Background

Unfinished nursing care (UNC), which is becoming increasingly more of a concern in worldwide healthcare settings, involves the skipped, delayed, or incomplete delivery of nursing interventions needed for the patient and/or the patient’s family [1, 2]. The prevalence of UNC, which ranges from 55 to 98% globally [1], is considered as an accurate indicator of both patient safety and nursing care quality [3, 4]. The primary reasons for UNC are issues in communication, labour, and material resources [5]. The occurrence of UNC has also been associated with staff shortage and factors at both the structural level (e.g., nurses’ roles and experiences) and the process level, such as the stressful work environment, some negative managerial practices, the amount of overtime, and the high and/or complex demand for patient care [6,7,8,9,10,11]. In terms of consequences, UNC is linked to poor patient (e.g., pressure sores), nurse (e.g., moral distress), and organisational outcomes (e.g., increased length of stay) [5, 12,13,14]. Given these unfavourable outcomes, it is crucial to continuously assess those nursing interventions that are commonly postponed or missed, as well as the underlying reasons and consequences, to inform evidence-based strategies aimed at decreasing the frequency of UNC.

The worldwide COVID-19 pandemic has made it difficult for health facilities to maintain their sustainability and continuity of care due to the dramatic call to increase the care capacity with limited resources [15,16,17]. The staff sector most impacted by the pandemic — especially due to concerns regarding infection — has been recognised as nursing staff delivering direct patient care and thus representing the most crucial element of the health system infrastructure [18]. In addition to the need to increase the amount of care, nurses have also been impacted by unfamiliar work settings due to changes in the layout of the hospitals, sickness exposure, and urgent deployment from one department to another without the required skills. Therefore, various components (e.g., communication) of nursing care have been compromised by the limited interaction required during the pandemic and the need to be distanced. Nurses’ care capacity has also been negatively impacted by feelings related to the pandemic triggering anxiety, depression, and burnout [19, 20]. A rise in the number of nurses layoffs, the increased shortage of nurses, poor working circumstances, negative feelings, and imbalances in the nurse–patient ratio may all have increased the occurrence of UNC during the pandemic [21, 22] by further eroding the quality of care [23, 24]. Gurkovà et al. [25] stated that UNC may have increased the risk and adverse effects of the COVID-19 pandemic, resulting in ethical issues and a widespread mistrust in health systems [26]; moreover, Nash et al. [27] also stated that healthcare disparities were the consequences of UNC.

However, while the pre-pandemic occurrence of UNC has been well established, with several primary studies and systematic reviews (e.g., [28]) also investigating the underlying reasons (e.g., [29]), no summary of the studies conducted during the pandemic has been provided to date. Summarising the evidence produced may highlight the issues experienced during the pandemic in order to prevent them in future epidemiological disasters. It may also provide information on the quality of care in dramatic circumstances and the variations, if any, in the routine care before the pandemic. Finally, it may also set a new baseline in the context of UNC given the profound disruption and changes affecting the healthcare systems, requiring a long-term recovery. Thus, the aim of this review was to systematically review the occurrence of, reasons for, and consequences of UNC among patients in healthcare settings in the face of the COVID-19 pandemic.

Methods

Design

To begin with, two researchers (AB, SC) performed a rapid literature search to establish whether any studies had been published on UNC occurrences, their reasons, and consequences among patients during the pandemic. The beginning of the pandemic period was defined as 11 March 2020, according to the declaration by the World Health Organisation [30].

According to the Population (P), Exposure (E), Comparator (C), Outcomes (O), and Study Design (S) framework [31], the following were considered: P, patients in any healthcare setting; E, the COVID-19 pandemic period, as started on 11 March 2020 up to 5 May 2023 [30]; C, none; O, occurrence, reasons, and consequences of UNC, as perceived by nursing staff; and S, any types of quantitative study designs. Consequently, the following research questions were identified: (1) What was the occurrence of the UNC phenomenon among patients during the pandemic? (2) What were the reasons for the UNC during the pandemic? (3) What were the consequences of the UNC among patients during the pandemic? (4) What were the main methodological features of studies designed/conducted during the pandemic?

The systematic review was reported in its methods and findings according to Preferred Reporting Items for Systematic Reviews and Meta Analysis (PRISMA) guidelines [32].

Ethical considerations

The researchers designed a systematic review protocol that was registered in PROSPERO (CRD42023422871).

