SNAP-2 (EpiCCS): FAQs

SNAP-2 patient recruitment will start on Tuesday 21st March 2017 at 00:00hrs and end on Monday 27th March 2017 at 23:59hrs. There will be a one week follow-up of patients that ends on Monday 3rd April 2017.

The study closes on May 26. This is Day 60 for the last patient recruited on March 27, which is the final day of recruitment.

Which patients are to be included in SNAP-2?
All adult (18 or above) patients, undergoing any type of surgery or procedure requiring the support of an anaesthetist, in an NHS hospital in the UK during the study period and are expected to require an overnight stay in hospital after their operation are eligible to be included. These would include all procedures taking place in an operating theatre, radiology suite, endoscopy suite or catheter laboratory for which inpatient (overnight) stay is planned, including both planned and emergency/urgent surgery of all types, Caesarean section, surgery for complications of childbirth, endoscopy and interventional radiology procedures.

Do we include patients undergoing surgical procedures on critical care when the patient is already an inpatient in critical care, eg percutaneous tracheostomy, ICP boltline insertion?
If they go to theatre for their procedure, yes. If they're having the procedure on ICU, no.

Do we include patients undergoing radiological intervention or endoscopy under local anaesthetic and or sedation eg angio, nephrostomy insertion, endoscopies etc that requires in patient stay post op?
If the patient has an anaesthetist supporting (potentially needing sedation or conversion to GA), these patients are included. If they don't have an anaesthetist and therefore the proceduralist is performing the procedure by administering sedation themselves, they are excluded. Anaesthetist involvement is crucial to our inclusion criteria.

Do the patients have to have a GA to be eligible?
No. GA is not necessary. As you can see from Question 4.3 in our Patient CRF, operations being done under just neuraxial block, sedation and regional/local anaesthetic are included.

Do the procedures have to be in theatre?
No, they can be in the endoscopy suite, or cath lab, or interventional radiology suite. So long as an anaesthetist is involved. It can be "remote anaesthesia".

Do we include all anaesthetic procedures that result in an inpatient stay, eg local anaesthetic etc.
Yes.

Can we co-recruit with the PQIP trial?
Yes: we are happy for any SNAP-2 patient to be co-enrolled into any study

The exclusion criteria are as follows:

• Patients who indicate they do not want to participate in the study
  
• Day-case or Ambulatory surgery; patientds <18 years old
  
• Non-surgical obstetrics
  
• ASA-PS grade VI
  
• Non-interventional diagnostic imaging (e.g. CT or MRI scanning without interventions)
  
• Emergency department or critical care interventions requiring anaesthesia or sedation but no interventional procedure

Would it be possible for the local research delivery team (research associates and data managers) to complete the CRF?
Yes, Clinical Research Nurses as well as the local research delivery team (research associates and data managers) can complete the CRFs retrospectively. The PI or other clinical member of staff should also be able to provide support to answer any questions which may arise from the CRFs that require clinical expertise.

How do I add a second record for the second trip to theatre for a patient?
We would like for each theatre visit to be documented with a separate CRF. When it comes to uploading the data onto the webtool:

We don't want to remove the validation for hospital numbers because we don't want multiple copies of the same CRF uploaded. But at the same time we want to capture patients who genuinely return to theatre and have a CRF for each theatre visit.

Therefore In this case, please append an "-a", "-b", "-c" suffix and so on to the hospital number to identify the patient as having gone back to theatre on multiple visits during the study recruitment week.

So, if the hospital number is "123456", then first visit to theatre, input that. Second visit to theatre input "123456-a", and so on.

How do you suggest we approach emergency cases and weekend admissions?
We hope that you will be able to collect data on all emergency and weekend cases. In terms of perioperative data collection, hopefully you can persuade your anaesthetists/surgeons to complete the CRFs at the time of surgery (this is the best way to collect the data for all cases, whatever time of day or day of week, especially as some of the questions on the CRF require their opinion, and therefore can't be completed retrospectively or by anyone else). In terms of D7 follow up data, we appreciate that for weekend cases this will be difficult. Options would include asking trainees on call to do this, as long as they follow the SOP for POMS data collection. The other alternative is to collect the D7 weekend data retrospectively from patient notes on the Monday. The POMS has been validated for retrospective data collection as long as SOPs are followed, so that should be OK, although it is likely to be harder to find the relevant data from the notes retrospectively.

