Keywords

FormalPara Overview

This chapter is presented in two sections.

The first section sets a professional context for nursing documentation, the impact of nursing data and the nursing imperative for efficient and appropriate processes for current and future data collection including the impact of dynamic knowledgeable and nursing leadership to leverage appropriate data collection.

The second section discusses capturing nursing data. It focuses on clinical data types namely unstructured data, structured data, coded data and semi-structured data, their application and the advantages and disadvantages of each type. Standardised nursing terminologies and their impact on practice are also discussed.

FormalPara Learning Objectives for the Chapter

Section 1—To gain an understanding of:

  1. 1.

    The imperative for and impact of effective professional record keeping practice in nursing.

  2. 2.

    The importance of nursing data to underpin and bring about services that will support global populations into the next decade and beyond.

  3. 3.

    The impact of Executive Nurses who provide and develop leadership in the informatics area of practice in their organisations for digital practice and the knowledge age.

Section 2—To gain an understanding of:

  1. 1.

    Clinical data types namely unstructured data, structured data, coded data and semi-structured data.

  2. 2.

    The application of clinical data types and the advantages and disadvantages of each type.

  3. 3.

    Standardised Nursing Terminologies and their impact on practice.

7.1 Recording and Evidencing Nursing Practice

In their study of innovation across 152 differing health systems globally, Braithwaite et al. (Braithwaite et al. 2018), identified five principles for optimising future health and social care underpinned by four success factors. Across those elements, digital practice and the meaningful use of data were significant levers for change including: development of sustainable systems; digital innovation; and recognition of the need to situate models of care as near to the person’s own home or community; underpinned by: evaluative activity using appropriate measures (Braithwaite et al. 2016); transformation of data into information and then intelligence to continually create an evidence base (Braithwaite et al. 2018); authentic collaboration between multiple stakeholders; and person-centredness (Braithwaite et al. 2017). Nursing is well placed to leverage the future service change required that will support global populations into the next decade and beyond. The levers and success factors for change identified by Braithwaite et al. (Braithwaite et al. 2018) should be the stimulus for nursing leadership to renew and redouble focus on meaningful use of data, efficient digital practice and interoperability across systems and places of care, recognising appropriate evaluative activity, across global health economies.

7.2 The Nursing Imperative

In this global context and that of nursing practice in the twenty-first century, where services are under increasing pressure (Kelly et al. 2016), populations are rising with predicted large numbers of over 65 year olds many living with complex co-morbidities and associated conditions of ageing (Amalberti et al. 2016), the profession is evolving, to take on a broader scope of practice, improving the quality and person centric nature of models of care (Kennedy and Moen 2017). The largest percentage of the professional healthcare workforce nursing occupies 50% of the total headcount for many countries (World Health Organisation 2016a). Simultaneously, there is a wealth of evidence demonstrating the impact of registered nurses on healthcare outcomes for populations across a range of roles (Dick et al. 2017; Griffiths et al. 2016). Working across programmes of care and service delivery environments, the need for high quality information from this significant section of the health care workforce is an imperative to drive future change harnessed through the professions’ ability to collect data at the point of care.

The imperative for registered nurses to make accurate records of their interactions with staff and people, regardless of where they work, falls under a number of obligations within statutory, regulatory and contractual frameworks, and not least the requirement to evidence the professional impact and quality of their daily work. More usually, advice and guidance across organisations emanate from local policy and standards set by national or regional bodies, which have led to a range of views often differing in emphasis, rather than a clear national strategy to nursing and midwifery data.

Like many nursing and midwifery regulatory bodies globally, those in the United Kingdom (UK) and Ireland make explicit reference to the necessity for high quality professional records in their codes of practice (Nursing and Midwifery Council 2018; Nursing and Midwifery Board Ireland 2014) drawing attention to the need for timely, accurate record keeping that demonstrates decision making and service provision to populations. Whilst it is well discussed that accurate and complete records providing a detailed account of a person’s journey through health and social care services can protect both patient and registrant (Prideaux 2011), a common cause of legal claim arises from a breakdown in communication between health professionals particularly related to incomplete or inadequate records (Wood 2010). Evidence exists of variance of the quality of nursing information reflected in records (Saranto et al. 2014); incomplete records demonstrating a lack of information on the effect of nursing measures (Jeffries et al. 2012) and considerable deficiencies in quality of information (Gomes et al. 2019). Additionally, studies have demonstrated that when staffing issues impact on a care environment, documentation of nursing care is one of the safety critical activities which is often missed (Thomas-Hawkins et al. 2020; Ball et al. 2018). ‘Good’ records put the patient at the centre of care, demonstrate clinical decision-making and goals of care, allow audit of practice, support quality improvement and evidence co-production, safety and efficacy over time.

