Administration and data collection

Summary

Management of data relating to interventional procedures includes the documentation and reporting of important aspects of care pertinent to procedure details including outcomes, and has implications on reimbursement and quality assurance. The continuous collection of data for the purpose of quality assessment is an essential part of clinical governance, and the standardisation of data definitions allows the integration between various quality improvement initiatives, including clinical registries. Quality indicators (QIs) have emerged as tools to measure performance and patient outcomes, because they facilitate the public-reporting of data in an intuitive, yet scientific format designed to benchmark quality of care.. Datasets for percutaneous coronary interventions (PCI) are standardised on a national level, with ongoing efforts to harmonise registries internationally. The collection of ‘real world’ registry data provides important additional information to that garnered from randomised controlled trials, and both underpin research and efforts to improve patient care.

Introduction

The collection of data that arise from any interventional procedure is a key element of quality assurance. In addition, there are important administrative and practical issues within any interventional department. In most centres there will be software designed to capture interventional data running on the catheter laboratory computers. The extent and functionality of this software however varies, ranging from complete solutions that provide a reporting function as well as a database and patient administration systems, to very focused systems which only record the baseline parameters and procedural details. Often, the relevant data will have to be retrieved from the electronic health records (EHRs) so that they can be entered into a separate interventional database.

Data collection and recording is important for several reasons:

DOCUMENTATION OF THE PROCEDURE

A basic requirement is the documentation of the procedure that has been performed. The record must include the patient identifier, the name of operators and participating staff as well as the type of procedure that was performed. A procedure log should provide information on the time when essential steps of the procedure were accomplished and should record the equipment that was used. Documentation of the outcome and complications is important. Most software will provide a reporting function that provides the essential data while offering the option of adding specific details concerning the case as well as recommendations for post-interventional care.

REIMBURSEMENT

Increasingly, reimbursement for procedures is processed through coding of procedures and assigning a tariff by linking with patient related factors. The impact of presentation (emergency vs. elective) and comorbidities can be very relevant. Lack of appropriate data can result in inaccurate remuneration. It is essential therefore to collect the required data at the time of the procedure. Advanced hospital software systems can link with the catheter laboratory software and enable an exchange between the databases.

QUALITY ASSURANCE

An evaluation of key components of performance is an important tool for assuring the quality of care. The data required comprise more than the procedure-related data. Risk stratification tools rely on parameters from the patient’s history, acute presentation, comorbidities as well as procedural outcomes and follow-up information.

MEDICO-LEGAL PERSPECTIVE

If an operator or a department should be legally challenged concerning a procedure, the documented data will play a crucial role in the process. Procedural data logged in the database will be scrutinised. It is advisable to check the quality of the procedure related information, and discuss the relevance of the information provided for legal purposes.

REGISTRIES

The analysis of data in multicentre, national or international registries enables the comparison of processes and outcomes of care across centres. However, merging data from different systems requires the standardisation of definitions and basic parameters (see below). The design of the registry and the selection of its data variables determine the ability of the registry to provide meaningful information relevant to the clinical questions asked.

RESEARCH

Research, whether performed locally or in cooperation with other centres, relies on accurate data entry. The database can be used to identify patients that fulfil entry criteria for studies, or to analyse procedural outcomes.

It is therefore apparent that a good system of data collection and administration will have a direct influence on the quality of care, efficiency and performance of a department. It is desirable to implement a system in which all the relevant data along the patient pathway are captured, as well as to provide an infrastructure that ensures good data quality. Data management is a prerequisite for a good interventional department, and a marker of quality.