Are patients lacking consent included?
The Research Ethics Committee and the Confidentiality Advisory Group in England, as well as the Patient Benefit and Privacy Panel in Scotland have looked at our Opt-Out consenting process in great detail.

All three panels were satisfied that the patients who are awake and have capacity should be given a copy of the Patient Information Leaflet, and prominent posters should be displayed in preoperative areas. There is no need to retrospectively give PILs to patients after their surgery if they lack capacity. Because of the nature of the study, patients with the highest chance of benefiting from Critical Care are also the most likely not to be able to give consent to a study. Therefore to reduce the risk of selection bias, they are included in the study without their consent. This is similar to the approach adopted by the National Emergency Laparotomy audit and the EPOCH study, so there is precedent. Furthermore, the study is non-interventional, and the issue over consent/non-consent is the recording of identifiable information in order to link their hospitalisation to outcome data that is collected by Hospital Episode Statistics/Office of National Statistics mortality data. The lay representatives who have been involved in the study (the RCoA's Lay Committee, as well as a patient representative on our Steering Group) have all indicated that this approach is acceptable to them, and we have included their testimony about these issues in all our applications throughout the study design stages.

What do we do if a patient decides to opt out of the study?
This depends on when you are notified by the patient.

If the patient tells you after you have collected data about them, and the data has been entered onto the webtool, the PI should find the patient's SNAP-2: EpiCCS Case ID number on the webtool, and notify us of this Case ID number by email. We will then remove that patient from the database. We will not be able to identify the patients by their name, NHS number or other identifiable information as we will be blinded to this.

If the patient tells you after you have collected data about them on the paper CRFs but before this has been transcribed into the webtool, then you can simply dispose of the paper CRF as you would any confidential waste in the NHS, and take care not to enter any information about them onto the database.

Lastly, if the patient tells you that they want to opt out before you have taken any information about them, then just avoid recording any data about them from the start. You can make a note within the patients' clinical records so as to avoid data being recorded about them further down the line.

We will not be keeping a log of the patients who have decided to opt out of the study, but PIs are free to do so for their local records.

Who will be responsible for providing the information sheet to patients?

It can be anyone that the Principal Investigators choose to delegate this to.

In our Pilot study, both sites had different approaches to this: One site tasked the anaesthetists who were visiting the patients preoperatively to make their anaesthetic assessment with distributing the PIS to the patients they were seeing at the time. The other site gave some Trainee Anaesthetists and Clinical Research Nurses the specific role of distributing the information sheets. From feedback, the second method was superior, because the ones who had specific responsibility to distribute the PIS were more able to answer any questions the patients may have had about the study.

Our REC (Research Ethics) and CAG (Confidentiality Advisory Group) support did not stipulate that the Information Sheets needed to be given to patients postoperatively (retrospective notification). They were happy with the provision of PIS preoperatively to those who had the capacity and that patients who were incapacitated will be included in the study because of the risk of selection bias in the study.

How do we handle Case Record Forms with missing data?
Some paper CRFs have data missing which makes it difficult to transcribe completely onto the webtool. If you are unable to fill in the data fields completely, leave them blank. You will not be able to "Lock" the record in the database, however, we will still include the records in our analysis, and apply statistical treatments for missing data. The cases will also still count towards accrual. Locking the record only preserves the record as it is and prevents further tampering.

We have some warning messages that appear when we leave out missing data, and it is preventing us from locking the record
In our system there are 2 types of "required data", one that flashes amber, and one that flashes red. The amber ones are warnings, but you can still progress with data entry. The red ones are errors which may prevent you from progressing. Having a red error means that that data item is critical and essential if missing, or the entry given is illogical or out of range. Amber warnings only prevent the data from being "locked", but the rest of the data from the record is still usable. Locking the record will prevent anyone else from altering the record in future. However your data from that record will still be analysed from our point of view, and will still count towards accruals.

For example, if you are missing the NHS number in your record, this is flags up an amber warning. We know there will be people without NHS numbers (foreign visitors for example), so don't worry about this. You will still be able to enter other data about the case and progress to the other tabs. In contrast, a red error in this example would be an NHS number that doesn't fit the rules (needs to be 10 digits long, and the last digit is a check digit against the previous 9 digits). Another example of a red error is when you enter a value that is outside the ranges we have specified (like a preoperative Systolic BP of 500mmHg).

