>> Guidelines can influence the quality of care in various ways. One possibility is the development of quality indicators (QI) based on guidelines, expert opinions or scientific evidence (Kötter et al. 2011). These help to measure, present and compare quality as part of the “Deming cycle” (Plan-Do-Check-Act) (Taylor et al. 2014). For this purpose, QI are developed as defined, measurable items or adopted from guidelines (plan) (Kötter et al. 2011) - an indicator can, for example, be asking for preventive care documents within the first 24 hours after admission to a facility. If health care is now provided (Do), it is checked how many of the patients treated in a facility (with a specific clinical picture) were asked for their health care documents in the 24 hours after admission (check). This is compared to the number of all patients admitted to the facility. A QI is therefore not a direct measure of quality, but rather a tool for performance monitoring (Cohen and Leemans 2014). This makes it possible to show whether the current quality of care is sufficient or needs to be adjusted (Act). There are only isolated QIs for PCs, but they have weaknesses such as low quality (Roo et al. 2013). One article identified the integration of QI into daily work as a major challenge (Cohen and Leemans 2014).
Electronic health records offer one opportunity to integrate QI into daily work. One study found that data from a cancer registry linked to electronic health records can be used to assess the quality of care using QI (Caldarella et al. 2012). A similar approach to measuring quality in the field of PC in Germany is the national hospice and palliative care register. Electronic health records in the PC are used to fill out a patient-related core data set and transfer it to the national register (Radbruch et al. 2014). There is currently an interface for sending data to the register in two software systems for outpatient PCs (PalliDoc, ISPC) and two documentation systems for standard care (Agfa Orbis, ID DIACOS) (German Society for Palliative Medicine/National Hospice and Palliative Register 2019). These data have already been used to examine and compare the implementation of AAPV in two project regions (Mühlensiepen et al. 2019).
If a low quality of care has been identified using QI, it is necessary to act in these areas (Act). In many areas (communication, decision-making, symptom management), electronic health records also offer good support, for example by integrating reminder functions for an assessment (Finucane et al. 2021). This potential is currently only being used to a limited extent in Germany, as documentation is often still analogue (Gonçalves et al. 2018) or existing systems are not yet adapted to the special requirements of PCs (Krüger-Brand 2015). In addition, the documentation for PC in Germany only exists for outpatient care or as an isolated solution for the exchange between outpatient and inpatient areas (Krumm et al. 2014). If treatment is planned and documented in the PC using standard care templates, information is often not stored in a designated location and patient preferences or decisions cannot be easily found (Tai-Seale et al. 2012). This highlights the need for the development of specific electronic health records for PC that incorporate the definitions and concepts of PC to advance evidence-based practice and improve quality of care.
The aim of this article is to clarify how electronic patient records (EHRs) in the PC area can contribute to measuring and increasing the quality of care. First, the basics of electronic health records will be considered. This is followed by an examination of the domains named by the NCP and the QI available for measuring quality. The areas whose quality can be actively influenced by electronic health records are examined in more detail below. For this purpose, the requirements of palliative-specific documentation and its content are considered based on the needs of adult palliative patients and their relatives.
Electronic health records
Against the background of the following statements, it is important to make definitions with regard to the different types of electronic health records. In Germany, a distinction is made between three types (Deiters/Houta 2015; Haas 2017; Meister/Houta 2020): electronic patient records (EPA), electronic health records (EGA) and electronic case files (EFA). These types can be differentiated according to the two core properties “access authorization” and “data”.
An EHR is patient-led, meaning that patients decide which data is visible to which actors. The purpose of EHRs is medically intended and aims at an exchange with those involved in primary care. In contrast to the EPA, an EGA has a broader purpose for collecting health-related data. Citizens in the role of health insured persons or patients are entitled to access. The quality of the data cannot be guaranteed at either the EPA or the EGA. Likewise, the completeness of the file cannot be guaranteed. In contrast to the EPA and EGA, the EFA is a doctor-managed file, which means that patients authorize doctors to access it. To ensure quality, (a) data may only be entered by them and (b) all documents must always be available to all doctors. The “all or nothing” principle applies with regard to the access permissions granted.