Inclusion and exclusion criteria

Studies were considered if they (1) regarded the nursing field; (2) focused on UNC occurrence, its reasons, and/or consequences during the pandemic, as perceived by nurses and nursing aides; (3) were published in English, Italian, or Turkish; (4) collected the data using a validated tool/instrument in the UNC field; (5) were conducted after 11 March 2020 during the COVID-19 pandemic up to 5 May 2023 [30]; and (6) used any types of quantitative designs (randomised controlled trials, non-randomised controlled trials, cohort studies, prospective or retrospective observational studies, cross-sectional studies, longitudinal studies).

Studies were excluded if they (1) did not address UNC data and/or did not involve nurses/nursing aides or care workers in the nursing field; (2) used non-validated tools/instruments measuring UNC or interviews; (3) were conducted in a paediatric setting, due to its specificity not being comparable with the adult field; (4) were designed as qualitative studies, reviews, commentaries, editorials, or books; (5) were written in other languages; or (6) had an abstract/full text that was not accessible.

Search method

MEDLINE-PubMed, the Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Scopus were searched to identify the eligible studies as sources on 5 May 2023. According to the uniqueness of this research, where no MeSH terms have been established and different key words are used [1, 2], all synonymous and equivalent keywords established in the field of UNC were used to access the databases. Specifically, the following keywords were used: “nurse”, “nursing”, “missed care”, “missed nursing care”, “unfinished nursing care”, “unfinished care”, “implicit rationing of nursing care”, “implicit rationing”, “rationing of nursing care”, “rationed care”, “prioritization process”, “omitted nursing care”, “task left undone”, and “task undone” using “OR” and “AND” operators (Supplementary Table 1).

Quality Appraisal

The Joanna Briggs Quality Appraisal Tool for analytical cross-sectional studies was used in the quality assessment for all eligible studies when they were based on cross-sectional designs [33]. This tool contains eight items with response options of yes, no, unclear, and not applicable. These items regarded inclusion criteria, subjects and setting description, exposure, standard criteria for measurement of the condition, confounding factors, strategies to deal with confounding factors, outcomes measurement, and statistical analysis. Two researchers (AB, SC) independently assessed the quality of the studies as “Rater 1” and “Rater 2”. In the case of a disagreement, the senior researcher (AP) was consulted to reach a consensus, as summarised analytically in Supplementary Table 2.

Besides the quality appraisal, to prevent bias, the following strategies were applied: (a) all researchers contributed to the writing of the review protocol; (b) at least two researchers searched the literature, chose the studies, and extracted the data, independently; (c) the senior researcher oversaw the data extraction; and (d) agreement was required before moving on to each next step.

Data extraction and synthesis

All studies that met the inclusion criteria, regardless of the results of their methodological quality, underwent the data extraction and data synthesis. The studies were divided into two groups and shared between two researchers (AB, SC). In primis, the data extraction grid was piloted in one study, and the findings agreed: no changes were required. Then, researchers independently extracted data from the remaining studies by populating the grid with the following data: (1) author(s), year, and country; (2) study aim(s) and design; (3) sample and setting; and (4) period of data collection and tool(s). Then the findings of the quality appraisal were provided (Table 1). At the end of data extraction, the researchers rechecked the data. Disagreements were solved with the consultation of the senior researcher (AP) until consensus was reached.

Table 1 Main characteristics of the included studies (= 25)
Full size table

A narrative synthesis process was used to summarise the findings [57] according to the review questions, applying the following methodology:

  1. (1)

    Studies conducted during the pandemic and their methodological quality: the researchers conducted a preliminary synthesis to provide an initial description of the main characteristics of the studies and their methodological quality, and similarities and differences across studies were presented by using textual explanations [57].

  2. (2)

    The occurrence of UNC: Findings were tabulated according to the tools used in each study, namely the MISSCARE Survey, the Basel Extent of Rationing of Nursing Care (BERNCA) and the Revised BERNCA (BERNCA-R), the Perceived Implicit Rationing of Nursing Care (PIRNCA), the Basel Extent of Rationing of Nursing Care for Nursing Homes (BERNCA-NH), the Intensive Care Unit Omitted Nursing Care instrument (ICU-ONC), and the Unfinished Nursing Care Survey (UNCS). In all tools, participants are required to rank the nursing interventions missed and/or postponed from always to never. Then, according to the following considerations,

  • the tools used different metrics (Likert from 1 to 5 for MISSCARE Survey and UNCS, from 0 to 4 for BERNCA, from 0 to 3 for PIRNCA, from 1 to 4 for BERNCA-NH, from 1 to 4 for ICU-ONC) and differed in the direction of measures (e.g., from always missed to never missed, e.g., [43], or the opposite, e.g. [50]); and

  • UNC interventions reflect an order [58, 59], such as first, second, and third, of interventions missed, expressing a prioritisation process (what should be actualised first and what can be delayed).