How do we transcribe procedures from free-text to the webtool?

Please select the procedure that best describes the operation that the patient planned to undergo. If there is more than 1 planned procedure, select the primary procedure.

The paper Case Record Form would have the procedure written in free-text format. Please try to classify the procedure as best you can according to the classification system in this web form.

For example, for an entry on the paper CRF which reads: "Trial of Instrumental delivery: Ventouse vaginal delivery", select Obstetrics >> Pregnancy and confinement >> Rotational vaginal delivery.

If you cannot find the primary procedure from within the list, then you may have to look under different Main Group or Subgroup headings.

Lastly, if you still cannot find the correct procedure under all the options, then select the closest analogous procedure. There may not be a code available for the procedure under the classification system used. If you still find it difficult to find the procedure, do give us a phone call and we will try to help!

What is the timeframe for data entry?
The pilot sites managed to transcribe everything as they went along. Both Pilot sites finished uploading data a few days after Day 7 follow up.

So it shouldn't last more than a week after the Day 7 follow up. We will then ask sites at Day 60 to finish off and they should be done with transcription soon after that.

if the patient was discharged from theatre ventilated and continue to stay ventilated then that would mean there is no new requirement for o2/ventilation do we say no to this then?
"New requirement" refers to compared to their baseline (before coming in for surgery). So in this case, you will say there is a new requirement for O2 or ventilation, unless the patient is on home oxygen or receiving CPAP normally at home and the form of respiratory support they are receiving at Day 7 is still CPAP.

Our PACU and Recovery are one and the same. How do we record this?
Our definition of PACU is found in our SOP-annotated version of the CRF, to quote:

We understand that there may be many hospitals which are unfamiliar with PACUs/OIRs. If these facilities exist in your hospital they are short-term post-operative critical care beds for surgical patients developed to an acceptable standard appropriate for the management of an artificially ventilated patient overnight.

The PACU or OIR concept is well-described here:
bja.oxfordjournals.org/content/92/2/164.full.pdf+html

Therefore if your recovery area just adopts the name of "PACU" but is unable to manage an artificially ventilated patient overnight or deliver full Level 2 or 3 care, then it would not meet our definitions.

For recording Cancellations data, do you want us to record all cases that get cancelled? Or just cases that met the inclusion criteria?
Please record just the cases that would meet the inclusion criteria which get cancelled using the cancellations spreadsheet  SNAP-2 Cancellations (11 KB), then uploading it onto the cancellations tab in the Webtool. So if any Day Case surgery gets cancelled, you don't have to record these in the Cancellations dataset.

Are Obstetric Patients and/or HDUs included in this survey?
Obstetric patients are included in the main study. This includes all Caesarean sections, elective or emergency, trials of instrumental delivery, surgical procedures performed in obstetric theatres. A rule of thumb is: if the anaesthetist is involved and the patient is undergoing a procedure in the operating theatre, then include that case. If the anaesthetist is not involved and the patient is undergoing a procedure in a room in the labour ward or delivery suite, eg suturing a perineal tear after delivery, then exclude that patient.

Also for the organisational survey would like it to include the women's hospital so that we know what the Obstetric HDU and ICU space availability is.

Obstetrics and Gynaecology is not listed as a specialty on the Clinician Perceptions CRF
Unfortunately, this was an error that was only picked up once the forms had been printed. Obstetrics/Gynaecology is an option on the webtool so please ask your colleagues to note their specialty as free text on the form and whomever is transcribing these to the webtool can ensure that their specialty is recorded correctly.

Are labour epidurals included?
Labour Ward epidurals for labour analgesia are not included.

What is your definition of non-surgical obstetrics?
Non-surgical obstetrics: Ventouse or Forceps delivery in the labour room (as opposed to theatre), episiotomies, small perineal tear repairs that are done by the midwives or obstetricians in the labour room under local anaesthetic without anaesthetists suppoting.

A rule of thumb is: if the anaesthetist is involved and the patient is undergoing a procedure in the operating theatre, then include that case. If the anaesthetist is not involved and the patient is undergoing a procedure in a room in the labour ward or delivery suite, eg suturing a perineal tear after delivery, then exclude that patient.

