A Computerized In-Hospital Alert System for Thrombolysis in Acute Stroke
Background and Purpose—An effective stroke code system that can expedite rapid thrombolytic treatment requires effective notification/communication and an organized team approach. We developed a stroke code program based on the computerized physician order entry (CPOE) system and investigated whether implementation of this CPOE-based program is useful for reducing the time from arrival at emergency departments (ED) to evaluation steps and the initiation of thrombolytic treatment in various hospital settings.
Methods—The CPOE-based program was implemented by 10 hospitals. Time intervals from arrival at the ED to blood tests, computed tomography scanning, and thrombolytic treatment during the 1-year period before and the 1-year period after the program implementation were compared.
Results—Time intervals from ED arrival to evaluation steps were significantly reduced after implementation of the CPOE-based program. Times from ED arrival to CT scan, complete blood counts, and prothrombin time testing were reduced by 7.7 minutes, 5.6 minutes, and 26.8 minutes, respectively (P<0.001). The time from ED arrival to intravenous thrombolysis was reduced from 71.7±33.6 minutes to 56.6±26.9 minutes (P<0.001). The number of patients who were treated with thrombolysis increased from 3.4% (199/5798 patients) before the CPOE-based program to 5.8% (312/5405 patients) afterward (P<0.001). The CPOE implementation also improved the inverse relationship between onset-to-door time and door-to-needle time.
Conclusions—The CPOE-based stroke code could be successfully implemented to reduce in-hospital time delay in thrombolytic therapy in various hospital settings. CPOE may be used as an efficient tool to facilitate in-hospital notification/communication and an organized team approach.
The efficacy of intravenous (IV) tissue plasminogen activator in acute ischemic stroke is time-dependent.1,2 However, a recent systemic review indicated that the average time from a patient’s arrival at the emergency department (ED) to the initiation of thrombolytic treatment exceeded 60 minutes in most studies.3 There have been several efforts to reduce in-hospital time delays, including reorganization of the ED,4 use of point-of-care international normalized ratio testing,5 and use of an acute stroke triage pathway.6 Stroke code systems and stroke team activities based on care protocols may expedite rapid thrombolytic treatment.4 However, operation of a stroke code system requires many resources, effective communication between staff members of various departments, and adequate monitoring with feedback to continually improve the system.
One promising approach for an effective stroke code system is using computerized physician order entry (CPOE). CPOE is a process that physicians use to enter medical orders electronically. These medical orders are communicated over a computer network linked to a hospital information system with physicians, nurses, technicians, and other staff in various departments.7 Therefore, CPOE allows physicians to provide accurate and rapid medical order entry and enables relevant staff to access necessary information immediately. Because CPOE permits capture of time data for individual steps more easily and objectively, it is useful to monitor the program’s efficacy and to provide feedback. These advantages of CPOE may improve critical care pathways for diverse emergent medical conditions.
We developed a stroke code program based on a CPOE system, the brain salvage through emergent stroke therapy (BEST) program.8 A pilot study was conducted in 1 hospital and demonstrated that the implementation of the BEST program could reduce time from arrival at the ED to evaluation steps and thrombolytic treatment.8 In concert with increasing demand of nationwide implementation of a well-structured critical pathway for thrombolysis in acute ischemic stroke, the present BEST Generalization (BEST-G) study was planned to investigate whether the BEST program can be successfully implemented more widely and can effectively reduce in-hospital time delay in IV thrombolytic treatment.
Materials and Methods
This was a multicenter prospective study that evaluated the efficacy of a stroke code system utilizing CPOE to reduce the time between arrival at the ED and various evaluation steps and IV thrombolysis.
The BEST program is a CPOE-based stroke team activation/notification system that enables activation, communication, notification, entering of predetermined standing order sets, providing of protocols and guidelines, and deactivation online. The CPOE was also used to evaluate the program’s efficacy because time data for each evaluation step could be obtained via CPOE.