Data regarding the position (= order) of each nursing intervention according to the averages documented in the studies were extracted and then ranked according to the position: for example, the average of 3.23 with the MISSCARE Survey [35], indicating that this was the most unfinished activity, was ranked as first. Then, according to Blackman and colleagues [60], the first three interventions of high occurrence of being unfinished were identified; from the fourth to the sixth, those of intermediate occurrence; and from the seventh to ninth, those of a low occurrence of UNC.

  1. (3)

    The UNC reasons: Reasons were summarised based on the following considerations:

  • Studies using the MISSCARE Survey and the UNCS reported the reasons for UNC item by item, according to the structure of the tool;

  • Other studies documented the relationships (as correlations, associations) indicating a significant role of some factors in increasing/hindering UNC during the pandemic.

In the first case, the reasons were extracted and analysed in the same manner as UNC activities; in the second, studies (22 out of 25) documenting a statistically significant relationship of given factors with the UNC were extracted and categorised as organisational, work, or individual factors according to the literature in the field [29]. Of the remaining three studies, which were not focused on the reason for UNC, one was a methodological study that was focused on the psychometric assessment of the tool [56], one was a comparative study that was focused on the comparison between the data from a COVID-19 sample and a reference sample [54], and one was a study in which conditions were identified affected by the consequences of UNC [48].

  1. (4)

    UNC main consequences: if any, were described narratively.

All researchers were involved in the data analysis and synthesis process to ensure rigour in the process.

Findings

The results regarding the included studies are described below, including an exploration of their characteristics and quality and the occurrence of, reasons for, and consequences of UNC.

Search outcomes

In total, 1,389 articles were identified from the electronic databases. The search results were transferred to a reference manager (Mendeley) to organise the data extraction process. First, three steps were followed for the study selection: in the first stage, titles, in the second stage, abstracts, and in the third stage full text of the retrieved studies were screened for their eligibility by two reviewers (AB, SC), independently. In the case of any disagreement, the opinions of a third senior researcher (AP) were consulted during the entire process. Consensus between the researchers was essential for study inclusion.

In the first stage, 726 studies were excluded; from 1,389 studies, 663 articles were retained for abstract screening. Thus, in the second stage, 298 studies were excluded. At this stage, 365 studies met the criteria for next-step screening. Before the full-text screening, 219 duplicated studies were removed, and a visual inspection was conducted by two researchers (AB, SC) to check for duplicates. Then, 146 studies remained for full-text screening, and 122 of them were excluded for different reasons, as reported in Fig. 1. The references of the excluded reviews were screened by two researchers (AB, SC) to check their eligibility in an independent fashion and then agreed upon. In total, 38 articles were checked, of which 33 were already included, three were not related to UNC, and one was a qualitative study design. At the end of the screening process, 25 studies were included (Fig. 1).

Fig. 1
figure 1

PRISMA flow chart

Full size image

Included studies and their quality

Out of the 25 studies included (Table 1), 20 used a descriptive cross-sectional design (e.g., [34]) and five a comparative cross-sectional design confronting the data (a) before and during the pandemic [35]; (b) or before the pandemic, and the second/third wave [38]; and (c) of the COVID-19 sample and the reference sample [37, 48, 54]. Most studies were conducted in Europe (= 12, e.g., [50]) and Asia (= 11, e.g., [45]). Of the remaining, one was carried out in Africa [47] and one in Canada [53]. Study locations ranged from a hospital (e.g., [35]) to specific hospital settings (tertiary [55], district [51], government [56], private [34], teaching [50]) in various types of units (e.g., medical/surgical [54], urology [43], cardiology [48]). In addition, COVID-19 units were included in two studies [22, 37, 41] and nursing homes in another two [22, 40].

