Impact of a Modified Early Warning Score on Rapid Response and Cardiopulmonary Arrest Calls in Telemetry and Medical-Surgical Units

January-February 2017 • Vol. 26/No. 1 15

William M. Parrish, DNP, RN, CCRN-K, CPHQ, is Coordinator, Nurse Residency Program, Providence Health and Services, Everett, WA. Marilyn Hravnak, PhD, RN, CRNP, FCCM, FAAN, is Professor, University of Pittsburgh, School of Nursing, Department of Acute and Tertiary Care, Pittsburgh, PA. Linda Dudjak, PhD, RN, FACHE, is Associate Professor, University of Pittsburgh, School of Nursing, Department of Acute and Tertiary Care, Pittsburgh, PA. Jane Guttendorf, DNP, RN, CRNP, ACNP-BC, CCRN, is Assistant Professor, University of Pittsburgh, School of Nursing, Department of Acute and Tertiary Care; and Acute Care Nurse Practitioner, Department of Critical Care Medicine, UPMC Presbyterian, Surgical Intensive Care Unit, Pittsburgh, PA.

Impact of a Modified Early Warning Score on Rapid Response and Cardiopulmonary Arrest Calls in Telemetry and Medical-Surgical Units

I n response to the Institute forHealthcare Improvement’s (2014)call for hospitals to reduce the number of cardiac arrests and sud- den life-threatening patient events, more than 1,500 hospitals across the United States have implement- ed rapid response teams (RRTs). Healthcare systems implementing these teams empower staff, and in some cases patient family members, to call a group of clinicians with advanced assessment skills quickly to the bedside to evaluate a change in patient condition. Many organi- zations implementing RRTs have demonstrated reductions in num- ber of cardiac arrests, mortality rate, and critical care unit length of stay (Maharaj, Raffaele, & Wendon, 2015).

For RRTs to be effective, organi- zational leaders must equip clini- cians with the knowledge and tools to recognize patients at risk for dete- rioration in condition. According to Chen and colleagues (2015), RRTs often are mobilized to the patient bedside when an extreme alteration occurs in one parameter of vital sign and, in many cases, after the patient is already in distress. Early warning scores (EWS) or modified early warning scores (MEWS) repre- sent a way to detect smaller changes in multiple patient parameters over time, giving clinicians an earlier indication of potential change in condition. Early recognition of an untoward change in patient condi- tion is vital to early intervention, which then can improve patient outcomes (Mapp, Davis, & Krowchuk, 2013).

Literature Review A literature search was conduct-

ed using the PubMed and Ovid databases and search terms early warning score and modified early warning score, limited to articles published 2012-2014. The search generated 361 results. Articles were selected for inclusion if they were available in full text in English, and specifically addressed impact of an EWS on RRT activations or patient outcomes (e.g., cardiac arrest, mor- tality rate).

A systematic review on the impact of EWS on patient outcomes conducted by Alam and colleagues (2014) found an evidence gap exists due to few multicenter trials using a standardized scoring system. How – ever, overall positive impact was demonstrated on clinical outcomes. Authors conducted a search of the

PubMed,, and Coch – ran Library databases using the terms early warning score, hospital, hospital setting, and adults. Seven of 532 identified references were included because they were con- trolled, addres sed the im pact of EWS on patient outcomes, and included only subjects over age 16. Two of the seven studies demon- strated significant reduction in pre- implementation versus post-imple- mentation mortality rates (5.8% vs. 2.8%; p=0.046; and 1.4% vs. 1.2%; p<0.0001). Two additional studies in the systematic review also dem – onstrated positive trends in survival that did not reach statistical signifi- cance. One study reported an 8% increase in survival to discharge (p=0.47) and the other a 0.9% reduction in mortality (p=0.092).

McNeill and Bryden (2013) per- formed a systematic review evaluat-

Continuous Quality ImprovementContinuous Quality Improvement

William M. Parrish, Marilyn Hravnak, Linda Dudjak, Jane Guttendorf

To reduce the number of cardiac arrests in telemetry and medical- surgical units, a 70-bed community hospital integrated a weighted, aggregate, electronic modified early warning score into the elec- tronic medical record. Impact was evaluated via a quality improve- ment initiative.