The work flow for activation and deactivation of the BEST program appears in Figure 1. Briefly, candidates for thrombolytic treatment were identified in the triage area on arrival at the ED. Patient screening was based on stroke warning signs as prepared by the American Heart Association Stroke Council for the general population.9 When a patient had at least 1 warning sign of stroke, an ED physician/triage nurse activated the BEST program by clicking a check box on the patient’s order entry window and selecting the activation icon. Immediately after the BEST activation, an ED physician/triage nurse is notified by a neurologist by a cellular phone. In some hospitals participating in this study, on activation the program automatically sent a short message service to cellular phones of on-call neurologists and staff. Once the program was activated, the patient’s name was highlighted in orange in the patient list. Therefore, a candidate for thrombolytic treatment could be easily recognized by all medical personnel in a hospital. By entering predetermined order sets via CPOE, personnel could rapidly notify and communicate with appropriate medical staff. Administrative authorizations, which are often required before proceeding with tests and treatment and may potentially delay care process, were waived until the BEST program was deactivated. Entering medical orders for a CT scan and blood tests automatically activated an alarm such as a beeping sound and a pop-up window on the computer screens of staff members who were responsible for fulfilling the physician’s orders. These processes allowed technicians to receive orders at the same time when the physicians were entering medical orders; therefore, technicians could prepare examinations, wait for a patient or for blood samples, and perform examinations without delay. The BEST program was deactivated after patients received tissue plasminogen activator or when thrombolytic treatment was not indicated. Neurologists deactivated the program by selecting the deactivation icon. On deactivation, the patient’s name changed from orange back to the original background color so that every team member could recognize that the patient no longer needed thrombolytic treatment.
The CPOE system also was used for introduction of the protocol for thrombolysis and the BEST program. Essential portions of the protocol were incorporated into standing orders in the form of messages, and a full manual could be easily referenced at any time via the computer by clicking the guide menu bar.
Implementation of the BEST Program in Study Hospitals
Eligible hospitals for this study required: (1) an available CPOE system; (2) a written care protocol for IV thrombolysis; (3) obtainable time data via a CPOE system; and (4) predetermined standing order sets for thrombolysis. The thrombolysis protocol in each hospital was based on those of major trials and guidelines.10,11 IV tissue plasminogen activator was considered for patients between 18 and 80 years old and those who could be treated within 3 hours after symptom onset. Ten university or general care hospitals participated in this study. The Institutional Review Board of each hospital approved this study.
The BEST program, including protocols and a manual, was introduced and provided to investigators at the study hospitals. Each hospital designed and implemented its own program and protocol based on the BEST protocol and manual provided. Medical information and technology department personnel in each hospital joined the team to adapt the CPOE system to the BEST program because each hospital had different CPOE systems.
A workshop for the introduction of the BEST program was held in March 2008. Interim meetings were held in April and May 2008. The implementation of the program in each hospital was completed between June and September 2008.
Patient Group and Data Analysis
Assessed time data included the intervals from arrival at the ED to BEST activation (when the icon for the BEST program was clicked on the computer), neurological evaluation (the time when a neurologist assessed a patient), blood tests (the time when the results were entered into the CPOE), brain CT scan (when the CT scanning was initiated: door-to-CT time), and thrombolysis (when the bolus of IV recombinant tissue plasminogen activator was injected: door-to-needle time). The time interval from the symptom onset to arrival at the ED (onset-to-door time) was also assessed. Time data for blood tests, BEST program activation, and CT scan appeared on the CPOE, which is set at a universal time.
To investigate the effects of the BEST program, time intervals from arrival at the ED to various evaluation steps and IV thrombolysis were compared between the before BEST-G group and the after BEST-G group. The before BEST-G group included patients who were treated with IV thrombolysis during a 1-year period before initiation of the program (the day when the BEST program was initiated on the CPOE system). The after BEST-G group included patients who were treated during the 1-year period after initiation of the program. Data for the before BEST-G group were retrospectively obtained soon after initiation of the BEST program, and those for the after BEST-G group were obtained after the end of the 1-year study period.
Statistical analyses were performed with the Statistical Package for the Social Sciences (SPSS 15.0; SPSS, Chicago, IL). Categorical variables were analyzed by χ2 test, and continuous variables were analyzed by an independent sample t test. The relationships between door-to-needle time and other time interval-related variables (including onset-to-door time) were analyzed by Pearson test. Multiple linear regression analysis was performed to assess the independent predictors of door-to-needle time. A value of P<0.05 was considered statistically significant.
