Were the unfinished nursing care occurrence, reasons, and consequences different between COVID-19 and non-COVID-19 patients? A systematic review

We included studies published since 2020 by carefully inspecting the data collection period considered by each: COVID-19 cases and deaths have been reported since December 2019 in some countries, while the formal announcement of the pandemic occurred on 11 March 2020 [20]. However, the health service capacity was expanded at the beginning of the escalation phase [32] and this may have also affected nursing care [33]. With the accumulation of evidence clarifying the whole history of the pandemic, updating this review regarding the right period of inclusion might be important to describe its effects on UNC over the waves comprehensively, as the continuing restructuring of the health service, on the one hand, and the fatigue of nurses, on the other, may have also influenced the perceptions.

We decided to include only comparative studies aimed at confronting data in the same settings to minimise the role of other factors on UNC data documented at the unit levels [9]. This decision ensures increased homogeneity of the external factors affecting nurses during the care of both groups of patients, whereas it should be acknowledged that some additional internal factors (e.g., psychological stress) generated by the COVID-19 pandemic may have influenced the perception of nurses involved in the pandemic period, as well as across waves, compared to those involved in the pre-pandemic period. Summarising all studies performed during the pandemic without comparative intents may inform on the methodological challenges encountered in doing research during a pandemic and on UNC occurrence and reasons in challenging times when care routines are disrupted [34]. Having a comprehensive baseline may also constitute a new reference point for the post-pandemic studies in the field of UNC, given that the healthcare settings in the pre-pandemic period were characterised by different priorities, work conditions, and issues, threatening the comparison [35].

Some missed data of the studies were collected from authors or estimated by the research team. We decided to adopt a co-constructive approach by involving authors, given the difficulties in doing research during the pandemic [27] that may have affected its quality. However, the quality of the included studies, as evaluated with the Joanna Briggs Institute [25] checklist, was appropriate in all items, except for those regarding the confounding factors that have been less precisely described. The work environments and the nursing care processes were chaotic (e.g., [36]); thus, taking into account all confounding factors was a challenge. Moreover, while the exposed group (COVID-19 patients) was substantially homogeneous, the comparison group (e.g., medical surgical, critical care patients) may have been affected by different health issues requiring different elements of care. Imbalances in the main profile of patients may have introduced biases in the nurses’ perceptions.

A few UNC comparative studies have been performed in different countries. In the attempt to establish differences, if any, three main patterns of comparison have emerged: (1) UNC before and during the pandemic (first wave), (2) before and during different waves of the pandemic (second and third), and (3) simultaneously during the pandemic (second wave) comparing COVID-19 and COVID-19-free units. The first and second patterns were prevalent given that four studies based their comparison on data collected before the pandemic (e.g., October 2019 [29]) and then involved the same settings to detect changes: in these cases, a well-established line of research (e.g., [29]) may have allowed the rapid development of the research action by repeating the same protocol used at the baseline and across different waves (e.g., [31]). Differently, only one study [26] compared at the same time the perceptions of nurses caring for COVID-19 patients and of those caring for patients in COVID-19-free units. It was likely difficult to conduct simultaneous comparisons, not only for the extremely critical conditions of the nursing staff where the priority was to prevent any form of additional burden (e.g., filling in a questionnaire), but also because of the continuous adaptation of the unit’s missions, with some of them exclusively dedicated to COVID-19 patients [26].

One study collected data early, in March 2020 [24], thus describing the first wave and suggesting the capacity of researchers to be ready to detect an emerging issue. One collected data mirroring a longitudinal approach, with repetitive data collections considering the second and third waves [31], while one used a simultaneous approach by collecting data in the second wave [26]. Data were collected up to May 2021 for the exposed and in 2019 or 2021 for the comparison group: as a consequence, the available findings reflect substantially the first year of the pandemic, when some strategies to increase the service capacity were implemented. Therefore, long-term consequences of these changes have not been evaluated [37]. Some studies may have been performed later and/or submitted to a journal for peer review. In the initial stage of the pandemic, the reviewing process of journals was speedy, but it became slower in recent times because of the increased number of manuscripts submitted and the progressive “normalisation” of the pandemic [38], rendering studies in the field less important. Moreover, studies were mainly monocentric in nature and conducted in hospitals [39]: medical, surgical, and critical care settings were considered, reflecting the main settings already included in this field of research [23]. Data regarding UNC in the community or residential settings have not been provided, likely due to the lack of instruments measuring the issues in these settings [23].