7.3 The Information Future of Nursing Practice

Increasingly nurses are being invited to expand their scope of practice in service models, to purportedly enable a holistic response to evolving needs, rurality of populations and Social Determinants of Health (SDH) (Mason et al. 2015; Nelson-Brantley et al. 2018). Whilst the nursing profession is often well placed to implement new models of service delivery, a change in ways of working is required to engage with the increasing complexity of care for populations and impact of SDH (Mason et al. 2015). Enhanced ways of working for nurses to match future service models linked to demographic trends include: promoting population health literacy (McMurray et al. 2018); greater use of technologies and Electronic Health Care Records (EHCRs) with advanced coordination of multi-professions across care settings (Zaworsky and Bower 2016; Erikson et al. 2017); and collaborative partnership with service users and their families to negotiate individualised person-centred outcomes. Whilst such service models obviously require investment to maximise the shift in demand from populations through development of the profession to support them (Leahy-Warren et al. 2017), early findings from global nursing exemplars have demonstrated positive outcomes such as reduction of hospital admissions, improved experience of care and cost reductions (McMurray et al. 2018; Maeung et al. 2013). In these practice prototypes, meaningful real-time data has been identified as a crucial enabler for efficient services to make best use of available nursing time (McMurray et al. 2018).

Evidence on the impact of documenting in EHCRs in terms of contribution to sustainability and efficacy of global health services varies, ranging from positively preventing hospitalisations (Burnel 2018) to reported ‘weak empirical evidence in relation to increasing efficiency and improving medical care’ (Bolous-Rødje 2019, p. 3). Furthermore, the necessary integration across organisational boundaries to communicate and share data seamlessly has not been achieved (Fitzpatrick and Ellingsen 2013). These issues are of particular importance when considering the intention to focus efforts on providing future nursing practice models, co-producing plans of care and supporting individuals in their own homes (Braithwaite et al. 2018), as movement of data across primary, secondary and community locations will be required to facilitate nursing and other members of professional teams. Accessible, reliable integrated systems are essential to the success of maximising digital nursing record keeping practice across organisations. Inefficient design and technical issues have been evidenced to impact on care delivery (Staggers et al. 2018), where nurses struggle to access systems to input relevant data across organisational boundaries and buildings. Poorly designed technology that fails to capture the essence of nursing practice and decision making, along with systems wrought with technical issues have been demonstrated to frustrate nurses, who will resort to printing documents to manually include information (Staggers et al. 2011). All too often, design flaws arise from the simple translation of paper documents into digital forms, rather than investment in system designs that map nursing workflows, capture nursing decision making through appropriate terminologies and enable point of care nursing data entry.

In addition, undoubtedly, as person-centred approaches evolve along with a focus on co-production, the use of citizen portals linking to EHCRS will increase, as populations become more digitally enabled across health economies. The sparse evaluative evidence on the efficacy of citizen portals demonstrates the value of a mixed approach model of both professional and digital contact (Zanaboni et al. 2018) and as the profession with the highest level of constancy with populations (All Party Parliamentary Group on Global Health (APPG) 2016), nursing has undoubtedly the greatest opportunity to influence the uptake and therefore utility of this digital partnership, therefore unlocking the potential of new approaches to health outcomes for populations , including those underpinned by self-care.

7.4 Evidencing Impact and Supporting Nursing Practice

The importance of evidencing the impact of registered nurses on population health outcomes, given the opportunity from the breadth and scale of the workforce, should be a driver for a strengthened professional focus in a world that is increasingly digitally driven and data saturated. Nurses, as the largest regulated profession with continuous contact with the public to impact on population health (All Party Parliamentary Group on Global Health (APPG) 2016) generate the greatest volume of healthcare data (Englebright and Caspers 2016). It is imperative, therefore, that this data is meaningfully structured and captured (Hussey et al. 2015), for the purposes of analysis and sharing (Ricciardi and Boccia 2017), evidencing and assuring the contribution of nursing.

The development of the emerging discipline of nurse informaticians has been growing on a global scale over the last 30 years (Liu et al. 2015), if somewhat driven by interested enthusiasts rather than a determined workforce decision to value and grow such roles. This piecemeal development of the discipline appears counterintuitive, the digital knowledge and skills of the nursing workforce, including those required to meaningfully use and evaluate information, evidently requiring a focus for investment in the current global context of services to create a workforce that is appropriately equipped to expertly handle their own professional data.

Big data capabilities from EHCRs can now provide healthcare organisations with information that enables understanding and evaluation of care decisions, processes and outcomes. Founds (Founds 2018) welcoming the big data era, asserted that the individualised and public facing ethos of nursing practice positioned the profession well to adopt and promote this future enabler for sustainable service models. This in turn can help quantify the nursing contribution, underpinning evidence-informed practice with real-time data (Reid Ponte et al. 2016). Sadly, EHCRs can sometimes include scant data describing nursing interventions and outcomes, failing to adequately describe critical decision making. Learning over the years of implementation would suggest there is a danger in trying to compare nurse and physician use of an EHCR to apply the same terminologies (Rogers et al. 2013). Evidence demonstrates that the ‘checklist’ medical system of recording within current EHCRs does not always capture relevant nursing information, leading to a lack of nursing narrative, particularly around psychosocial needs, often devoid of critical data on the performance of the largest professional workforce (Green and Thomas 2008).

