Publications

The World Health Organization (WHO) estimated 660,000 new cervical cancer cases and 350,000 deaths globally in 2022, with over 90% occurring in low- and middle-income countries (LMICs). Global elimination targets call for vaccination of >90% of girls against human papillomavirus (HPV) by age 15 by 2030. While Botswana's national policy aligns with these targets, progress monitoring is constrained by fragmented, predominantly paper-based health information systems. We describe our experience verifying individual-level HPV vaccination status among adolescent girls and young women living with HIV (AGYWLWH) in Botswana. This activity was conducted to determine eligibility for the HOPE II trial, a randomized controlled study evaluating the efficacy of a single-dose nonavalent HPV vaccine. Verification followed the national rollout of school-based HPV vaccination in 2015, targeting girls aged 9-13 years. Participants provided pre-screening consent for review of vaccination records held in school health systems, Ministry of Health (MoH) registers, and participant-held vaccination cards where available. We sought HPV vaccination records for 80 potential participants across 80 schools and 68 clinics. Only 15 participants (18.8%) had verifiable records: 2 (2.5%) retained vaccination cards, 4 (5.0%) had documentation in school records, and 9 (11.3%) had entries in MoH registers. Most school records were incomplete or missing, and several sites reported loss or disposal of records due to relocation, fire damage, or storage constraints. These findings highlight critical gaps in HPV vaccination documentation. Strengthening electronic immunization registries, standardizing record transfer, and improving data integration could enhance monitoring and support equitable progress toward cervical cancer elimination.

BACKGROUND: In the era of widespread antiretroviral therapy (ART), the relationship between HIV and hypertensive disorders of pregnancy remains unclear. This study compares the prevalence, severity, and neonatal outcomes of pre-eclampsia in women living with HIV (WLWH) and women without HIV in Botswana, a country with high ART coverage.

SETTING: The study was conducted at Princess Marina Hospital, Botswana's largest public tertiary referral center, where national ART guidelines support universal treatment and high rates of viral suppression among pregnant WLWH.

METHODS: We analyzed data from 11,754 women delivering at Princess Marina Hospital (November 2021-December 2023), combining 2 prospective surveillance cohorts. Pre-eclampsia was defined using American College of Obstetrics (ACOG) criteria. ART use, CD4 counts, and viral load data were collected. Primary outcome was pre-eclampsia prevalence; secondary outcomes included disease severity and adverse neonatal outcomes.

RESULTS: Pre-eclampsia prevalence was similar among WLWH and women without HIV (8.4% vs. 8.6%). However, after adjusting for age, parity, and hypertension history, WLWH had a lower risk (adjuste risk ratio 0.84, 95% CI: 0.71 to 0.98). Neonatal outcomes among those with pre-eclampsia did not differ significantly by HIV status. Among WLWH, initiating ART preconception and achieving undetectable viral loads were associated with reduced pre-eclampsia risk.

CONCLUSIONS: In settings with high ART coverage, WLWH may have a lower risk of pre-eclampsia than HIV-negative peers. HIV status does not seem to affect neonatal outcomes when ART is effective. These findings support early ART initiation and sustained viral suppression to optimize maternal health.

BACKGROUND: Depression significantly impacts cancer patients globally, complicating treatment outcomes by negatively affecting quality-of-life (QOL), self-care and treatment adherence. However, little is known about the prevalence of depression and its effects on QOL among cancer patients in Botswana.

AIM: This study aimed to assess the prevalence of depression and its relationship with QOL among cancer patients at Princess Marina Hospital's oncology unit.

SETTING: Patients seen at the oncology unit of Princess Marina Hospital, Botswana's largest tertiary referral hospital.

METHODS: A cross-sectional study was conducted from February 2023 to August 2024. The sample size was 302. Socio-demographic data were collected using a researcher-designed questionnaire. Depression was assessed using the PHQ-9, with a score of 9 or higher indicating depression, and QOL was measured using the WHOQOL-BREF.