January-February 2017 • Vol. 26/No. 116

ing the impact of RRTs on adult patient survival. Authors originally performed a search of Ovid Medline, CINAHL, Cochrane Library, Web of Science, NHS National Research Register, NHS UK Research Network Study Portfolio, and EMBASE for studies evaluating the impact of an EWS and RRTs on patient outcomes. Studies were included if they exam- ined adults and focused on one of the following outcome criteria: inci- dence of cardiac arrest, unplanned intensive care unit (ICU) admissions, ICU mortality, length of ICU stay, length of hospital stay, or inpatient hospital survival. Six of the reviewed studies also evaluated EWS. Four included aggregate weighted scoring systems with multiple patient parameters; two others used a single parameter approach. Only one of the included studies demonstrated improved hospital survival rates; however, both studies using a single parameter system demonstrated reduction in cardiac arrest rates, and two of the four using multiple parameter EWS demonstrated posi- tive impact on unexpected deaths or cardiac arrest calls. Although some of the data in the review were identi- fied as poor quality, authors conclud- ed a whole system of scoring cou- pled with clinical action is needed. In addition, aggregated early warn- ing scoring systems seemed to be more effective than single parameter triggering systems.

A retrospective observational study in an academic medical cen- ter evaluated the potential effective- ness of a MEWS based on the vital signs of medical-surgical patients collected for the 48-hour window prior to an adverse event (Ludik – huize, Smorenburg, de Rooij, & de Jonge, 2012). Authors used an aggregate weighted scoring system for vital signs. Patients were includ- ed in the study only if they experi- enced emergency surgery, unplan – ned admission to the ICU, car- diopulmonary resuscitation, or unexpected death. Of 204 patients meeting these inclusion criteria, 81% (n=166) experiencing an adverse event achieved an EWS of 3 or more at least once during the 48 hours preceding the event.

Ludikhuize and colleagues (2014) also conducted a quasi-experimen- tal study to evaluate the degree to which a MEWS used three times daily, versus an EWS obtained only when prompted by a clinical event, impacted RRT activation at an aca- demic medical center. Data were collected for 3 consecutive months on adult patients to include vital signs and adverse events, such as unplanned admission to the ICU or cardiopulmonary arrest. These data were analyzed for 372 patients who were monitored using an EWS pro- tocol and 432 control patients. Authors concluded scoring occur – red more often when used by proto- col (70%, n=2,513) than when trig- gered by a clinical event (2%, n=65). Addi tionally, the rate of RRT calls per admission on the units using protocols were double those of units not using protocols.

Huh and co-authors (2014) pub- lished a retrospective cohort study of RRT activation through the use of an electronic aggregate weighted EWS, compared to a clinician- or provider-activated call prompted by established RRT activation criteria. Scores were calculated based on scoring criteria built into the elec- tronic medical record. Authors col- lected data on all inpatient and out- patient RRT activations in an aca- demic medical center for 24 consec- utive months (n=3,030), including outcome information, such as trans- fer to a higher level of care or death during resuscitation. RRT activa- tions prompted by the electronic medical record EWS calculation resulted in fewer ICU transfers (192 vs. 400; p=0.0000) and deaths dur- ing cardiopulmonary resuscitation (2 vs. 8; p=0.0000) than RRT activa- tions by a clinician or provider.

Improvement Needs In an attempt to decrease the

number of cardiopulmonary arrests occurring outside the ICU, the pri- mary author’s medical center formed a group of nurse educators, leaders, and nursing informatics staff to establish a MEWS that would provide clinical staff with an early alert to changes in patient

condition. Before facility imple- mentation of an electronic MEWS, cardiopulmonary arrest and RRT calls were initiated by a care provider when a patient was found to be in distress or when a signifi- cant alteration in vital signs was noted. For example, an RRT call may have been initiated when a patient care technician reported to the primary nurse a patient was severely short of breath during rou- tine vital signs. While suggested activation criteria were provided to clinicians during introduction of the facility RRT in 2006, no formal warning system existed to prompt staff of a potential change in patient condition.