Current data capture processes globally have varying degrees of efficacy to realise the future potential of the profession. Much of the vast volumes of nursing data captured, even when in digital format, are not structured or coded in a standardised way to allow linkages to be made across data elements (Khokar et al. 2017). There is also recognition that nurses spend considerable amounts of time recording high volumes of data of questionable utility (Lodhi et al. 2016; Leary et al. 2017). Great possibilities exist through the wealth of information gathered via the constancy of nursing including the potential to monitor real-time nursing assessments, interventions, processes and outcomes across care continuums (Welton and Harper 2016). Coupled with the opportunity to connect with and gather health data from populations through wearable technologies, cloud computing, smartphone mobile technologies and social media (Lokuge et al. 2018), there has never been such an era of opportunity to study and evaluate linkages to create knowledge for the purposes of improvement of professional practice, Personalised Medicine and population health outcomes (Higgins et al. 2018). This presents unprecedented prospects to advance the scientific knowledge and discover new opportunities for emphasis on aspects of nursing practice that effect improved predicted outcomes of care (Lodhi et al. 2016). The reality however is that adoption is slow, systems not fully understood (Gee et al. 2012) and funding to support development hard won, mainly due to the risk of failure which has been significant in the past within the United Kingdom (UK) (Campion Awaad et al. 2014). Barriers to technology adoption and therefore meaningful use include: misconceptions about evidence based practice; lack of leadership support; lack of time and mentorship; poor information literacy; and challenges with financial investment (Fulton et al. 2014).

7.5 Leading Digital Transformation of Nursing Information Practice

Some of the challenges to adoption could be positively leveraged through the influence and voice of nursing leadership globally, particularly those in executive roles. Whilst Kennedy-Page and Simpson (Kennedy-Page and Simpson 2016) asserted: ‘Big data has the potential to elucidate the phenomena of nursing’ (p. 272), other authors are more cautious in their prediction, understanding that it is dependent on strategic nursing leaders to recognise this potential and provide the necessary leadership for digital innovation to enhance nursing and develop an appropriately skilled and future facing workforce that includes the competence to manage information in an increasingly digital world (Jaimet 2016). Worryingly, the evidence suggests a lack of digital competence for nurse leaders in executive positions (Simpson 2013; Remus 2016). Those executive nurses who understand the value in providing and developing leadership in this area of practice in their organisations however, positively impact professional development processes for digital practice and the knowledge age (Remus 2016; Clarke and Mitchell 2014). Understanding and representing the experience of nurses, ensuring that technology appropriately captures, rather than impairing, burdening or eroding critical thinking and practice, is a new and important area of competence for executive nurses (Burkosi 2019).

Linking executive responsibilities, Kennedy-Page and Simpson (Simpson 2013) asserted that a critical component of the Executive Nurse leadership role is the meaningful capture and use of nursing data to drive quality practices across organisations. This included ability to transform raw data into information and then subsequently knowledge that could underpin the ‘art and science of nursing practice’ (Simpson 2013, p. 271). It is imperative, therefore, that executive nurses lead the development of nursing capabilities to capture, mine, collate and study data leading to relevant knowledge emerging directly in real-time rather than that gathered retrospectively (Khokar et al. 2017; Westra et al. 2017), including awareness of high volume data machine learning processes for predictive outcomes modelling (Obermeyer and Emanuel 2016). A critical future step for executive nurses is to push beyond current acceptance of physician-led EHCR design to data capture (Byrne 2012; Kerfoot 2015) reframing the purpose of nursing data collection from a focus on the legal and regulatory responsibilities to an opportunity to exponentially increase underpinning knowledge, advancing quality and the scientific practice of nursing (Kennedy-Page and Simpson 2016).

Key Points of Learning

  1. 1.

    Nursing is well placed to leverage the future change required to bring about services that will support global populations into the next decade and beyond.

  2. 2.

    Occupying 50% of the professional healthcare workforce globally, the need for high quality information is an imperative to drive future change harnessed through the professions’ ability to collect data at the point of care.

  3. 3.

    Effective record keeping practice puts the patient at the centre of care, demonstrates clinical decision-making and goals of care, allows audit of practice, supports quality improvement and evidences co-production, safety and efficacy over time.

  4. 4.

    Future facing service prototypes have identified meaningful real-time data as a crucial enabler for efficient services to make best use of available nursing time.

  5. 5.

    The digital knowledge and skills of the nursing workforce, including those required to meaningfully use and evaluate information, requires a focus for investment to create a workforce that is appropriately equipped to expertly handle data.

  6. 6.

    Generating the greatest volume of healthcare data, it is imperative that nursing data is meaningfully structured and captured for the purposes of analysis and sharing evidencing and assuring the contribution of nursing.

  7. 7.

    Executive nurses who understand the value in providing and developing leadership in the informatics area of practice in their organisations, positively impact professional development processes for digital practice and the knowledge age.

7.6 Capturing Nursing Data in a Digital Environment

Development and implementation of Electronic Health Care Records (EHCRs) and other information technology (IT) offers tremendous opportunities to enhance nursing practice and capacity to capture and utilise patients’ clinical information to improve healthcare. This enables multiple benefits including more timely access to health information which reduces duplication in work, improved end-user efficiencies and enables retrieval of pertinent information (e.g., patients on a specific medication) and aggregation of large data sets to enable service development and planning and new potential for research (Department of Health 2017; Nguyen et al. 2014; Kouroubali and Katehakis 2019; HIMSS 2019).