For the study, which category of c-section equate to which Operative urgency level?
The anaesthetists who will be on hand should be able to convert the Obstetric categories into NCEPOD categories, the obstetric categories were originally designed as a performance target for labouring women entering theatre, with each category having a specific time to delivery target. They have since been replaced by new definitions: www.oaa-anaes.ac.uk/assets/_managed/editor/File/RCoA/Good%20Practice%2011%20CS%20FINAL.pdf the new classification system may have calibration issues for risk adjustment and so we are not using it for our study.

Practicing anaesthetists completing the forms would be able to quickly classify the labouring women's Caesarean section into NCEPOD classifications instead. Most Category 1 C-sections would fall under the "Immediate" NCEPOD classification. Category 2-4 could be either "Urgent" or "Expedited", but they do not correspond directly. Therefore we expect the clinicians completing the CRFs at the time to make a clinical judgement.

How do we record Manual Removal of Placenta?
Obstetric -> Curettage of the delivered uterus.

How to we record tear repairs?
Female reproductive organs -> vagina/perineum -> operations on the female perineum.

For the BMI, the most recent BMI calculated for our obstetric will be at 32 weeks, is this the BMI we should document on the CRF? Or should this be documented as not done?
For BMI: the nearest recorded weight and height (and hence BMI calculation) would be acceptable.

For the Blood results, the most recent urea/creatine/ sodium/ potassium will be bloods taken at 28 weeks. Are these suitable to be used? Or should we use postoperative bloods which would be taken within hours of the operation? Or should we document as not done?
Blood results: likewise the latest blood results available within 3 months before surgery are acceptable, so 28 weeks for a 40 week gestation delivery are ok. Please do not provide post op blood results.

How will ICU bed data be collected? It is not uncommon for a bed to be empty but 'closed' due to staffing or other issues. This means that the unit may be recorded as at 'full capacity' when in fact beds are empty!
That is the intention of the wording of the question. If the bed is physically empty, but "closed", it means that there is something that needs to be done to open that bed up for use, be it call in for additional nursing staff to man the bed or have the bedspace deep cleaned, etc.

These are the sorts of pressures we think that influences clinician's decision-making around allocating that bed to the patient. You might do so for the higher risk patients (mortality risk > 30% for example), but be disinclined to do so for someone whose risk is lower (mortality risk ~5%) for example.

It is our argument that an ICU at full capacity is one that doesn't have a bed physically empty and staffed, ready to take a patient.

How do we record data for mixed ICU/HDU?
We know several sites have mixed ICU/HDUs with the ability to flex beds up or down between Levels 1-3 depending on what type of patient needs a bed. For this question, please answer in this way:

We want to record the number of physically beds AND staffing for these beds. We want the limiting factor when it comes down to capacity.

-If you have 2 physical beds but 1 nurse, the answer is 1
-If you have 2 physical beds and 2 nurses to staff them, the answer should be 2.
-If you have extra nurses for the day, but only 1 physically empty bed, then the answer should again be 1.

Our PACU and Recovery are one and the same. How do we record this?
Our definition of PACU is found in our SOP-annotated version of the CRF, to quote:

We understand that there may be many hospitals which are unfamiliar with PACUs/OIRs. If these facilities exist in your hospital they are short-term post-operative critical care beds for surgical patients developed to an acceptable standard appropriate for the management of an artificially ventilated patient overnight.

The PACU or OIR concept is well-described here:
bja.oxfordjournals.org/content/92/2/164.full.pdf+html

Therefore if your recovery area just adopts the name of "PACU" but is unable to manage an artificially ventilated patient overnight or deliver full Level 2 or 3 care, then it would not meet our definitions.

Protocol asks for the PI to complete the CCU occupancy form but presumably this can be done by CCU staff?
Yes - it can be done by CCU staff, but the PI has responsibility for ensuring that this happens.

How do we measure day 7?
Day of surgery is Day 0, so a patient that was seen Tuesday 21st March will have a 7 day follow up on Tuesday 28th March, and so on.

For the 7 day follow up - how will we get some of the data for patients who are at home and haven't given consent to be telephoned?
Our protocol says that patients who have been discharged home by the time it reaches Day 7 will be assumed to have no POMS-defined morbidity. So you don't have to call patients up for this.

We have some uncertainty about how to answer question 5.4 about whether a discharged patient has returned to their preoperative level of mobility.
We should have been a bit clearer about the mobility question. If the patient has been discharged by Day 7, you actually do not need to fill in the answers for mobility. The webtool will not even let you submit a response for that (it becomes greyed out). It's fine to answer "not known" on the paper CRFs, because when it comes to entering it on the webtool, it actually won't be uploaded.

