With the widespread use of nursing information systems (NISs) in hospital settings, electronic documentation has become the mainstream form of information storage, exchange, and analysis.1 Embedded with a combination of standardized nursing language (SNL) and software algorithms based on an abundance of scientific evidence, the nursing clinical decision support system (Nu-CDSS) mimics expert thinking, representing the advanced stage of NIS.2 The nursing CDSS provides nurses with suggested nursing diagnoses, interventions, and care evaluations based on the clients' clinical information,3 which foster clinical reasoning and evidence-based practice.4 The adoption of nursing process-based CDSS has been reported to reduce nurses' documentation burden,5 improve documentation completeness,6 and nursing-sensitive outcomes.6
Standardized care plans (SCPs) are an integrated component of the Nu-CDSS, representing nurses' decision-making based on clinical thinking and professional judgment together with decision support by the CDSS.3 It has been pointed out that "care plans provide a course of direction for personalized care tailored to an individual's unique needs,"7 which suggests the formulation of care plans according to nursing diagnoses as central to the nursing process and the foundation of quality care.
The introduction of SCPs increased the visibility and transparency8 of nursing as well as the content and completeness of documentation.9 However, because of the lack of clarity of statements inherent in SNLs,10,11 nurses are finding it difficult to formulate care plans using SNLs, thus undermining the clinical usefulness of SCPs. Concerns have been raised about the potential of deindividualization and fragmentation of care with the use of SCPs.12,13 Nurses' perceptions and attitudes toward SCPs also affect their use.14,15 Bedside nursing has long been practiced as a task-oriented profession with little focus on the nursing process or care planning. Studies16,17 have shown that nurses lack the ability to develop care plans based on nursing diagnoses because of insufficient understanding of the nursing process. Moreover, some nurses view care plans as an additional documentation requirement to serve administrative purposes rather than a guide to practice,14,18 which hinders SCP implementation. A 2018 study by D'Agostino et al19 identified large variations among nurses regarding attitude toward, intention to, and actual behavior in making nursing diagnoses in clinical documentation. Deficits in care plan formulation were pervasive with various biases, omissions, and lack of long-term perspectives identified in aged care settings according to a US study.20 Despite extensive research into this area, incorporation of SNLs and SCPs into the NIS remains unsatisfactory even in countries with advanced stages of nursing informatization21,22 and nurses' perceived usefulness of information provided by SNLs are important determinants of their adoption.
The SCP within the CDSS, which is supported by expert thinking, should guide the provision of quality nursing care. Recently, Park et al23 investigated the consistency between nursing practice and documentation in electronic nursing records using the importance-performance analysis (IPA), indicating the high consistency between nurses' self-appraised importance and performance among various nursing interventions. However, research has also shown that much of the nursing activities were left undocumented in the NIS, which can in part be attributed to nurses' lack of in-service training on standardized nursing process24 and the documentation templates not meeting clinical needs.25 Compared with countries leading in health information technology development, SCPs are relatively new to health institutions in China, and there is little research on nurses' experience with SCPs. Based on previous findings, the extent to which nursing practice is reflected in the SCP is worth exploring.
STUDY AIM
This article aims to explore how nurses develop SCPs under a commercial Nu-CDSS by investigating the importance and performance of nursing interventions corresponding to each nursing diagnosis in the SCP followed by qualitative interviews to delve deeper into nurses' experience with SCPs.
METHODS
Design
This is a mixed-methods study with the quantitative part using the IPA23 and the qualitative part using semistructured interviews.
Setting
This study was conducted in a 2200-bed tertiary teaching hospital in Shanghai, China, which employs more than 2000 nurses. From 2020 to 2021, the Nu-CDSS was implemented in 20 general wards, and each ward employs an average of 17 nurses.