These benefits are extended via adjunct technologies such as clinical decision support (CDS) software and the Internet of Things (IoT) which respectively match patient clinical information with a computerised knowledge base to provide recommendations to the healthcare professional (HCP) and provide more comprehensive and accurate patient-generated information to the EHCR (Gartner 2019).

Access to more clinical data does not necessitate higher quality data and patient care, and we need to consider what are best practices to drive value from an EHCR as well as the adjunct technologies (Gartner 2019). At the same time, we must retain the overarching aim of capturing clinical information which is to track a patient’s condition and communicate this to other members of the nursing and midwifery team in addition to the wider healthcare team to inform clinical decision-making (Kuhn et al. 2015; Mehta et al. 2016).

With nurses and midwives being responsible for a large amount of the data being entered into the EHCR or other digital systems it is vital that that nursing data is represented in a shareable manner which preserves its complexity, context and richness of patient care. These data can usually be entered in an unstructured, structured, coded or semi-structured format (HSE 2019).

7.7 Unstructured Data

Unstructured text refers to free or narrative text generated using a single window (i.e. similar to a word processing programme) which is often included in clinical notes, surgical records, medical reports or discharge summaries (Linder et al. 2012; Helgheim et al. 2019; Lardon et al. 2015). Unstructured free text entry of clinical data allows freedom of speech and expressivity (Lardon et al. 2015; Johnson et al. 2008), which facilitates documentation of complex presentations or impressions of a diagnosis which do not fit into predictable templates or quantifiable values (Rosenbloom et al. 2011; Crampton et al. 2016). It is also a critical factor in assisting management decisions and reflecting the training and perspective of the professional recording the data (Johnson et al. 2008; Siegler and Adelman 2009). Free text is often preferred and valued by HCPs due to its familiarity, speed and ease-of-use (Johnson et al. 2008; Rosenbloom et al. 2011; Shachak et al. 2013). For the reader, narrative text provides a greater and more comprehensive understanding of the patient compared with highly-structured data (Johnson et al. 2008; Rosenbloom et al. 2011). However, narrative text often contains large amounts of text, much of which may be redundant, which can obscure key information (Wrenn et al. 2010; Huang et al. 2018). Due to its unstructured format, it can also lead to omission of important information (Johnson et al. 2008; Wilbanks et al. 2018) and makes it difficult to effectively retrieve and use information for preventive care, disease management and quality improvement purposes (Shachak et al. 2013). These challenges may also be amplified where copy-and-paste or some autofill functions are utilised to duplicate unstructured narrative data from one note into the new current note (Wilbanks et al. 2018; Weis and Levy 2014).

Many of the intended benefits of EHCR systems such as clinical decision support (CDS) and automatic pull of data from one section of the EHCR to another (e.g., Smart Form), require automatic processing of clinical information which necessitates the use of controlled vocabulary as opposed to free text. Therefore, whilst free text may be more familiar to end-users (Lardon et al. 2015; Weis and Levy 2014; Bush et al. 2017; Campion et al. 2014; Joukes et al. 2018), it limits the extent and reliability to which computers can interpret and re-use the data (Johnson et al. 2008; Kalra 2006; Joukes et al. 2019). Conversion of free text into a structured format can be a time-consuming and difficult task (Kreimeyer et al. 2017) and thus, development of automated mechanisms for interpreting free text is of utmost importance (Helgheim et al. 2019). Artificial intelligence (AI) such as natural language processing (NLP) is a promising method for data extraction and retrieval from unstructured text (Rosenbloom et al. 2011; Ford et al. 2016). AI could be used by the HCP at the time of data entry to identify key terms from unstructured text and formulate structured text from it (Loui and Hollinshead 2016), or for secondary purposes (e.g., data retrieval for audits). However at present, challenges exist with the portability of NLP systems between clinical settings and its ability to recognise improper grammatical use, misspellings, local dialects, short phrases (e.g., BID) and clinical shorthand (e.g., D2M) (Rosenbloom et al. 2011; Helgheim et al. 2019; Sohn et al. 2017; Pomares-Quimbaya et al. 2019). Overall, whilst unstructured data facilitates more comprehensive and flexible clinical documentation, it also comes with many challenges to optimising EHCR use which could affect patient safety and HCP productivity.

7.8 Structured Data

Structured data entry at the point-of-care, as opposed to post hoc structuring using NLP discussed above, includes: (1) Inputting data into structured forms/templates which divide components of the note into different sections (e.g., history of presenting illness); and (2) Selecting options from drop-down lists, tick boxes or radio buttons (Rosenbloom et al. 2011; Vuokko et al. 2017; California Healthcare Group 2010; Murray and Berberian 2011). Structured documentation templates often lend themselves to less complicated patient presentations (Mehta et al. 2016; Linder et al. 2012), computerised provider order entry (CPOE) systems (Siegler and Adelman 2009; Seidling and Bates 2016), registry forms, research forms dates (Krumm et al. 2014), social information, biological data measures and biological investigation results (Lardon et al. 2015; Helgheim et al. 2019). Whereas check boxes, radio buttons and drop down lists suit aspects which have limited options (Rosenbloom et al. 2011; Kreimeyer et al. 2017; Murray and Berberian 2011) such as yes/no and patient-reported outcome measures (Busack et al. 2016; Zhang et al. 2019). Until large scale NLP can accurately produce structured data from dictated and free text reports , structured data entry will be an essential input method to enable data retrieval for reports and analytics as well as CDS software (Linder et al. 2012).

