Impact of order set use on outcome of patients with sepsis

Abstract

In an effort to improve outcome of patients with sepsis, we developed and implemented a disease-specific alert and order set for our computerized physician order entry system. This alert and order set were implemented in 2015. We have produced a progressive decrease in mortality for patients at our hospital with diagnosis of sepsis. We see a significant decrease in mortality for patients who had the sepsis order set used compared to those who did not have the order set used. We recommend use of an order set for patients with sepsis.

Background

We evaluate the care of patients in our hospital in an effort to continuously improve. Sepsis is a major focus for patient care improvement. There are established standards of care (Novosad et al. 2016; http://www.sccm.org/Documents/SSC-Guidelines.pdf 2012). These standards include several acute care activities including clinical and laboratory evaluation, rapid implementation of treatment, and follow-up evaluations. These multiple and multiphase steps are an excellent opportunity for a checklist to remind physicians of required tests and treatments to provide safe, timely, and effective care (Gawande 2009). A computerized physician order entry (CPOE) system is an ideal system to present physicians in real-time evidence-based efficient checklists of orders to provide care for specific patients. (Winterbottom et al. 2011; Hanzelka et al. 2013).

With our CPOE implementation our clinicians developed order sets to allow physicians to provide ideal care in the most efficient manner. One order set (Fig. 1) is for patients with signs and symptoms of sepsis. We also developed a best practice alert (BPA) (Fig. 2) for nurses and physicians that is activated in the computerized medical record when a patient has vital signs and white blood count suggestive of systemic inflammatory response syndrome (SIRS) or sepsis. The alert would be seen when nurse or physician open the chart. Specific parameters of vital signs, white blood cells, and serum lactate levels are shown in Table 1. The alert has a link to the sepsis order set. Physicians can select to use the order set from the link to best practice alert (BPA) or not. The order set can also be selected without the BPA.

Fig. 1

Order set used by physician

Fig. 2

Best practice alert seen by physician when patient meets criteria for sepsis

Table 1 Criteria for alerting physicians and nurses

During the time period of our data collection, there were no other new evaluations, medications or therapies implemented for our physicians and nurses. A specialized nurse who is our sepsis coordinator would see each time a BPA was initiated and follow-up with the nurse and physician about patient assessment and treatment.

Results

We report patients with sepsis as a principal or secondary diagnosis. Our patient mortality has decreased from 25% in 2013 to 19.4% in 2015 and in the first quarter of 2016 to 15.05% (P = 0.005) (Table 2). In 2016 from May 1 2016 to August 16, 2016, we evaluated the use of the sepsis order set after the best practice alert was presented. In those patients with the BPA our sepsis order set was used in 111 of 183 (60.6%) patients in the emergency department and 233 of 592 inpatients (39.4%) (P < 0.0001). Overall sepsis order set utilization was 344 of 775 patients (44%) (Table 3). For our patients during the measured period of May 1, 2015 through April 30, 2016 where the order set was used (n = 218) our mortality was 11.57% (95% CI 7.32–15.82). During the same period, patients in whom the order set was not used (n = 905) the mortality was 18.19% (95% CI 16.3–20.1) (P = 0.015) (Table 4). Length of stay was 11.68 days (95% CI 10.03–13.33) in patients who had the order set used and 13.31 days (95% CI 12.2–14.41) in those who did not (P = 0.18). Readmissions were 11.7% (95% CI 7.45–15.95) for those who had order set and 16.69% (95% CI 14.2–29.1) (P = 0.097) in those who did not.

Table 2 Patient mortality for all patients with sepsis diagnosis

Table 3 Use of sepsis order set after best practice alert May 1, 2016—August 16, 2016

Table 4 Sepsis order set utilization and patient outcome May 1, 2015- April 30, 2016

Discussion

We evaluate the care of patients in our hospital in an effort to continuously improve. We measure, analyze, plan, implement, and then re-measure as processes for continuous improvement. Sepsis is a major focus for patient care improvement. There is a national focus to improve diagnosis, treatment, and outcomes. The assessment of patient’s signs and lab tests can now reliably be done by computerized health record systems. Alerts can be generated to nurses and clinicians. The next phase for improvement is being sure that every patient receives complete and timely care. This should be done in the most efficient manner for nurses and physicians.

Conclusions

We report an improvement in mortality with use of an order set designed to be complete, timely, and easy to use. Length of stay and readmissions were improved in patients who had order set used but did not reach statistical significant difference. We believe we will see ongoing improvements in patient outcomes with clinician education, specific case reviews, nurse implemented protocols, and perhaps computer requiring use of the order set for all patients with sepsis. We recommend order sets be evaluated and used for sepsis and other diseases.

Methods

We use our electronic health record (EHR) to measure incidence and mortality. Patients who were coded as a principal or secondary diagnosis of sepsis were included. Coding is done by professionally trained, evaluated, coders who review every patient chart. Problem lists and physician notes are used to assign diagnostic codes. We analyzed the results for patients with sepsis coded for overall mortality, use of order set, and mortality based on use or non-use of the order set. We also analyzed major quality metrics including length of stay (LOS) and readmissions within thirty days. We compared outcome in those patients in whom the order set was used to those in whom it was not. Results were expressed as number of cases and as mean percentages (rates) ±95% confidence intervals (CI) of the mean. Statistical significance for mortality, readmission, and use of sepsis order sets in Emergency Department vs. inpatients was analyzed with the Fisher Exact test. The Student T test was used to analyze differences in LOS. Chi-Square Statistic was employed to estimate reduction in mortality from 2013 to 2016. P ≤ 0.05 was accepted as significant. No data was collected from humans or animals, only chart review was performed. This is a retrospective chart review of patients in an urgent clinical condition requiring rapid evaluation and management. We did not apply randomization as we believe the alert and order set should be applied to all patients.