The Nursing Clinical Decision Support System
The Clinical Care Classification (CCC) developed by Dr Saba et al26 was chosen as the SNL to guide development of the Nu-CDSS. Using the nursing process as the theoretical framework, CCC comprises 21 care elements, 176 nursing diagnoses, and 201 core nursing interventions divided into four action types: assess, perform, teach, and manage.27 The CCC was translated into Chinese in 2018 and updated to version 2.5 by the time Nu-CDSS development began.28 Because CCC covers three stages of the nursing process (diagnosis, intervention, and evaluation), it better represents nursing practice compared with other recognized SNLs.29
Based on the CCC coding principles, nursing informaticists together with nursing experts in China expanded the core interventions in CCC version 2.5 to develop a comprehensive nursing knowledge base containing 1825 nursing interventions.30 Association algorithms between nursing assessment, nursing diagnoses, and nursing interventions were also established by 16 nursing experts to form a nursing process-based CDSS. Our hospital introduced the Nu-CDSS for pilot use in 2020. The system has been in use for 2 years and has accumulated nursing records for thousands of inpatients. Refer to supplementary materials for the care planning interface (Supplementary Figure 1, http://links.lww.com/CIN/A229) and a sample file of the care plan exported from the system (Supplementary Figure 2, http://links.lww.com/CIN/A230).
Sample
Four medical-surgical wards were randomly chosen for this study. All nurses working on the wards were deemed eligible to participate except for those who had been employed for less than 1 year at the time of recruitment because of their limited access to the Nu-CDSS. A total of 52 nurses participated in the quantitative part (response rate, 86.7%), including 12 in general surgery, 14 in thoracic surgery, 12 in endocrinology, and 14 in neurology. The average years in nursing were 10.28 +/- 8.49 (range, 1-31). Six of the participants were head nurses/specialist nurses/education nurses, and the rest had no administrative positions. Ten nurses were purposively selected to participate in semistructured interviews, among whom three were head nurses, and the rest were staff nurses. Their work experience as a nurse ranged from 3 to 28 years.
Data Collection
Importance-Performance Analysis
Importance-performance analysis is a business research methodology proposed by Martilla and James in 1977.31 Importance-performance analysis later evolved into a common management tool for organizational performance development and has been applied to various disciplines, which has implications for quality improvement. In nursing, IPA has been used as a tool to evaluate the effect of hospital information system implementation32 and to guide the improvement of care quality.33-35
Performance of Nursing Interventions Corresponding to Each Nursing Diagnosis
The nursing records of discharged patients, dating back from August 30, 2021, were retrospectively extracted from the Nu-CDSS. We determined the sample size to be 100 for each department, and a total of 400 SCPs were reviewed. To ensure there was enough content in the SCP, only clients with a minimum 3-day length of hospital stay were included. The performance of nursing interventions is represented by the number of times they appear corresponding to a certain nursing diagnosis. For each SCP, multiple occurrences of an intervention corresponding to the same nursing diagnosis were not accumulated. Data extraction from the SCPs was carried by two investigators to ensure trustworthiness of the data.
Importance of Nursing Interventions Corresponding to Each Nursing Diagnosis
According to the most prevalent nursing diagnoses and corresponding nursing interventions in the SCPs we reviewed, an evaluation form was constructed regarding the perceived importance of the nursing interventions linked to each diagnosis. Each item was scored 1 to 4 from "not very important" to "a bit important," "important," and "very important." Content validity was evaluated by five nursing experts who were involved in Nu-CDSS implementation and had clinical teaching experience related to the nursing process and care planning. Nursing interventions regarded by the experts to have poor correlation with each nursing diagnosis were deleted. Then, the evaluation form was distributed to all eligible nurses in October 2021. The average score was used to determine the importance of nursing interventions. In addition, interventions required by a medical order such as "administering enteral nutrition" and "preoperative nutritional support" were collected for response but not included in the final analysis.
Semistructured Interviews
From November 2021 to January 2022, semistructured individual interviews were conducted by the first author (Y. Zhai) to gain deeper insights into their experiences with SCPs in their practices using a topic guide informed by previous literature,36,37 which included four questions: (1) What is your perspective on using the SCP? (2) What is your experience with care planning in the Nu-CDSS? (3) Do you have any difficulties associated with care planning in the Nu-CDSS? Please elaborate. (4) What have been/can be done to improve care planning in the system? The interviews were audiotaped and transcribed verbatim before sending back to the participants for verification. The length of the interviews ranged from 22 to 47 minutes, averaging 33 minutes.