For the author, entering data into structured templates in the EHCR can reduce data omission, as checklists can act as ‘memory joggers’ to assist HCPs to comply with best practice (Johnson et al. 2008; Rosenbloom et al. 2011; Linder et al. 2012; O’Donnell et al. 2018; Lorenzetti et al. 2018; Saranto et al. 2014). Additionally, structured data facilitates automated population of data fields (i.e., autofill) from other sections of the EHCR (Johnson et al. 2008; Helgheim et al. 2019; Linder et al. 2012) and from EHCR-integrated devices (Plastiras and O’Sullivan 2018) (e.g., bar code medication administration), improving overall efficiency in clinical documentation as well as reducing errors in the transfer of data between systems (Lawrence et al. 2018). Additionally, for the reader, structured templates are easier to read and locate information, whilst administrative staff benefit from the ability to easily aggregate and retrieve structured data (Joukes et al. 2018). However, whilst end-user efficiency may be improved when taking into account secondary uses of clinical data such as content importing technology (e.g., SmartForms), entering structured data at point-of-care requires more effort on the part of the end-user (Johnson et al. 2008; Rosenbloom et al. 2011; Bush et al. 2017; Campion et al. 2014; Joukes et al. 2018) and can negatively affect system usability (O’Donnell et al. 2018; Ajami and Bagheri-Tadi 2013; Kruse et al. 2016). It also imposes restrictions on HCPs in terms of how they document (Kalra 2006; Vuokko et al. 2017) and how they critically think and make decisions, which can risk the depersonalisation of healthcare (Nguyen et al. 2014; Saranto et al. 2014) and the incorrect identification of patients as having a certain condition due to lack of room for ambiguity (Johnson et al. 2008; Crampton et al. 2016).

To negate these risks, it has been recommended that the EHCR system does not mandate end-users to check a box if not appropriate, and that structured templates should never replace the clinical narrative (Kuhn et al. 2015). Unprecedented challenges have also been identified with structuring and standardising certain types of data such as psychosocial and emotional information, and whilst their importance is recognised, according to the literature, the best format for recording these data needs to be explored further (Busack et al. 2016). It is recommended that the design of structured templates involves a multi-disciplinary task force, workflow analysis (including downstream effects) and ongoing evaluation and comparisons of pre and post templates (Cao et al. 2017). Fundamental to any discussion of structured documentation is patient care (California Healthcare Group 2010), as well as recognition of the minimum dataset which needs to be collected to support patient care (Hakonsen et al. 2019). Additionally, whilst structure needs to be balanced with flexibility, developers should be mindful that the addition of too many options within the structured template could result in no meaningful data being collected (Rosenbloom et al. 2011). Even after following this process, a structured template will not suit every patient presentation, especially the more complicated patients (California Healthcare Group 2010). Therefore, personalisation which enables end-users to customise how data is input and viewed is recommended to allow some flexibility and improve end-user satisfaction with structured templates (Parent 2017; KLAS 2018; Hine et al. 2008).

7.9 Coded Data

Clinical information can often be tacit, context-bound, and ambiguous (Ben-Zion et al. 2014), and without a ‘shared tongue’, communication between HCPs can be significantly impaired (Sundling and Kurtycz 2019). Therefore, standardised terminologies (STs) have been developed which are associated with codes and represent defined aspects of clinical practice (Williams et al. 2017; World Health Organisation 2016b; Bronnert et al. 2012). For example, traditionally several terms are utilised to describe high cholesterol but with STs everyone uses the same term, and these are mapped to a code (e.g., ICD-10 code E78.0 represents Hypercholesterolemia). This multiples the benefits of structured data, as definitions are understood and synonyms can be aggregated (e.g., heart attack, myocardial infarct and MI) (California Healthcare Group 2010). Additional benefits include:

  1. 1.

    Improved data quality (Sundling and Kurtycz 2019; World Health Organisation 2016b; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2017).

  2. 2.

    Terminology understood by all HCPs across organisations and geographical boundaries (irrespective of language) (Sundling and Kurtycz 2019; World Health Organisation 2016b; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2017).

  3. 3.

    Patients benefit from HCPs utilising same term across clinical documentation to describe their condition (Sundling and Kurtycz 2019).

  4. 4.

    Improved quality of care (Sundling and Kurtycz 2019).

  5. 5.

    Semantic interoperability between information systems (Cao et al. 2017; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2017).

  6. 6.

    Accurate and comprehensive searches to identify patients requiring follow-up or changes to treatment based on revised guidelines (Sundling and Kurtycz 2019; World Health Organisation 2016b; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2017).

  7. 7.

    Monitoring of treatment effectiveness, patterns and trends (Vuokko et al. 2017; California Healthcare Group 2010; Saranto et al. 2014).

  8. 8.

    Use of CDS software (Vuokko et al. 2017; California Healthcare Group 2010; Saranto et al. 2014).

  9. 9.

    Additional research opportunities (Sundling and Kurtycz 2019; World Health Organisation 2016b; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2017).

