with care standards. Am J Med. 1997; 102:89–98. 14. Lomas J, Anderson GM, Domnick-Pierre K, Vayda E, Enkin MW, Hannah WJ. Do practice guidelines guide practice? The effect of a consensus statement on the practice of physicians. N Engl J Med. 1989; 321:1306–11. 15. Musen MA, Shahar Y, Shortliffe EH. Clinical decision-support systems. In: Shortliffe EH, Perreault LE, Wiederhold G, Fagan LM (eds). Medical Informatics: Computer Applications in Health Care and Biomedicine. New York, NY: Springer-Verlag, 2000, pp 600–39. 16. Morris AH. Developing and implementing computerized protocols for standardization of clinical decisions. Ann Intern Med. 2000; 132:373–83. 17. Croskerry P, Shapiro M, Campbell S, et al. Profiles in patient safety: medication errors in the emergency department. Acad Emerg Med. 2004; 11:289–99. 18. Frison L, Pocock SJ. Repeated measures in clinical trials: analysis using mean summary statistics and its implications for design. Stat Med. 1992; 11:1685–704.
