Study design and sample
We conducted a descriptive study involving the screening of all 90 incarcerated women at the Nsawam Medium Security Prison in Ghana on September 27-28, 2019 to assess the prevalence of high-risk HPV among them.
After obtaining written informed consent and psychological counseling, women underwent a structured interview based on a questionnaire we routinely use for our work at the Cervical Cancer Prevention and Training Centre in Catholic Hospital, Battor (additional file 1). Their responses were recorded directly into a Microsoft Excel spreadsheet database developed specifically for recording all data associated with cervical screening and treatment. Both the questionnaire and electronic data capture tool were developed at the CCPTC at the Catholic Hospital, Battor. These tools have been in use since 2017 for capturing all data on cervical screening and treatment conducted by the team from CCPTC.
To ensure the utmost privacy of clients, all clients were identified by unique alphanumeric identifiers and none of their personal identification data such as names and previous addresses were obtained during the screening.
Ethical considerations
Ethical clearance was obtained from the Ethical Review Committee of the Ghana Prison Service. All women whose images were used in the manuscript gave us consent to use the images.
HPV testing with the AmpFire HPV detection system
HPV testing was performed using the AmpFire HPV detection technology (Atila BioSystems, Mountain View, CA, USA). It is an isothermal polymerase chain reaction (PCR) assay that uses minimum instrumentation and detects15 high risk HPV (16, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 68) in a single tube reaction and simultaneously identifies specifically the presence of types 16 and 18. Dry brush samples were submitted immediately at room temperature to the laboratory for testing. Samples could be processed individually or batched (1 to 94 samples per run). DNA extraction was not required and sample processing to final results took less than two hours.
The AmpFire HPV screen kit contained the reaction mix (with buffer, enzymes, and dNTPs), primer mix (with primers and probes), positive control, and negative control. Twelve microliters of reaction mix was mixed with 11 µl of primer mix in a 0.2 ml reaction single tube, 8-well strip, or 96-well plate. One milliliter of lyse buffer was added to the dry sample brush in a 5ml empty tube and left to sit at room temperature for 20 minutes after it had been vortexed. After this, 2 µl processed samples were added to the reaction tube to bring the total volume to 25 µl. The reaction tubes were incubated at 60°C for 60 minutes in the Atila Power96 real-time PCR system (the PCR program is set with denaturation step at 60°C for 30 seconds, followed by extension step at 60˚C for 30 seconds while taking fluorescence reading. A total of 60 cycles were run with fluorescence recorded once per minute from the FAM/HEX/ROX/CY5 channels. The results of the Ct values for each amplification curve in all fluorescence channels were automatically reported by the thermocycling software system. For each sample, an exponential amplification curve in the CY5, ROX, FAM, and HEX channels indicated the presence of the DNA of HPV16, HPV18, non-16/18 HR-HPV types, and internal control, respectively. The lack of exponential amplification curve in the HEX channel was interpreted as an invalid result. The negative and positive controls were included in each assay to ensure the quality of the assay and avoid possible contamination.
Mobile colposcopy with the Enhanced Visual Assessment system
After passing a speculum and taking a sample with a dry brush for HPV DNA testing, each of the two trained nurses, working in different rooms, independently performed colposcopy with a mobile colposcope, the Enhanced Visual Assessment (EVA) system (MobileODT, Tel Aviv, Israel). The EVA System consists of a mobile colposcope built around a smartphone, and an online image portal for storing images. A smartphone app is used to control the mobile colposcope, and upload pictures to the image portal or to store images on the phone for review. The adequacy of colposcopy, the transformation zone (TZ) type, and any lesions found on the cervix or in the vagina were recorded by the nurses. The images were anonymized by assigning them special codes. A gynecologist subsequently reviewed the images.
Key definitions
Transformation zone types
Type 1 (TZ 1): The entire circumference of squamocolumnar junction visible; fully ectocervical.
Type 2 (TZ 2): The entire circumference of squamocolumnar junction visible; partly or fully endocervical.
Type 3 (TZ 3): The entire circumference of the squamocolumnar junction is not visible; partly or fully endocervical.
Categories of prisoners
Remand prisoner: A person confined in prison whose case is awaiting hearing or pronouncement of sentence in a lower court (community tribunal, magistrate court, or circuit court).
Life prisoner: A person sentenced to spend the whole of his or her life in prison custody by a court of competent jurisdiction.
Condemned prisoner: Any person who is found guilty by the court of justice and sentenced to death by hanging or firing squad by a court of competent jurisdiction.
Convict prisoner: A person who has committed a crime and has been tried and found guilty by a court of competent jurisdiction and has been sentenced to serve a specific term in prison.
Statistical analysis
Percentages were calculated for categorical data and means and standard deviations were used to describe continuous data. Skewed continuous data (such as period of incarceration) were summarized using medians and interquartile ranges. For our primary objective of estimating the prevalence of high-risk HPV among female inmates, we used simple percentages and reported 95% confidence intervals (CIs). For the secondary objectives, we explored the association between age of inmates and period of incarceration and high-risk HPV positivity using grouped box plots. In addition, we compared the period of incarceration between those with high risk HPV positivity and those without using a Wilcoxon rank-sum test. This was done because the period of incarceration was highly right skewed. A student’s t-test was used to compare the means of two populations in order to explore further the relationship between high risk HPV positivity and the age of inmates.
We reported 95% CIs for all key estimates obtained. All statistical analyses were performed using STATA version 15 (StataCorp LLC, College Station, Texas, USA). We considered a statistical significance level of 5% and reported all p-values for all our hypothesis tests.