Whilst many STs have been developed, no single terminology has been accepted as a universal standard (Rosenbloom et al. 2006). Three different types of coding sets exist:

7.9.1 Aggregation Terminologies (or Administrative Code Sets)

Enable classification of concepts using simple hierarchy relationships for administrative purposes such as reimbursement (Williams et al. 2017; Bronnert et al. 2012; Health Information and Quality Authority (HIQA) 2014). As these codes were designed to either group diagnoses and procedures or to contain broad categories with administrative technical terms, aggregation terminologies can be restrictive and prevent concepts from having multiple parents (Williams et al. 2017). Where HCPs are forced to use these code sets to capture clinical data, there is potential for inaccuracies and loss of the clinical intent (Williams et al. 2017; Bronnert et al. 2012). Examples of aggregation terminologies include International Classification of Diseases and related health problems (ICD), International Classification of Primary Care (ICPC), Read Codes and Office of Population Censuses and Surveys Classification of Interventions and Procedures (OPCS) .

7.9.2 Reference Terminologies (or Clinical Code Sets)

Enable more sensitive and specific terms to be collected as they are concept-based and controlled clinical terminologies which maintain a common reference point in healthcare (Bronnert et al. 2012). Unlike aggregation terminologies, reference terminologies facilitate the combination of concepts (i.e., post-coordination) to create a more detailed or complex concept from a simple one (Williams et al. 2017; Rosenbloom et al. 2006). For example, the following terms may coexist: chest pain, substernal chest pain and crushing substernal chest pain. Reference terminologies are less restrictive, considered more usable and meaningful for HCPs , reduce time spent searching for terms and enable use of CDS software as well as aggregation of data (Bronnert et al. 2012; Rosenbloom et al. 2006; van der Kooija et al. 2006). Reference terminologies utilised at point-of-care include the Systematized Nomenclature of Medicine-Clinical Terms (SNOMED-CT) which capture all clinical notes including allergies, vitals, past history, family history, symptoms, clinical findings and diagnosis; the Logical Observation Identifiers Names and Codes (LOINC) which captures laboratory and clinical observations; and RxNorm which captures medication names (Bronnert et al. 2012; Health Information and Quality Authority (HIQA) 2014; SNOMED International 2019). To balance the more usable reference terminology with the more rigorous aggregation terminologies which may be needed for national audits or reimbursement, reference terminologies can be mapped to an aggregation terminology e.g., ICD (The Office of the National Coordinator for Health Information Technology 2017).

7.9.3 Interface Terminology

To capture more granularity and clinical intent in the documentation, a third type of standardised terminology has been developed referred to as interface terminologies (Bronnert et al. 2012). These interface terminologies are often discipline-specific such as standardised nursing terminologies (The Office of the National Coordinator for Health Information Technology 2017; Vivanti et al. 2018), institution-specific (Rosenbloom et al. 2013) or speciality-specific (Sundling and Kurtycz 2019). Whilst large-scale reference terminologies attempt to represent every possible entity, interface terminologies reduce the need for post-coordination (e.g., combination of “acute” and “pain”) as they represent the common terms utilised in the specific practice its employed in (Bronnert et al. 2012; Rosenbloom et al. 2013; Berger 2013). Additionally, this decreases time spent searching for codes and facilitates documentation of more comprehensive, accurate and relevant clinical information (Bronnert et al. 2012; Rosenbloom et al. 2013; Berger 2013). Interface terminologies can be also be used to gain a deeper understanding of care approaches during evaluations, as well as having potential to improve patient outcomes (Saranto et al. 2014; The Office of the National Coordinator for Health Information Technology 2017; Macieira et al. 2017; Tayyib et al. 2015). Therefore, interface terminologies are important for problem lists (Bronnert et al. 2012) and these can then be mapped to the reference and aggregation terminologies where required for health information exchange, administrative or secondary use (Bronnert et al. 2012; The Office of the National Coordinator for Health Information Technology 2017; Westra et al. 2008).

7.10 Standardised Terminologies in Nursing Practice

Since the 1970s , there has been a concerted effort to promote STs within nursing and midwifery practice (Hardiker 2011) with the pioneering of the first standardised nursing language or terminology NANDA International (NANDA-I), formerly known as North American Nursing Diagnosis Association (Jones et al. 2010; Oreofe and Oyenike 2018). These standardised nursing terminologies commonly systematically group, define and encode nursing care as nursing diagnoses, interventions and outcomes (Kieft et al. 2018; Warren et al. 2015; Hellesø 2006; Bernhart-Just et al. 2010) and link nursing diagnoses with evidenced-based interventions and outcomes (Clancy et al. 2006). This is seen as the pathway for making the nursing process more useable and visible (Oreofe and Oyenike 2018) which promotes good communication, provides the basis for care planning and identification of patient problems (Dykes et al. 2009) and improves data quality for research and service development planning (da Costa and da Costa Linch 2018).

