Strengthening tRansparent reporting of reseArch on uNfinished nursing CARE: The RANCARE guideline

Abstract Unfinished, rationed, missed, or otherwise undone nursing care is a phenomenon observed across health‐care settings worldwide. Irrespective of differing terminology, it has repeatedly been linked to adverse outcomes for both patients and nursing staff. With growing numbers of publications on the topic, scholars have acknowledged persistent barriers to meaningful comparison across studies, settings, and health‐care systems. The aim of this study was thus to develop a guideline to strengthen transparent reporting in research on unfinished nursing care. An international four‐person steering group led a consensus process including a two‐round online Delphi survey and a workshop with 38 international experts. The study was embedded in the RANCARE COST Action: Rationing Missed Nursing Care: An international and multidimensional Problem. Participation was voluntary. The resulting 40‐item RANCARE guideline provides recommendations for transparent and comprehensive reporting on unfinished nursing care regarding conceptualization, measurement, contextual information, and data analyses. By increasing the transparency and comprehensiveness in reporting of studies on unfinished nursing care, the RANCARE guideline supports efficient use of the research results, for example, allowing researchers and nurses to take purposeful actions, with the goal of improving the safety and quality of health‐care services.

Up to 98% of nurses working in hospital inpatient care report having left necessary care activities unfinished or undone during recent shifts (Jones et al., 2015;Recio-Saucedo et al., 2018). Across various settings, the activities reported left unfinished most often include emotional support, patient education, coordination, and discharge planning and care planning. Yet, meaningful comparison between studies is hindered by the phenomenon's complexity, paired with a lack of transparency and comprehensiveness in reporting (e.g., on care task list, scoring cutoffs, or handling of missing values) and lack of consistency of terminology (e.g., use of different terms, care left undone, and missed nursing care, for the same instrument [cf. Ausserhofer et al., 2014;Ball et al., 2018]).

