Comparison of clinical outcomes between nurse practitioner and registrar-led medical emergency teams: a propensity-matched analysis

The NPMET study was identified as a quality assurance activity and was approved by study site Research Ethics Committee (approval no. QA17PH44).

Retrospective study of all MET calls between June 1, 2016, and March 9, 2018, in an Australian tertiary ICU.

Admitted patients who had acute clinical deterioration to warrant a MET call during the study period.

All rapid response calls (RRC) are entered in Riskman database. All patient medical records are scanned in a digital medical record system. Both these databases will be interrogated to find MET calls, eligible for inclusion. Data regarding physiological observations are extracted from RRC observation form, scanned into digital medical record system. Demographic data, diagnoses, discharge destination and hospital outcome will be derived from iPM. Health information services (HIS) at Peninsula health, scan all inpatient medical records to determine comorbidities and Charleston Comorbidity Index (CCI). This information will be extracted from HIS database. Acute Physiological and Chronic Health Evaluation score II and III recorded in COMET database, for all patients admitted to ICU. These data along with admission and discharge time, for patients admitted to ICU, will be extracted using COMET database. Data for treatment provided, outcome and alteration in goals of care are entered into Riskman database and were extracted for the study.

RRS at the study site comprises two-tiered efferent limb, (1) comprising single criterion-triggered MET response for patients meeting physiological observation thresholds or staff concern triggers (Additional file 1: E-Table 1), and (2) code blue response for patients who were unresponsive, had respiratory or cardiac arrest, threatened airway or anaphylaxis.

The two clinical leads from the Critical Care Liaison nursing team participate in all MET calls and code blue responses when on duty. They each have greater than 20 years of critical care nursing experience and have completed postgraduate study in Critical Care Nursing and Master of Nursing (Nurse Practitioner). These two members have received endorsement from the Nursing and Midwifery Board of Australia to practice as Nurse Practitioners. The nurse practitioners have an advanced scope of practice which enables them to prescribe medication within a hospital approved formulary, order and interpret diagnostic investigations, refer to specialist teams and commence treatment. Their scope of practice includes advanced life support interventions including insertion of intra-osseous access, laryngeal mask airways and code team leadership. NP service was provided by two nurse practitioners and was available to be rostered for 8 h between 08:00 and 22:00 h (08:00–16:00 or 14:00–22:00), all days of the week including weekends. When NPs were not rostered, ICUR were assigned the role of attending and leading MET calls. All ICURs had at least four years of work experience after completing under-graduation, at least 9 months of which was spent either in ICU or emergency department resuscitation area under supervision of specialist consultants. All ICURs were either advanced emergency department trainees or registered as a trainee with College of Intensive Care Medicine (CICM). All ICURs received formal training toward recognizing and responding to critical patient deterioration, vascular access procedures and basic airway management techniques. Majority of ICURs were trained to perform advanced airway management including endotracheal intubation. Both NPs and ICUR could receive guidance from an ICU consultant physically present during business hours (08:00 to 18:00 h) and were on call out of hours (18:00 to 08:00 h).

MET calls were included in the analysis if.

MET calls were excluded from the analysis if both NP and ICUR attended, neither NP nor ICUR attended, no record of attendees in MET call record or patients not admitted to hospital.

Continuous variables were summarized using mean and standard deviation (SD) or median and interquartile range (IQR) depending on the underlying distribution of the data. Categorical variables were reported as counts and percentages. Comparisons between groups (NP vs. ICUR) were made using the Student’s t test for normally distributed continuous variables, Wilcoxon rank-sum test for non-normally distributed continuous variables and Chi-square or Fisher’s exact test as appropriate for categorical variables.

Propensity score matching was used to reduce selection bias from confounding factors between the ICUR and NP group. The individual propensities for being in the ICUR-led group were estimated with the use of a multivariable logistic regression model that included age, Charlson Comorbidity Index (CCI), weekend (Saturday and Sunday) occurrence of MET call, out of hours occurrence of MET call, “not for resuscitation” plan prior to MET call, triggering of MET call due to oxygen saturation < 90% and diagnoses (medical vs. surgical) as the predictor variables. This propensity score was used to match patients managed by ICUR to those managed by NP using a one-to-one nearest neighboring matching with a caliper width of 0.15 times the standard deviation. Standardized differences were calculated to assess the balance of covariates between ICUR and NP groups. Primary and secondary outcomes were compared between ICUR- and NP-led groups using conditional logistic regression taking into account the matched design with results reported as odds ratios (OR) and 95% confidence intervals (95% CI). Sensitivity analysis was also performed using a multivariable logistic regression model that included the same covariates as the propensity score model to ensure robustness of the primary analysis.

All reported p values are two-sided, and a p value less than 0.05 was considered to indicate statistical significance.

The propensity-matched analysis included 263 pairs of patients. The distribution of propensity scores appeared to be similar in both groups. Furthermore, the average propensity score was also similar [mean (standard deviation) 0.446 (0.161) in registrars and 0.435 (0.155) in nurse practitioners] in each group. A comparison of propensity-matched groups with regard to baseline characteristics and the characteristics of MET calls and the standardized differences are presented in Tables 1 and 2, respectively. Primary and secondary outcomes were compared between the matched pairs, and the results are presented in Table 5. There was no difference in the primary composite outcome between NP- and ICUR-led MET calls. However, the risk of hospital mortality was lower in NP-led MET call group (OR 0.57, 95% CI 0.35–0.91, p = 0.02). The likelihood of discharge home was also higher in NP-led MET calls as compared to ICUR-led MET (OR 1.55, 95% CI 1.09–2.2, p = 0.015). NP-led MET calls showed a trend toward decreased mortality within 24 h of index MET call, but this did not reach statistical significance (3.4% vs. 7.2%, 0.47 (0.21–1.05); p = 0.06).