The American Nurses Association (ANA) have approved twelve terminology sets that support nursing practice for use within the ECHR which includes both nursing-specific and multi-disciplinary terminologies (Table 7.1) (Gencbas et al. 2018; The Office of the National Coordinator for Health Information Technology 2017). Additionally, a recent review commissioned by the Five Country Nursing and Midwifery Digital Leadership Group identified that in addition to the STs outlined in Table 7.1, nurses and midwives have also used locally-controlled and other medical and/or multi-disciplinary STs (Fennelly et al. 2020). These STs are in utilisation across at least 26 different countries (Fennelly et al. 2020) but although both the UK and Ireland have adopted the use of SNOMED-CT (Health Information and Quality Authority (HIQA) 2014; Sheerin 2003; Arnot-Smith and Smith 2010), their utilisation of nursing-specific STs has been more sporadic (Sheerin 2003; Murphy et al. 2018).

Table 7.1 American Nurses Association (ANA)-approved terminology sets
Full size table

Initially, the deployment of STs in nursing and midwifery practice has been uncoordinated with little convergence towards a unified nursing language system that can be integrated within the wider health-care language arena (Cho and Park 2006). It is now acknowledged that these STs need to be integrated and understood within the broader healthcare system to support interoperability and data continuity across community and acute settings (Oreofe and Oyenike 2018; Martin et al. 2011; Kim et al. 2014). However, many of the nursing-specific STs have been developed, utilised and evaluated in a specific clinical setting such as the Omaha System and CCC in primary care, and the ICNP and PNDS in secondary care (Fennelly et al. 2020). Consideration, therefore, needs to be given to the advantages and disadvantages of a large comprehensive multi-disciplinary ST versus a discipline or institution-specific ST, and the purpose of using the ST. Although the multi-disciplinary ST may facilitate communication between professions and settings (Jukes et al. 2012), searching for the correct code to match a patient’s diagnosis from a long list can also be time-consuming (Vuokko et al. 2017). Whereas nursing-specific STs offer more granularity, help to distinguish nursing care (Estrada and Dunn 2012) and enable the linkage of nursing diagnoses with evidence-based interventions which have potential to improve clinical decision-making skills (Wuryanto et al. 2017) and patient care (Zhang et al. 2018). However, they may not always fully comprehensive of the nursing and midwifery care provided (Cho and Park 2006) and have, at times, resulted in the availability of too many terms representing the same patient presentation (i.e., content duplication) (Morais et al. 2015; Monsen et al. 2011). Therefore, irrespective of the type of ST being used, its applicability, validity and reliability should be considered for the specific context it is being implemented in. Otherwise, it could jeopardise patient care and safety as well not as driving the full potential of using the ST.

Although the type and content of the ST is important, its implementation also drives its success. A usable ECHR interface in which the ST is searched or enters influences the use and acceptability of the ST amongst nurses and midwives (Cho and Park 2003; Hariyati et al. 2016). An interface which supports shortcuts such as searching mechanisms, display of the most frequently selected codes (for the given user) at the top of the list, i.e., ‘favourites’ (California Healthcare Group 2010) and/or use of NLP techniques to suggest appropriate codes and expression (SNOMED International 2019; Hodge and Narus 2018) have been recommended. Additionally, when the ST is being used within a unstructured template compared to a drop down list, the education of nurses to use the ST has been shown to impact on the quality of the nursing documentation (Müller-Staub et al. 2007, 2008; Müller-Staub 2009). Although STs have been shown to facilitate retrieval and aggregation of data from clinical records and health information exchange (Tastan et al. 2014; Topaz et al. 2014), this also depended on the reporting and analytical capabilities built into the ECHR system and the format and file types used (Westra et al. 2010).

Overall, use of STs or coded data within the EHCR provide several benefits to end-users, patients, healthcare organisations and policy-makers, and it is likely that more than one type of terminology will be required in the EHCR to facilitate both administrative and clinical purposes. Decisions regarding the selection of these STs should be made prior to EHCR design as otherwise adaptations to the terminology in the EHCR are expensive and labour-intensive (Sundling and Kurtycz 2019). Each of these terminologies will come with a license fee and mapping of terminologies to one another will need to be maintained as changes and updates are made to the ST by the software developer, third-party vendor or the individual healthcare organisation (Kalra 2006; The Office of the National Coordinator for Health Information Technology 2017; Houser et al. 2013).

7.11 Semi-structured Data

Use of STs do not always easily accommodate for diagnostic uncertainty (Ford et al. 2016) and are not always sensitive and specific to the condition in question (e.g., depression could present symptomatically as insomnia, fatigue, malaise) (McBrien et al. 2018). Therefore, the option to enter free text in conjunction with the ST is often allowed within the ECHR (Ford et al. 2016) to allow additional context or further clinically-relevant information to be added (Wilbanks et al. 2018; California Healthcare Group 2010; Sundling and Kurtycz 2019). This hybrid model of unstructured, structured and coded data is known as semi-structured data (California Healthcare Group 2010; Murray and Berberian 2011). Additionally, within the structured elements of the EHCR, options to input narrative data are often provided where end-users cannot find an appropriate structured concept or code (Rosenbloom et al. 2011). However, this can risk end-users overusing the free text box rather than searching for the appropriate code and thus, end-users need to understand the benefits of using coded and structured data in combination with unstructured data. Overall, semi-structured clinical data combines the benefits associated with the flexibility of unstructured data with the downstream benefits of using coded and structured data.