The sample size was limited mainly in the exposed groups of nurses, with a variable participant rate, from very low (15.7%) to high (86.7%), suggesting difficulties in involving nurses during the pandemic that may be due to the dramatic times experienced by them and the prevailing clinical priorities. In addition, the methods of data collection based mainly on paper and pencil (e.g., [24]) may have prevented participation due to the fear of being infected by touching a surface contaminated with the virus [40]. Safety reasons and green healthcare research approaches suggest improving the use of online surveys.

The profile of participant nurses is in line with that documented in previous studies in the field (e.g., [16]), where mainly females, experts, and those educated at the bachelor level were involved. Most of the individual characteristics were homogenous between the exposed and the comparison nurses, except for education in the study by Cengia et al. [26], where fewer nurses with a bachelor’s degree were in the exposed group compared to the comparison group. Individual variables have been underlined as affecting the perception of UNC among nurses; therefore, their homogeneity between the groups may have prevented confounding factors. In the case of education, higher levels have been documented to increase UNC [9], but more studies are needed to increase the understanding of how nurses set priorities while coping with complex conditions.

Data collected with the MISSCARE Survey tools were used in all studies, allowing comparison of the findings. Cengia et al. [26] used the UNC Survey that was developed from the MISSCARE Survey, thus sharing the main items [41] in both Section A, measuring the elements of unfinished care, and Section B, measuring the underlying reasons as perceived by nurses.

Interesting trends emerged in the UNC occurrence that can be interpreted under three main lines. First, the overall score was significantly different at the study level in only one study [24], where several items, all in the same direction, indicated a higher UNC occurrence among COVID-19 patients. The study was performed in the first wave when nurses were unprepared, exposed to unprecedented organisational changes, emotionally burdened, fearful, and in high levels of uncertainty [42]. Conversely, no differences, except for one item indicating less UNC among COVID-19 patients, were found by Cengia et al. [26] in their study conducted in the second wave. Similarly, Falk et al. [31] mainly found less UNC among COVID-19 patients in the second and third waves. After the first wave, nurses were more prepared, and healthcare services were prioritised towards COVID-19 units, putting non-COVID-19 patients at risk to receive less care [43]. In those studies [29, 30] reporting mixed directions (in some items increased and in others decreased UNC among COVID-19 patients), this may be related to the different settings of care (intensive care vs. medical and surgical units) but also to the prioritisation of different needs (e.g., mouth care, administering medications on time for COVID-19 patients) that may have changed the process of care.

At the item level, less care was given to COVID-19 patients across studies in “turning patients every two hours”, “providing skin/wound care”, and “responding to call light within 5 minutes”, while increased care was given in “setting up meals for patients who feed themselves” and “medication administration within 15/30 minutes of the scheduled time”. Care was mixed in other items (“ambulation three times a day”, “mouth care”, “assisting with toilet needs within 5 minutes of request”). Researchers have documented similarities in the elements of unfinished care and in their order, suggesting a stable pattern across countries and over the years [44]. This order seems to be altered during the pandemic. The model of care delivered in the units assisting COVID-19 patients was based mainly on intervention clusterisation (“I enter the room and take care of all the patient’s needs”) to avoid cross-contamination and decrease the fatigue level related to the use of PPE [45], and this may have changed the patterns of care (e.g., “being able to respond rapidly to a call light when I assist another patient”). Moreover, the context of care was changed, limiting the spread of the virus, and this may explain why in some studies “ambulating the patients three time a day” was missed [24, 29] while in others it was ensured [31], possibly due to the care setting allowing or limiting movements. Surprisingly, relational care (e.g., “emotional support”) was more significantly unfinished among COVID-19 patients only in one study [24] conducted in the first wave and not in the others; this may indicate the effects of strategies used by nurses to compensate for the potential lack [46, 47] as well as a sort of “normalisation” in both groups regarding the limited emotional care provided due to the PPE, family visiting restrictions, and the limited personal contact recommended to prevent virus diffusion.

Some interpretations of the findings may also be offered at the instrument level measuring UNC. For instance, COVID-19 patients were recommended to be kept in the prone position [48]. Thus, the item “turning patients every two hours” mainly to prevent pressure sores may have been misinterpreted or attributed less meaning. Moreover, the list of items included was not able to reflect the practices required to prevent the spread of infection as infection prevention/control programmes [49]. Thus, instruments may have missed the capacity to capture some specific elements of care required during the pandemic.

