NHS Digital Data Release Register - reformatted

Liverpool Heart and Chest Hospital NHS Foundation Trust

Project 1 — DARS-NIC-303379-H4C8H

Opt outs honoured: No - consent provided by participants of research study (Consent (Reasonable Expectation))

Sensitive: Non Sensitive, and Sensitive

When: 2020/05 — 2020/05.

Repeats: One-Off

Legal basis: Health and Social Care Act 2012 – s261(2)(c)

Categories: Identifiable

Datasets:

  • Hospital Episode Statistics Outpatients
  • Hospital Episode Statistics Critical Care
  • Civil Registration - Deaths
  • HES:Civil Registration (Deaths) bridge
  • Hospital Episode Statistics Accident and Emergency
  • Hospital Episode Statistics Admitted Patient Care
  • Bridge file: Hospital Episode Statistics to Diagnostic Imaging Dataset
  • Diagnostic Imaging Dataset

Objectives:

RIPCORD (Does routine pressure wire assessment influence management strategy at coronary angiography for diagnosis of chest pain?) was a proof of concept study designed to assess the feasibility and management impact of routine assessment of fractional flow reserve (FFR) in patients undergoing angiography for diagnosis and management of stable chest pain. FFR is an invasive measure of the change in pressure across any given narrowing (or ‘stenosis’) of a coronary artery, acquired using a pressure wire placed in the vessel. It has been shown that using this measurement to determine whether a stenosis is clinically significant, and requires treatment, results in improved clinical outcomes for patients. Despite this weight of evidence, FFR is still not routinely used in many centres across the UK and the world. There are multiple reasons for this, including: cost implications, extra procedural time/complexity and lack of expertise. In RIPCORD (compromising a total of 200 patients) when FFR data were available to cardiologists who had previously only had angiographic data, there was a change of the management plan (between options: optimal medical therapy (i.e. tablet treatment), Percutaneous coronary intervention (i.e. balloons/stents) or coronary artery bypass graft surgery (CABG)) in 26%. This was the result of a change in the classification as to whether coronary arteries had a “significant” stenosis in 32% between angiographic assessment alone versus FFR data. These data suggest that routine FFR measurement in such patients at the stage of diagnostic angiography would have profound impact on improving the management strategy by tailoring therapy. It is likely that a strategy of systematic FFR-guided assessment of coronary artery disease will be associated with more effective resource utilisation, improved patient-reported quality of life and better clinical outcome when compared with angiographic guidance alone. Based upon this hypothesis, a large scale and definitive randomised trial of this strategy was designed (RIPCORD 2). RIPCORD 2 is an open-label, prospective, dual-arm, multi-centre, randomised controlled trial across 17 percutaneous coronary intervention (PCI) centres in the UK (16 in England/Wales and 1 in Scotland). These are all specialised local or regional tertiary cardiac centres where patients with heart attacks or angina can undergo invasive coronary angiography and, if necessary, treatment using balloon/stent type technology. Full Research Ethics Committee and local site approvals were gained. Patients were screened for inclusion/exclusion criteria at two time points (before, and again at the time of angiography). Patients eligible and willing to participate underwent a formal informed consent process before the point of angiography. Patient information sheets and consent forms were approved by NRES and local hospital process prior to implementation. The consent process included gaining permission to access subsequent data from Hospital Episode Statistics (HES) for follow up purposes. Eligible patients were randomised 1:1 to either conventional angiography alone or routine pressure wire assessment (FFR) in all main vessels. Study Outcome measures are as follows: - Primary Economic Outcome Measure: The primary economic outcome measure will be a comparison of health care costs, observed over the 12 month follow-up period. The case record form will be designed to capture key elements of resource utilisation during the index procedure and up to the confirmation of the initial management strategy. Costs for each patient will then be calculated for hospitalisation events, reported by the UK Hospital Episode Statistics. The study will capture details of the index and all subsequent hospitalisation events, elective procedures, outpatient appointments and investigations. A standardised UK cost model will be applied for reported diagnostic and procedural codes. The primary analysis will describe the range of observed total costs and compare the mean total costs incurred in the two groups created at randomisation (or median, depending on the distribution of observations). - Primary Quality of Life Outcome Measure: This analysis will compare the mean (or median), patient reported quality of life scores using the EQ-5D-5L health questionnaire. Patients will complete the questionnaire at 12 months from randomisation (plus or minus one month). - Secondary Outcomes: Clinical Events Clinical event rates at one year derived from HES; • All-cause mortality • Number of hospitalisation events and total hospital days • Hospitalisation events coded as: - Cerebro-vascular accident (CVA) - Myocardial infarction Coronary revascularisation. This analysis will involve a pre-specified subgroup analysis of: • Planned revascularisation - if declared as the index strategy • All additional revascularisation events Additional descriptive statistics will be provided (for the two randomised groups) in terms of the overall pattern of hospitalisation events in terms of diagnostic classification and procedures performed. A total of 1100 patients were recruited into the study (1062 in England/Wales, 38 in Scotland). The study is now at the end of the 12 month follow up period for the last of those 1100 patients. As such the study team now intend to request the HES data for 12 month follow up of the complete cohort, which all the participants consented to. Given the primary and secondary outcome measures of the study, researchers clearly require a complete set of pseudonymised HES data relating to all hospital admissions anywhere in the UK with related codes, as well as mortality data (both dates and causes of death) for the 12 month follow up period. Data on all hospital admissions for any purpose are required because ultimately it is very hard to define what is 'cardiovascular' or 'non-cardiovascular' when it comes to admissions for various reasons. There are some cases that are obvious: e.g. an admission with a severe heart attack, or at the opposite end of the spectrum an admission with a stubbed toe. The reality is however is that life is not always that black and white. Frequently people are admitted with multiple issues, some may be related to cardiac causes and others not. Similarly trying to define exact causality is not always that obvious. An extreme example would