Making surgery safer: Testing a wireless monitoring patch on general surgery wards

1. Recruitment is determined by recording the number of patients eligible, approached, consenting and randomised. Recruitment rate will be calculated as the number of patients randomised out of the number of patients eligible. The proportion of ineligible patients will be calculated as the number of patients ineligible out of the number of patients approached.
2. Adherence to protocol, and reasons for non-adherence as defined by the number of patients who do not receive the correct type of monitoring as per randomisation (and reasons for this) and the number of patients who do not wear the patch for their entire hospital stay or at least five days during their admission (and reasons for this including failure to repatch after return from critical care).
3. The amount of missing data for a data item will be calculated as the proportion of missing data for that item out of the number of patients randomised. Loss-to-follow-up will take into account withdrawal and death. The proportion of patients who are classed as ‘drop-out’ by design (i.e. never being admitted to a participating ward) will be calculated using the number of patients randomised as the denominator.
4. Optimal outcome measures to test effectiveness. This will be determined by observing effectiveness endpoints (see below) such as time to administration of antibiotics in cases of sepsis, critical care admission rate, length of hospital stay and assessing their potential as primary outcome measures for the definitive study. Assessment of the optimal outcome measures will take into consideration the amount of missing data and summary statistics for each potential outcome.
5. Estimation of sample size for definitive RCT will be calculated using relevant effect size(s) as seen in the effectiveness endpoint(s).

1. Time to antibiotics in cases of sepsis will be calculated as the time in minutes between the first evidence of sepsis on either or both monitoring tools and the first administration of antibiotics to the patient, and determined using the electronic patient record. Clinical suspicion of sepsis is defined by the presence of a likely source of infection and 2 or more criteria from a collection of clinical signs and laboratory investigations as follows:
1.1. Temperature >38.3°C or <36.0°C
1.2. Tachycardia >90 beats per minute
1.3. Tachypnoea >20 breaths per minute
1.4. pCO2 <4.3 kPa
1.5. Hyperglycaemia (blood glucose >6.6 mmol/) in the absence of diabetes mellitus
1.6. Acutely altered mental status
1.7. WBC count >12×10^9/L or <4×10^9/L
2. Number of HDU/ICU admissions defined as any admission to Level II/III care after stepdown to the general wards (i.e. non-perioperative admission from a participating ward) following surgery, and determined using the electronic patient record
3. Length of stay in HDU/ICU in days per admission, calculated as the difference in days between date of admission to either HDU/ICU and date of discharge from HDU/ICU, and determined using the electronic patient record. HDU and ICU lengths of stay will be amalgamated into a total HDU/ICU length of stay (excluding any peri-operative crtitical care admission)
4. Total length of stay in hospital in days per admission, including perioperative and postoperative Level II/III care, calculated as the difference in days between the date of admission and date of discharge, and determined using the electronic patient record
5. Number of postoperative complications, defined as any complication occurring after the patient has left the theatre complex and returned to wards J44 or J45 (i.e. not including perioperative HDU/ICU post-operative complications), and determined using the electronic patient record
6. Number of re-interventions, defined as the number of medical, radiological and surgical interventions required to treat postoperative complications, and determined using the electronic patient record. The proportion of patients receiving at least one re-intervention will be presented using as the denominator the number of patients receiving on-trial monitoring (i.e. admitted to a participating ward at some point during their stay)
7. Patient acceptability, as determined by the patient questionnaire and semi-structured interviews, and focus groups throughout the study
8. Nursing acceptability, as determined by the Modified System Usability Score and semi-structured interviews throughout the study
9. 30-day readmission rate, defined as the number of patients who are admitted to hospital for any reason within 30 days of discharge from their index admission, and determined using the electronic patient record