2.3.8. Barriers

Even if administrators and employees might have the best intentions to create and share knowledge, there are often obstacles complicating their efforts. Most studies exploring the subject have a qualitative orientation [46–48]. Likewise, we chose an open-ended question of AKMI to reveal more barriers to implementing KM in hospitals. Still, we also entered three items for the quantitative part of the scale.

#### *2.4. Pilot Study*

We initially developed a pilot questionnaire of 38 questions. Then, we asked the opinion of three experts from the Hospital. The first expert was a medical doctor, who was the manager of a public Health Center, and responsible for the continuous education of physicians who undertake their internship. The second expert was the administrative manager of a General Hospital, an expert in Hospital Administration. The third expert was a registered nurse with a Master's degree in special education [49]. After the discussions, the number of questions increased to 64.

The extensive questionnaire of the 64 questions was pilot tested in a sample of 31 employees (physicians, nurses, midwives, health visitors, and administrative staff) who work at two public health centers and a public Physical Medicine and Rehabilitation Center. The participants (24 females and 7 males) had a mean of 19.82 years of working experience (SD 7.99) and a mean age of 45.42 (SD 6.72) years.

The questionnaire was completed for a second time 15 days later by the same group of people. Test-retest was measured with the intraclass correlation coefficient (ICC, two-way mixed model on absolute agreement) [50] for all the questions except the demographics. The results of ICC are interpreted according to the scores as follows: <0.40 poor, 0.40–0.49 adequate, 0.60–0.74 good, 0.75–1.00 excellent [51]. The ICC of our study was excellent (ICC average measures: mean 0.904, min 0.717, max 1.000). For the questions answered by the Likert scale, we measured Cronbach's Alpha coefficient wıth mean value equals to 0.905 [52]. Figure 1 shows the dimensions of AKMI.

**Figure 1.** The architecture of Applied Knowledge Management Instrument (AKMI).

#### *2.5. Data Analysis*

Data were collected and entered on Microsoft Excel and analyzed with SPSS v22 (SPSS Inc, Chicago, IL, USA). Missing values were less than 3.2% for each item (overall missing values less than 4%). We conducted Little's MCAR test, which indicated that values were missing completely at random. Therefore no entry was excluded from the analysis, and missing values were not replaced because sometimes imputation techniques for handling missing data result in biased estimates [53,54]. Reverse coding questions were recoded into different variables before further analysis.

To test whether data were appropriate for factor analysis, we measured the Kaiser–Meyer–Olkin (KMO) coefficient (KMO = 0.696). Furthermore, we carried out Barlett's test of sphericity, which showed a significant p-value <0.001. Both tests indicated that the dataset was suitable for factor analysis [55]. Furthermore, the sample size of the dataset was larger than 250, which is a prerequisite for obtaining reliable results [56].

#### **3. Results**

#### *3.1. Study Population*

This study took place from February to June of 2015 at General Hospital of Kalamata, which is a medium-sized public hospital with 300 beds, and the biggest (out of two) from a rural area of 200,000 inhabitants in Messinia, Greece. In 2015, the hospital had approximately 700 employees, of which 30% were males and 70% females. We asked 300 employees to participate in the study, and 261 employees agreed and completed the questionnaire (87% response rate). Even if there is no rule of thumb for the ideal sample size for testing a newly developed scale, a sample size of more than 200 people is acceptable [57].

Demographics of participants as regards gender, age, professional status, and working experience resembled the rest of the employees, who did not participate (Table 1). The completion time was 10–15 minutes.


**Table 1.** Demographic characteristics of the 261-employee sample.

#### *3.2. Validity*

Face validity is a measure for the suitability of the project. It concerns the appropriateness, sensibility, or relevance of a test and its items and evaluates how it appears to the people who undertake it. Even if face validity seems an ambiguous term, it is essential for the success of a test or scale [58]. Many participants reported that our questionnaire was exciting and comprehensive. Additionally, they realized how effective knowledge management is and stated that they could participate more in knowledge sharing in the future. Content validity is a characteristic associated with the scale's adequacy for the measurement of the concept under consideration. It can only be checked subjectively through its approval by connoisseurs [59]. Our questionnaire was a subject of extended discussions at the pilot phase with three experienced health professionals with various educational and professional backgrounds, to ensure content validity. Additionally, we performed a factor analysis to establish construct validity [60].

#### *3.3. Exploratory Factor Analysis*

We conducted factor analysis, with extract method Alpha factoring, which resulted in 19 components with an eigenvalue greater than 1.0 that explained 52.8% of the variance. Due to low scoring, we removed a group of questions regarding "facilities" and nine more items. Most of the single items we excluded were reverse coded, and that means that they confused the participants. Following principal component factoring and varimax rotation, we repeated the analysis with a forced nine-factor solution. This time, the solution explained 56.87% of the variance. Table 2 illustrates results from factor analysis.

The estimate for the internal consistency of the entire questionnaire (Cronbach's alpha) was 0.802. Each dimension had the following Cronbach's α: perceptions—0.724, intrinsic motivation—0.626, extrinsic motivation—0.739, knowledge synthesis—0.652, knowledge sharing—0.570, cooperation—0.567, leadership—0.717, culture—0.821, barriers—0.664. Median, interquartile range, and outliers of the results from the nine dimensions of AKMI are presented in Figure 2, and the final version of AKMI with preliminary results is shown in Table 3.

**Figure 2.** Boxplots of the nine dimensions.


**Table 2.** Exploratory factor analysis of the knowledge-management (KM) dimensions.



We further estimated the polychoric inter-correlations among the AKMI subscales and reported significant correlations between factors. Table 4 shows the estimated polychronic intercorrelations.


