Evaluation of COVID-19 Patient Safety Compared to Non-COVID-19 Patients and Predisposing Factors of Nursing Errors

Objectives: The present study aims to determine Patient Safety in hospitalized patients with COVID-19 compared to non-COVID-19 ones and find predisposing factors of nursing errors according to nurses’ perceptions. Methods: This descriptive-comparative research em-ployed data from 800 nurses in eight Iranian hospitals in 2021 using three researcher-made questionnaires of sociodemographic, patient safety indicators, and predisposing factors of nursing errors, with high reliability and validity. The collected data were analyzed using ANOVA, independent t-test, and the SPSS22 software. Results: The mean patient safety scores in patients with and without COVID-19 were 3.42 ± 0.17 and 3.74 ± 0.06, respectively. The highest differences in patient safety were attributed to infection control (0.66) and patient fall (0.56) dimensions. The most common causes of nursing errors were related to management (2.67 ± 1.39), and the most common predisposing factors of nursing errors were high workload, low ratio of nurses to patients, and fatigue. Conclusions: COVID-19 patients have lower safety than non-COVID ones. Also, improper management and high workload lead to nursing errors. Therefore, the authorities must devise appropriate strategies to reduce the nurses’ workload and improve patient safety, especially in COVID-19 patients.


Introduction
Since March 2020, the world has faced a severe threat called COVID-19. According to the WHO, the COVID-19 pandemic is currently the most critical health crisis in the world and the most significant challenge and threat facing the world and humanity (1,2). This disease is a public health problem that has claimed the lives of many men, women, and children worldwide. Until August 6, 2021 (when this article was written), over 201 million people worldwide had been diagnosed with the disease, and 4,284,467 had died, as confirmed by the WHO (3). The sudden increase in the demand for health care led to overload and the collapse of health systems (4). COVID-19 is an emerging disease with unknown clinical and therapeutic symptoms (5,6) transmitted through close person-to-person contact. Infection control experts consider patient isolation an essential and straightforward way to control the infection and spread of the disease; therefore, hospitalized COVID-19 patients are isolated in separate wards. However, using such a precaution may not be totally safe (7). Isolation for infection control is a process that makes the patients subject to medical errors and side effects (8).
In addition, unmeasured organizational factors, such as the prevalence of new diseases, high patient turnover in medical centers, and unit-level workload, may contribute to the variance in missed nursing care, increased medical errors, and threaten safety of the patients (9). Nursing error is a type of error by healthcare team members that can result in irreparable and irremediable damage (e.g., permanent disability or death) (10).
Nurses comprise the largest group of healthcare professionals who interact most with patients. Promoting a safe environment is an essential nursing task and plays a significant role in care management (11). During the COVID-19 pandemic, nurses have been at the frontline of fighting against the disease, and critically-ill patients rely on them more than ever (12,13). Studies have shown a significant relationship between the prevalence of COVID-19 and the problems such as stress, depression, burnout, anxiety, and reduced productivity in health workers (13)(14)(15)(16)(17). Medical workers' mental health problems impair their attention, cognitive functioning, and clinical decision-making (18), consequently increasing medical errors and incidents and, ultimately, putting the patients at risk (19).
Health worker safety and patient safety are inseparably connected domains. Health and safety risks to health workers can lead to risks for patients, patient harm, and adverse patient outcomes. The risks to patients and healthcare workers during this pandemic are more significant than usual (20,21). Thus, health workers cannot provide quality and safe patient care in environments where their safety is threatened (22).
Error is inevitable in all professions, particularly health-and treatmentrelated jobs. Care provision in clinical settings is associated with var-ious concerns, including adverse effects of treatment, accidents, and medical errors (23). The most common nursing errors in a hospital setting are falls, pressure ulcers, infections, medication errors, documenting errors, and equipment injuries (24). Every year, 134 million side effects occur in hospitals of low-and middle-income countries due to unsafe care, leading to 2.6 million deaths (25). In high-income countries, it is estimated that onetenth of patients get injured while receiving hospital care (26), about 50 % of which are preventable (27). Failure to maintain patient safety significantly increases the cost of care, morbidity, and mortality. Patient safety is a priority in the healthcare system to improve treatment outcomes and prevent complications (23).
Patient safety is a critical element of quality in healthcare and is considered an indicator of the quality of care (28). Nurses' priorities include patient safety and improving the quality of delivered care to patients. Improving patient safety involves identifying the incidents, analyzing the trend of events, and developing corrective solutions for promoting the system. There are several mechanisms to reduce adverse events and improve patient safety, but implementing them requires studies to identify the current situation. Due to the novelty of COVID-19, the studies in this field are limited. To plan for preventing and decreasing the rate of nursing errors and increasing patient safety, especially for COVID-19 patients, evaluation of fundamental patient safety indicators and predisposing factors in nursing errors can reveal the gaps in this area in the health system and might be very helpful.
Therefore, the present study was conducted to determine the patient safety in hospitalized patients with COVID-19 compared to non-COVID patients and identify the predisposing factors of nursing errors according to the nurses' perceptions.