| BACKGROUND
Unfinished nursing care is "a problem of time scarcity that precipitates the process of implicit rationing through clinical priority setting among nursing staff resulting in the outcome of care left undone" (Jones et al., 2015). Introduced by scholars as an umbrella term, it covers multiple closely related, partly overlapping phenomena found in the literature often referred to as missed care (Kalisch & Williams, 2009), implicit rationing of nursing care (Schubert et al., 2007), care or tasks left undone (Aiken et al., 2001), or omitted nursing care (Vincelette et al., 2019). In accordance with previous research (Jones et al., 2019; RANCARE COST-Action 15208 & EU-COST Association, 2016) and to ensure inclusivity of all related concepts, we refer to unfinished nursing care throughout this article.
The quantity of unfinished nursing care is traditionally measured using multi-item scales in self-report surveys filled out (mostly) by nursing staff. To address some of the known limitations of self-reported measures, more novel approaches have recently been reported, such as the use of a single-item measure (Hamilton et al., 2017) or identification of missed care activities through the use of data from electronic health records (Dall'Ora et al., 2019).
While the research community now widely acknowledges the common underlying phenomenon of unfinished nursing care, differences in its key elements are still reflected in publications: quantitative studies on the topic face a variety of methodological and conceptual challenges (Vincelette et al., 2019) that are thus reflected in measurement and analyses as well as the transparency and comprehensiveness of their reporting. High levels of variability are seen not only between studies investigating different conceptualizations (e.g., missed care vs. implicit rationing) but also within studies using a single concept and corresponding instrument (e.g., different cutoffs in the scoring of multi-item surveys; different terminology used for the same instrument and sample; differing task list within the same approach [tasks undone scales  (Moher et al., 2010). An international four-person steering group led a consensus process featuring a modified online Delphi survey to identify, refine, and agree on relevant items for the guideline (Hsu & Sandford, 2007). This group was comprised of three researchers experienced in the field (Patti Hamilton, Dietmar Ausserhofer, and Michael Simon) and a graduate student (Catherine Blatter). The steering group has experience in investigating the concepts rationing of care (Dietmar Ausserhofer and Michael Simon), nursing care left undone (Dietmar Ausserhofer), and missed care approaches (Patti Hamilton), as well as applying conceptual comparison (Patti Hamilton). To identify experts for the Delphi survey, we performed an unpublished update of the literature review on studies investigating unfinished nursing care from Jones et al. (2015), with extension to all settings. Participants were considered eligible for invitation if they were listed as first or corresponding authors on any of these papers. Second, all members of the RANCARE 15208 COST Action were considered eligible for participation. COST Actions are networking activities aiming to connect researchers from academia, industry, public and private sectors to drive innovation in a specific field. The COST Action is funded over a course of 4 years by the European Cooperation in Science and Technology (COST; https://www.cost.eu/). Lastly, the list of eligible participants was screened for by the authors and potentially extended by hand with a snowballing approach. All eligible participants were invited via email.
Participation in this study was voluntary; informed consent was obtained before each Delphi round. All collected data were treated as confidential and presented anonymously in subsequent rounds.
In accordance with Swiss law, the responsible ethics committee  Consensus on items was defined a priori according to the RAND/ UCLA Appropriateness Method (Fitch et al., 2001). In this method, the following measures are calculated for every item: median, interpercentile range (IPR; 30th−70th), asymmetry index (AI; absolute difference between central point of IPR and central point of measurement scale, i.e., 5), and IPR adjusted for asymmetry (IPRAS = 2.35 + (AI × 1.5)). Those are used to calculate the absolute disagreement index (DI = IPR/IPRAS). A DI < 1 indicates no extreme variation of answers, also stated as the absence of disagreement.
Finally, consensus on relevance/clarity of an item is classified as the median score of an item lying within the 3-point range of 7-9 in the absence of disagreement (Fitch et al., 2001;Van Grootven et al., 2018). Participants were invited to leave suggestions for improvement at any time via comments.
Twenty-six also joined the face-to-face workshop. The majority of participants were females (76%) and had a background in nursing (70%) or psychology (11%), followed by educational sciences (5%), business and economics (5%) or midwifery, sociology and health systems (each 3%). Seventy-five percentage of participants had more than 5 years' professional experience; 30% had >20 years. Eighty percentage held doctoral degrees, with two-thirds holding positions as assistant, associate, or full professor.
The initial item list of n = 61 items reached consensus (i.e., median rating ≥7 without disagreement) regarding relevance and clarity on 98% of the items in the first Delphi round. Fifty percentage of participants provided additional feedback through comments.
With respect to the comments' significance, the steering group agreed to adjust and refine the items for further discussion during Based on results from the first Delphi round and the workshop, we sent a list of 38 revised guideline items in Delphi Round 2 for rating on relevance (1-9), clarity (1-9) and we added the self-developed question, whether the item should be included (yes/no), to allow for a clear decision. For each item, the median rating was ≥7 without disagreement on either relevance or clarity and therefore included in the guideline. Furthermore, 17 (45%) items were rated by >90% of participants as to be included. The remaining 21 (55%) items were rated for inclusion by ≥75% of participants. With all items' acceptance confirmed, the Delphi survey was terminated. Substantial comment-based feedback was implemented (namely, clarification of wording and splitting two items), resulting in the final version of the guideline.

| RESULTS
The components of the final RANCARE guideline consist of 40 items that address key elements influencing the reporting of quantitative research on unfinished nursing care (cf. Table 1). It is structured according to common sections of a research paper.
Many items are self-explanatory and applicable to most observational research; others target primarily self-report measures or routinely collected data.
The RANCARE guideline on Section 1 addresses the key aspect of conceptualization and terminology within the overall phenomenon of unfinished nursing care. Researchers should use the terminology of the concept studied throughout their paper (e.g., Item 1) but we recommend that they link their work to the umbrella term by including "unfinished nursing care" additionally as a keyword (Item 2 Please consider the following points when using the RANCARE guideline: • RANCARE guideline provides guidance for transparent reporting of quantitative research on unfinished nursing care. • Unfinished nursing care is used as an umbrella term throughout this guideline for all commonly used concepts related to the topic (e.g., missed care, implicit rationing, task undone, care left undone, care omitted).
• The guideline addresses issues commonly encountered in communicating quantitative research on the topic-however, some aspects may be pertinent for qualitative or mixed-methods research.
• Please use the RANCARE in addition to the reporting guideline relevant to your study's design (e.g., STROBE or CONSORT) • All items are offered as points to consider-they may not be applicable to every study or need to be left out due to limited space for publications.