Studies were published between 2020 and 2023; however, nine of them completed the data collection in 2020 (e.g., [52]), 10 in 2021 (e.g., [47]), two between 2020 and 2021 [37, 50], one in 2022 [55], two between 2019 and 2020 [35, 54], and one between 2019 and 2021 [38]. Participants were mainly nurses, and their sample size ranged from 130 [42] to 672 [34] in 21 studies; in others, participants were generally identified as “care workers”, ranging from 374 [22] to 2,700 [40], while those including nursing assistants and registered nurses together ranged from 43 [48] to 287 [54] participants. The MISSCARE Survey tool was the most used (= 14, e.g., [44]), followed by BERNCA (= 1, [46]), Revised BERNCA (BERNCA-R) (= 2, [51, 52]), BERNCA-NH (= 2, [22, 40]), PIRNCA (= 4, e.g., [42]), ICU-ONC (= 1, [53]), and UNCS (= 1, [37]) (Table 1).

All studies reported a good methodological quality with minimal bias (Supplementary Table 2). Most were ranked positively for at least six (“yes” responses) out of eight questions (= 11; e.g., [39]), nine studies for at least seven questions (e.g., [44]), and five for at least five questions (e.g., [41]). Four studies failed to clarify the strategies to deal with confounding factors (e.g., [56]), while seven described these strategies unclearly (e.g., [51]). The settings and study subjects were stated as being unclear in eight studies (e.g., [52]). Additionally, in one study, the sample inclusion criteria were not detailed, while in another study, the confounding factors were not reported. The objective, standard criteria used to measure the condition were not assessable in any of the qualified studies, since the condition was considered the COVID-19 disease. At the overall level, all except six studies [25, 34, 42, 43, 46, 55] documented the occurrence of and reasons for UNC activities.

The occurrence of UNC

In the 14 studies based on the MISSCARE survey, the most frequent UNC activities were “Ambulation 3 times per day or as ordered”, “Turning patient every two hours”, “Attending interdisciplinary care conferences whenever held”, “Providing mouth care”, and “Patient teaching about procedures, tests and other diagnostic studies”. In particular, “Ambulation 3 times per day or as ordered” was the activity most missed in three studies [35, 38, 39]; it was the second unfinished activity in the study by Al Muharraq et al. [36] and the third in another three studies ([48]; in both the COVID-19 sample and the reference sample of von Vogelsang et al. [54]) (Table 2, Supplementary Table 3). “Turning patient every two hours” was the most frequent UNC activity in two studies (in the COVID-19 sample of Nymark et al. [48]; in the reference sample of von Vogelsang et al. [54]) and the second in another three ([35]; in the reference sample of Nymark et al. [48]; in the third wave sample of Falk et al. [38]). This activity was third in another four studies ([35, 36, 38]; second wave [47]) (Table 2, Supplementary Table 3). However, the first unfinished activity in five studies was “Attending interdisciplinary care conferences whenever held” ([36, 44, 49]; in the reference sample of Nymark et al. [48]; in the COVID-19 sample of von Vogelsang et al. [54]) and “Monitoring patient” in one study [45] (Table 2, Supplementary Table 3). Conversely, the least frequently unfinished activities were “Monitoring intake/output”, “Vital signs assessed as ordered”, “Bedside glucose monitoring”, and “Patient assessments every shift” (Table 2, Supplementary Table 3).

Table 2 The occurrence of and reasons for unfinished nursing care in studies using the MISSCARE survey [59]
Full size table

Considering the studies using the PIRNCA tool, the most frequent unfinished interventions were the “Coordination of care and discharge planning” and the least common the “Implementation of prescribed treatment plan” in Schneider-Matyka et al. [50]. Contrarily, Yuwanto et al. [56] discovered that “Coordination of care and discharge planning” were the least frequently unfinished activities. The other most frequent UNC activities were listed in Schneider-Matyka et al. [50] and Yuwanto et al. [56], respectively, as (i) “Offer emotional or psychological support”, (ii) “Converse with team members”, (iii) “Converse with external agency”, and (i) “Routine skin care”, (ii) “Converse with external agency”, and (iii) “Assist with bowel and bladder elimination”, while the least unfinished were, respectively, (i) “Medication administration”, (ii) “Enteral and parenteral nutrition”, and (i) “Converse with patient regarding discharge”, (ii) “Infection control practices” (Table 3, Supplementary Table 4).