Data Analysis
The quantitative data were imported into IBM SPSS Statistics version 21 (IBM Inc., Armonk, NY, USA) for statistical analysis. Descriptive statistics (mean, SD) were used to analyze the importance scores. The correlation between performance and importance of nursing interventions corresponding to each nursing diagnosis was analyzed with bivariate Pearson's correlation.
Two researchers independently analyzed the interview transcripts and negotiated if they disagreed until a consensus was reached. We used Braun and Clarke's38 thematic analysis to code the text step by step: (1) We read through the text to become familiar with the content, (2) extracted meaningful statements from the text to form the initial codings, (3) compared the initial codings and summarized similar concepts to form the prototype of theme, (4) compared and adjusted the prototype of the theme and the coding, and (5) confirmed the theme and named it.
Ethical Considerations
This study was approved by the ethics committee of Zhongshan Hospital affiliated with Fudan University (project number B2021-481R). All research subjects signed an informed consent form before participating in the study.
RESULTS
Core Nursing Diagnoses and Corresponding Interventions
According to the prevalence of nursing problems and the relevance of nursing interventions in the SCPs reviewed, six core nursing problems were determined, corresponding to a total of 58 nursing interventions. Each nursing problem corresponds to 7 to 15 nursing interventions (Table 1). For instance, for the diagnosis bleeding risk, "Notify doctor of changes in patient condition," "Monitor vital signs," "Watch for signs of bleeding," "Provide local compression hemostasis," "Assess factors associated with bleeding risk," "Monitor coagulation lab results," and "Provide health education on bleeding-related knowledge" were identified as its associated interventions.
The Importance and Performance of Nursing Interventions
The mean scores of importance of nursing interventions were 2.67 to 3.83, and their frequency in the SCPs reviewed ranged from 3 to 168. The performance (frequency in the SCP) and importance (scores based on the questionnaire) of interventions linked to activities of daily living (ADL) alteration are listed in Table 2. Refer to supplementary tables, http://links.lww.com/CIN/A231, for IPA of interventions linked to other nursing diagnoses.
The correlation between performance and importance of nursing interventions corresponding to each nursing diagnosis was -0.241 to 0.413 (P > .05) (Table 3).
Qualitative Results
Three themes emerged from the analysis: struggling to adapt to SCPs, SCPs poorly adapted to clinical needs, and gaining new routines with care planning.
Struggling to Adapt to Standardized Care Plans
Nurses in our setting rarely did care planning before the introduction of the Nu-CDSS. Although the nursing process-based framework embedded in the Nu-CDSS improved the completeness of nursing documentation, nurses struggled to adapt their routines to the Nu-CDSS, especially for experienced nurses who already had a deep mindset with the old system, which is built upon task-oriented nursing.
Those of us who have been working for so many years have already formed a working pattern and know what to do every day, so we may not follow this system[horizontal ellipsis] and changing this inherent thinking is quite difficult. If you ask us to completely follow the care plan, we feel that maybe we won't be able to do it. (Staff nurse, 8 years)
[horizontal ellipsis]it (the CDSS) is really confusing, which may be inconvenient for us, and we haven't completely accepted it. (Staff nurse, 8 years)
Because of lack of perceived usefulness, some nurses viewed the care planning process as a meaningless task and may blindly follow the CDSS guidance without much thought as to whether the care plan addresses the patient's care need, which in turn caused confusion to nurses on subsequent shifts.
Some nurses may be in a hurry to get off work and don't bother to go through a careful screening of the nursing interventions during care planning. They just click and save[horizontal ellipsis] then you will see a patient with as many as 14 schedules within an hour when you take over. (Staff nurse, 4 years)
Standardized Care Plans Poorly Adapted to Clinical Needs
The nursing process is well-recognized as conducive to delivering patient-centered care according to the care needs identified through assessment. However, the suitability of using a care plan in acute care settings where most patients have a short length of hospital stay was questioned by nurses. Nurses felt that individualized care was hardly possible under the task-based routines and time constraints.