Number of Patients Treated With IV Thrombolysis
Gender and age were not different between the groups. In all, 15,713 patients were admitted to hospitals via the ED because of cerebrovascular diseases (ICD code G45, G46, I63, I64, I60, or I61) during the study period. Patients with ischemic stroke comprised 71.5% (5798/8103) of all patients admitted during the 1-year period before implementation of the BEST program and of 71.0% (5404/7610) during the 1-year period after the implementation of the program. The total number of stroke patients was not different between the 2 periods (P=0.453). However, the proportion of patients who were treated with IV thrombolysis among ischemic stroke patients increased significantly from 3.4% (199/5798) to 5.8% (312/5404) after implementation of the BEST program (P<0.001).
Time Intervals From Arrival at ED to Evaluation, Treatment, and Outcome
Onset-to-door time was not different between the groups after implementation of the BEST program: 62.0±35.9 minutes in the before BEST-G group and 67.3±38.6 minutes in the after BEST-G group (P=0.120). The time from arrival to neurological evaluation was also similar (14.2±16 minutes in the before BEST-G group and 12.2±16.8 minutes in the after BEST-G group; P=0.262). However, the time intervals from arrival to evaluation steps and treatment were decreased after implementation of the BEST program (Figure 2). Time to initiation of a CT scan was reduced from 24.7±18.2 minutes to 17.0±13.9 minutes (P<0.001). Mean time intervals to reports of complete blood counts and prothrombin time/activated partial thromboplastin time were reduced by 5.6 minutes (from 32.2±19.5 minutes to 26.6±19.2 minutes; P=0.001) and 26.8 minutes (from 66.5±68.9 minutes to 39.7±25.7 minutes; P<0.001), respectively. Door-to-needle time was shortened from 71.7±33.6 minutes to 56.6±26.9 minutes (P<0.001). Onset-to-needle time was slightly reduced by 9.8 minutes (from 133.6±39.6 minutes to 123.8±42.2 minutes; P=0.009).
Factors Associated With Time Intervals
Univariate and multivariate analyses revealed that the implementation of the BEST program (β=−14.067; standard error=2.604; P<0.001) and onset-to-door time (β=−0.210; standard error=0.034; P<0.001) were significant factors that were associated with door-to-needle time. There was an inverse relationship between onset-to-door time and door-to-needle time, which suggests that time delays to evaluations and treatment occur more frequently in patients who arrived at the ED earlier after symptom onset. This delay was reduced after implementation of the BEST program (Figure 3). However, there were no relationships between onset-to-door time and time to complete blood counts (r=−0.050; P=0.258), prothrombin time/activated partial thromboplastin time (r=0.000; P=0.993), and CT scan (r=0.010; P=0.882).
Baseline National Institute of Health Stroke Scale scores were not different between the 2 groups (13.6±6.6 in the before BEST-G group and 12.7±6.5 in the after BEST-G group; P=0.128). Data for modified Rankin scale score at 3 months were available in 191 patients in the before BEST-G group (96.0%) and 302 patients in the after BEST-G group (96.8%). Proportions of patients with favorable outcome (modified Rankin scale score 0 or 1) were not different (39.9% in the before BEST-G group and 45.5% in the after BEST-G group, P=0.230; Figure 4). Additionally, no difference in mortality was observed between the 2 groups (15.2% in the before BEST-G group and 14.6% in the after BEST-G group; P=0.870).
This study showed that a CPOE-based stroke code system (BEST program) can be successfully implemented in various hospital settings. Implementation of the BEST program reduced time for a CT scan and blood tests. Door-to-needle time was reduced by ≈15 minutes (21% relative reduction of in-hospital delay). The proportion of patients treated with IV tissue plasminogen activator increased by 70.5% after the BEST program implementation, which might be partly ascribed to reduced in-hospital time delay.
Stroke code protocols, which are used to reduce in-hospital time delay, require input of many human resources, including neurologists, ED physicians, radiologists, nurses, technicians for blood tests and CT scans, and administrative personnel. A well-organized team approach is important for the implementation of a functionally active stroke code protocol.12 In addition, for the efficient stroke code protocols, the following conditions are essential: (1) early identification of a candidate for thrombolysis and activation of a stroke code program; (2) prompt notification and response of the staff members who are responsible for fulfilling the orders; (3) accurate and rapid physician’s orders; (4) up-to-date protocols, which should be easily accessible to and familiarized by participating staff members; and (5) time logs for monitoring. In this respect, CPOE can be a nice tool for effective stroke code protocols.
