Design, settings and participants
This comparative cross-sectional study was conducted on patients admitted to two major university and non-university hospitals. These hospitals had 350 and 320 beds respectively, and they are known as the main centers providing medical specialty and subspecialty services in Bushehr Province, southern Iran.
The patients were selected using the stratified random sampling method. The sample size was determined according to the number of beds in each hospital, number of beds in each ward, and the whole sample required in this study. With reference to the study conducted by Mossadegh-Rad [16] and the total quality ratio of the university hospital by 0.85, assuming an acceptable difference of 0.1 between university and non-university hospitals as well as α=0.05 and β=0.2, and comparison between wards using the correction formula , the sample size in each hospital was calculated to be 410 persons. Nevertheless, 1000 patients finally received the questionnaires due to the possibility of incomplete cases. In line with the survey conducted by Charalambous et al. [27], considering the trust score of nurses and the differences in the comparison group by d=1, the previously calculated sample size could cover the trust score difference among nurses.
The inclusion criteria in this study were at least 24 hours after admission, written patient consent to participate in research, ability to respond to the questionnaires, consciousness, and no psychiatric disorders based on self-reporting. According to the inclusion criteria, the pediatric and maternity wards as well as the Intensive Care Units (ICUs) wherein the patients were unable to complete the questionnaires were excluded. Therefore, the patients admitted to emergency departments for adults, women’s wards, surgical wards for women, internal wards for women, surgical wards for men, internal wards for men, oncology wards, and Coronary Care Units (CCUs) were included. The study samples were selected from the patients admitted to the above-mentioned wards within all three shifts. The exclusion criteria were unwillingness to continue the research, along with incomplete questionnaires. The patients also became informed of the voluntary basis of the study, and they were assured that their personal information and names would not be entered into the questionnaires to respect the confidentiality policy.
Data quality control
Four data collectors were recruited from postgraduate nursing students. These students were not the staff of the two selected hospitals. To ensure accuracy, consistency and completeness, one member of the research team closely supervised the data collection process. Before data collection, in a session, the chief researcher introduced the data collectors to the questionnaires and the data collection process.
Data collection instruments and procedure
Questionnaires were directly distributed among the patients by the data collectors and were delivered to them upon completion. For each participant, it took 10 minutes to complete the questionnaires. To collect the data, a demographic characteristics information form, the HEALTHQUAL questionnaire, and the Trust in Nurse Scale were used.
HEALTHQUAL
The HEALTHQUAL questionnaire was designed by Mossadegh-Rad (2018) to measure the quality of healthcare services using 30 items evaluating customers’ perceptions and expectations within four dimensions of “environment” (11 item), “empathy” (12 items), “efficiency” (3 items), and “effectiveness’ (4 items) (16). All of the items could evaluate patients’ viewpoints on service quality in two parts, i.e., patients’ perceptions (real quality) and patients’ expectations (ideal quality). The options for each item were also set on a five-point Likert scale. In addition, a four-part item could assess the “importance” of the dimensions of service quality. In this questionnaire, the quality gap was equal to the difference between expectations and perceptions multiplied by importance. Accordingly, a negative gap indicated unacceptable quality, a zero gap represented acceptable quality, and a positive gap showed quality beyond customers’ expectations. Scores 1-1.80, 1.81-2.60, 2.61-3.40, and 3.41-4.20 denoted very poor, poor, moderate, and good service quality, respectively. Additionally, scores above 4.20 specified very good service quality. In the study conducted by Mossadegh-Rad, the Cronbach’s alpha coefficient to endorse the reliability of the HEALTHQUAL questionnaire was determined to be 0.94 [16]. The content validity of the given questionnaire was similarly confirmed with the content validity index (CVI)=0.71 and the content validity ratio (CVR)=0.71 (28).
Trust in Nurses Scale
The Trust in Nurses Scale was developed by Radwin and Cabral (2010) to measure patients’ trust in nurses. It contained five items, each one with six options, namely, 1= never, 2= rarely, 3= some of the times, 4= often, 5= usually, and 6= always. Accordingly, higher scores could show higher levels of trust among patients. The content validity of this instrument was carried out recruiting a panel of experts by CVI=0.90. Furthermore, the reliability of the questionnaire concerned was determined via internal consistency with the Cronbach’s alpha coefficient of 0.93 [29].
Ethical Consideration
This study was approved by the Ethics Committee of Bushehr University of Medical Sciences, Bushehr, Iran, under the code of ethics: IR.BPUMS.REC.1397.054. It was also implemented by observing ethical considerations, including obtaining informed consent and acting in accordance with the principles of confidentiality and privacy.
Data analysis
To analyze the data, the SPSS Statistics software (version 22) was employed. Descriptive statistics (frequency and frequency percentage, mean and standard deviation), text narration, tables and a figure were used to present results. For inferential statistics, Chi-square test, independent-samples t-test, and the univariate general linear model (GLM) were used. For all cases, values less than 0.05 were considered statistically significant.