Study design and settings
We conducted a cross-sectional study in Bhaktapur provincial hospital which has a 75-bed capacity serving approximately 1000 patients in the emergency department and 9000 in the outpatient department every month [14]. The hospital provides a range of services including outpatient, inpatient, emergency, laboratory, and radiology services.
Sample size and recruitment
We recruited 204 patients from the Out-Patient Department (OPD) of Bhaktapur Hospital from May 7, 2019, to May 22, 2019. We excluded the patients if they were severely ill, were 10 years or younger, or were unable to communicate verbally. The sample size was calculated using 95 % confidence interval, 5 % margin of error, 86 % of patient satisfaction- as observed in a previous study from a similar setting [15], and a 10 % non-response rate. Our sampling frame included patients receiving OPD services. We recruited 15 patients per day until the required sample size was achieved. Using systematic random sampling, we recruited every 20th participant every day from the hospital registration, starting with the first OPD patient. If the selected patient refused to participate or was ineligible, we recruited the next patient.
Ethics approval of the study was obtained from the Institutional Review Committee of Kastamandap School of Public Affairs Management, Kathmandu, Nepal. Written consent was obtained prior to the collection of data. For illiterate patients, we read the informed consent form and obtained their verbal consent since the risks associated with our study were low and the potential harm for participants was unlikely.
Data collection
We collected data through face to face interviews using a patient satisfaction questionnaire (PSQ-III) developed by RAND Corporation [16]. Interviews were conducted in the Nepali language. Each interview took about 20-25 minutes to complete. The questionnaire consists of 18 items probing seven dimensions of patient satisfaction: general satisfaction, technical quality, interpersonal manner, communication, financial aspects, time spent with the doctor, and accessibility and convenience. Each question in the PSQ-III has a 5-point Likert Scale ranging from “Strongly Disagree,” “Disagree,” “Neutral,” “Agree” to “Strongly Agree.” After translating the questionnaire into the Nepali language, we pre-tested in 20 patients of Bhaktapur Hospital and made the necessary modifications.
Further, we also collected sociodemographic information of the participants, including age, gender, ethnicity, religion, marital status, occupational status, educational status, religion, the distance of health facility from their residents, and their enrollment into government health insurance. These data were used as predictors in the analyses.
Statistical analyses
Sociodemographic characteristics of the respondents were described in frequency and percentages. We calculated the mean and standard deviation of the Likert scale of each item of PSQ-III. Further, we calculated the frequencies and percentages of satisfied, neutral, and dissatisfied patients.
According to the guidelines of the patient satisfaction questionnaire (PSQ-III) [see Additional File 1], we classified the satisfaction in each item as follows:
- ‘Strongly agree’ or ‘Agree’ = Satisfied for Items: 1, 2, 3, 5, 6, 8, 11, 15 and 18
- ‘Strongly disagree’ or ‘Disagree’ = Satisfied for Items: 4, 7, 9, 10, 12, 13, 14, 16 and 17
- For all items, the score ranges from 1 (strongly dissatisfied) to 5 (strongly satisfied). The mean score for each item was calculated in the manner that higher the score more the satisfaction level for all the items in the PSQ-III.
To calculate the overall score in each domain, we averaged the score of designated items for each domain as guided by PSQ-III, which is as follows [see Additional File 1].
- General Satisfaction: Item 3 + 17
- Technical Quality: Item 2 + 4 + 6 + 14
- Interpersonal Manner: Item 10 + 11
- Communication: Item 1 + 13
- Financial Aspects: Item 5 + 7
- Time Spent with Doctor: Item 12 + 15
- Accessibility and Convenience: Item 8 + 9 + 16 + 18
After categorizing the satisfaction score of each domain into three ordinal variables: satisfied, neutral, and dissatisfied, we utilized ordinal logistic regression (OLR) to assess sociodemographic variables associated with satisfaction level. The variables that were significant in univariable analysis (p<0.05) were included in the multivariable OLR model. We reported adjusted odds ratios with a 95% confidence interval and a p-value. A p-value of less than 0.05 was considered significant.