Postoperative quality and safety using Efficacy Safety Score (ESS) and a wireless patient monitoring system at the ward: A randomised controlled study

Postoperative pain, side‐effects and time to mobilisation are indicators for the quality of postoperative recovery. The aim of this randomised controlled study was to investigate if efficacy safety score (ESS) combined with a wireless patient monitoring system would improve these clinical outcomes for patients at a general surgical ward.

Opioids are still standard care for management of acute postoperative pain, but the risk of severe adverse events can limit optimal dosing for analgesia. 7 Although over-dosing of analgesics after non-complicated anaesthesia and surgery is rare, it still represents a major safety challenge. 8,9 In a post-anaesthesia care unit (PACU) there are staffing and monitoring available to ensure quality and safety of the treatment. After discharge to an ordinary ward, there is less available capacity for frequent patient observation.
Postoperative patient side-effects, such as pain, nausea and vomiting, are often not documented in a standardised manner at the ward or put into a comprehensive overall evaluation or overview. 6,10 Early warning scores, such as the National Early Warning Score (NEWS), have been developed to standardise surveillance of vital functions, 11 but do not include patient's quality parameters. At the project hospital, NEWS was based on cumbersome registration using pen and paper.
Efficacy Safety Score (ESS) is a validated clinical decision tool for the first 24 postoperative hours including both the PACU and the ward periods. 12 It covers multiple components of patient safety and quality status while monitoring the patient's experiences of the quality of care, including pain at rest and at movement. The Patient Status Engine (PSE) from Isansys Lifecare Ltd. is a wireless, semiautomatic registration system of vital patient parameters from wearable medical sensors. 13 The aim of this study was to test the hypothesis that quality of care could be improved by combining ESS and PSE during the ward period of the first 24 hours after surgery, with reduced time to mobilisation and improved pain management, while ensuring patient safety.

| Design and setting
This single-centre, randomised controlled trial, with two parallel groups, was conducted at the Orkdal Department, St. Olavs Hospital, Trondheim University Hospital, Norway. Design and description of the study adhered to the Consolidated Standards of Reporting Clinical Trials statement (CONSORT). 14 The study was approved by the Regional Committee for Medical and Health Research Ethics (reference number 2017/1903/REK South East A) and registered at clini caltr ials.gov (NCT03438578).

| Eligible patients for the study
Eligible patients were identified from operating theatre lists of mixed surgery at a medium-sized surgical unit and recruited during pre-admission clinic or when prepared for surgery. Inclusion criteria were patients undergoing acute or elective surgery expected to be hospitalised more than 24 hours postoperatively. Exclusion criteria were patients <18 years of age, poor communication capabilities or when planned surgery was incompatible with mobilisation during the first 24 hours. All the patients were subjected to the same post-operative prescriptions and PACU discharge criteria, according to the hospital protocols.
The intervention and study observation period started when patients returned to the ward from the PACU.
The ward nurses obtained patients' written informed consent preoperatively and performed study enrolment. The patients were then randomly assigned to one of two groups-a standard care group (SC-Group) or an intervention group (INT-Group)-using a random number generator and sequentially numbered, opaque sealed envelopes set up a priori by the study personnel. 15 The ward nurses had to pick up the envelope at a restricted office after enrolment. Due to the clinically obvious monitoring system, neither the staff nor patients could be blinded to group allocation at the ward.

| Intervention group
In the INT-Group, the ward nurses assessed ESS in parallel with electronic automatic retrieved vital signs from the wireless monitoring platform PSE. The PSE is a class IIa CE-marked medical device (Isansys Lifecare Ltd.,) for hospital use monitoring heart rate, ECG, ventilation rate, axillar skin temperature, blood pressure and finger pulse oximetry; all from wireless and wearable sensors. 13 The PSE gives an updated NEWS every minute which is calculated from the sensors, except for blood pressure measurements which are initiated manually. Registration of ESS was recorded for storage on the bedside PSE device, done hourly during the first 4 hours after PACU discharge and then every second hour except when the patient was confirmed sleeping. In this study, we used bedside monitoring and visual warnings were displayed on the bedside device. Information on given medication was extracted manually from the patients' charts.

| Standard care group
In the SC-Group, NEWS was documented on paper formularies at were registered together with given medication, extracted from the patients' charts.