The reason why this question was included was mainly to capture potential reasons for patients who remaining in hospital at Day 7, but did not have any POMS-defined morbidity. If they had not returned to baseline mobility, then that could explain their still remaining in hospital.

How do we record re admissions within 7 days?
Ignore readmissions please. When the patient is discharged from hospital after the initial stay following their surgery, that is the end of the study involvement for them. We will tracking readmissions through other mechanisms (mainly data linkage with HES/ONS).

If, for example, the patient presented at the first hospital on Day 1 of the study (Tuesday 21/3/17), was discharged the next day and then re-presented to this or other hospital on say Day 4 and required another operation, they would then have had another CRF filled in. Therefore, do not worry about readmissions, we have other mechanisms to pick up these patients in the study.

A patient has been transferred to another hospital to continue care , how do I complete the day 7 review?
For patients who have been transferred to another hospital, please do class them as a discharge, and then on the webtool, we have added an option for Question 5.3 (Status at discharge) for "Patient transferred to another hospital".

If you are able to still track the progress of the patient for Day 60 follow-up if the patient is still in your trust, then that would be wonderful. However, we don't expect you to be able to track the Day 7 follow-up data.

We have some warning messages that appear when we leave out missing data, and it is preventing us from locking the record
In our system there are 2 types of "required data", one that flashes amber, and one that flashes red. The amber ones are warnings, but you can still progress with data entry. The red ones are errors which may prevent you from progressing. Having a red error means that that data item is critical and essential if missing, or the entry given is illogical or out of range. Amber warnings only prevent the data from being "locked", but the rest of the data from the record is still usable. Locking the record will prevent anyone else from altering the record in future. However your data from that record will still be analysed from our point of view, and will still count towards accruals.

For example, if you are missing the NHS number in your record, this is flags up an amber warning. We know there will be people without NHS numbers (foreign visitors for example), so don't worry about this. You will still be able to enter other data about the case and progress to the other tabs. In contrast, a red error in this example would be an NHS number that doesn't fit the rules (needs to be 10 digits long, and the last digit is a check digit against the previous 9 digits). Another example of a red error is when you enter a value that is outside the ranges we have specified (like a preoperative Systolic BP of 500mmHg).

With Questions which refer to morbidities "in the past 24 hours", does the past 24 hours relate to Day 6 post op as the survey is being carried out on Day 7 post op?
Yes, for these questions you make the assessment on Day 7, with Day 0 being the day of surgery, and ask about the 24hr period prior to the assessment timepoint. Therefore you will need to know if the patient's situation on Day 6 to answer these questions.

Question 5.4: Does the patient's preoperative level of mobility yrefer to before their surgery or on the day of?
Preoperative level or mobility should mean their pre morbid state: so if a patient came in moribund on the day of surgery but was otherwise walking normally the day before, and then underwent emergency surgery, their preoperative level of mobility would correspond to the day before they were admitted. If the patient came in for elective surgery and were normally mobile anyway, this was their normal premorbid level of mobility and it was unchanged on the day of surgery as well.

Ultimately you have to ask "what is the level of mobility that is normal for the patient"?

Question 6.6: Do the new focal deficit, coma or delirium have to be in the last 24 hours or anytime in the past 7 day?
If they have had a focal deficit or been in coma or delirious since the surgery, and are still exhibiting this on Day 7, answer yes.

If they had a deficit or had been delirious or comatose anytime during the days since the procedure, but are no longer comatose at the moment you are assessing the patient, answer no.

Clinician Perception CRF Question 1.1a asks about the primary workload, but it does not list specialties exhaustively?
Please write down the specialty by hand in free text on the paper CRF, when it comes to transcribing it online, there will be a "Others: " box to select for this response.

Obstetrics and Gynaecology is not listed as a specialty on the CRF
Unfortunately, this was an error that was only picked up once the forms had been printed. Obstetrics/Gynaecology is an option on the webtool so please ask your colleagues to note their specialty as free text on the form and whomever is transcribing these to the webtool can ensure that their specialty is recorded correctly.