be someone who gets dizzy and blacks out in the middle of the road, hit by a bus and breaks their leg. The cause for admission might seem to be orthopaedic but actually it all can be traced back to the heart medication they received that gave them low blood pressure. Another example would be if one patient in a particular arm of the study gets inferior heart treatment that might mean they get depression because they have much worse angina, as a result they could be admitted with depression or even pneumonia because they are less active, etc. Additionally the nature of the study means that the group a participant is in would make it more or less likely that they underwent a coronary bypass operation. It is very common to get non-cardiac complications after these operations (e.g. mobility problems, wound infections etc) which are also very important to identify. There is a lot of 'grayscale' so trying to distinguish accurately between those admissions that will be directly related to activity in the trial and those that are not is essentially impossible. By far the cleaner and more scientifically sound methodology is to compare all healthcare costs incurred in both groups. This is the only method free of any potential confounders or error. This would be the best and only good way of assessing for differing financial implications of the two treatment strategies in the study. Secondly, a reduction in total healthcare costs is much more meaningful as a study outcome as it translates into real world cost savings that can be interpreted by the scientific community and related to daily practice. Due to the all encompassing nature of the primary health economic outcome measure and the requirement for all hospital admissions and certain outpatient activities (including procedures/investigations) the study will require the following datasets: HES Admitted Patient Care, Civil Registration (mortality), HES Accident and Emergency, HES Critical Care, HES Outpatients, Diagnostic Imaging. The trial resource utilisation model requires researchers to take into account any patient hospital contact that relates to their heart health: defining this requires a broad knowledge of their outpatient and inpatient contact with the NHS and tests generated. All admission and outpatient appointments will therefore be components of the economic model, since the resource utilisation model the study has chosen does not pre-specify specific datapoints that trigger costs, but takes account of complete NHS interactions. Further, admissions including revascularisation, angiograms, heart attack, heart assessment, etc not only go into overall cost analysis, but are part of pre-specified secondary endpoints for the trial, and these will be identified in HES Accident and Emergency, HES Critical Care and HES Outpatients. Digital Imaging Dataset (DIDs) will provide valuable resource data for investigations including MRI and CT, etc many of which will be referred via Outpatients. DIDs data are required in order to inform the uptake of all the digital imaging tests (most commonly MRI/CT) by patients in the randomised trial. This is completely consistent with the trial primary outcome which is complete resource utilisation within the NHS. Thus, imaging utilisation is as important as any other aspect of interaction with the NHS to feed into the model. The nature of the imaging test is not specifically important, because the resource model (one of several options for resource models commonly employed in research trials) requires information on all DID. Mortality data will provide cause of death details which will also contribute to the overall cost analysis. The admission related codes for each patient will be used in conjunction with UK tariffs to calculate the overall incurred healthcare costs for the primary outcome measure. This is not achievable by any other means than access of the HES coded data. The nature of each admission (i.e. specific codes) is required for this costing exercise and also to calculate the incidence of other clinical secondary outcome measures. Mortality data is also required as an outcome measure and is integral to understanding any important differences observed in the two groups. The study requires a one-off extract of data for patients for a 12 month period from the date of randomisation. The study team will minimise the data requested to that which is necessary to calculate outcome measures; datasets not relevant to this will not be asked for. For example, the study is not requesting maternity or mental health datasets. Within the requested datasets the fields requested have been limited to essential items. For instance, within the APC dataset, the study has not requested patient level, geographical, socio-economic or psychiatric data. The organisation involved is primarily the ICECAP research group based at Liverpool Heart and Chest Hospital NHS Foundation Trust. The study had 17 recruitment sites in the UK, but the recruitment sites are not involved in the processing of the data being requested. The chief investigator and sponsor for the study are based at University Hospital Southampton, however the ICECAP research group at Liverpool Heart and Chest Hospital (where the co-principal investigator resides) is the designated site/agent responsible for collection, storage and analysis of the study data. This is the only site where data returned from NHS Digital will be stored or accessed. Liverpool Heart and Chest Hospital NHS Foundation Trust and University Hospital Southampton NHS Foundation Trust are joint data controllers and both organisations will process the data. However, Liverpool Heart and Chest Hospital will be the only site where data will be accessed or processed. No work on this study has or will be undertaken outside the UK. The study was funded by the Sponsor- the University Hospital of Southampton. Explicit written informed consent was given by trial participants for the access to their electronic healthcare records. The study cites GDPR Articles 6(1)(e) and Article 9 (2)(j) as lawful bases to process the data. The study requires this data on one occasion, at one time point. The purposes of the study are clearly outlined and were explained to participants during the consent process. These have not, and will not, change during the study. All patients were given contact information should they wish to withdraw from the study. To date, no participants have withdrawn their consent. The study has already received equivalent datasets from NHS Wales Informatics Service in Wales. The HES data from England/Wales etc will be processed separately. The key outcome data (e.g. healthcare spell costs and specific clinical outcomes e.g. MI codes) will then be extracted into a master pseudonymised database. This is the point at which data from different sources will be linked; this therefore does not involve the entirety of the HES datasets, just the key information required to achieve the study results. The study team take measures to ensure there is no data protection breach including pseudonymising the data, by removal of patient identifying information other than the study ID.