RESULTS: The prevalence of depression was 35.8% (95% confidence interval [CI]: 30.3%, 41.5%). Patients with depression had a higher mean pain score (6 vs. 4, p < 0.001) and poorer QOL across all domains: physical health (40.9 vs. 59.7, p < 0.001), psychological health (52.3 vs. 69.3, p < 0.001), social relationships (49.0 vs. 68.3, p < 0.001) and environmental health (48.1 vs. 58.8, p < 0.001). A new diagnosis of cancer was associated with an increased risk of depression (p = 0.03).

CONCLUSION: The prevalence of depression among cancer patients is high in Botswana, and it is associated with poorer QOL, highlighting a significant public health concern.

CONTRIBUTION: Treatment of depression in this group of patients, along with pain management, may be crucial in improving their QOL.

OBJECTIVE: To determine histopathologic outcomes one year after LEEP for CIN2/CIN3 in women living with HIV on ART in Botswana and assess whether a 12-month follow-up interval is appropriate.

METHODS: This was a prospective cohort study of women living with HIV who previously underwent LEEP for confirmed CIN2/CIN3. Participants underwent follow-up approximately 12 months later, either through Ministry of Health or a standardized sub-study protocol, with colposcopy or VIA and tissue sampling performed per clinical algorithm. The primary outcome was histopathologically confirmed CIN2+ at follow-up. Associations between clinical factors, demographics, and CIN2+ were evaluated; chi-square and two-sample T-test were used to compare data and risk ratios (RR) were calculated.

RESULTS: Of 127 eligible participants, 90 attended follow-up at a median of 16 months (IQR 13 - 20). Almost all were on ART with a mean CD4 count of 728 (SD ± 310) cells/mm3. Of 66 participants with histopathology results, 29% had CIN2+ and one participant had invasive cancer. Positive endocervical margins at baseline were associated with increased risk of CIN2+, while age, time on ART, and CD4 count were not. Follow-up interval of ≥24 months was associated with more than double the risk of CIN2+ at follow-up (crude RR 2.3, 95% CI 1.1-4.7).

CONCLUSION: In a cohort of women living with HIV on ART, 29% had persistent/recurrent CIN2+ at follow-up. Follow-up interval and endocervical margin status were the strongest predictors. These data support the WHO recommendation for a 12-month follow-up interval in women living with HIV who are well-controlled on ART.

OBJECTIVE: Cervical cancer is the leading cause of cancer deaths among women in Botswana, yet uptake of cervical screening remains suboptimal. Understanding knowledge and barriers to screening is essential to inform strategies to achieve cervical cancer elimination.

METHODS: We conducted high-risk human papillomavirus (hrHPV) testing and an accompanying cross-sectional survey at health clinics and in the community in South East District, Botswana (February 2022-July 2023). Eligible participants were women aged ≥25 years with an intact cervix and no prior diagnosis of cervical cancer. Interviewer-administered structured questionnaires assessed demographics, prior screening attendance, knowledge of cervical cancer, and perceived barriers to screening. Open-ended responses were thematically categorized. Logistic regression was used to examine factors associated with knowledge and screening barriers.

RESULTS: Of 3000 women screened, 2994 (99.8%) had known HIV status and were included. Median age was 43 years (interquartile range [IQR] 36-51), and 1500 (50.1%) were women with HIV (WWH). Overall, 67.2% (2005/2994) reported prior cervical cancer screening, with substantially higher uptake among WWH than women without HIV (80.0% vs 54.5%, P < 0.001). Knowledge of cervical cancer etiology was limited, with only 27.1% (788/2905) correctly identifying hrHPV as the cause. Younger age, HIV-positive status, higher education, formal employment, and prior screening attendance were independently associated with better knowledge. Among women who had never been screened, the most commonly reported barriers were lack of prioritization (38.3%; 376/981), fear (17.0%; 167/981), lack of awareness (17.0%; 167/981) and service provision issues (12.0%).