We have such a large workload, to be honest, I think it is more appropriate for us to complete tasks in batches, rather than provide customized interventions for each single patient. There is no time to do it at all. (Staff nurse, 2 years)
Some nurses thought the preset standardized interventions in the Nu-CDSS to be generic and lacking context so that the SCPs had little value in guiding subsequent practice.
I don't think the plan is like a plan, it's just an intervention. If the system really had a better plan, it would help us achieve nursing goals and that would be pretty good. But it doesn't, so it's useless. (Staff nurse, 2 years)
Because of the large number of interventions in the built-in knowledge base of CDSS, nurses were confused about how to choose between them, and there was a certain randomness among nurses when developing care plans.
I feel like those interventions should all be checked, but if you check so many, the workload will be too much, one person has 20-30 schedules[horizontal ellipsis] I don't know how to check the frequency of some measures[horizontal ellipsis] Anyway, it's all prn[horizontal ellipsis] it also does not have a fixed pattern about how to choose[horizontal ellipsis] (Staff nurse, 9 years)
Lack of a standard of practice at the organizational level led to inconsistent system use patterns among nurses, which affected the confidence and motivation of nurses to make SCPs.
Because we really don't have a so-called right way. In the end, what kind of thing is right is all based on our own groping, and everyone has a set of ideas for record-keeping; there is no standard. (Staff nurse, 8 years)
Moreover, some nursing tasks in our setting could not be properly represented with the standardized interventions recommended by the Nu-CDSS, leading to perplexities among nurses. There was no place for nurses to type in additional interventions by free text in the care planning interface, thus compromising the accuracy of the care plan.
You will see a lot of the interventions the Nu-CDSS recommends that do not correspond to our clinical tasks[horizontal ellipsis] We felt that this (nursing) problem should not be solved with these interventions. (Head nurse, 29 years)
Gaining New Routines With Care Planning
Before the introduction of the Nu-CDSS, nurses had limited experience in care planning. However, they gradually formed their own record-keeping habits during prolonged use of the system. To tackle the uncertainty during system use, nurses sought to form a unified use pattern within the ward as much as possible by communicating with each other or under the guidance of the head nurse.
Just discuss among ourselves, like "Hey, how did you do this?", then everyone did it the same way[horizontal ellipsis] (Staff nurse, 14 years)
As nurses have different understandings and sensitivities to nursing diagnoses and interventions because of varying education levels and clinical experience, performance standards on system use are set by management in some wards to standardize the record-keeping pattern, which is also conducive to quality control.
I will make some ward-level specifications. For example, if the patient's body temperature is >38[degrees]C, then I clarify to the nurse which nursing interventions this patient should have or which interventions she should choose when the patient is taking oral laxatives. In this way, I try to standardize practices within the ward as much as possible. (Head nurse, 22 years)
DICUSSION
Characteristics of Nurses' Care Planning Practices
The results show that many interventions within the action type assessment, such as assess vertigo symptoms, assess medications that increase the risk of fall, and ADL scale, are of high importance but low performance, which means they are not fully documented in the care plan. Our finding is not in line with that of Yang et al39 showing that nursing activities that fall into the action type assessing/monitoring/evaluating/observing accounted for almost half of all interventions that appeared in the SCPs. This may be related to the nursing process-guided workflow embedded in the Nu-CDSS where nurses tend to balance the care plan with other parts of nursing documentation. As these assessments are completed in the assessment stage, when developing care plans, nurses tend to avoid selecting items related to assessment. It is worth noting that not all the interventions can be properly fit into the care plan. Interventions such as monitor BP changes and monitor heart rate/rhythm changes all belong to vital signs and can be documented in specific modules in batches rather than a single patient at a time. When sudden changes in a patient's condition occur, nurses are accustomed to documenting the vital signs and other interventions directly in the progress notes rather than updating the care plan. Therefore, in order to avoid duplication, the nurses include them less frequently in the care plan. These are consistent with a recent Australian study showing that nurses placed more value on the nursing progress notes or comment sheet as a record of nursing action instead of the care plan.18