Expected Benefits:

The results of this study could have a profound impact on recommended practice in cardiology. This has the potential for significant clinical improvements for patients in the future, in both prognostic terms and with regards to overall quality of life. It is therefore important for patients that the data and results are disseminated to enable a change in practice. Patients gave their consent for use of this data and it is therefore the obligation of the investigative team to complete the trial and publish its findings for the betterment of evidence based clinical practice in the future. Findings have the potential to change clinical practice guidelines on a national and international level; if positive the trial may lead to the recommendation of routine pressure wire usage in patients undergoing coronary angiography. Given the high prevalence of coronary artery disease and the large numbers of patients undergoing invasive procedures worldwide, this would potentially impact the care of thousands of patients. Demonstration of a clinical benefit to patients could substantially increase the use of pressure wire technology on an international level. There are also potentially large health economic implications. If the study's hypothesis is proven to be true then a switch to a routine FFR guided approach in the investigation of coronary artery disease could lead to significant savings in healthcare costs. This is of particular importance in a constrained health care system of limited resources and funding such as that in the UK. The potential health economic benefits are hard to quantify until the results of the study are available. However, if a significant cost saving per patient is proven, given the huge number of patients treated for coronary artery disease worldwide, the potential overall benefits would be large. For example, in 2016 there were over 260,000 invasive coronary angiograms and over 100,000 PCI (percutaneous coronary intervention) procedures performed in the UK alone. Therefore, even a small cost saving per patient will equate to a large cost reduction overall. This could financially benefit health care systems locally and nationally. Given the frequency of guideline production with cardiology, any alterations as a result of this trial's results may have the potential to significantly change practice within a short space of time (e.g. within 5 years). Although not the primary purpose of this study, the formulation of results will also contribute towards a post graduate MD thesis for one of the main researchers.

Outputs:

The output expected would be at least one, or potentially multiple, high-quality scientific papers on the results of RIPCORD 2. As there are different types of outcomes in the trial (traditional clinical event end points compared to health economic analyses) this may lend itself to multiple different publications of results. The expectation is for the primary publication of the randomised control trial results to be in a high impact journal. As a reference example, the original RIPCORD study was published in Circulation: Cardiovascular Interventions. Due to the larger nature of this second study with potentially greater clinical impact the study anticipate publication in an equivalent or even higher impact journal. Due to the large amount of ancillary data captured the study may aim for additional smaller papers on specific aspects e.g. relating to procedural details. The target data for the primary paper would be April 2020 but depends on the speed at which the data can be extracted and subsequently analysed. The final article will not include any patient identifiable information. The final article will be reviewed by the research team prior to submission. As stated, it is not yet clear which journal will be targeted for publication as this will depend on circumstances nearer the time of publication. The data in outputs will be anonymised, aggregate level data with small numbers suppressed in line with HES analysis guide. It is also likely, given the potential large clinical impact, that results will be presented during at least one large international cardiology conference, either ahead of or in conjunction with publication. Depending on timing of our final results, the study will most likely target the American College of Cardiology (ACC), or European Society Cardiology (ESC) conferences in 2020. In this aspect the results may be cited on cardiology related websites/social media/news platforms surrounding those conferences. The research may be cited in future publications by other authors. Again, no patient identifiable information will be included in the results presented in this domain. Results of the study will be available to the public and presented in appropriate forums. This may include presentation to patient groups (such as the SURE group in Merseyside), public news reports and on websites such as clinicaltrials.gov. There is a clear and real potential for these results to affect clinical practice guidelines in cardiology, both on a national and international level. If significant the results could lead local, national and international policy making bodies e.g. National Institute of Clinical Excellence or European Society of Cardiology to update relevant clinical practice guidelines.