CONCLUSION: Despite relatively high screening uptake, important gaps in knowledge and persistent individual- and system-level barriers remain. Strengthening cervical screening delivery through HPV self-collection alongside targeted education, would address current barriers to cervical screening in Botswana.

OBJECTIVE: To evaluate the impact of antiretroviral therapy (ART) duration and CD4 + cell count on risk for high grade cervical dysplasia in women with HIV (WWH) compared to women without HIV in the treat-all era with integrase strand inhibitors (INSTIs).

DESIGN: A prospective longitudinal cohort study in Botswana.

METHODS: From February 2021 to July 2022, baseline HPV self-sampling was offered to women with and without HIV. Those HPV+ underwent biopsy for histopathological diagnosis. Using women without HIV as reference, risk ratios were calculated for HPV, cervical intraepithelial neoplasia (CIN) 2 or worse (CIN2+), and CIN3+, stratified by ART duration and CD4 + cell counts.

RESULTS: Of 3000 women enrolled, 2953(98.4%) underwent HPV testing, which was positive in 823(55.7%) WWH and in 654(44.3%) women without HIV. Histopathology was available for analysis in 1291(87.4%) women (709 WWH, 582 women without HIV). Over 99% of WWH had an undetectable HIV viral load and 94.4% were on a dolutegravir-based ART regimen. WWH had a higher risk of HPV [risk ratio 1.27, 95% confidence interval (95% CI): 1.18-1.37], CIN2+ (risk ratio 1.52, 95% CI: 1.16-1.98), and CIN3+ (risk ratio 1.75, 95% CI: 1.25-2.45) compared to women without HIV. There was attenuation of risk for CIN2+ with higher recent CD4 + cell count, and those with higher nadir CD4 + cell count had similar risk to those without HIV (nadir CD4 + ≥500 CIN2+ risk ratio 1.15 [95% CI: 0.56-2.37], CIN3+ risk ratio 1.81 [95% CI: 0.86-3.79]; nadir CD4 + 350-499 CIN2+ risk ratio 1.23 [95% CI: 0.71-2.12], CIN3+ risk ratio 1.34 [95% CI: 0.68-2.64]).

CONCLUSION: Although some attenuation of risk for CIN2+ was observed with higher recent and nadir CD4 + cell counts, WWH continue to have a higher risk of CIN2+/CIN3+ compared to women without HIV. These findings support tailored cervical screening algorithms for WWH.

OBJECTIVE: The aim of this study was to evaluate the performance of HPV type restriction and cycle threshold (Ct)-limit setting to optimize detection of cervical intraepithelial neoplasia (CIN) with primary HPV testing.

METHODS: Baseline cervical screening at time of entry into a prospective longitudinal cohort of women with and without HIV was conducted from February 2021 to July 2022 in Botswana. All women underwent HPV testing of 15 individual types using the AmpFire assay; all HPV-positive and a random subset of HPV negative had histopathology collected. Performance parameters of HPV type restriction groupings were calculated, and sensitivity by individual HPV type Ct-value limits were plotted.

RESULTS: Among 2964 women who underwent primary HPV screening, 1293 (43.6%) tested HPV-positive. Among women with HIV (WWH), HPV types 16/18/33 were associated with the greatest burden of CIN2+/CIN3+ (53%/56%). In WWH, grouping by HPV types separately reported in commercial assays (16/18/45) had low sensitivity (44% [CI: 36%-52%]) but high specificity (86% [CI: 84%-88%]) for CIN2+; 8-type HPV restriction (16/18/31/33/35/45/52/58) improved sensitivity (79% [CI: 72%-86%]) and maintained reasonable specificity (67% [CI: 65%-70%]) for CIN2+. Similar results were seen in women without HIV. Ct-limit setting for medium oncogenic HPV types (31,33,35,52,58) maintained a sensitivity of 72% in WWH while reducing over-detection of non-pathogenic HPV.