With the implementation of the electronic MEWS on the medical- surgical and telemetry units at the institution, a quality improvement initiative was launched to provide a framework for evaluating the impact of the electronic MEWS. Three evaluation metrics were iden- tified: the number of RRT calls, the number of cardiopulmonary arrests occurring on the noncritical care units, and survival to discharge fol- lowing an RRT call or a cardiopul- monary arrest. Implementation groups before and after MEWS implementation were compared.

This project was conducted in a 70-bed suburban acute care facility, which included a 36-bed medical- surgical unit, a 26-bed telemetry unit, and an 8-bed ICU. The MEWS was implemented on the medical- surgical and telemetry units, but not in the ICU. The facility dis- charged an average of 538 patients per month (June-November 2014). Approval for this quality improve- ment project was obtained from the hospital vice-president and admin- istrator.

MEWS criteria (see Table 1) were culled from the literature and included graded parameters for heart rate, systolic blood pressure, respiratory rate, temperature, neu- rologic assessment, oxygen satura- tion after initial intervention, and use of a supplemental oxygen delivery device (Subbe, Kruger, Rutherford, & Gemmel, 2001). A plan for integration into the elec-

Continuous Quality Improvement

January-February 2017 • Vol. 26/No. 1 17

Impact of a Modified Early Warning Score on Rapid Response and Cardiopulmonary Arrest Calls in Telemetry and Medical-Surgical Units

tronic documentation system for the facility was devised. Hospital policies were updated to reflect the score at which a care provider should receive a real-time, on- screen pop-up prompt by the elec- tronic MEWS system to consider calling the RRT (score of 4 or greater, or 2 or more points above patient’s baseline). If the pop-up was to be received by a member of patient care services other than the primary nurse assigned to the patient, that person would be responsible for notifying the pri- mary nurse of the alert. While the pop-up notification directs the care provider to consider calling the RRT, facility policy did not mandate RRT activation. Education on the electronic MEWS and the scoring system used within the electronic medical record was provided to medical-surgical and telemetry patient services staff by the primary author. Critical care nurses who respond to RRT calls also were pro- vided education on the system to give them insight into the criteria; physician providers were educated separately in medical staff meet-

ings. In collaboration with nursing informatics staff, the electronic MEWS was implemented on the medical-surgical units in September 2014 and on telemetry units in November 2014.

Data Collection The project consisted of a

descriptive review of pre- and post- implementation data surrounding implementation of the electronic MEWS. Three months of pre-imple- mentation data were collected for the baseline comparison group (June-August 2014). These data were compared to post-implemen- tation data collected October- December 2014 for the medical-sur- gical unit and December 2014- February 2015 for the telemetry unit. De-identified information on RRT and cardiopulmonary arrest calls was obtained from aggregate data collected by the facility for institutional reporting. All data were de-identified and stored on the facility’s private network. Period incidence rates for RRT calls and cardiopulmonary arrests (number

per 1,000 discharges) were calculat- ed before and after implementa- tion, and the percentage change from baseline was reported.

Patient survival to hospital dis- charge after RRT activation was reported as a percentage of all patients having experienced an RRT call in the medical-surgical and telemetry units; survival-to-dis- charge for patients after a cardiopul- monary arrest call was reported as a percentage of all patients having experienced cardiac or pulmonary arrest in the medical-surgical and telemetry units. Patients with a do- not-resuscitate order were excluded from the survival analysis.

Results and Limitations The period incidence of RRT calls

before and after electronic MEWS implementation did not differ sig- nificantly for the medical-surgical units or the telemetry units (see Table 2). For the individual unit subgroups and the combined units, the number of RRT calls was some- what lower in the post-implementa- tion phase. The rate for combined

TABLE 1. Electronic Modified Early Warning System Scoring Criteria

Score 3 2 1 0 1 2 3

Heart rate: Beats/min <40 40-50 51-100 101-110 111-130 >130

Systolic BP mm Hg <70 70-80 81-100 101-159 160-199 200-220 >220 if acute or change from baseline