7.12 Key Findings for Nursing on Clinical Data Types

No one type of clinical data will accommodate nursing and midwifery documentation of every clinical scenario and/or secondary use of the data and thus, it is likely that a combination of those will be utilised (Arrendale 2018). In determining the most appropriate type of data or combination of data types, the advantages and disadvantages of each should be considered (Table 7.2), as well as the workflows and downstream effects of capturing data in this format.

Table 7.2 Summary of clinical data types in the electronic health record
Full size table

Key Points of Learning

  1. 1.

    With nurses and midwives being responsible for a large amount of the data being entered into the EHCR or other digital systems it is vital that that nursing data is represented in a shareable manner which preserves its complexity, context and richness of patient care. These data can usually be entered in an unstructured, structured, coded or semi-structured format.

  2. 2.

    No one type of clinical data will accommodate nursing and midwifery documentation of every clinical scenario and/or secondary use of the data and thus, it is likely that a combination of those will be utilised

  3. 3.

    In determining the most appropriate type of data or combination of data types, the advantages and disadvantages of each should be considered as well as the workflows and downstream effects of capturing data in this format.

  4. 4.

    Use of Standardised Terminologies or coded data within the EHCR provide several benefits to end-users, patients, healthcare organisations and policy-makers, and it is likely that more than one type of terminology will be required in the EHCR to facilitate both administrative and clinical purposes.

  5. 5.

    There has been a concerted effort to promote Standardised Terminologies within nursing and midwifery practice. These standardised nursing terminologies commonly systematically group, define and encode nursing care as nursing diagnoses, interventions and outcomes and link nursing diagnoses with evidenced-based interventions and outcomes

  6. 6.

    Decisions regarding the selection of these STs should be made prior to EHCR design as otherwise adaptations to the terminology in the EHCR are expensive and labour-intensive

Review Questions

Questions

  1. 1.

    Describe what you consider to be the core requirements for delivery of good records in nursing care documentation?

Answer

‘Good’ records put the patient at the centre of care, demonstrate clinical decision-making and goals of care, allow audit of practice, support quality improvement and evidence co-production, safety and efficacy over time.

  1. 2.

    What do you consider to be the key barriers to technology adoption in your area of clinical practice?

  2. 3.

    Do they reflect all, one or some of the key barriers listed in this chapter. Select one of the barriers from the list of items below and expand using supporting evidence.

    1. (a)

      Misconceptions about evidence based practice;

    2. (b)

      Lack of leadership support;

    3. (c)

      Lack of time and mentorship;

    4. (d)

      Poor information literacy;

    5. (e)

      Challenges with financial investment

    Ref https://indiana.pure.elsevier.com/en/publications/faculty-and-organizational-characteristics-associated-with-inform

  3. 4.

    This chapter explains the importance of representing nursing data in EHCRs or other digital systems in a shareable manner which preserves its complexity, context and richness of patient care. Discuss the different types and formats of data that can usually be entered in an EHCR.

Answer

EHCR data can usually be entered in an unstructured, structured, coded or semi-structured format. We briefly provide some of the key points from the chapter on the different type of data formats.

Unstructured text refers to free or narrative text generated using a single window. Unstructured free text entry of clinical data allows freedom of speech and expressivity, which facilitates documentation of complex presentations or impressions of a diagnosis which do not fit into predictable templates or quantifiable values. However, narrative text often contains large amounts of text, much of which may be redundant, which can obscure key information. Due to its unstructured format, it can also lead to omission of important information and makes it difficult to effectively retrieve and use information for preventive care, disease management and quality improvement purposes.

Structured data entry at the point-of-care, as opposed to post hoc structuring using Natural Language Processing (NLP), includes: (1) Inputting data into structured forms/templates which divide components of the note into different sections (e.g., history of presenting illness); and (2) Selecting options from drop-down lists, tick boxes or radio buttons. Structured documentation templates often lend themselves to less complicated patient presentations, computerised provider order entry (CPOE) systems, registry forms, research forms dates, social information, biological data measures and biological investigation results. For the author, entering data into structured templates in the EHCR can reduce data omission, as checklists can act as ‘memory joggers’ to assist HCPs to comply with best practice. It also imposes restrictions on HCPs in terms of how they document and how they critically think and make decisions, which can risk the depersonalisation of healthcare and the incorrect identification of patients as having a certain condition due to lack of room for ambiguity.

Coded data and standardised terminologies (STs) have been developed which are associated with codes and represent defined aspects of clinical practice. For example, traditionally several terms are utilised to describe high cholesterol but with STs everyone uses the same term, and these are mapped to a code (e.g., ICD-10 code E78.0 represents Hypercholesterolemia). This multiples the benefits of structured data, as definitions are understood and synonyms can be aggregated (e.g., heart attack, myocardial infarct and MI).

A semi structured hybrid model of unstructured, structured and coded data is known as semi-structured data. Additionally, within the structured elements of the EHCR, options to input narrative data are often provided where end-users cannot find an appropriate structured concept or code. However, this can risk end-users overusing the free text box rather than searching for the appropriate code and thus, end-users need to understand the benefits of using coded and structured data in combination with unstructured data. Overall, semi-structured clinical data combines the benefits associated with the flexibility of unstructured data with the downstream benefits of using coded and structured data.