Procedure report

The results of a diagnostic or interventional procedure need to be reported. Ideally, the report is generated by an integrated system within EHRs, using previously entered data incorporating data captured during the procedure. While templets for standardised reports have been developed [11. Sanborn TA, Tcheng JE, Anderson HV, Chambers CE, Cheatham SL, DeCaro MV, Durack JC, Everett AD, Gordon JB, Hammond WE, Hijazi ZM, Kashyap VS, Knudtson M, Landzberg MJ, Martinez-Rios MA, Riggs LA, Sim KH, Slotwiner DJ, Solomon H, Szeto WY, Weiner BH, Weintraub WS, Windle JR. ACC/AHA/SCAI 2014 Health Policy Statement on Structured Reporting for the Cardiac Catheterization Laboratory. Circulation. 2014 129(24):2578-609. ], there is no uniform report structure and no agreed guidance on the core set of variables required for a procedural report. The extent of the existing reports varies considerably, with each department having a customised structure tailored to their specific setup.

It is important that a procedure report captures the data variables which are needed to assess the quality and the appropriateness of the procedure, as well as to allow risk adjustment and outcomes evaluation. As such, the data required can be grouped under the following headings:

• Patient demographics, including unique patient identifiers, date of birth, past medical history and details of comorbid conditions.
• Clinical presenting features and indications for any procedures performed.
• Procedural details (including date and time of the procedure, procedures performed, equipment used and adjunctive medication).
• Hospital identifier and operator names, including other personnel involved in critical aspects of the patient’s care (such as anaesthetists, if applicable).
• Outcomes, including adverse events, complications, and patient experience.
• A report should summarise these features and include recommendations for post-procedural care, a discharge plan and any planned subsequent strategies.

Additional information that may guide the post-interventional care can be included in the report, alongside a visual illustration of the relevant angiographic findings. The operator should write the report in such a way that makes it possible to understand the clinical relevance of the findings and outcomes. For example, the report should clearly state the description of any bystander lesions, whether any investigations or treatment is still needed, and a detailed strategy for future medical therapy.

Audit and registries

CLINICAL GOVERNANCE

The systematic approach to maintaining and improving the quality of patient care in a health system is referred to as clinical governance. It is a framework that embodies three key attributes: recognisably high standards of care, transparent responsibility and accountability for those standards, and a constant dynamic of improvement. Clinical governance is composed of several service improvement processes that work together to improve patient care, these include:

• Education and training: such as continued professional development to ensure physicians are aware of current medical literature.
• Clinical effectiveness: the appropriateness, efficacy, cost effectiveness and safety of different therapies.
• Research and development: the application of new research findings into clinical practice and guideline development.
• Openness: poor practice can thrive if it occurs out of the scrutiny of peers, and while openness is important, it must respect appropriate individual patient and practitioner confidentiality.
• Risk management: addressing and minimising risks to patients, physicians and organisations.
• Clinical audit.

CLINICAL AUDIT

Clinical audit is the component of clinical governance that offers the greatest potential to assess the quality of routine care provided to patients. It provides mechanisms for evaluating the quality of care against recognised standards, and identifies areas for improvement. Quality indicators (QIs) have been increasingly used to monitor and report important aspects of healthcare delivery. They can assess the quality of clinical services, and evaluate the effectiveness of quality improvement interventions.

Clinical audit can be divided into four domains, each addressing different aspects of patient care:

• Structure: This defines the environment in which treatment is being delivered, including issues such as the number of PCI labs for a given population, the staffing of such labs, the level of services available and the volume of activity undertaken.

• Appropriateness: was a particular patient treated with the appropriate therapy? Given the clinical setting, including, for instance, the presentation characteristics and comorbidities, was the intervention indicated? Were correct adjunctive therapies used?

• Process: this is the mechanisms by which care was delivered; for example, the speed of delivery of primary PCI, the length of stay, the completeness of revascularisation, the staging of the procedures, etc.