Respiratory rate/min <8 8 9-17 18-20 21-29 >30

Temperature ˚F <95 95-100.4 100.5 – 101 >101

Central nervous system, if changed from baseline

Acute confusion Awake and responsive or chronic confu- sion


Responds to verbal commands only


Responds to pain only




O2 saturation after intervention

<85% 85%-89% 90%-94% >94%

O2 delivery method High-flow nasal cannula or trach collar

Simple face mask or nasal cannula

Room air Face tent or aerosol

Non-rebreather or Venturi mask

Note: Adapted from Subbe et al., 2001

January-February 2017 • Vol. 26/No. 118

units was 12.54 before vs. 10.78 after implementation, representing 14% reduction in calls.

During the evaluation period, no cardiopulmonary arrests occurred on the medical-surgical units for the pre- or post-implementation periods. The period incidence of cardiopulmonary arrest calls (num- ber of RRT calls per 1,000 hospital discharges) on the telemetry units was slightly lower after implemen- tation (1.19 vs. 1.16), which repre- sented 2.5% reduction from base- line (see Table 2). Hospital survival to discharge following an RRT call (see Table 3) did not differ signifi- cantly for pre-implementation and post-implementation groups for medical-surgical units, telemetry units, or combined unit rates. Survival to discharge for patients after a cardiopulmonary arrest call

(see Table 3) was the same in pre- and post-implementation periods for the telemetry unit (50% vs. 50%). These data could not be reported for the medical-surgical unit as no arrests occurred during either phase of the project.

Sample size for the project was small due to the type of facility and current patient volume. Also, the evaluation period after implemen- tation of the electronic MEWS was limited and the relatively small numbers of events (RRT calls and arrests) limited the ability to make broad assumptions from the com- parison group data. Ongoing, extended collection of data sur- rounding RRT and cardiopul- monary arrest calls will be critical to evaluate the impact of the electron- ic MEWS system.

Nursing Implications Several important points should

be considered by organizations implementing an electronic MEWS. First, pop-up notifications within the electronic medical record to alert staff to an elevated score may not be sufficient to prompt the clini- cian to activate the RRT. Monitoring of follow-through of RRT activation following clinician notification of an elevated electronic score would be essential in determining if the pop-up notification is effective. Second, data collection on how often nurses may initiate an alterna- tive intervention or consultation with a medical provider in lieu of activating the RRT in response to notification of an elevated electron- ic score would be helpful in deter- mining if the electronic MEWS impacts clinician decision making.

Continuous Quality Improvement

TABLE 2. Rapid Response Team (RRT) Calls and Cardiopulmonary Arrest Calls per 1,000 Hospital Discharges

Event Unit Pre-Implementation Post-Implementation Result

RRT Calls

# RRT Hospital


# RRT Calls/1,000 Discharges # RRT

Hospital Discharges

# RRT Calls/1,000 Discharges

% Change from Baseline

Medical-Surgical 11 1,674 6.57 9 1,615 5.57 15% ↓

Telemetry 10 1,674 5.97 9 1,727 5.21 13% ↓

Total 12.54 10.78 14% ↓

Cardiopulmonary Arrest Calls

# Arrests Hospital


# Arrest Calls/1,000 Discharges # Arrests

Hospital Discharges

# Arrest Calls/1,000 Discharges

Medical-Surgical 0 1,674 0 0 1,615 0 ——

Telemetry 2 1,674 1.19 2 1,727 1.16 2.5% ↓

Total 1.19 1.16 2.5% ↓

TABLE 3. Hospital Survival: Patients* Experiencing Rapid Response Team (RRT) Calls or Cardiopulmonary Arrest

Event Unit Pre-Implementation Post-Implementation

RRT Calls* Number (%) of patients surviving to discharge after RRT call/Total patients experiencing RRT call

Medical-Surgical 10/11 (91%) 7/8 (88%)

Telemetry 9/9 (100%) 7/9 (78%)

Total 19/20 (95%) 14/17 (82%)

Cardiopulmonary Arrest Calls* Number (%) of patients surviving to discharge after arrest call/Total patients experiencing arrest

Medical-Surgical N/A (no arrests) N/A (no arrests) Telemetry 1/2 (50%) 1/2 (50%)

Total 1/2 (50%) 1/2 (50%)

* Excluding do-not-resuscitate patients

January-February 2017 • Vol. 26/No. 1 19

Finally, given the relatively low incidence of events (RRT calls and cardiopulmonary arrests) during the 3-month duration of this proj- ect, designing a continuous quality improvement initiative to collect data over a longer period with a larger sample may yield more useful benchmarking data.

