Study design, setting and location
This study adopted a cross-sectional analytical study design in data collection, analysis and presentation. The design was appropriate for this study to test for associations between demographic characteristics and satisfaction [16]. The study was conducted at a teaching hospital with a 1655 bed capacity and offers various specialist services to the public and is the primary trauma center in Lusaka [17, 18]. It was carried out from June to October 2016. The bed occupancy rate at the teaching hospital is about 85%, while the average length of stay in hospital for patients ranges from 5 to 6 days. Some wards at the hospital are designated low cost, while others are high cost [19]. Patients in low cost wards receive free medical services compared to those in high cost wards who pay for health services [19]. Low cost wards are subsidized by the government [19]. Ethical approval for this study was obtained from Eres Converge in Zambia.
Study population
The study targeted adult orthopaedic patients admitted in low cost surgical wards on standard hospital diet. Eligible to participate in this study were adult surgical orthopaedic inpatients, 18–64 years, able to communicate and on standard hospital diet ≥3 days. This period of time is long enough for patients to form an opinion about hospital foodservices. The study’s exclusion criteria were paediatrics, cognitively impaired patients, patients with critical conditions and those who declined to participate.
According to the food service staff at the teaching hospital, orthopaedic patients like other patients in low cost wards are served three main meals comprised of breakfast, lunch and supper. The hospital does not provide any snacks to orthopaedic patients. With regards the food service system used to serve patients in low cost wards, the hospital uses a centralized system whereby food is prepared in the main kitchen by cooks. Once the food is ready, it is held hot in covered buckets and using trolleys, delivered fresh, plated on the ward and served as soon as possible by waiters. However, the food service system in high cost wards is different in the sense that food is prepared on demand for patients who can pay for food services. Furthermore, patients in high cost wards can choose food from a paper menu provided and food is plated right in the kitchen prior to distribution by waiters. In low cost wards, patients are offered soya porridge at breakfast and beans is served almost on a daily basis for lunch or supper. Beef is the only animal product that is provided by the hospital and is supplied irregularly because the hospital lacks a fixed menu. Soya pieces are served fortnightly to orthopaedic patients and that the only vegetable supplied by the hospital is cabbage and is served once or twice per week. Nshima (cooked maize meal cereal) is the only cereal that is served to patients at lunch and supper. The hospital does not offer any fruits to orthopaedic patients.
Sample size and sampling procedure
The teaching hospital was purposively sampled out of the four main referral hospitals in the country because it is the primary trauma center with a relatively higher number of adult surgical orthopaedic patients [18]. Low cost surgical wards (male and female) were also purposively sampled because that is where orthopaedic patients who eat regular hospital meals are admitted. Participants were drawn from all low cost (three male and two female) surgical wards. The nurses-in-charge of each of the selected low cost surgical wards were requested for the admission records from which a sampling frame was generated [20]. Comprehensive sampling was used to include all participants who satisfied the inclusion criteria and consented to participate in the study [21, 22].
Research instruments
The researcher-administered questionnaire and data collection procedures were pre-tested at a similar referral hospital in Zambia (Additional file 1). The pre-test was done on a selected sample of 10 adult surgical orthopaedic patients with conditions similar to those of the main study. The pre-test participants were not included in the main study. The procedures employed in pre-testing the instruments were identical to those used in the main study. Validity was ensured by use of already validated tool from a similar study [6]. The test-retest method was used to determine the reliability of the instruments during pretest. Data was collected twice during the pretest at an interval of 2 days from 10 participants. A test re-test correlation coefficient of 0.72 (CI: 0.61–0.82) was computed from the two sets of data and found to be adequate [23]. In order to avoid bias in patient behaviour, attending doctors, nurses and support staff were not informed about the study except in unavoidable situations [24].
Recruitment and training of two research assistants
Recruitment of research assistants was done competitively on the basis of: possession of a diploma in nutrition, conversant with English and two other languages widely spoken in Lusaka (Bemba and Nyanja), prior experience in research and a resident of Lusaka. Successful candidates were trained by the researcher for four days with emphasis on the following thematic areas; research ethics, study purpose and objectives, interviewing techniques, and correct recording of data in the questionnaires. The training was conducted through role plays, lectures, demonstrations and a pre-test.
Data collection procedures
The instrument used in this study had 9 dimensions of diet satisfaction. They were type, portion size, variety, taste, appearance, time of food distribution, temperature, overall quality, and attitude of staff serving hospital food. Patients were requested to indicate their level of satisfaction by selecting responses on a five-point Likert rating scale. The points associated with each scale were as follows; 1 = very dissatisfied, 2 = dissatisfied, 3 = fairly satisfied, 4 = satisfied, 5 = very satisfied. Patients who choose “very dissatisfied” and “dissatisfied” were considered dissatisfied, while those who selected “fairly satisfied”, “satisfied” and “very satisfied” were regarded satisfied. Satisfied participants were associated with a mean score of 3 and above, while dissatisfied patients had mean scores of 1 and 2. The instrument also contained questions on socio-demographic characteristics of the patients. The questionnaire was administered for 30–35 min.
Statistical analyses
Statistical Package for Social Sciences (SPSS) version 21.0 was used to analyze data. Descriptive statistics in terms of means, frequencies, percentages, and standard deviations were generated. Chi-square test was used to determine associations between variables such as age categories, length of stay in hospital, education level, income categories and satisfaction. Test of quantitative variables for normal distribution was done using the Kolmogorov-Smirnoff test. A p-value of less than 0.05 was considered statistically significant.