• Outcome: while there is often a focus on major adverse cardiac and cerebrovascular events, measures of quality of life as well as patient reported outcomes are increasingly used. (Please see Chapter: ‘Quality of life assessment’ by Mattie Lenzen, Ron Van Domburg, Susanne S. Pedersen)

The process of quality assessment has been described as a loop or spiral. Following the identification of an area with gaps in care, key domains relevant this area of interest are agreed, QIs are developed from literature review and consensus development [22. Aktaa S, Batra G, Wallentin L, Baigent C, Erlinge D, James S, Ludman P, Maggioni A, Price S, Weston C, Casadei B, Gale CP. European Society of Cardiology methodology for the development of quality indicators for the quantification of cardiovascular care and outcomes. European Heart Journal - Quality of Care and Clinical Outcomes. 2020. ]. Clinical practice is then measured against these indicators using either prospectively or retrospectively collected data. If there are areas where practice falls short of accepted standards, then a strategy to improve care is defined and instituted. Closing the ‘loop’ involves a repeat audit to assess if suggested improvements to practice have occurred, and that standards have improved ( Figure 1)

The process implied here is a cycle or spiral where each cycle aspires to a higher level of quality. The National Institute for Clinical Excellence in the United Kingdom have published a guide entitled, “Principles for Best Practice in Clinical Audit” [33. Principles for Best Practice in Clinical Audit. National Institute for Clinical Excellence (NICE), 2010.
A description of the principles of clinical audits with detailed explanations about setting up and running local clinical audit projects] which describes this process in some detail (http://www.nice.org.uk/media/796/23/BestPracticeClinicalAudit.pdf).

An audit does not mandate a cycle. Benchmarking components are important, even without overt direct feedback loops.

STANDARDISATION OF DATASET DEFINITIONS

The widespread implementation of electronic health records (EHRs) for clinical care, research, and quality assurance initiatives creates an opportunity to transfer data seamlessly between systems (interoperability), and thus reduce the burden of data collection [44. Bhatt DL, Drozda JP, Shahian DM, Chan PS, Fonarow GC, Heidenreich PA, Jacobs JP, Masoudi FA, Peterson ED, Welke KF. ACC/AHA/STS Statement on the Future of Registries and the Performance Measurement Enterprise. Journal of the American College of Cardiology. 2015 66(20): 2230-2245. ]. However, the heterogeneity of data definitions within interventional cardiology across various settings generates a significant barrier to developing an infrastructure in which continuous research and quality improvement activities are embedded into routine healthcare delivery [55. Bothwell LE, Greene JA, Podolsky SH and Jones DS. Assessing the Gold Standard — Lessons from the History of RCTs. New England Journal of Medicine. 2016 374:2175-2181. ]. Moreover, this variation hampers the integration and utilisation of data between the components of clinical care (e.g EHRs, procedure reports, primary care datasets), and limits the opportunity to improve the efficiency of patient care.

The American College of Cardiology (ACC) and the American Heart Association (AHA) have recently published their data standards for coronary revascularisation [66. Dehmer GJ, Badhwar V, Bermudez EA, Cleveland JC, Cohen MG, D’Agostino RS, Ferguson TB, Hendel RC, Isler ML, Jacobs JP, Jneid H, Katz AS, Maddox TM, Shahian DM. 2020 AHA/ACC Key Data Elements and Definitions for Coronary Revascularization: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Clinical Data Standards for Coronary Revascularization). Circulation: Cardiovascular Quality and Outcomes. 2020 13(4): e000059. ]. In addition, the European Society of Cardiology (ESC) has launched the European Unified Registries On Heart Care Evaluation and Randomized Trials (EuroHeart) project (https://www.escardio.org/Research/euroheart) which aims to develop standardised data variables for a number of cardiovascular disease (CVD) conditions and interventions, including PCI and valvular heart disease [77. Wallentin L, Gale CP, Maggioni A, Bardinet I, Casadei B. EuroHeart: European Unified Registries On Heart Care Evaluation and Randomized Trials. European Heart Journal. 2019 40(33): 2745-2749. ]. These efforts not only provide harmonised vocabularies and common definitions in interventional cardiology, but also describe the methods by which data are collected. Furthermore, the EuroHeart variables will be implemented into an IT-platform which may serve as the means for clinical registries, and adopt sets of QIs that may help quantify the quality of care for CVD [22. Aktaa S, Batra G, Wallentin L, Baigent C, Erlinge D, James S, Ludman P, Maggioni A, Price S, Weston C, Casadei B, Gale CP. European Society of Cardiology methodology for the development of quality indicators for the quantification of cardiovascular care and outcomes. European Heart Journal - Quality of Care and Clinical Outcomes. 2020. ].