Our study showed the poor correlation between nurse-perceived importance and performance for interventions linked to each nursing diagnosis, with some contents in the care plan being in discordance with actual practice and some activities not adequately documented. This finding reflects the defects in care planning among nurses, such as randomly selecting all the interventions recommended by the Nu-CDSS, likely due to the care planning module within the Nu-CDSS not meeting nurses' expectations to guide clinical practice. Because of poor integration of the SCP with clinical practice, nurses lack the motivation to develop individualized care plans, and care plans are not referred to as a source of information exchange, as demonstrated in the previous studies.14,18 Our study found that only 11 of the 58 nursing interventions have a performance of >=40 of the 400 SCPs we reviewed, indicating the highly homogeneous pattern among care plans. This is perhaps because nurses have formed a habit regarding care planning after routinized system use without adequate consideration of individual needs of each patient. These findings suggest that there is room for adaption of SCPs for it to be successfully integrated with clinical workflow.21 Moreover, effective leadership from nurse administrators is warranted to engage individual nurses with SCP implementation.36,40
Controversial Issues With Standardized Care Plans
The restrictive documentation framework under which SCPs are developed may constrain the clinical reasoning of nurses and lead to homogeneity in care planning.10 Nurses tend to blindly select the same content within the drop-down boxes or base their choices on records from previous shifts when updating the care plan without consideration of the current conditions and the individualized needs of patients.11,41 Bruylands and colleagues'42 study showed that the decision support of providing nursing diagnosis recommendations based on nurses' assessment records did not improve the quality of nursing records compared with manual input, indicating that CDSS cannot replace nurses' clinical reasoning at any stage of the nursing process. Clinical decision support-guided documentation has long been criticized as jeopardizing clinicians' professional autonomy with few perceived benefits for experienced nurses.43
An inherent shortcoming with SCPs is the lack of expressivity associated with the use of SNLs.44 Despite the broad scope covered by SNLs, they cannot meet the dynamic care needs of patients and nurses still relying on free-texting during record-keeping11 to communicate information in an accurate and understandable manner, which was a common practice in our hospital, indicating the need for continuous refinement and local adaptation of SNLs to meet the practice needs in our setting. Another important finding in our study was nurses' uncertainty about how the care plan should be developed, indicating a need for quality standards across the institution. Ostensen et al15 also found that nurses tend to balance between overview and detail when working with the SCP. Despite extensive studies into quality criteria of nursing documentation and the value of SNLs, no consensus has been reached regarding what to include in the SCP.15 In our study, nurse leaders in pilot wards closely engaged in quality monitoring specific to the care planning practices among nurses, which is imperative to successful implementation of digital health tools in healthcare settings.45 As an imperative part of nursing documentation, how SCP should serve clinical practice warrants further study, and it is important that clear and visible leadership should be in place to guide nurses to use the SCP properly.13
Limitations
This study has several limitations. First, the Nu-CDSS we investigated is still in pilot use, and its partial implementation may have led to nurses' random use, which influences the measurement of performance of nursing interventions. Second, The IPA was carried out based on the interventions linked to each nursing diagnosis without adequate consideration of the related factors (etiologies) and defining characteristics (signs/symptoms), which also affects the relevance of the corresponding interventions. Finally, research subjects were limited to four medical-surgical wards and may underrepresent the whole picture of how nurses work with SCPs in our institution.
CONCLUSION
This study explored how nurses develop care plans in a Nu-CDSS. The results showed that there was a certain degree of difference between the performance and importance of nursing interventions, indicating that nurses' care planning may not reflect their professional judgments. Nurses had been working with the new system for a limited period and had not fully incorporated care planning into their workflow and held negative attitudes toward it. Management has a vital role to play in fostering nurses' clinical thinking inherent in care planning to guide the provision and evaluation of nursing care. The Nu-CDSS also needs to be constantly optimized to adapt to clinical needs to better represent nursing care. Standardized care plans should be remodeled to better integrate with other parts of nursing records and clinical practice, improving the efficiency of documentation and quality of care.
References