How do we handle Clinician Perception Forms with missing data?
Some paper CRFs have data missing which makes it difficult to transcribe completely onto the webtool. If you are unable to fill in the data fields completely, leave them blank. You will not be able to "Lock" the record in the database, however, we will still include the records in our analysis, and apply statistical treatments for missing data. The cases will also still count towards accrual. Locking the record only preserves the record as it is and prevents further tampering.

Do all patients count towards accruals?
All participants providing implied consent (including clinicians taking part in the survey study) will count towards accruals. Patients who do not provide written consent for the 7-day study, but are offered the opt-out option, via distribution of patient information sheets, will count as "informed" and therefore count towards accruals.

When will accruals be counted?
Accruals will be counted on the date of recruitment (date of surgery for patients, and date of data entry for the clinicians' perceptions).

How do we ensure that accruals are added in time for this year's deadline?
The deadline for uploading data to our webtool is 09:00 on Monday 10th April. The data we require to be uploaded before these deadlines are:

• Patient demographic and intraoperative details (sections 1 and 2 of the patient paper CRF); and
  
• The completed clinicians' perceptions CRFs, which also need to be locked.

This is so that we can use this information to upload details of all the patients you have recruited to count towards the NIHR 2016-17 accruals targets. We have checked with NIHR and all patients recruited to SNAP-2 must be uploaded in time for their 2016-17 deadline or they will not count towards any accruals targets. If you have any questions about this, please get in touch with us.

How many investigators can register for each hospital site?
There should be a maximum of one Principle Investigator (PI) for each hospital site; this may mean that there are multiple PIs within a trust/health board depending on how many sites they cover.

There is no limit to the number of Local Investigators (LIs) that can register at each site. The number of LIs required will vary widely depending on how many operations are performed over the study period and the layout of services within individual hospitals.

Who can register as an PI?
This must be a consultant, and may be the Quality Audit and Research Coordinator (QuARC), Clinical Director, Audit Lead or another nominated consultant within the department.

Who can register as an LI?
Anaesthetists at any stage of training (including Foundation Years) or consultants, those in non-training posts and research nurses are all eligible to register as an LI.

Can medical students take part?
Yes Medical Students can definitely take part. They will be acknowledged in the same way as all other collaborators. We also had some Medical Students participate in SNAP-1. We would however encourage additional clinical supervision if Medical Students are completing the Case Record Forms, as they may lack sufficient clinical knowledge in some instances to answer the questions correctly. They are most welcome to help with data transcription from paper CRFs onto the webtool, ensuring that all the forms have been completed and collected correctly, and other administrative tasks that the PI may deem suitable, for example.

Can I register to take part in SNAP-2 if I work in a paediatric hospital?
SNAP-2 is only including patients 18 years and over. As occasionally non-paediatric patients undergo procedures in paediatric hospitals, you may still register to take part in SNAP-2 and collect data if any adult patients are operated on in your hospital over the study period.

Are there a minimum number of patients that local sites are required to recruit?
No. We would encourage any hospital with an anaesthetic department that performs surgery in patients over the age of 18 to take part. Even if there are only a small number of eligible cases over the two study days this information will still be of value and contribute to the project as a whole.

Will site initiation visits be going ahead?
We will not be undertaking site initiation visits but if local teams have queries we can answer them by email or arrange a phone call with the chief investigator or a member of the study team

Will site files be provided?
We will not be providing site files so would be grateful if local teams could prepare their own

Will study documentation be provided?
Local sites will be responsible for adding their logos to patient information sheets and posters and printing locally. We will provide paper copies of all study questionnaires by post and will confirm the postal address with sites once capability and capacity has been confirmed. To clarify, we will print:

• Appendix 1: CRF for the main study
  
• Appendix 3: CRF for the Critical Care Bed Occupancy
  
• Appendix 4: CRF for the Clinician Perceptions substudy

The sites will have to print the following patient-facing documents, because these are the ones which would need localising:

• Appendix 5: The PIS for the main study will need sites to enter their Local lead name and contact details. Again they can include trust logos if they wish.
  
• Appendix 7: The Poster for the whole study. They can enter their Local lead name and contact details and include trust logos if they wish. They are welcome to print this out in whatever size they see fit, but this has to be prominently displayed in perioperative areas of the hospital (e.g. pre-admissions lounges, patient day rooms on surgical wards, within the anaesthetic pre-assessment clinic, etc).

Appendix 2 and 6 have now been removed from the study following a Substantial Amendment to the protocol.