A survey by the Münch Foundation in 2018 showed that Germany is in 13th place out of 20 countries surveyed in Europe when it comes to implementing the EPA and has therefore fallen by two places in the last two years (Gonçalves et al. 2018) . However, at the beginning of 2021, based on the EHR concept, the electronic patient record (ePA) was introduced as an application of the telematics infrastructure in accordance with §291 a SGB V, which collects cross-case information in a patient-led manner and can also be filled out by doctors (PDSG, SGB V). In addition to these digital file folders, the question arises as to which content is required for specific supply purposes. Medical information objects (MIO) represent such (technically) standardized information and are part of the legal requirements. These universally usable and combinable information modules can be used in electronic health records for documentation (e.g. in the form of the vaccination certificate). In the spirit of interoperability, MIOs are readable by any system and thus facilitate the exchange of information between different suppliers with different systems (Krüger-Brand 2020).
In the context of this article, the term EPA is used as an equivalent for the terms Electronic Health Records and Electronic Medical Records, which are often used synonymously in the international literature.
Quality indicators for palliative care
QI is intended to measure, display and compare quality in a specific area. But how can EHRs contribute to measuring the quality of care in PC? It is essential to determine which areas require a special quality check. Since these so-called domains in PC differ from those in standard care, the NCP examined which domains are particularly important in PC: (1) structure and process of care (training and further education, continuity of care), (2 ) Physical aspects (measuring, documenting and managing physical symptoms and side effects), (3) Psychological and psychiatric aspects (measuring, documenting and managing psychological symptoms, assessing and managing the psychological reactions of patients and relatives), (4) Social aspects (discussing care goals, offering support to patients and relatives), (5) Spiritual, religious and existential aspects (providing information about spiritual services), (6) Cultural aspects (incorporating cultural assessments, taking preferences into account, language and rituals) , (7) Care of the imminently dying person (recognizing and documenting the entry into the active dying phase, recording and documenting wishes regarding the place of death), (8) Ethical and legal aspects (documentation of patient preferences, drawing up living wills, Strengthen advance care planning) (Ferrell 2005; Ferrell et al. 2018). Based on this, a review by De Roo et al. (2013) collected existing QI that can be assigned to the respective domains (Roo et al. 2013). This resulted in 17 sets with a total of 326 QIs in all domains named by the NCP, which often focused on specific patient groups (e.g. older people), settings (e.g. hospices) or national characteristics (e.g. all settings in the Netherlands). For example, a QI was the documentation of the offer of psychosocial support within the first 72 hours after admission to an intensive care unit. For the calculation, the total number of patients who or their relatives received the offer of psychosocial support from a team member within 72 hours of admission was divided by all patients who were admitted to the ward for 72 hours. Patients who had a score of 2 or 3 on the Glasgow Coma Scale and had no (identified) relatives were excluded. The majority of QIs related to the process or outcome of care. Only three of the sets considered both process, outcome and structure, although from the authors' point of view this seemed advisable in order to carry out a comprehensive assessment of the care. QI was increasingly identified with a view to the domains (1) structure and process, (2) physical aspects, (3) psychological and psychiatric aspects, (7) care of the imminently dying person and (8) ethical and legal aspects. Only one quality indicator was found for domain (6) cultural aspects (Roo et al. 2013).
The authors came to the conclusion that the existing sets need to be further developed and tested in practice. In addition, adaptation to different countries, health systems and settings is required for successful use (Roo et al. 2013).
The collection of the underlying data for the application of QI can be done using, among other things, EPAs. But this integration of QI into EHRs can be challenging. Campbell et al. therefore developed a model that supports this in order to promote an improvement in care: (1) selection of the measure, (2) definition of the measurement criteria, (3) validation of the criteria and validation of the measurement process, (4) improvement of the recording activities related to measures, (5) initiation of quality improvement processes (Campbell et al. 2021). QI can be so helpful in uncovering areas where the quality of care is not satisfactory.