| Ordinary postoperative protocol
Both groups followed the same protocols and indications for medication according to the hospital's clinical guidelines. At the study hospital, post-operative pain management is provided by prophylactic multimodal regular prescriptions according to instructions for the individual procedure: paracetamol, non-steroidal anti-inflammatory drugs, steroids, opioids, local anaesthesia in the wound, perineural blocks and regional anaesthesia. Extra medications beyond these are mainly intravenous or oral administration of oxycodone 2.5-5 mg.
According to hospital guidelines, extra opioid medication may be administered when pain is >3 on an 11-point verbal numeric rating scale (VNRS) based on nurse judgement. Further, the guideline is to provide supplementary oxygen if saturation is <94% on pulse oximetry for patients not having chronic obstructive pulmonary disease and/or when needed as judged by the nurse.

| Outcomes
The primary endpoint of this study was time to full mobilisation, defined as being able to walk more than one step with or without support. 16 For both groups, mobilisation was attempted as early as possible after surgery, following the protocols for nursing plan and physiotherapy plan.

| Sample size calculations
A validation study addressing ESS in a total of 207 patients 12  http://clinc alc.com/Stats/ Sampl eSize.aspx.). A sample size of 200 (100 in each group) included patients was considered sufficient to identify meaningful effect of the intervention on the primary outcome, allowing for some dropouts and missing data.

| Statistical methods
The Kaplan-Meier method and a log-rank test were used to analyse time to mobilisation, and group differences were analysed using the Mann-Whitney test. Additionally, Cox-regression was used to estimate hazard ratios for time to mobilisation between the intervention and control group, adjusting for age, American Society of Anesthesiologists (ASA) classification and sex. We used linear regression to estimate mean differences between the groups for opioid medication doses, numbers of pain assessments, NEWS performed and pain intensity assessments adjusting for age, ASA classification and sex. The precision of estimated effects is given by a 95% confidence interval. Categorical variables, such as reported oxygen therapy, postoperative nausea and retching/vomiting, were analysed using Chi-square tests. Patient satisfaction was analysed with Fischer's exact test based on a Chi-square test, due to low numbers of expected patients in some groups.
All collected data were registered in Microsoft® Excel® for PC, version 16. Data were analysed using SPSS version 25.0 (SPSS, Chicago, IL, USA).  (Table 1).

| Primary outcome
As the study progressed, it turned out to be a high number of prostatec-

| Secondary outcomes
The difference in mean average intensity of pain on a 0-10 VNRS  Table 3 for distribution.

Premedication (n)
Paracetamol 96 99 Dexamethasone 93 98 Opioids 84 81 Non-steroidal inflammatory drugs 33 27 The same authors recommended that postoperative on-demand medication should be given if pain intensity is over 3 on an 11-point scale, which also was determined as a cut-off for moderate-to-severe pain. 18 The reason for less pain in the INT-Group may both be more attention and more opioid medication. The nurses performed a higher number of NEWS and pain documentations in the INT-Group, and hence probably gave more attention to these patients. They also gave more opioids to patients in this group. This suggests improved communication and that the nurses were comfortable in giving a higher opioid medication. 7 The association between more opioids used and less subsequent pain is to be expected.