Table 3 Unfinished nursing care occurrence in studies using the perceived implicit rationing of nursing care [1]
Full size table

In accordance with Tomaszewska et al. [51] and Uchmanowicz et al. [52], who used BERNCA-R, the most common first, second, and third UNC activities were “Education and training”, “Necessary disinfection measures”, and “Monitoring patients as the nurse felt necessary”. The studies identified “Change of the bed linen”, “Skin care”, and “Assist food intake” as the least frequent UNC activities [51] (Table 4, Supplementary Table 5).

Table 4 Ufinished nursing care occurrence in studies based on the basel extent of rationing of nursing care (= 1), revised basel extent of rationing of nursing care (= 2) [61] and basel extent of rationing of nursing care-nursing homes tool (= 2) [62]
Full size table

In two studies that used the BERNCA-NH tool, “Social care” and “Emotional support” reported the highest occurrences [22, 40]. The most frequent UNC activities were listed in Hackman et al. [40] as (i) “Cultural activity for residents with contact outside of nursing home”, (ii) “Scheduled single activity with a resident”, and (i) “Scheduled group activity with several residents”; in contrast, the most frequent unfinished activities in Zhang et al. [22] were (i) “Activating or rehabilitating care”, (ii) “Emotional support”, and (iii) “Scheduled group activity with several residents”. On the other hand, “Assist dressing/undressing”, “Drinking”, “Food intake”, and “Sponge bath/partial sponge bath/skin care” were listed as the least frequent UNC activities [22, 40] (Table 4, Supplementary Table 5).

In the remaining two studies, recent tools were used. In the study conducted using the ICU-ONC tool, the most common unfinished activities were “Mobilization every two hours”, “Mouth care for intubated patients”, and “Document treatments and procedures”; those least frequent were “Cardiac monitoring surveillance”, “Flag the presence of signs or symptoms of infection”, and “Titrate intravenous perfusions for hemodynamic targets” [53] (Table 5, Supplementary Table 6). In the study using the UNCS [37], the most frequent UNC for both the COVID-19 sample and the reference sample were “Performing bedside glucose monitoring as prescribed”, “Performing clinical handover to adequately inform the next shift nursing team about patients’ conditions”, and “Recording vital signs as planned”, while the least frequently unfinished activities were “Helping patient in need in ambulation”, “Providing passive mobilization/changing position in bedrest patient”, and “Providing mouth care to patients who need it” (Table 6, Supplementary Table 7).

Table 5 Unfinished nursing care occurrence in the study using the intensive care unit omitted nursing care instrument [53]
Full size table
Table 6 The occurrence of and reasons for unfinished nursing care in the study using the the unfinished nursing care survey [58]
Full size table

The reasons for UNC

Among the studies using the MISSCARE Survey, four [39, 45, 49, 55] did not report the reasons item by item. In the remaining, “Inadequate number of staff” (e.g., in Wave 1 and Wave 2 sample of Falk et al. [38]; [25]) was reported as the most significant reason in six studies, “Unexpected raise in patient volume and/or acuity” as the first or second reason in four studies (e.g., [38, 48]), and “Urgent patient situations” as the first, second, or third in six studies (e.g., [41, 47]) (Table 2, Supplementary Table 3). The reasons for UNC that were given least were “Other departments did not provide the needed care”, “Inadequate hand-off from previous shift or sending unit”, “Caregiver is off unit or unavailable”, and “Tension or communication breakdowns with the medical staff/other support departments” (Table 2, Supplementary Table 3).

Regarding the findings from the UNCS [37], “Priority setting” and “Supervision of nursing aides” were reported as the most frequent factors causing UNC, followed by “Communication”. In particular, the most frequent reasons were “Inaccurate initial priority setting”, “Tension/conflicts within the nursing staff”, and “Inadequate nursing care model (e.g., functional task-oriented model of care)”. The reasons given least were the material and human resources as well as the unpredictability of the workflows (Table 6, Supplementary File 7).

In 22 studies, UNC has been linked to other, additional factors. Among these, organisational factors, insufficient resources, and large hospital facilities were reported as increasing UNC [40, 45]; other factors (e.g., adequate staff, the quality of care, the safety of the patients in the unit, a favourable nursing work environment, and the perceived accountability, organisational support, and leadership) hindered the occurrence of UNC (Table 7). Among the work-related factors, the type of shift work (afternoon shift [35]; 12-hour shift [41]; both day and night shift (not only night shift) [47]), overtime work, the type of unit, the workloads, and other factors increased the occurrence of UNC, whereas having a few patients to each nurse or COVID-19 patients, or better staffing levels, all decreased the occurrence of UNC (Table 7). Moreover, at the individual level, less than 10 years of experience and several other factors close to the nurses’ emotional state and well-being all decreased the occurrence of UNC (Table 7).