**Table 4.** Polychronic inter-correlations between factors.

Correlations are significant as follows: \*\*\* *p* < 0.001, \*\* *p* < 0.05, \* *p* < 0.01.

#### **4. Discussion**

This research aimed to develop a questionnaire to understand the concepts of knowledge management and to investigate the organizational factors that affect all aspects of the knowledge creation process within hospitals.

Knowledge management is related to sustainability, organizational learning, knowledge transfer, quality of care and safety, type of motivations, and barriers, all of which will affect the level of service.

#### *4.1. Knowledge Management and Sustainability*

The application of knowledge management can lead to a sustainable healthcare system, and leaders can achieve the goals of their organizations [61]. It is important to note that the knowledge management process can be significantly related to improvements in the quality of healthcare as well as the organizational-level of social and economic outcomes, as stated by Popa [10,62]. Doctors may process the information related to the healthcare industry, and based on their experience and knowledge, can improve the quality of the system and the management of their patients. Moreover, patients can increase their knowledge from various sources like the internet, social media, and other medical staff. In this way, patients can determine or change their behavior and thoughts and demand the best possible service. The optimal management of the knowledge process affects the quality of a system.

Social stainability issues in healthcare facilities is another aspect which is explained by [63,64]. An organization with collaboration can apply knowledge management to share information to make healthcare organizations sustainable.

#### *4.2. Knowledge Management and Human Resources*

Knowledge is also regarded as organizational culture, skills, reputation, intuition, and codified theory that influences human behavior and thoughts [65,66]. There is also a concern about the current and future status of human resources management in healthcare organizations [67] and the impact of human resources information systems technology. Each organization will need to use HR practices that will balance evidence from data, its objectives, individual factors, and Human Resources Information systems. Organizations are becoming increasingly aware of the importance of employees in gaining and maintaining competitive advantage.

The competitiveness of a healthcare organization depends on the effectiveness of its knowledge management [62], and the knowledge-sharing process helps sustainable engagement in healthcare.

#### *4.3. Knowledge Management and Organisational Learning*

With knowledge management, healthcare leaders can understand how collective learning enhances the quality and safety improvement of hospitals. Organizations can support the process of internal learning if the goal is the improvement of their services. External knowledge acquisition often occurs through processes involving people. Knowledge management can help to reduce errors. For example, effective control is achieved using a clinical decision-support system. As a result, the potential reduction of medical errors can affect the improvement of healthcare delivery.

For example, research suggests that collective learning plays a role in improvement [68]. Specifically, cooperative learning is the process of gaining information which helps the capabilities in groups and organizations. Another process is collective learning, which has to do with the understanding and skills in groups and organizations [68,69]. Collective learning differs from individual learning because it requires individuals to analyze and interpret organizational experience [68].

The implementation of knowledge management can be thought of in two different ways [70,71]. The first is that there is a possibility that knowledge management to increase the autonomy of the medical staff by enhancing knowledge access. Knowledge sharing can lead to knowledge creation. On the other hand, controlling activities of the team can decrease collective intelligence. The excess of autonomy can encourage individuals to destabilize the organization, and there is a chance for them to act against the interests of the organization.

#### *4.4. Knowledge Management and the Developed Questionnaire*

Scientific interest in the various aspects of knowledge management can allow the connection of past results and the creation of knowledge. The findings and their implications should be addressed in the broadest context possible. Future research directions may also be highlighted. Perceptions of knowledge management were examined for another group of professionals, such as librarians in India [72] and other sectors, like construction and design companies in Spain [73]. Comparisons have been made between the perceptions of employees about knowledge management from small and large organizations in the United Kingdom [74]. Intrinsic and extrinsic motivations of KM were explored by researchers from various scientific fields [30,75]. There is still a debate in this field if external rewards can be considered as drivers for knowledge sharing, and our questionnaire aspires to clarify this issue. Knowledge creation, sharing, and cooperation are amongst the most researched topics in this area. However, in the healthcare sector, the focuses were mainly qualitatively analyzed [16], even if there are a small number of surveys, e.g., [37]. As regards leadership, studies have indicated individual styles of leadership to be significantly associated with the art of KM practices [76]. Zheng et al. [77] suggest that KM fully mediates the impact of organizational culture, and Leidner et al. [78] claim that organizational culture influences knowledge management initiatives. Based on these findings, we will subsequently create a model to determine the correlation structure of KM dimensions using a structural equation modeling procedure.

We think that self-efficacy plays an essential role in knowledge sharing. Until now, self-efficacy is mainly correlated with computer skills and knowledge-management systems [79] and less with occupational self-efficacy. With our dataset, we could check for significant connections between occupational self-efficacy and intentions to create or share knowledge.

The barriers of knowledge management procedure will be studied using the information we have collected with a closed and open-ended question. We asked health professionals to name the three most essential barriers according to their experience about the implementation of knowledge management in their organizations. The rationale of the task is to reveal existing barriers, especially in their working environment, and understand the correlations of barriers with the rest of the dimensions of the set-up questionnaire, e.g., leadership, and organizational culture.

The main advantage for the use of a specific knowledge-management instrument for healthcare units concerning a standard KM questionnaire is that the former takes into account the sui generis nature of the healthcare environment and the particular type of working relationships among health professionals. Additionally, the design of AKMI was done cautiously, with carefully examined methodological steps of an exhaustive literature review, pilot testing and retesting extended discussions with health professionals, and item reduction with factor analysis according to the main findings. The completion time was acceptable, and the dropouts were practically non-existent. Finally,

participants spontaneously expressed their content after completed the questionnaire by stating that "this was their first step to actively participating in the knowledge-management process."

In terms of limitations, there are some caveats about specific dimensions of the questionnaire due to a just fair Cronbach's alpha score. Furthermore, our study does not permit premature generalization of the results obtained.