Methods
This descriptive-comparative research used convenient sampling to collect information from 800 frontline nurses of different levels in eight hospitals in Iran from April 1 to August 7, 2021. According to Cochran's formula and the studies conducted on limited populations using convenient sampling, one way to determine the sample size for confirmative factor analysis (CFA) is to use the number of parameters and assign at least five cases per parameter. Because the questionnaire has 100 parameters, we estimated 800 samples (eight cases per parameter) to be appropriate, but 880 individuals were selected as the sample size to be closer to the community distribution and random sampling, considering a loss rate of 10 % (29,30). The exclusion criteria were staff reluctance to participate in the study, having less than six months of experience, and having a managerial position. To follow ethical considerations, necessary permission was obtained from the hospital authorities to conduct the study. Anonymous questionnaires were provided to the nurses once the researcher explained the research objective, the optionality of participating in the study, and how to cooperate after obtaining their written consent. The confidentiality of the information was emphasized when distributing and collecting the research data. In this study, the nurses were divided into two groups: working in COVID-19 units (COVID-19 general units, COVID-19 intensive care units) and other units (inpatient non-COVID-19 units).

Data collection tools
This study used three questionnaires, the first dealing with the nurses' sociodemographic characteristics, and included nine items. The second questionnaire was on the patient safety indicators (self-designed) extracted upon reviewing the literature (31,32). The patient safety indicators questionnaire included nine scales: pharmacological measures (14 questions), nursing care (22 questions), patient injury during care (seven questions), patient identification (seven questions), patient fall (seven questions), patient training (14 questions), diagnostic-therapeutic methods (seven questions), infection control (ten questions), and safe blood transfusion (12 questions). Five of the scales and their 60 subscales were adapted from Fakhrodin's research, four and 36 subscales were taken from the validation booklet of patient safety friendly hospitals (PSFHI), and four questions on oxygenation of patients were added. The questionnaire was designed to assess the number of nursing care errors during the previous month, and the items were measured based on the negative four-point Likert scale: never (score 1), 1-3 times (score 2), 3-5 times (score 3), and more than five times (score 4).
The third questionnaire addressed the predisposing factors of nursing errors (self-designed) extracted upon reviewing the literature (33). The questionnaire included five items: mental and physical conditions of nurses and their skills (six questions), team coordination (four questions), management (nine questions), patients (three questions), unit's physical and environmental conditions (seven questions), and the shifts in which medical errors occurred. The scoring method was as follows: very low (score 1), low (score 2), medium (score 3), high (score 4), and very high (score 5).
Ten faculty members confirmed the face validity and qualitative content of the questionnaire in this study. Cronbach's alpha was also used to measure the questionnaires' overall reliability, dimensions, and variables. To this end, the questionnaires were first provided to 30 people to measure overall and component reliability.
The reliability of the predisposing factors of the nursing errors questionnaire (0.87) and that of the Patient Safety indicators questionnaire (0.85) were obtained as well ( Table 1). The Cronbach's alpha was more significant than 0.7, indicating the high reliability of the questionnaires. The Kolmogorov-Smirnov test was also used to evaluate the normality of the research data. The significant numbers obtained were greater than 0.5, showing the normality of the data distribution.
The collected data were analyzed using the SPSS22 software. The analysis of variance, independent t-test, Spearman's correlation, and descriptive statistics such as frequency, mean, and standard deviation with a significance level of p < 0.005 were also used to analyze the data.