Opportunities for quality improvement
through EPAs
If quality deficiencies have been identified in a certain area - for example with the help of QI - or if a facility would like to take preventive action, approaches to improving quality make sense. Here, too, it is helpful to use EHRs because they are an integral part of care. But how can EHRs help improve the quality of care? For example, guidelines can be integrated into EPAs (e.g. references to the necessary collection of certain data or exceeding limit values), which can support users in applying evidence-based care (Shah et al. 2021). In addition, clinical decision aids can be integrated into EHRs, which can support users in decision-making, for example when prescribing medication (Sutton et al. 2020). In addition to the advantages described here, some key aspects of PCs that can be positively influenced by the content or functions of EPAs are listed below as examples.

coordination
The complex structures of care in PC as well as the increasingly complex service requirements lead to the collection of information in different sectors by different professional groups, the dissemination of which among providers is essential for patient-centered care. Nevertheless, a study showed that such information is often exchanged by email or fax, among other things, and is therefore not always available in electronic form across all sectors (Pringle et al. 2014).
PC providers identified the timely and sufficient flow of information between all parties involved as a key challenge when moving palliative patients between different sectors. For this purpose, the providers found it useful to provide a cross-institutional EHR to exchange information and coordinate care (Mertens et al. 2018).
In Great Britain, the web-based Electronic Palliative Care Coordination System (EPaCCS) is used for this purpose. This allows providers (including family doctors, emergency call services, hospitals, nursing services, specialized palliative services, hospices) to collect, exchange or view information on palliative patients (Petrova et al. 2018). This includes master data, precautionary documents, the desired place of death and care, the disease prognosis, data from the main responsible providers, assessments and medication plans (Allsop et al. 2017). In Leeds, an EPaCCS data set was detected for 21% of all deceased people (n=1229), which was created on average one month before death (Allsop et al. 2017). Patients with an EPaCCS data set visited the emergency room less often, were admitted to hospital less often and died there less often than people without an EPaCCS data set. In addition, the proportion of patients with an EPaCCS data set who died at their preferred place of death was above average (55% – 79%). When evaluating these results, however, it must be taken into account that patients who use EPaCCS may be people who are actively committed to dying at home (Leniz et al. 2020) and EPaCCS is almost exclusively used by family doctors and outpatient nursing services are used, which could also be related to death in one's own home (Wye et al. 2016).

Advance Care Planning
Advance care planning, also known as advance care planning (ACP), is a conversational process between patients and providers about values, desires, and treatment options that is documented in the EHR. Here, patients can develop their own preferences for their future treatment and support and name authorized representatives (Cheung et al. 2020). ACP helps to improve quality of life, reduce unnecessary treatments, reduce costs and increase relative satisfaction (Turley et al. 2016).
The problem here is that these conversations are often held too late, not continuously or not at all (Cheung et al. 2020). In a review, 16 studies were identified that the EPA automated patients for ACP using the criteria age (n=3), diagnoses with a limited life expectancy (n=2), and admission to an inpatient care facility (n=1 ) or stating that a person refuses resuscitation (n=1). This led to a request from providers to talk to patients about ACP and to document this (Huber et al. 2018).
What is also crucial is the form in which the conversations are documented. A review showed that (prevention) documents are perceived by providers as more helpful than pure conversation notes (Bush et al. 2018). However, conducting and documenting the discussions is not sufficient to ensure that the patient's wishes are taken into account. Another review showed that utilities do not know where to find documentation about ACP in the EPA or whether this information is valid. A central location for ACP in the EPA could support this (Bush et al. 2018). What is crucial here is that this information is available to all providers and that care is coordinated according to the treatment goals.

Symptom control
The providers from different professions and sectors work together to care for and treat the symptoms (Radbruch et al. 2020). The symptoms are often complex (e.g. refractory pain) and can only be alleviated through multimodal control approaches that take into account the holistic nature of the psychological, physical, social and spiritual components of the symptom development (Quill/Abernethy 2013). What is crucial here is the commitment of all providers in terms of advice, consultation and personal support to patients (Doody et al. 2018). This results in a further requirement for the EPA: In existing EPAs, there is often only the possibility of documenting physical symptoms (Stewart et al. 2017).