When compared with conventional stroke code systems using telephone calls, the CPOE-based system has several advantages. In the BEST program, activation of the stroke code system is simple, with the user clicking an icon on the computer, and every relevant staff member could easily recognize a candidate for thrombolysis instantly at the time of activation. Therefore, prompt and multiple notifications and rapid actions of the relevant staff participating in thrombolytic treatment were possible. This reduces burden for notification and communication between diverse medical personnel and allows physicians or nurses to spend more time at the patient’s bedside. CPOE is also useful for maintaining the quality of care.13,14 The BEST program could facilitate implementation and maintenance of standard stroke care by incorporating predetermined standing order sets, evidence-based protocols, and manuals into the system. In the BEST program, accurate and rapid order entry was ensured and protocols/manuals could be easily referenced online. Time logs for quality control were automatically captured from a computer server. These features of the CPOE-based program may be useful for reducing in-hospital time delays and improving the quality of care with ongoing up-to-date education, monitoring, and feedback.
Previous studies indicated an inverse relationship between onset-to-door time and door-to-needle time.16,17 Although this inverse relationship may be, in part, because of patients who arrive very late and only those with short door-to-needle-time get included in the statistics, it also may be related to physicians’ behavior. With a 3-hour time limit for IV tissue plasminogen activator treatment, physicians are pressed for time with patients who arrive at the ED later; however, physicians might not be so prompt to administer treatment to those with earlier arrival. This study also demonstrated that the onset-to-door time was negatively correlated with the door-to-needle time. However, the inverse relationship was lessened after the CPOE-based system implementation. This finding suggests that the CPOE system helps to prevent negative physicians’ behavior and helps to provide consistent emergent care.
In contrast to the door-to-needle time, door-to-CT time and door-to-blood test reporting times were not correlated with the onset-to-door time, which suggests the presence of behavioral differences between physicians and technicians. Although physicians’ behaviors are governed by clinical necessity,18 technicians’ behaviors are more likely to be determined by the system and protocol. Therefore, to achieve a maximal reduction of time delay, different strategies might be necessary, depending on the team members/departments.15
The goal of a quality improvement of care pathway is to improve the patient outcome. A pooled analysis demonstrated that intravenous tissue plasminogen activator improves stroke outcome in a time-dependent manner.2 In this context, reducing in-hospital treatment delay is likely to improve stroke outcome in individual patients. However, our program leading to ≈10 minutes of earlier treatment did not show improvement in outcome. Several explanations for this are possible. First, reduction of 10 minutes, which was a 7.3% reduction of time interval, might be insufficient to lead to a clinically evident improvement. Second, the traditional outcome analyses might fail to detect subtle changes of clinical improvement induced by the 7.3% savings of time. Last, this study was not designed to investigate whether the BEST program could improve clinical outcomes and might be underpowered to detect a clinical efficacy.
This study has several limitations. Most of the participating centers were educational hospitals and equipped with computerized network systems. Thus, the generalizability of our findings may be limited. However, because hospitals with a computerized network system are increasing exponentially,16–18 the CPOE program is expected to be more widely applicable in the near future. The BEST program does not include a prehospital stroke code system. This might be responsible for similar time intervals to neurological evaluation between the before BEST-G and after BEST-G groups. Combined utilization with prehospital notification would further improve hospital responses.
In conclusion, this study demonstrates that the CPOE-based stroke code can be successfully implemented to reduce in-hospital time delay in thrombolytic therapy in various hospital settings. However, there is room for further reduction of time interval to thrombolysis. Additional measures have to be taken to further shorten the delay. Reduced time intervals appeared to contribute to a substantial increase of patients who received IV thrombolysis. The use of CPOE has been encouraged as a means to improve care quality and safety by reducing medical errors, supporting clinical decisions, and cutting costs.7,19–23 Our findings suggest that CPOE can be further utilized as a tool for efficient notification/communication. The CPOE is widely applicable to other urgent medical conditions that require rapid notification/communication and organized team approaches.
The authors thank Young Ah Kim, RN, PhD, and Hyungil Lee in the Division of Information and Technology, and Jahae Chun, RN, and Hyunjoo Shin, RN, in the Department of Quality Improvement of Severance Hospital, Yonsei University Health System, for their efforts in developing the BEST program. The authors also thank Hye Sun Lee, a biostatistician in the Department of Biostatistics, Yonsei University College of Medicine, for her assistance in the statistical analysis.