Analysis of nurses' sociodemographic factors on patient safety
The results of the study in terms of the effect of the nurses' gender on patient safety showed no significant difference between the safety of the patients and male (3.58 ± 0.22) and female (3.57 ± 0.22) genders.
Considering the effect of the nurses' disease on patient safety, the safety of the patients cared for by the nurses who previously contracted COVID-19 and those who did not contract the disease was 3.45 ± 0.25 and 0.65 ± 0.16, respectively. The patients receiving care from the nurses whom COVID-19 did not infect had higher safety than those under care by the nurses previously infected by the disease.
Patient safety was higher in the units where the number of patients under the care of a nurse was ≤ 5 (3.63 ± 0.15) compared to the units where more than five patients were cared for by a nurse (3.01 ± 0.23), and the independent t-test (DF = 413, p = 0.00, t = 4) showed a significant difference. According to the results, patient safety was affected by the number of patients cared for by a nurse, and it was higher in the units with fewer patients.
The results also showed that working extra shifts affected patient safety, and the patients cared for by the nurses who worked overtime had lower safety (3.43 ± 0.23) than those under the care of the nurses who did not (3.61 ± 0.05).  The nurses determined that the highest mean differences between the patient safety scores of COVID-19 and non-COVID-19 patients were those of infection control (0.66) and patient fall (0.56), respectively.

Analysis of patient safety indicators in hospitalized patients with COVID-19 and patients without COVID-19
The most significant mean differences between the patient safety scores of the patients admitted to non-COVID-19 units and intensive care units of COVID-19, determined by the nurses, were those of infection control (0.65) and pharmacological measures (0.33) The results of the ANOVA test in Table 3 show a significant difference between the mean scores of pharmacological measures, provision of nursing care, injury to patients during care, patient identification, patient falls, nurse-to-patient training, diagnostic-therapeutic methods, infection control, and safe blood transfusion among COVID-19 and non-COVID-19 patients admitted to the hospitals.

Analysis of common causes of nursing errors
As shown in Table 4, the most and the least common predisposing factors of nursing errors determined by the nurses were mental and physical conditions of nurses and their skills (category 1), i.e., fatigue and lack of acquaintance with professional rules and regula-tions, team coordination (category 2), i.e., inappropriate relationships between team members and lack of patient involvement in the care process, management aspects (category 3), i.e., high workloads for nurses and lack of policies and guidelines, patient conditions (category 4) , i.e., a large number of patients and improper behavior of patients and their relatives, and physical and environmental conditions of the unit (category 5), i.e., working night shifts, drowsiness, and poor lighting.
From the nurses' viewpoint, the most and the least common causes of nursing errors were related to management aspects (2.67 ± 1.39) and the environmental conditions of the unit (2.38 ± 1.25), respectively. the most common reasons for nursing errors were high workload, low ratio of nurses to patients, and fatigue. According to the results, most errors occurred during night shifts (71 %).

Correlation analysis of patient safety questionnaire dimensions with each other and with the result of the study (patient safety)
In Table 5, the results of the Spearman's test show that all dimensions of the patient safety questionnaire significantly correlated with each other and with the study outcome (patient safety). The correlation scores of the questionnaire dimensions and patient safety ranged from r = 0.46 to r = 0.87. Table 5. Findings on the correlation between the scores of safety dimensions and the total score of patient safety