As part of a study, existing nursing documentation was adapted that only focused on the management of physical symptoms. While previously more physical symptoms were documented, after the adaptation a comprehensive approach to documentation and recording the patient's perspective was identified. The study also reports an improved exchange of information within the team due to the increased clarity of the documentation (Stewart et al. 2017).
But even if the symptoms can be documented holistically: symptom control is often inadequate, even in patients who could be treated in accordance with guidelines (Rutten et al. 2020). In a study, alerting providers when a predetermined level of severity is exceeded was shown to be helpful. It was found that severe symptoms were more likely to be subsequently relieved when providers were alerted than when they were not (Gustafson et al. 2017). In addition, symptom control is made more difficult by the fact that providers document the symptoms and not patients themselves. In a study, the information provided by oncology patients about their symptom burden was compared with the documentation provided by the providers. A weak to slight agreement was found for nine of the symptoms considered, and for three others the agreement was moderate. If no medication was available for the symptoms, the agreement was significantly worse (Sikorskii et al. 2012). In addition to the possibility of training providers to minimize this bias, another option may be to include documentation of symptoms by patients or their relatives. In the area of mHealth, there are various apps for palliative patients, which unfortunately rarely include the option of integrating the information collected into existing EHRs. A review found that 48% (n=15) of PC apps focus on managing symptoms. In addition to the general recording of symptoms, many apps relate to pain recording for various target groups, such as patients with cancer, heart failure or COPD as well as those taking oral oncological drugs (Portz et al. 2020). The data from such apps enable improved management of the disease by identifying possible connections between activities and symptoms or other variables (Slev et al. 2017).
discussion
The goal of this article was to illustrate how PC EHRs can help measure and improve quality of care. It was important here that existing QI for PCs often have low quality, are not usable for all areas and their integration can be difficult. There are already international approaches to integrating content and functions into EHRs that can help improve care.
It was shown that a holistic symptom observation and control as well as the possibility of entering ACP in a structured manner in an EHR have positive effects on the quality of the PC. For Germany, there is a need for increased support for the standardization of content documentation through the development of PC-relevant MIOs. This enables all providers to read, edit and exchange relevant information on the PC, even between different EHRs.
There is also a need to develop a cross-institutional EPA for PCs in Germany. Exchange across professional groups and sectors is essential to ensure high-quality treatment.
But it's not just the exchange between suppliers that is crucial; Capturing the patient's subjective perspective is also crucial in PC, including ACP, and must be considered and understood in EHR. This corresponds to the requirement that the common quality dimensions in PC (structure, process and result quality) be expanded to include a subjective perspective of relationship quality. Relationship quality includes a focus on respect, dignity, contact at eye level as well as reciprocity and helpfulness and also influences the quality of structure, processes and results (Dirkschnieder et al. 2016). By assessing the needs of patients regarding ACP, the requirements for the quality of the relationship are met by an exchange between providers and patients and the treatment is planned and carried out according to the wishes of the patients.
However, the use of EPAs by utilities is not exclusively positive. Files that have low usability can lead to interruptions in the workflow and heavy cognitive strain on users (Sieck et al. 2019). It is therefore crucial to develop EHRs that are adapted to the reality of care and the work processes of the users. Participatory design approaches are often used in technology development. This creates a bidirectional learning process between developers and users, in which the necessary content from all professions for joint development is exchanged, but specific knowledge is contributed by the actors (Vandekerckhove et al. 2020). In the area of developing EHRs, it has been found that this approach greatly improves the user experience (Sieck et al. 2019).
It is also crucial to meaningfully integrate EPAs into care. Users need to be trained to ensure, for example, that patients can see the screen on which the EHR is displayed during consultations. A review found that this leads to a significantly more positive assessment of the use of EPAs (Alkureishi et al. 2016). <<