Conclusion Although authors expected RRT

calls would increase in number fol- lowing electronic MEWS imple- mentation, a notable decrease occurred. The small number of calls during the pre- and post-implemen- tation period made it difficult to draw conclusions. No change in numbers was noted for cardiopul- monary arrest calls. Before imple- mentation of the electronic MEWS system, however, the project facility did not have a formal early warning system to guide clinicians in deci- sion making regarding activation of the RRT. While the project revealed information useful to facility and system leaders in determining the future of the electronic MEWS sys-

tem, ongoing evaluation of the sys- tem is needed. Develop ment and implementation of a formal quality improvement process for collecting, evaluating, and benchmarking data will be instrumental in this evalua- tion. Addi tionally, more evidence is needed on staff culture, education, and hospital policy surrounding implementation of an electronic MEWS to determine if its use impacts patient outcomes.

REFERENCES Alam, N., Hobbelink, E.L., van Tienhoven,

A.J., van de Ven, P.M., Jansma, E.P., & Nanayakkara, P.W. (2014). The impact of the use of the early warning score (EWS) on patient outcomes: A systemat- ic review. Resuscitation, 85(5), 587-594.

Chen, J., Bellomo, R., Flabouris, A., Hillman, K., Assareh, H., & Ou, L. (2015). Delayed emergency team calls and associated hospital mortality: A multicenter study. Critical Care Medicine, 43(10), 2059- 2065.

Huh, J.W., Lim, C., Koh, Y., Lee, J., Jung, Y., Seo, H., & Hong, S. (2014). Activation of a medical emergency team using an electronic medical record-based screen- ing system. Critical Care Medicine, 42(4), 801-808.

Institute for Health Care Improvement. (2014). Early warning systems: Scorecards that

Impact of a Modified Early Warning Score on Rapid Response and Cardiopulmonary Arrest Calls in Telemetry and Medical-Surgical Units

save lives. Retrieved from http://www.ihi. org/resources/Pages/ImprovementStor ies/EarlyWarningSystemsScorecards ThatSaveLives.aspx

Ludikhuize, J., Borgert, M., Binnekade, J., Subbe, C., Dongelmans, D., & Goos – sens, A. (2014). Standardized measure- ment of the modified early warning score results in enhanced implementation of a rapid response system: A quasi-experi- mental study. Resuscitation, 85(5), 676- 682.

Ludikhuize, J., Smorenburg, S.M., de Rooij, S.E., & de Jong, E. (2012). Identification of deteriorating patients on general wards; measurement of vital parameters and potential effectiveness of the modi- fied early warning score. Journal of Critical Care, 27(4), 424.e7-424.e13.

Mapp, I., Davis, L., & Krowchuk, H. (2013). Prevention of unplanned intensive care unit admissions and hospital mortality by early warning systems. Dimensions of Critical Care Nursing, 32(6), 300-309.

McNeill, G., & Bryden, D. (2013). Do either early warning systems or emergency response teams improve hospital patient survival?: A systematic review. Resusci – tation, 84(12), 1652-1667.

Maharaj, R., Raffaele, I., & Wendon, J. (2015). Rapid response systems: A systematic review and meta-analysis. Critical Care, 19, 254.

Subbe, C.P., Kruger, M., Rutherford, P., & Gemmel, L. (2001). Validation of a mod- ified early warning score in medical admissions. QJM, 108(7), 519-522.

Copyright of MEDSURG Nursing is the property of Jannetti Publications, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder’s express written permission. However, users may print, download, or email articles for individual use.

"Looking for a Similar Assignment? Get Expert Help at an Amazing Discount!"