2.1. Study area
The study was conducted in three selected hospitals in southwest of southern nation nationalities and people regional state (SNNPRS) of Ethiopia. The southwest part is the farthest and marginalized area of the region. There are four hospitals in the area: Tercha hospital, Mizan Aman hospital, Gebretsadik Shawo hospital, and Teppi hospital. Only the three hospitals were selected because there is no any study conducted about the nursing process in the hospitals, and Tercha hospital was not chosen because there was one study conducted about the nursing process and also the pretest was conducted there to prevent diffusion of information. Mizan Aman Hospital is located in the town of Mizan Aman, 568 km southwest of Addis Ababa, established in 1968 and providing medical, surgical, pediatric and gynecological services with 122 beds and 81 nurses. Gebretsadik Shawo hospital is located in the town of Bonga, 460 km south-west of Addis Ababa. It was founded in 2006, and had 77 nurses. Teppi hospital is located in Teppi town, 614 km southwest of Addis Abba, which was established in 2014 and has 32 nurses.
2.2. Study design and period
Hospital based cross-sectional quantitative descriptive study was conducted from March 10 to April 1, 2015. The source population was all nurses working at the three hospitals. Nurses in the three hospitals with a diploma / above level of education and served for six months or more were included.
2.3. Sample size
The sample size was calculated using a single population proportion formula, taking the margin of error (d) as 5 percent, the confidence level of 95 percent, α = 0.05, the non-response rate of 10 percent and the nursing process implementation proportion (P) as 52.1 percent(15).
Where n is the minimum sample size, z is the normal standard distribution (z=1,96) at a confidence level of 95 percent and α=0,05, p is the prevalence / population proportion and d is the tolerable margin of error.
A total of 140 nurses were included in the study after considering the formula for finite population correction and 10 percent non-response rate. The correction formula was used because the calculated sample size is greater than the total population and the total population is less than ten thousand.
2.4. Sampling procedure
The total sample size was allocated proportionally to the three hospitals based on the number of nurses working in each hospital i.e. 24 nurses for Teppi hospital, 60 nurses for Mizan Aman hospital and 56 nurses for Bonga hospital, and sampling frame was prepared for each hospital by receiving list of nurses from each hospital’s human resource department. Finally from the sampling frame 140 nurses were selected using simple random sampling after the purpose and procedure for data collection was clarified, and confidentiality and privacy were guaranteed.
2.5. Data collection techniques and tools
Data was collected using self-administered questionnaires that were adapted and modified from various studies (2, 15, and 16) and had three parts: sociodemographic issues, nursing process implementation, and factors affecting the nursing process (see additional file 1). To ensure the validity of tool, the questionnaire was presented to three nursing expert, checked it for double, confusing and leading questions, tool validity determined using content validity and also the instrument was pre-tested before final data collection. Cronbach's alpha cofficent for nursing process implementation scale was 0.73. The tools included open as well as close-ended questions prepared in English. One supervisor and three Bsc nurses collected the data at the workplace of the nurses after they received one day’s training .
2.6. Study variables
2.6.1. Dependent variable
Implementation of the nursing process
2.6.2. Independent variables
Sociodemographic (age, sex, educational level, religion, year of service, marital status, ethnicity, an institution where education award is obtained and monthly salary), organizational factors such as non-supportive hospital administration, resource scarcity, shortage of staff training, time shortage, monitoring and evaluation.
2.7. Operational definition
Implementation of nursing process: in this study nurses who responded yes to the question did you follow the steps of the nursing process during provision of care? Were considered as implemented
2.8. Data quality assurance
Before the actual data collection pre-test was done on 5 percent (seven) nurses working in Tercha hospital, which is not part of the main study, and some modification was made to the tool and procedure based on the pre-test findings. Collectors and supervisors were given training. The data was checked by the supervisor on each day of collection for completeness and consistency. Before entry in to the computer and again before analyzing; the principal investigator rechecked for missing values.
2.9. Data processing and analysis
The data was first checked for completeness and consistency then categorized coded and entered into EP info version 3.5.3 and exported to SPSS version 20 for analysis. Univariate analysis such as percentage and frequency distribution was used. Bivariate analysis was used; variables with P-value < 0.2 from the binary logistic regression analysis were moved to multivariable logistic regression analysis in order to control the effect of confounders and to identify predictors of nursing process implementation. Significance of association was tested using 95 per cent confidence level and p< 0.05.