First, all reasons were different between COVID-19 and non-COVID-19 patients at the overall level only in two studies [24, 26], but in the item analysis, the specific factors triggering UNC were opposite: several and composite in the first wave [24], suggesting that the units were unprepared to support nursing care, and limited and specific (regarding nursing aides and nursing care models) in the second wave [26]. Although studies were conducted in different countries, this may suggest that facilities increased their support for COVID-19 units and nursing care over the waves. This seems to be true also in the case of two specific items - “medications were not available when needed” and “supplies and equipment were not available when needed” - which were significant reasons in the first wave [24] but not in the second and third waves [31]. This may confirm that in the initial phases of the pandemic, the resources were lacking and then prioritised and provided to COVID-19 patients. Statistical differences emerged when comparing the pre-pandemic period and the following waves [31], suggesting that resources were devoted mainly to COVID-19 patients.

Two studies did not find any differences in the reasons [29, 30], which may be interpreted at the instrument level, as the instruments may have been unable to capture all factors triggering UNC, given that they were developed before the pandemic. For example, the lack of experience among nurses deployed from one unit to another to increase the care capacity [15] is not contemplated among the reasons for UNC, which instead reflects the routine deficiencies and not those in extraordinary circumstances like during the pandemic. The development of tools measuring UNC reasons in specific situations may be recommended for future studies.

No UNC consequences have been documented to date suggesting and area of further investigation. The limited time available and the critical conditions of the healthcare services [50], where follow-ups have been interrupted, may have prevented any attempt.

This review is affected by several limitations. Six studies were excluded from the inclusion process because no answer was obtained despite attempts to contact the authors to assess when data were collected. We included studies comparing UNC data collected during the pandemic with data collected before: in all these studies [24, 29‐31], the pre-pandemic period ranged from October to November 2019, just before the outbreak. Other excluded studies (e.g., [14]) could have had data from the pre-pandemic period available, and they could have been asked for the database to evaluate the differences, if any. However, according to the aims of the review, we decided not to perform any secondary analysis. Furthermore, while assessing the reasons for UNC, only those collected with validated instruments were considered without considering additional variables explored in the included studies at the different levels (e.g., nurse-to-patient ratio) due to the variability of data collected across studies preventing any form of comparison and summarisation of the findings to detect trends, which was the main intent of this review.

UNC is universally used as an indicator of the quality of nursing care. Findings indicate that the first wave, that of the escalation in which hospitals urgently expanded their capacity to ensure care for COVID-19 patients, was the wave in which nurses perceived higher UNC among these patients. In later stages, non-COVID-19 patients were more exposed to UNC. A similar trend also emerged among the reasons, suggesting that the prioritisation of the system towards COVID-19 patients caused missed care of non-COVID-19 patients by reducing resources. This suggests that nurses need support in transiting and adapting the models of care in the escalation phase of healthcare services while dealing with a pandemic, and later those nurses caring for other patients should be supported in providing the expected care. Furthermore, with respect to UNC reasons, the long-term effects on the nurses of both cohorts generated by the turbulent organisational conditions and the nursing care routine disruptions can cause increased fatigue and UNC, suggesting the need for additional support also when the pandemic is over. Moreover, because decisions regarding UNC minimisation must be timely, a standardised electronic tracking system assessing UNC occurrence in the e-health record systems is suggested. Raising awareness among both clinical nurses and managers might help them to recall the need for interventions and might also inform decisions regarding the supports/resources needed. Additionally, developing care protocols facilitating the decision-making processes of nurses against the unknown, especially when deployed from one unit to another, without appropriate training, might be helpful. Conducting a continuous internal evaluation for UNC would minimise its occurrence: in this context, a systematic intentional rounding intervention can be useful.

Above all, it is important to continue to investigate UNC during the pandemic, especially in this phase in which the return to the normality allows researchers to proceed with more accuracy in their investigations. Many studies could be carried out as secondary analysis if researchers who collected data during the pandemic (even in the waves following the second to which the included studies refer) compared with those collected before, in the same settings, are available. Furthermore, studies investigating long-term effects of UNC would be needed to allow a fuller understanding of the phenomenon among both COVID-19 and non-COVID-19 patients.