The potential of standardised dataset definitions is increasingly recognised. However, there remain substantial variations in the definitions of important data variables relevant to interventional cardiology, including those where universal agreement on their definition exist [88. Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, White HD. Fourth Universal Definition of Myocardial Infarction (2018). Journal of the American College of Cardiology. 2018 72(18):2231-64. ]. For instance, there is marked heterogeneity in the definition of myocardial infarction after revascularisation across major clinical trials [99. Gregson J, Stone GW, Ben-Yehuda O, Redfors B, Kandzari DE, Morice M-C, Leon MB, Kosmidou I, Lembo NJ, Brown WM, Karmpaliotis D, Banning AP, Pomar J, Sabaté M, Simonton CA, Dressler O, Kappetein AP, Sabik JF, Serruys PW and Pocock SJ. Implications of Alternative Definitions of Peri-Procedural Myocardial Infarction After Coronary Revascularization. Journal of the American College of Cardiology. 2020 76:1609-1621. ]. This variation has important repercussions on the interpretation of the findings of these studies, and, subsequently, on knowledge development, and highlight the challenges involved in the widespread adaptation and uptake of developed definitions in various clinical and research activities.

Another challenge is the accomplishment of machine-interpretable nomenclature. The development of consensus terminologies and definitions for interventional cardiology variables enables systems to have shared data models, and therefore, homogeneously understand exchanged data at the level of medical concepts [1010. Gliklich R, Dreyer N, Leavy M, eds. Registries for Evaluating Patient Outcomes: A User’s Guide. Third edition. Two volumes. (Prepared by the Outcome DEcIDE Center [Outcome Sciences, Inc., a Quintiles company] under Contract No. 290 2005 00351 TO7.) AHRQ Publication No. 13(14)-EHC111. Rockville, MD: Agency for Healthcare Research and Quality. April 2014. ]. However, a translation of these variables into a computational language that can provide syntactic representation of information is needed. This translation maintains the ‘meaning’ of the data (semantic interoperability), and requires the establishment of a framework of standardised ontologies and coding systems [1111. Liyanage H, Krause P, de Lusignan S. Using ontologies to improve semantic interoperability in health data. BMJ Health & Care Informatics. 2015 22. ]. Consequently, data that are routinely collected in EHRs can be amenable for use in quality registries and clinical trials [1212. Fitzpatrick T, Perrier L, Shakik S, Cairncross Z, Tricco AC, Lix L, Zwarenstein M, Rosella L and Henry D. Assessment of Long-term Follow-up of Randomized Trial Participants by Linkage to Routinely Collected Data: A Scoping Review and Analysis. JAMA Network Open. 2018 1:e186019-e186019. ].

PROSPECTIVE REGISTRIES AND RANDOMISED CONTROLLED TRIALS

The evidence on which decisions about optimal care are based come from many sources. The highest quality information is generally accepted as being derived from the results of randomised controlled trials (RCTs), but there are important limitations in the interpretation of such data. Registries provide a different way of trying to understand medical therapies, and in many ways are complimentary to randomised trials. Extremely important data have been derived from registries, and they have a crucial role for evidence development. Registries come with their own set of methodological issues that must be understood to avoid pitfalls in the interpretation of results. A comparison of some of the key features is given in Table 1.