Is a site specific study agreement required?
For sites in England, the Statement of Activities (included in the document pack) replaces a site specific study agreement. Sites just need to sign this and return to Suzanne Emerton at the UCL Joint Research Office, and keep a copy for their own records.

For sites in Scotland, Wales and Northern Ireland, there is a model agreement that needs to be signed, this should be returned to Gosia Zborowska at the UCL Joint Research Office.

Is there an NIHR Portfolio Adoption Letter?
The Portfolio Adoption doesn't issue a letter, but all you should need is the CPMS number:
CPMS ID: 31913.

Why is my site not on the IRAS form?
We are not adding sites onto the IRAS form. HRA approval was given on the basis of the sites not all being listed in the IRAS form as it would be unfeasible to do so, since the study is being conducted in over 200 sites across all 4 devolved nations. The sites already listed on the form are example sites which gave the Research Ethics Committee and HRA an indication of the early interest in the study.

To quote our HRA approval letter (page 7):

• "There is one site type participating in this study. The applicant confirmed that all NHS organisations will be undertaking the same activities, as detailed in the Schedule of Events. All NHS hospitals in the UK which undertake inpatient surgery will be invited to take part.

• Because of the number of eligible NHS organisations, Part C of the IRAS form is not listing all sites. Hospitals will be recruited using the National Institute for Academic Anaesthesia Health Services Research Centre (NIAA HSRC's) Quality Audit and Research Coordinator (QuARC) network. A very similar study was undertaken recently by the same research group (SNAP-1 study); the researchers will be approaching the PIs who participated, with the assumption that the majority will also be participating in this study. A list of SNAP-1 PIs, as well as a list of QuARCs, has been provided."

We advise the R&D departments to read the HRA approval letter closely (this is included in the study document pack). The approval letter further states:

• "If chief investigators, sponsors or principal investigators are asked to complete site level forms for participating NHS organisations in England which are not provided in IRAS or on the HRA website, the chief investigator, sponsor or principal investigator should notify the HRA immediately at hra.approval@nhs.net. The HRA will work with these organisations to achieve a consistent approach to information provision."

If we are only able to recruit a minority of the available patients, would our data still be acceptable to be included in the trial?
Our intention is to gather a full and comprehensive sample of each hospital's work. It would be preferable for the study to get complete data. But if we are unable to then we may have to address missingness through statistical means. We would however love for your site to participate and recruit as many patients as able.

Who will cover archiving costs?
Unfortunately the research costs for archival will have to be borne by the trust. We will not be providing funding to cover this. We had originally applied to the NIHR Portfolio for archival to be a cost paid for by the NIHR CRN, however this was revised by the ACORD specialist. We hope that the accruals that this study will generate will provide sufficient incentive for sites to participate despite this. One solution that other trusts have adopted is to ask the anaesthetic departments in their sites to bear the cost of archival.
We hope that this will not be an obstacle for your site's participation."

Can we archive yet?
We are currently doing data-linking with NHS Digital to obtain the ONS-linked mortality for all the patients who were included in the study. We kindly ask sites to delay the archival until the 1 year anniversary of the data collection: March 27th 2018

Should I complete a Delegation Log?
We will not be asking for a delegation log. We will instead keep a separate record of all the individuals involved in the study at the sites, and we are happy to provide this separately to the sites to confirm with their own records at the end of the study.

Do the Anaesthetists need to have up to date GCP Certificates?
Just the PIs need to have GCP, but we encourage all involved be GCP trained as well.

Is Caldicott approval required?
Caldicott approval is not required because of our Section 251 exemptions and the CAG approvals.

Can we archive yet?
We are currently doing data-linking with NHS Digital to obtain the ONS-linked mortality for all the patients who were included in the study. We kindly ask sites to delay the archival until the 1 year anniversary of the data collection: March 27th 2018

Who will cover archiving costs?
Unfortunately the research costs for archival will have to be borne by the trust. We will not be providing funding to cover this. We had originally applied to the NIHR Portfolio for archival to be a cost paid for by the NIHR CRN, however this was revised by the ACORD specialist. We hope that the accruals that this study will generate will provide sufficient incentive for sites to participate despite this. One solution that other trusts have adopted is to ask the anaesthetic departments in their sites to bear the cost of archival.
We hope that this will not be an obstacle for your site's participation."