Table 7 The additional unfinished nursing care-related factors emerged in the studies (= 22)
Full size table

The Main consequences of UNC

No studies reported the consequences of UNC.

Discussion

At the overall level, a total of 25 studies conducted mainly in European and Asiatic countries were produced during the pandemic, around 10 studies a year, continuing the tradition of this research field during difficult times for both nurses and healthcare settings. All tools available in the field were used, mostly the MISSCARE Survey, but also, on fewer occasions, BERNCA, also in its revised forms. As previously, mostly cross-sectional studies along with a few comparative studies were produced, suggesting the likelihood of a merely descriptive intent due to the challenging times. The order of UNC interventions that emerged across studies is substantially in line with pre-pandemic data, while some interesting variations emerged at the country and inter-country levels. Labour resources and reasons close to the emotional state and well-being of nurses were mentioned as most affecting UNC during the pandemic. However, none of the studies investigated the consequences of the phenomenon.

The discussion section follows the results structure and includes a reflection on the methodological quality of the studies and UNC occurrence, reasons, and consequences.

Included studies and their methodological quality

Studies released after the World Health Organisation declared the COVID-19 pandemic [30] as a period characterised by altered working conditions, workloads, and processes compared to those of the pre-pandemic era were included. No UNC differences between COVID-19 and non-COVID-19 patients emerged [63, 64], suggesting that the pandemic affected the whole system. Moreover, given the substantial disruption of the routine care processes in the health systems, which may require time to recover, and with the likelihood of not reaching the same levels of the pre-pandemic era, a comprehensive review may contribute to providing a new reference point for future studies in the field of UNC.

Fewer than 10 studies a year were produced, in line with the pre-pandemic era [64, 65]; moreover, data collection was performed mainly in 2020 and 2021, suggesting that available findings reflect the first phases of the pandemic. The leading continents in these studies were Europe and Asia, unlike in the past when the United States was the leading country, given that the missed care/left undone concepts were developed there [2]. Asian and European countries were those firstly and dramatically hit by the pandemic, thus triggering researchers to measure the UNC. However, the setting of the data collection has remained the hospital, as in the pre-pandemic era [66]: this finding is in line with the expanded capacity required in the hospitals and the recognition of their key role, especially in some waves, in facing the pandemic. Interestingly, several studies involved more units in very different institutions (e.g., [47]), which seems to suggest that this research line was scaled up during the pandemic from unit-based studies to large healthcare systems, thus embodying a reasonable health service research perspective because the whole system was changed to provide the care, and no one single part was left unaltered.

The study designs were cross-sectional with some comparative examples, as documented in the pre-pandemic era (e.g., [29]). The turbulent environments may have prevented longitudinal studies (e.g., to discover UNC outcomes). Forty-three [48] to 2,700 [40] nurses, nursing assistants, and care workers were involved, the sample sizes mirroring those of the pre-pandemic era [66]. However, no studies involved midwives, which suggests a lack of evidence in terms of what happened in maternal and paediatric departments.

Four different tools have been used to measure UNC, from those most validated across the world, namely the MISSCARE Survey [39] to more recent instruments, such as the ICU-ONC [53]. The different instruments used reflect the trends in this research field, characterised by a range of validated tools, thus preventing comparisons across studies. On the one hand, the utilisation of classic, well-validated tools may have provided accurate data and increased the comparison with pre- and intra-pandemic studies, whereas on the other hand, tools designed for a non-pandemic situation may have failed in their capacity to detect UNC in extraordinary conditions. Moreover, all tools collected UNC data as perceived by nurses, and their perceptions may have been influenced by the stress and the dramatic working conditions they were experiencing, as well as by the desire to do the best for the patients.

The overall quality of the studies was methodologically good: the extraordinary difficulties posed by the pandemic also required new strategies (e.g., to promote study participation among nurses, design protocols, and initiate studies while other priorities are perceived) in conducting research and seem to have been faced appropriately by researchers.