| DISCUSSION
Regarding the reporting of quantitative research targeting unfinished nursing care, the RANCARE guideline addresses the key elements of conceptualization, measurement, data analyses, and presentation of study findings. It increases the potential for research on this phenomenon by facilitating comparison between studies investigating closely related concepts. To maximize its usability across multiple study designs, the RANCARE guideline overlaps with elements of other guidelines; therefore, we recommend using it in addition to any design-focused reporting guidelines (e.g., STROBE and CONSORT).
With 446 entries currently listed in the EQUATOR network library for reporting guidelines (December 14, 2020), there is arguably a need to justify the development of yet another guideline. In recent years, reporting guidelines have increasingly been developed to improve the reporting of content-related aspects for specific research fields, rather than just targeting a study design. For instance, the TIDieR-checklist extends CONSORT (Consolidated Standards of Reporting Trials) for improved reporting of interventions (Hoffmann et al., 2014). In medical research, a recent example includes the RECOvER-checklist that aims to improve reporting of ERAS (enhanced recovery after surgery)-related studies (Elias et al., 2019).  (Moher et al., 2010).
A strength of the RANCARE guideline is the development within the ongoing RANCARE COST Action, which allowed us to integrate new knowledge of the topic without the usual delays inherent in the publication process. Still, the sampling approach for the consensus process was oriented primarily toward objective criteria such as authorship of publications and was not limited only to the affiliation with the COST Action. This approach ensured raters' research experience and yielded a sample with a broad background.
The development of a reporting guideline requires an iterative approach, critical elements of which include a consensus process and, if possible, a Delphi survey (Moher et al., 2010). Therefore, as recommended for a Delphi survey, we defined consensus a priori in accordance with the RAND/UCLA Appropriateness Method (Diamond et al., 2014). Since the quantitative measurement was not T A B L E 2 Contextual elements to describe the setting in studies on unfinished nursing care

Level Examples
System level • Country including cultural traditions • City including population size • Health-care system within country (e.g., funding system, for profit) • Scope of practice of nursing staff Organizational/Provider level • Setting (e.g., hospital, nursing home, and home care) and specific area within setting (e.g., ICU and dementia care units) • Size/volume of the facilities/units (no. of beds), classifications (e.g., Magnet® status), ownership (e.g., public) • Model of care (e.g., patient allocation, primary nursing, team nursing, lean management) or describe task allocation between (nursing) staff • Skill/grade mix on unit Individual level a • Characteristics of nursing staff (e.g., age, gender, and level of education) • Characteristics of patient population (e.g., age, gender, primary, and relevant ancillary diagnoses) • Information on shift, weekday (weekends), months of data collection Throughout the RANCARE guideline, we use unfinished nursing care as an umbrella term to ensure inclusion of all related concepts.
While our inclusion of nursing within this term acknowledges nurses' crucial roles within health-care delivery, we also acknowledge that healthcare is multidisciplinary and encourage researchers of other disciplines to use the RANCARE guideline.
Rather than focusing on one specific concept or terminology, the RANCARE guideline can provide guidance to all research investigating unfinished, rationed, missed, omitted, compromised, or underused care (e.g., Glasziou et al., 2017).

| CONCLUSION AND IMPLICATIONS
The RANCARE guideline's primary aim is to increase transparency and comprehensiveness in the reporting of quantitative studies on unfinished nursing care. More transparent and complete research findings are needed to better understand the link between unfinished nursing care and patient and/or nurse outcomes (Recio-Saucedo et al., 2018;Vincelette et al., 2019). Researchers who follow the RANCARE reporting guideline will be able to communicate their research findings more clearly and comprehensively. This will allow more meaningful comparisons across studies, settings, and countries, thereby supporting more efficient use of empirical findings. More comprehensive, transparent and detailed reporting of estimates of unfinished nursing care will also help researchers and nurse managers target purposeful actions within their settings with the goal of improving the safety and quality of health-care services. For further actions, we encourage researchers not only to follow the RANCARE guideline within publications but also to provide the authors with feedback on its usefulness.

ACKNOWLEDGMENTS
This article is based upon work from COST Action RANCARE CA15208, supported by COST (European Cooperation in Science and Technology). Dietmar Ausserhofer and Michael Simon report financial support for travel cost by the RANCARE COST Action. The authors would like to thank the participants of the consensus process for their constructive feedback and engagement and acknowledge the contributions of the following people, who participated in both