Processing:

Description of the data flow: 1) Liverpool Heart and Chest Hospital NHS Foundation Trust will securely transfer a file of identifying information (Study ID, NHS Number, Gender and date of birth) plus each participant's individual start and end dates to NHS Digital. For avoidance of doubt, participants' names will not be supplied to NHS Digital. 2) NHS Digital will link the cohort to HES, DIDs and mortality data using NHS Number, date of birth and gender. 3) NHS Digital will generate and transfer a file of linked data to the secure environment at Liverpool Heart and Chest Hospital NHS Foundation Trust. The data will be securely linked and held within Liverpool Heart and Chest Hospital NHS Foundation Trust and will be available for access only by the research team. The dataset will be used to identify all hospital admissions, related codes and mortality data (including dates and cause of death) up to 12 months after recruitment date for each individual in the cohort. Data management and analysis is to be conducted within the ICECAP research unit at the Liverpool Heart and Chest Hospital NHS Foundation Trust. The specific methodological activities involved in the processing of data are as follows: The HES Admitted Patient Care, A&E, Critical Care, Outpatients and Civil Registration data will be requested for all trial participants who provided consent and who have not subsequently formally withdrawn from the trial. A one off request for data will be made for all applicable participants after the 12 month follow up period has ended for all 1100 in the cohort. Each participant has a unique start date for follow up (the day at which they were randomised in the trial). The study is requesting 366 days of follow up for each individual (for example from 1/5/2017 to 1/5/2018 inclusive); this is to allow for the fact that participants were randomised at different times of day into the study and thus ensures a full 365 days of follow up is captured in all cases. Hospital activity (particularly in Admitted Patient Care datasets) may cross the start and end dates of follow up in many patients; the study seeks to capture all this data. The primary data search will be performed using hospital episode/spell admission dates, so that any admission starting between day 0 up until before, or on, day 366 of the follow up period is collected. At the beginning of the follow up period for each participant NHS Digital will also have to search by discharge date to ensure any hospital spell with a discharge date after day 0 but an admission date preceding it, is still collected. (This will pick up HES data for the index hospital admissions in the study). The research team will send Study ID, NHS Number, date of birth and gender to NHS Digital. The file will also contain the unique start and end dates for the follow up period for each participant (day 0 and day 366 respectively). Within the data file returned by NHS Digital, the research team need to know the study ID associated with each individual’s data within the time period of interest. No other identifiers other than the study ID are required in the returned file; the returned data will therefore be pseudonymised. The data file will be securely transferred back to the ICECAP research unit at the Liverpool Heart and Chest Hospital NHS Foundation Trust, again using NHS Digital’s Secure Electronic File Transfer SEFT system. The Study ID will be used to link the data received to that already recorded for each participant in the case record forms. Once the pseudonymised data has been downloaded and stored securely, there will be no further need to identify individuals. Data will be stored electronically on a password protected ICECAP group drive on the Liverpool Heart and Chest Hospital NHS Foundation Trust central servers, located in an access-controlled server room and connected to the main trust network, located behind a firewall. Physical access is limited to Computer Services Department staff. Data will be encrypted using industry standard techniques meeting the Data Security and Protection Toolkit (RBQ). The data will not be transferred to an additional location. Analyses will be completed using the pseudonymised data set and will be viewed in an encrypted Microsoft Excel/Access database on a network PC in the ICECAP research unit. Access will be limited to the research group involved in the study. All data will be stored and accessed at the Liverpool Heart and Chest Hospital NHS Foundation Trust at all times. All personal data in this trial is kept strictly confidential and is being handled, stored and destroyed in accordance with GDPR. The data in its pseudonymised form would not be identifiable alone. To be identifiable this study ID would have to be converted back to corresponding NHS numbers/identifiers- this information is kept in a completely separate database in a different secure location on a different server- again with limited access to certain individuals. Only individuals who are part of the ICECAP research group and part of this study, will have access to the data. The Chief Investigator of the study, who is employed by University Hospital Southampton NHS Foundation Trust, may need and will have access to view the data but only at the LHCH site (the data will not be accessible from the Southampton site). There will be no external data flow of identifiable information outside of LHCH. No other collaborator of this study will have access to the data received from NHS Digital (that includes the sites involved in the data safety monitoring etc). All organisations party to this Agreement will comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by Personnel (as defined within the Data Sharing Framework Contract - i.e. employees, agents and contractors of the Data Recipient who may have access to that data).