The occurrence of UNC

The different UNC activities, in their order, can be discussed around three main perspectives: (1) the instrument used; (2) the intercountry and intra-countries differences; and (3) the state of the evidence in the pre-pandemic era. The order of UNC interventions emerged across studies, for some countries are substantially in line with pre-pandemic data. The MISSCARE Survey studies highlighted that, during the pandemic, nurses firstly postponed or omitted interventions that call for proximity to the patient, such as oral care, or one-on-one interaction, such as ambulation. Studies using the ICU-ONC tool also showed the same trend, suggesting that these two tools can detect actions of care at the bedside. Nursing interventions related to organisation and communication were instead commonly unfinished in studies using the PIRNCA scale. Communication should also be seen as a fundamental care [67,68,69], as speaking and listening were most often seen as a nursing necessity during the pandemic. Differently, education, disinfection measures, and monitoring were the most frequent UNC activities in studies employing the BERNCA scale. Likewise, nursing interventions for patient follow-up were frequently unfinished in a study using the UNCS [37].

The most significant nursing interventions identified during the pandemic were monitoring, educating the patient, and implementing preventive measures against infections. Nurses may have felt that their usual applications were inadequate or incomplete given the growing demand for these interventions, or they may have believed that they would be unable to complete these applications out of fear of failing. Finally, social and rehabilitative nursing interventions were ranked first as unfinished activities in studies using the BERNCA-NH instrument. This reflects the contingencies of the COVID-19 pandemic, which forced residents of nursing homes to remain in their own rooms [70]. Therefore, at the overall level, it seems that nurses adopted the pre-pandemic patterns of prioritisation (e.g., failing in ensuring fundamental care) with the intent of reducing exposure in patients’ rooms for an extended period and to avoid the source of contagion [71], and/or due to the fatigue caused by the personal protective equipment worn (e.g., [72]). The rationed nursing activities did not turn out to be very different from those of the pre-pandemic period (e.g., [2, 73]), as also emerged in those studies that included comparative studies [35, 38].

However, interesting intra- and inter-country differences have emerged: at the intra-country level, two main patterns are evident. In Sweden, for example, Falk et al. [38] and von Vogelsan et al. [54] found that the three most unfinished activities are substantially the same, whereas in Jordan [35, 44] and Iran [41, 49], the first three unfinished activities differ (Table 2, Supplementary Table 3). Similarly, at the inter-country level, in those studies using the MISSCARE Survey performed across Europe, the unfinished activities seem to have similar trends in the order pattern. Comparing these countries with those where UNC has started to be measured (e.g., Iran, Jordan, Saudia Arabia, Indonesia, Sultanate of Oman), feeding the patient and offering emotional support were not missed immediately, while attending interdisciplinary meetings was unfinished at first. In the two studies using the BERNCA-NH tool, a similar divergence appeared: in the study by Zhang et al. [22] performed in China, some activities (i.e., providing emotional support and rehabilitation care) were the first to be unfinished, while in Hackman et al. [40] these were ranked as being missed less often. Examples can also be found in studies using the PIRNCA and performed in Poland [50] and Indonesia [56]. On the one hand, this seems to suggest that when the healthcare system is under tremendous pressure, as during the pandemic, the process of prioritisation is based on pre-established patterns (e.g., across Europe; [74]); on the other hand, different patterns seem to be enacted outside of Europe, mainly in Asiatic countries. Given that these countries are substantially new to measuring UNC, replicating studies to establish whether the emerged patterns are the same as those used in normal conditions is strongly recommended.

Above all, studies produced during the pandemic period report unfinished activities according to the tool used. For example, the MISSCARE Survey was developed in the early 2000s [59] and is able to measure “basic” nursing activities; therefore, its capacity to detect exactly what happened in the nursing processes during the pandemic should be debated.

The reasons for UNC

First, issues regarding human resources and the increased needs of patients were the most cited reasons in those studies using the MISSCARE tool, while issues among the staff or across departments impacted only a little. This is likely derived from the expanded capacity of the health systems under urgent circumstances [75] that increased the well-known shortages in resources, whereas facing the pandemic reduced tensions within the staff and across units, promoting a sense of collaboration [76, 77]. Moreover, nurses became infected and were not available when quarantined: all these situations seriously disrupted the capacity of nursing care [21, 22], threatening the patients’ needs [16, 17, 78]. Conversely, for Cengia et al. [37], human resources were not an issue in triggering UNC occurrence; however, this is a single study with the UNC survey tool, and although performed in several facilities, its findings may be interpreted from different perspectives: the units involved in the study may have been better equipped during the pandemic to deal with the situation, or nurses may have learnt for several years how to work under pressure, with limited resources, in a sort of “normalised” condition, where working under such conditions was not an issue [63].

