Study Area and Period
The study was conducted at Addis Ababa St. Paul’s and Zewditu Hospitals, from April 20, 2015 - May 10, 2015. Addis Ababa is the capital city of Ethiopia and it has 10 sub-cities. At the time of this study, there were a total of 13 public hospitals and 80 functional public health centers. The two hospitals were among the facilities selected in Ethiopia to provide Cervical Cancer screening service for HIV positive women two years before the study period.
Study design
Healthcare institution-based cross-sectional study was conducted.
Source and Study population
Source population: All women of reproductive age group (15-49 years) visiting Zewditu and St. Paul’s Hospitals for anti-retroviral therapy (ART) follow up and Pre-ART.
Study population: Those who were selected by using systematic random sampling technique.
Sample size
Sample size was determined using a single population proportion formula which is stated as, n=Z2p(1-p)/d2 (where ‘n’ is the desired sample size, ‘Z’ is the value of standard normal variable at 95% confidence interval, ‘p’ is the proportion of cervical cancer screening service uptake: which was assumed to be fifty percent and 'd’ is the marginal error which was assumed to be 5%). With adding a 10% non-response rate, the last sample size was calculated to be 424.
Sampling procedure
The total numbers of clients who were eligible for the test were estimated using the last quarter report of the two hospitals. Afterward, the samples were allocated proportionally to the estimated size of clients for the study period of each hospital. Accordingly, 201 samples from St. Paul’s and 223 from Zewditu Hospitals were taken. Finally, using a systematic random sampling technique, a sampling fraction of every 12th eligible clients from Zewditu and 11th from St. Paul’s Hospitals were included in the study.
Inclusion and Exclusion criteria
Inclusion criteria: HIV positive women in the reproductive age group enrolled for ART/Pre-ART follow up at Zewditu and St. Paul’s Hospitals during the data collection period were nominees of the study.
Exclusion criteria: Those who were newly enrolled during the data collection time were left out.
Variables
Dependent variable: Cervical cancer screening status.
Independent variables: Age, religion, marital status, occupation, income, level of education, awareness about cervical cancer, awareness about cervical cancer screening and duration of HIV infection.
Data collection procedures
Data was collected using a pretested and structured questionnaire through a face to face interview.
Data quality control
The questionnaire was first developed in English and translated to Amharic and it was again translated back to English. The review was made by different individuals to control the consistency of the translation. The tool was pre-tested on 5% of the sample size at Black Lion Hospital to see the accuracy of responses, the clarity of language and the appropriateness of the tool. Four data collectors and one supervisor were trained for a day about objectives of the study, data collection methods, and ethical considerations. Completeness and logical consistency of the collected data was checked by the investigators at the data collection site. Data coding, entry, and cleaning was performed by the investigators.
Data management and Data analysis
The data was cleaned, coded and entered using Epi Info version 7. Then, it was exported to SPSS version 21 for analysis. Frequencies and summary statistics were used to describe the study population with relevant variables and outlines. Bivariate logistic regression model for first events was used to study the association between cervical cancer screening status and other independent variables. Then, all variables with a liberal p-value of less than 0.20 in the bivariate logistic regression analysis were taken into the multivariable logistic regression model to identify the independent effect of different factors on cervical cancer screening status. A p-value less than 0.05 and adjusted odds ratio with a 95% confidence interval were used to determine the presence and degree of association between dependent and independent variables.