Not all patients in the INT-Group got opioids. The nurses identi-
fied those in need of more on an individual basis and provided these patients efficient on-demand pain relief, with maintained safety. Table 2 shows that some of these patients needed, and were given, far above-average opioid doses. The INT-Group was four times as likely to get opioid dosing within the three highest dose intervals. Extra opioids were given at lower VNRS score compared to patients in the SC-Group (Figure 3). The situation for the patients in the SC-Group with the highest average VNRS represents a major quality concern.
More opioid treatment in the INT-Group did not result in any increase in emetic symptoms or respiratory problems. Further, potential more drowsiness from more opioids in the INT-Group was not shown. In fact, the opposite seemed to be the situation, as these patients had earlier mobilisation and a tendency of shorter hospital stay.
Our interpretation is that the frequent surveillance with the ESS tool made the nurses able to pick out and surveil those patients with an individual higher need of opioids, and to treat them accordingly with better pain results and no increase in side-effects.
All our patients got optimal non-opioid multimodal premedication analgesic regimes, but despite this, most needed opioid pain medication, as will be expected for major surgical procedures staying overnight. 19 Opioid analgesic is still a cornerstone in postoperative pain treatment and best controlled in the early phase in the PACU with titrated intravenous agents for reliable and rapid action. 20 Adequate early pain management may be associated with a lower incidence of persistent postoperative pain. 21 Early appropriate treatment with higher opioid dosages of short duration which maximises analgesia, while minimising the risk of later abuse, has been advocated. 22 The pain assessment protocol in the INT-Group was more frequently scheduled and the documentation process itself was easier when compared to the hospital's standard protocol. This may explain some of the difference in the number of documented pain assessments.
The probability of receiving extra opioids for postoperative pain relief is significantly larger if a pain score is documented. 23 Absence of pain assessment and documentation were identified in a European survey report, which concluded with postoperative pain management being suboptimal. 24 The bedside documentation of  It may be argued that simply enforcing the implementation of better pain and analgesia protocols may be as useful as the ESS + wireless monitoring system. Although this simple advice has been known for some time, documentation show that postoperative pain care has not improved to any major extent during the last years. [4][5][6] Also, if a better and more rigorously applied pain protocol results in individualised use of more opioid analgesics, it will be necessary to have protocols on monitoring safety and side-effects. [7][8][9] The ESS combined with wireless monitoring, as in our study, may be a way for systematic implementation of better overall care, including pain care.
The increased satisfaction in the INT-Group may be explained by improved pain relief, but other factors like being continuously monitored, regularity in follow-up and interactions with health care providers are probably also important. When patients become active participants in their own care it is shown that they become more comfortable and able to function. 26 However, the difference in satisfaction in our patients was mainly in the scaling of being 'satisfied' vs 'very satisfied', and this nuance may not be clinically significant outside a study setting.
Clinical outcomes are shown to be improved by the deployment of automated notification systems for vital signs monitoring in hospital wards. 27 The finding of increased use of oxygen therapy for the INT-Group reflects an expected effect of continuous monitoring.

Patients in the INT-Group with hypotension and atrial fibrillation
were identified by the monitoring system, and the call-out algorithm initiated by a high ESS was used. The same algorithm was used for the two patients given nerve blocks due to insufficient pain control.
This contrasts the SC-Group patient who was accidentally found unconscious with seizures. We think this would have been discovered earlier with continuous monitoring. However, this study was not designed to have the power to make any conclusions about patient safety issues, but the case reports are noteworthy.
A strength of this study is that ESS is validated and easy to perform in daily clinical work. Also, we had two large and similar groups without major differences in baseline variables. The groups showed internal diversity in age, sex and surgery performed, but we chose this design with mixed patient characteristics to achieve good generalisability.
A weakness is that we did not test the two components of the intervention separately. We chose a pragmatic approach with a combination of two possible favourable interventions to maximise patient satisfaction and safety. Moreover, due to the monitoring system, neither the staff nor the patients could be blinded to group allocation.
However, the possible influence from the study induced focus on post-operative care may have resulted in better care also in the SC-Group, with potentially better results than in everyday care.
The relatively restricted observation period is another limitation. Whereas the ESS is designed also to be used in the PACU, 12 we F I G U R E 3 Mean distributed morphine milligram equivalents (MME) at the ward related to average pain on an 11-point verbal numeric rating scale (VNRS) (95% CI). Number of patients in respective columns. Differences between the groups were 10.3 mg for MME, P = .001, and 1.2 on the VNRS, P < .001 [Colour figure can be viewed at wileyonlinelibrary.com] TA B L E 3 Overall patient satisfaction for first 24 post-operative hours on a 1-5 scale, n (per cent). The difference in mean between the groups was 0.6 (95% CI 0.39-0.79), P < .001 chose to have a totally similar non-ESS protocol for all patients in the PACU period, in order to tease out the true benefit during the first day on the ward. This is a period when the patients have pain, sideeffects and need for frequent surveillance, more so than later in the postoperative course. Outcomes further on, for the next day and for clinical endpoints after discharge, will definitely be of interest, but were beyond the scope of this study.
Also, as the benefit of using ESS and wireless monitoring has been established, a further step will be to look at the time consume, nurse satisfaction and cost-benefit of using these methods in a daily routine, and also to look for potential simplifications in order to minimise the extra workload.
This study shows that important postoperative issues, such as pain and early mobilisation, were significantly improved with the use of the clinical tool ESS and PSE wireless monitoring at the ward.

ACK N OWLED G EM ENTS
We thank study nurse Lillian A. Opøyen. for help conducting the study. We also thank all health care professionals at ward B3, St Olavs hospital, Orkdal for their assistance in this study.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest. Isansys Lifecare Ltd. (UK) lent us the monitoring hardware and provided supplies at a reduced price.

AUTH O R CO NTR I B UTI O N S
E.S.: Study design, conduct, analysis and manuscript preparation.