Sources of Funding
This work was supported by a grant from the Korea Healthcare Technology Research and Development Project, Ministry for Health, Welfare, and Family Affairs, Republic of Korea (A060171, A085136).
J.H.H received funds from Boeringer Ingelheim for other research activities not reported in this research. The funds did not exceed $10 000/year. K.S.H. received funds from Boeringer Ingelheim for other research activities not reported in this research. The funds did not exceed $10 000/year. S.U.K received funds from Boeringer Ingelheim for other research activities not reported in this research. The funds did not exceed $10 000/year. G.M.K. received funds from Boeringer Ingelheim for other research activities not reported in this research. The funds did not exceed $10 000/year. The other authors reported no disclosures.
- Received March 5, 2010.
- Accepted June 1, 2010.
Hacke W, Donnan G, Fieschi C, Kaste M, von Kummer R, Broderick JP, Brott T, Frankel M, Grotta JC, Haley EC Jr, Kwiatkowski T, Levine SR, Lewandowski C, Lu M, Lyden P, Marler JR, Patel S, Tilley BC, Albers G, Bluhmki E, Wilhelm M, Hamilton S, Investigators AT, Investigators ET, Investigators Nr-PSG. Association of outcome with early stroke treatment: Pooled analysis of Atlantis, ECASS, and NINDS rt-PA stroke trials. Lancet. 2004; 363: 768–774.
Lindsberg PJ, Happola O, Kallela M, Valanne L, Kuisma M, Kaste M. Door to thrombolysis: ER reorganization and reduced delays to acute stroke treatment. Neurology. 2006; 67: 334–336.
Rizos T, Herweh C, Jenetzky E, Lichy C, Ringleb PA, Hacke W, Veltkamp R. Point-of-care international normalized ratio testing accelerates thrombolysis in patients with acute ischemic stroke using oral anticoagulants. Stroke. 2009; 40: 3547–3551.
Kuperman GJ, Bobb A, Payne TH, Avery AJ, Gandhi TK, Burns G, Classen DC, Bates DW. Medication-related clinical decision support in computerized provider order entry systems: A review. J Am Med Inform Assoc. 2007; 14: 29–40.
American Heart Association. Stroke warning signs. Available at: http://www.Americanheart.Org/presenter.Jhtml?Identifier=3053. Accessed May 2006.
Hacke W, Kaste M, Fieschi C, von Kummer R, Davalos A, Meier D, Larrue V, Bluhmki E, Davis S, Donnan G, Schneider D, Diez-Tejedor E, Trouillas P. Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second European-Australasian acute stroke study investigators. Lancet. 1998; 352: 1245–1251.
Yu FB, Menachemi N, Berner ES, Allison JJ, Weissman NW, Houston TK. Full implementation of computerized physician order entry and medication-related quality outcomes: A study of 3364 hospitals. Am J Med Qual. 2009; 24: 278–286.
Ellrodt G, Glasener R, Cadorette B, Kradel K, Bercury C, Ferrarin A, Jewell D, Frechette C, Seckler P, Reed J, Langou A, Surapaneni N, Multidisciplinary Rounds T. Multidisciplinary rounds (MDR): An implementation system for sustained improvement in the american heart association’s get with the guidelines program. Crit Pathw Cardiol. 2007; 6: 106–116.
CHo HJ, Yang JH, Yung YH, Kim YD, Choi HY, Nam HS, Lee KY, Heo JH. Quality improvement activity of stroke team to reduce time intervals from emergency department arrival to thrombolytic treatment. Stroke. 2009; 40: e209–e209.
Collin S, Reeves BC, Hendy J, Fulop N, Hutchings A, Priedane E. Implementation of computerised physician order entry (CPOE) and picture archiving and communication systems (PACS) in the NHS: Quantitative before and after study. BMJ. 2008; 337: a939.
Park RW, Shin SS, Choi YI, Ahn JO, Hwang SC. Computerized physician order entry and electronic medical record systems in Korean teaching and general hospitals: Results of a 2004 survey. J Am Med Inform Assoc. 2005; 12: 642–647.
Roy P-M, Durieux P, Gillaizeau F, Legall C, Armand-Perroux A, Martino L, Hachelaf M, Dubart A-E, Schmidt J, Cristiano M, Chretien J-M, Perrier A, Meyer G. A computerized handheld decision-support system to improve pulmonary embolism diagnosis. Ann Intern Med. 2009; 151: 677–686.