Other potential reasons documented among studies are in line with those documented by Chiappinotto et al. [29]. However, two new elements emerged at the overall level among studies performed during the pandemic. Firstly, in those cases where the same reason has been documented (e.g., the role of working overtime [25, 39, 47]), no conflicting findings have been reported across studies, suggesting an evident accumulation of knowledge in the same direction. Previously, conflicting findings emerged for the same reasons across studies, in some increasing and in others hindering the occurrence of UNC (e.g., working overtime [29]). The increased homogeneity of the findings that emerged in the pandemic studies may depend on the same circumstances experienced in all healthcare services across the world. Secondly, several emotional factors at the nurses’ level (e.g., satisfaction, burnout, satisfaction with economic situation, stress) have been investigated and associated with UNC. The focus seems to be the professional and personal well-being of the nurses, reasons that may have a role as antecedents of UNC but that also express the consequences of the unfinished care phenomenon itself as well as the consequences of the exacerbated working conditions during the pandemic.

The Main consequences of UNC

No UNC consequences have been documented to date confirming the tradition of this research field in which outcomes are under-reported [79]. In difficult times with turbulent environments, unstable staff, and disconnections between healthcare settings (e.g., hospital and community settings), it would be difficult to link the occurrence of UNC to the different potential outcomes at the patient, nurse, and organisational levels [5, 12,13,14]. However, the occurrence of UNC may have bolstered the negative effects of other widely observed phenomena, such as the decreased accessibility and continuity of care observed during the pandemic, thus indirectly affecting the health outcomes at both the individual and collective levels (e.g., reduced screening, reduced care for cancer patients) [80, 81].

Limitations

This review has several limitations. First, databases were searched using well-known established keywords in the field, strictly connected with the conceptual definitions in the field and with the tools measuring the phenomenon. Moreover, given that no MeSH terms have been established in the field, researchers used keywords. Consequently, some studies may have been missed. Second, studies whose data collection period was uncertain or ambiguous (e.g., started before or during the pandemic) were excluded. Moreover, studies not using validated instruments with available reliability and validity data were also excluded, and these decisions may have introduced a selection bias. Furthermore, grey literature was not assessed, introducing additional selection bias. Third, we included only articles written in English, Turkish, or Italian, so the comprehensiveness of this review could have been threatened by the exclusion of other languages. Fourth, in the data analysis and synthesis process, an approach was adopted aiming at ensuring accuracy given the different measurement tools used in the field. Moreover, the data analysis process was conducted in an innovative manner by considering each intervention or reason at the granular level (the order, according to the statistical values) instead of the global level (global scores). This may have provided clarity, but it may have compromised the depiction of a global picture of the phenomenon. No previous similar approaches have been used in this field. Accumulating evidence with additional studies, such as summarising findings in the post-pandemic era, may corroborate the analytical strategy used.

Conclusion

UNC studies produced during the pandemic documented the occurrence of the phenomenon and its reasons mainly in the first and second waves of the COVID-19 pandemic. These studies were conducted mainly in Europe and Asia, which were the first to be dramatically affected by the pandemic. The studies involved multicentre units in the attempt to measure the whole response of the healthcare settings, mainly using the MISSCARE Survey with descriptive intents and using quality, sound research methodologies.

At the overall level, those nursing care activities that were mostly unfinished during the pandemic are substantially the same as those reported in the pre-pandemic era, suggesting that nurses applied the same prioritisation responses in difficult times. However, interesting intra- and inter-country differences emerged: those countries new to measuring unfinished care reported different patterns compared to those seen in Europe and the US, where this research is well established; they also reported intra-country variations, suggesting an interesting new course of research in the field. The new patterns that emerged should be better investigated through post-pandemic studies to discover whether they reflected the decision-making process during difficult conditions or a different prioritisation process.

Across studies, the primary reasons for UNC were listed as labour resources, followed by other specific reasons related to organisational, work, and individual variables. Substantially, the evidence is in line with that previously documented. However, findings are consistent across studies, suggesting that health services experienced similar pressure worldwide. Moreover, several emotional factors have been investigated among nurses, revealing their important role in triggering UNC. This level should be investigated further, considering the long-term consequences of the pandemic on the well-being of the workforce. Given that no studies have attempted to measure the UNC consequences, more efforts are also required in this direction.