Discussion
This study mainly discussed the patient safety indicators in hospitalized patients with COVID-19 compared to those without COVID-19 and the most common predisposing factors of nursing errors. Unfortunately, due to the novelty of the disease, the researcher could not find any domestic or foreign articles or research on the safety of hospitalized COVID-19 patients compared to other admitted ones to compare results.
The findings of this study showed a significant difference between the mean scores of patient safety indicators (pharmacological measures, providing nursing care, injury to the patient during care, patient identification, patient fall, nurse-to-patient training, diagnostic-therapeutic methods, infection control, and safe blood transfusion) in COVID-19 patients compared to non-COVID-19 ones admitted to hospitals.
According to the results, the mean score of safety indicators of the patients admitted to non-COVID-19 units was higher than those admitted to COVID-19 units, showing better observance of patient safety indicators in non-COVID-19 patients compared to the other group. Similar studies on COVID-19 patients hospitalized and isolated in the under-pressure health care system showed that isolation could be associated with patient safety incidents (34)(35)(36). In line with the results of this study, the study by Henry Thomas Stelfox et al. (2003) found out that isolated patients had experienced twice as many preventable side effects as non-isolated ones and expressed greater dissatisfaction with treatment (37). The findings of the study by Jiménez-Pericás et al. (2020) showed more adverse effects in isolated patients than non-isolated ones (38). In their study, Bruyneel et al. (2021) stated that COVID-19 patients admitted to the intensive care unit (ICU) required more nursing time than those without COVID-19 in the ICU (39). In addition, the findings by Hamamoto al. (2021) indicated that more nursing staff was required to care for patients critically ill with coronavirus disease in intensive care units (40).
Health care workers have struggled to connect, touch, engage, and communicate with COVID-19 patients behind their protective equipment (41). Lower safety of COVID-19 patients compared to non-COVID-19 patients could be due to the need for more nursing time and nurses' high workload and fear of getting the disease.
Based on the results, the mean score of patient safety indicators in the patients admitted to COVID-19 intensive care units was higher than those admitted to COVID-19 available units, showing better observance of patient safety indicators in intensive care units compared with the inpatient ones. The results of the study by Marzban et al. (2013) showed that the mean rate of injury to patients due to hospital errors was lower in inpatient units than in ICUs, which is not consistent with the results of the present study on the safety of the patients hospitalized in inpatient units and special units (42). Lower safety of the patients admitted to general units of COVID-19 compared to those admitted to intensive care units of COVID-19 could be due to the lack of beds in ICUs and hospitalization of critically-ill patients in general units, and the inappropriate ratio of nurses to patients in general units compared to ICUs.
The results also showed that the number of patients under care by a nurse affected patient safety. In other words, patient safety was higher in the units where the number of patients under care by a nurse was smaller, and in ICUs, there was an optimal ratio between the number of nurses and patients. In their study, Saleh et al. (2015) stated that sufficient nursing personnel and appropriate nurseto-patient ratios were related to improved patient outcomes, including lower mortality rates, which is consistent with the results of the present study (43).
The results of this study showed that blood transfusion and pharmacological measures accounted for the highest and the lowest scores of patient safety indicators in all admitted patients, respectively. Aranaz-Andre et al. (2008) studied the adverse effects of health care in Spain and stated that most errors were attributed to pharmacological care, which is consistent with the results of the present study (44).
The most remarkable mean differences in patient safety indicators among the patients admitted to non-COVID-19 units and those admitted to general units of COVID-19 were attributed to infection control and patient fall, respectively, which showed more errors in infection control and more falls of the patients admitted to general units of COVID-19 compared to other units. In previous studies, nosocomial infections and patient falls were the most common medical errors (45). Regarding infection control, decreased safety of the patients admitted to COVID-19 units compared with those in non-COVID-19 units could be due to the highly contagious nature of the disease and the fear of health care providers of getting infected. Thus, they did not spend sufficient time caring for the patients to avoid catching the disease, did not observe hand hygiene, and did not change gloves when going from one patient to another. Most patients admitted with COVID-19 were elderly and unaccompanied (46). Thus, the risk of falls in COVID-19 patients was higher than in other patients due to age and lack of companions.
The results showed that the safety of patients cared for by nurses who did not contract COVID-19 was higher than that of patients cared for by nurses who previously contracted the disease. Thus, the safety of the health workers directly affected patient safety. Health and safety hazards threatening health workers could be associated with patient risks. Mohammad Nahal et al. (2021) showed high burnout and low productivity rates among the nurses infected with COVID-19 (13). The study by Magnavita et al. (2020) showed a higher rate of physical symptoms, anxiety, and depression in health care providers infected with COVID-19 (47). According to the results, if health care providers were not safe, they would not be able to provide quality and safe patient care.
Since all the processes of providing health services are associated with a degree of insecurity and risk, patient safety is one of the most basic principles of providing services in health centers. From nurses' viewpoint, the most and the least common causes of nursing errors were related to management aspects and the environmental conditions of the unit, respectively. Cause and effect models indicate that at least 98 % of accidents occur due to human error and poor management (48). Like the present study's findings, the most common cause of nursing errors determined by Zeighami et al. (2016) was related to management aspects. Managers played a crucial role in preventing nursing errors. They were usually responsible for guided and controlled use of protocols, policies, and standards to prevent nursing personnel errors (33).
In a similar study conducted in 2012, Baghaei et al. examined critical care nurses' perception of nursing errors and reported that environment was the most important predisposing factor in nursing errors. Their finding contrasts with the present study's results (49). The organizations have been created to achieve some goals, and the success rate in achieving the organizational objectives is directly related to the performance of labor, especially managers. As the prominent people deciding on facing various within and outside organization problems, managers play a considerable and determining role in the success and failure of the organization and performing their duty. The higher the governance and leadership in the hospital, the higher the patient safety performance will be.
The most common reasons for nurses' errors were high workload, low ratio of nurses to patients, and fatigue. In their study, Azarabadi et al. (2018) reported that high workload, a large number of patients, and fatigue were the most contributing factors to errors, which is consistent with the results of the present study (50). In line with the results of this study, Zarea et al. (2018) stated that the most common reasons for medical errors were low nurse-to-patient ratio, high workload, and fatigue caused by extra work (51).
According to the results, most nursing errors occurred during night shifts. In line with the results of the present study, Khammarnia et al. (2021) stated that the incidence of medical errors in night shifts was higher than in any other shifts (52).