Perhaps the most important difference between a randomised controlled trial and a prospective registry is that the characteristics of patients recruited into these trials, although similar in both arms of the trial, are usually quite different from the demographic features of the population being treated. The logistics of being involved in a trial, the requirement for informed consent, and the exclusion criteria for most trials means that the recruited patients are usually fitter, with less comorbid conditions than the patients treated in every day practice. This limits the ability for RCTs to assess treatment effects and safety in high-risk populations, and restricts the generalisability of their findings.

For example, in major cardiovascular society guidelines, level of evidence (LOE) A recommendations are usually derived from multiple RCTs or a single, large RCT. However the proportion of LOE A recommendations in the ACC/AHA and the ESC Clinical Practice Guidelines are around 8.5% and 14.2%, respectively [1313. Fanaroff AC, Califf RM, Windecker S, Smith SC Jr, Lopes RD. Levels of Evidence Supporting American College of Cardiology/American Heart Association and European Society of Cardiology Guidelines, 2008-2018. JAMA. 2019 Mar 19 321(11):1069-1080. ], highlighting the need for other sources for evidence development. In addition, the representation of acute myocardial infarction patients in RCTs is declining over the years, with better adherence to guideline-indicated care observed in centres enrolled in RCTs [1414. Fanaroff AC, Vora AN, Chen AY, Mathews R, Udell JA, Roe MT, Thomas LE, Wang TY. Hospital participation in clinical trials for patients with acute myocardial infarction: Results from the National Cardiovascular Data Registry. Am Heart J. 2019 Aug 214:184-193. ]. Previous snapshot registry survey showed that only about 11% of the patients enrolled would have been eligible for inclusion in major RCTs, and even in using only the major exclusion criteria, only 36% would have been eligible for these trials [1515. Hordijk-Trion M, Lenzen M, Wijns W, de JP, Simoons ML, Scholte op Reimer WJ, Bertrand ME, Mercado N, Boersma E. Patients enrolled in coronary intervention trials are not representative of patients in clinical practice: results from the Euro Heart Survey on Coronary Revascularization. Eur Heart J. 2006 27:671-8. ]. Taken together, registries form an essential part of clinical research and provide different perspective to the evidence derived from RCTs. Registries provide “real world” data, and may capture long-term trajectories of meaningful outcomes. Some clinical questions are best addressed by a well-designed, prospective registry when great care is invested in setting them and manage data collection throughout the study period.

Recently, the emergence of registry-based RCTs (R-RCTs) may provide an alternative to traditional RCTs [1616. Götberg M, Christiansen EH, Gudmundsdottir IJ, Sandhall L, Danielewicz M, Jakobsen L, Olsson S-E, Öhagen P, Olsson H, Omerovic E, Calais F, Lindroos P, Maeng M, Tödt T, Venetsanos D, James SK, Kåregren A, Nilsson M, Carlsson J, Hauer D, Jensen J, Karlsson A-C, Panayi G, Erlinge D and Fröbert O. Instantaneous Wave-free Ratio versus Fractional Flow Reserve to Guide PCI. New England Journal of Medicine. 2017 376:1813-1823. ], and enhance both the generalizability of RCTs and their cost effectiveness [1717. Ford I and Norrie J. Pragmatic Trials. New England Journal of Medicine. 2016 375:454-463. ]. Please also see chapter “Registry studies and post-marketing surveillance” by Stefan James, Ole Fröbert. R-RCTs may obviate the problem that many of the sickest patients are never recruited into trials, and the integration between registries, EHRs, and QIs may help improve the feasibility and efficiency of conducting high-quality traditional and registry-based RCTs. This integration may be facilitated by the following:

• An agreed dataset with harmonised definitions for data variables.
• Standardised sampling coverage. This should result in the participation of all appropriate centres, and be as inclusive and representative of the patients being treated as possible. It might also include a randomised selection of centres, but ideally will have 100% participation.
• Consecutive patient enrolment.
• Implementing well-designed user interface for data collection and reporting that includes immediate checks for internal consistency, plausibility (including range checks) and validation.
• Ready access to technical and clinical support.
• Systems for encryption and secure transfer of data to a central server with appropriate levels of data security.
• Once submitted, data submissions should be checked for completeness, accuracy and internal consistency,
• Centralised analysis of data should be overseen by professional statisticians.
• Linkage to national datasets, if possible (for example national mortality tracking).
• Appropriate consent for participation in registry studies must be obtained.
• Feedback to all data collecting centres to encourage timely data entry and high levels of data completeness.
• Where possible, there should be independent validation of at least a randomly selected sample of data.
• The names of all participating investigators should appear in the published registry report.
• Sponsorship of the registry should be clearly stated so that any commercial bias can be easily identified.

l One principle investigator and a small steering group should be designated to maintain administrative order, adjudicate disagreements and encourage timely submission of documents and data analysis.

Registries can identify disparities in care delivery and show existing variations in the application of evidence-based therapies [1818. Dondo TB, Hall M, Timmis AD, Gilthorpe MS, Alabas OA, Batin PD, Deanfield JE, Hemingway H, Gale CP. Excess mortality and guideline-indicated care following non-ST-elevation myocardial infarction. Eur Heart J Acute Cardiovasc Care. 2017 Aug 6(5):412-420. , 1919. Szummer K, Wallentin L, Lindhagen L, Alfredsson J, Erlinge D, Held C, James S, Kellerth T, Lindahl B, Ravn-Fischer A, Rydberg E, Yndigegn T, Jernberg T. Relations between implementation of new treatments and improved outcomes in patients with non-ST-elevation myocardial infarction during the last 20 years: experiences from SWEDEHEART registry 1995 to 2014. Eur Heart J. 2018 Nov 7 39(42):3766-3776. ]. The ability to link data from EHRs to national registries may reduce the burden of data collection and enhance the utilisation of collected data. However, collective efforts are needed to standardise the definitions of data variables relevant to interventional cardiology and develop a set of widely agreed outcome measures that are meaningful to patients and society. The role that national registries played during the coronavirus 2019 (COVID-19) pandemic was crucial in understanding changes in patterns of care delivery at time of crises [2020. Wu J, Mamas M, Rashid M, Weston C, Hains J, Luescher T, de Belder MA, Deanfield JE, Gale CP. Patient response, treatments and mortality for acute myocardial infarction during the COVID-19 pandemic. Eur Heart J Qual Care Clin Outcomes. 2020 Jul 30:qcaa062. , 2121. Wu J, Mamas MA, Mohamed MO, Kwok CS, Roebuck C, Humberstone B, Denwood T, Luescher T, de Belder MA, Deanfield JE, Gale CP. Place and causes of acute cardiovascular mortality during the COVID-19 pandemic. Heart. 2021 Jan 107(2):113-119. ], but also identified areas for improvement in the way data are collected, analysed, and interpreted in a way that influence current practise.

Personal perspective - PETER LUDMAN

The collection of routine clinical and procedural data as we treat our patients is essential if we are to improve the standard of care we provide. Unless we measure what we do, we cannot assess its quality. Quality assurance and quality improvement is a fundamental part of modern medicine, and can be used with various techniques to help individuals and organisations provide better care. Clearly these data can also be used to fuel observational research addressing broader questions, and are an important part of the landscape of evidence based medicine. Dataset linkage, not only between clinical datasets but also to administrative datasets further enhances the research potential, and it here that efforts to standardise dataset definitions are so important. The development of registry based randomised trials opens up the possibility of trials that can combine the inclusiveness of registries with the rigor of randomisation. The application of new mathematical techniques such as machine learning offer fresh insights into huge linked datasets. The accuracy and completeness of routinely collected clinical data is fundamental to all of these endeavours and needs to be led by committed and dedicated healthcare professionals. Political and healthcare systems need to recognise the importance of the collection and analysis of clinical data, affording them high enough priority to make sure they are appropriately resourced.