Conclusion
The COVID-19 pandemic is a global crisis that has caused some gaps in the health systems and threatened the quality of patient care and the safety of the staff and patients. Evaluation of fundamental patient safety indicators and predisposing factors in nursing errors in this study revealed the gaps in this area of the health system. The results of this study showed that the safety of the patients admitted with COVID-19 was lower than that of other hospitalized patients, and they were more likely than other hospitalized patients to experience the adverse effects of health services. Therefore, the safety of COVID-19 patients needs to be seriously improved. It is advisable to develop strategies for better interventions and transformation procedures targeting the promotion of patient safety, especially in isolated patients.
From the nurses' viewpoint, the most common causes of nursing errors were related to management aspects. The hospital managers' performance weakness wastes the financial and human resources and reduces efficiency. If the system for evaluating the performance of the hospitals' managers were efficient and there were tangible and computable indicators for evaluating their work, many costs in the hospitals could be reduced, and better services provided to covered hospitals since providing adequate services is related to sound and effective management.
The results showed that the safety of nurses directly affected patient safety, and from the nurses' viewpoint, the most common predisposing factors of nursing errors were high workload, low ratio of nurses to patients, and fatigue. In addition, most nursing errors occur on the night shift. Therefore, by improving work processes, creating safe and appropriate work environments, training the staff, and increasing nurses' motivation and safety, the authorities should pave the way to reduce nursing errors and increase patient safety. Also, to reduce nurses' workload, the number of staff required to care for patients should be proportional to the number of patients, level of disease, care environment conditions, and work shifts.
To increase the safety of the patients admitted with COVID-19 and other infectious diseases, health officials need to be prepared and provide special wards to isolate such patients in appropriate conditions. Finally, we hope that the authorities will use this research to reduce the predisposing factors of nursing errors and increase the safety of patients and nurses in hospitals.

Limitations of the study
One limitation of the present research was the data collection method through a self-report questionnaire that might lead to over-reporting or under-reporting. We suggest that future studies review patient records and report errors to measure patient safety.
The research tool (Patient Safety indicators questionnaire and predisposing factors of nursing errors questionnaire) has been developed by the researcher; we recommend that the tool be validated in different ways in future studies.

Conflicts of interest:
The authors declared no conflict of interest.