CONCLUSION: Eight-type HPV restriction and Ct-limit setting are promising strategies for improving the performance of primary HPV screening. A potential strategy to improve 8-type HPV restriction would be to treat all with HPV 16/18/45; treat HPV 31/33/35/52/58 if below the type-specific Ct limit and repeat HPV testing in 1-year for other positive HPV results.

PURPOSE: In Botswana, a see-and-treat approach to cervical cancer screening is taken. Our objective was to determine the number of loop electrosurgical excision procedures (LEEPs) performed for cervical intraepithelial neoplasia (CIN) 2/3 in Botswana, and follow-up rates and outcomes, among women with positive cervical margins.

METHODS: Data (patient age, HIV status, margin status, follow-up, and recurrence) from women who underwent LEEP with histologically confirmed CIN 2/3 between January 2014 and December 2015 were analyzed retrospectively. Histopathologic reports were reviewed at a central laboratory in Gaborone, Botswana. Follow-up and recurrence rates were summarized descriptively and compared according to HIV and margin statuses using chi-squared tests.

RESULTS: In total, 779 women (median age, 39.2 years) underwent LEEP showing CIN2/3; 638 (81.9%) women had CIN3 and 390 (50.1%) had positive LEEP margins (ectocervical, 186 [47.7%]; endocervical [including with ectocervical], 204 [52.4%]). Margin positivity was not associated with HIV status. Of women with positive endocervical margins followed at ≤1 and >1 year, 9.6% and 48.3%, respectively, had persistent CIN2/3 on repeat LEEP. Forty percent (90 of 204) of women with positive endocervical margins had no re-excision documented.

CONCLUSION: Most women who underwent LEEP had CIN3 and positive margins. Almost half with positive margins followed at >1 year after initial LEEP had CIN2/3 recurrence warranting further treatment; two thirds were not followed. Resources are needed to improve post-LEEP follow-up for women with margin positivity who require additional ablative/excisional procedures to reduce the cervical cancer burden in Botswana.

OBJECTIVES: To improve the quality of intrapartum care at Princess Marina Hospital, Gaborone, Botswana, through a champion-led educational intervention, introduction of the World Health Organization's Labor Care Guide (LCG), which highlights four key measures to promote respectful maternal care (RMC).

METHODS: In August 2022, we engaged medical and nursing staff in interactive training on the WHO LCG and implemented a mnemonic-based checklist-"COPE" (Companions, Oral fluids, Pain relief, Eliminate the supine position). Motivational posters were displayed in the labor ward. Surveys to assess client labor experiences were conducted in November 2022 (n = 204) and October 2024 (n = 211). An anonymous staff survey assessed personal practice changes before and after the training.

RESULTS: Over 2 years, significant improvements in reported maternal experience were observed. Reports of a "very bad" experience decreased from 116/204 (56.9%) to 13/211 (6.2%) and dissatisfaction with care reduced from 22/204 (10.8%) to 3/211 (1.4%). Adequate fluid provision increased from 108/204 (52.9%) to 159/211 (75.4%) and pain relief from 41/204 (20.1%) to 108/211 (51.2%). The reported presence of a labor companion increased from 14/204 (6.9%) to 133/211 (63.0%). Patient reports of exclusive supine positioning during the second stage of labor decreased from 106/204 (52.0%) to 11/211(5.2%). Staff questionnaires showed a consistent improvement in reported adoption of "COPE" practices compared with recollected practices before the intervention.

CONCLUSIONS: Deeply entrenched labor care practices are difficult to change; however, sustained, champion-led quality improvement interventions using the LCG and with sustained supportive supervision can yield meaningful improvement in women's birth experiences.

OBJECTIVE: This study aimed to evaluate maternal outcomes in a large cohort with high prevalence of human immunodeficiency virus (HIV) infection in Botswana after implementation of a treat-all policy.

METHODS: In this retrospective cohort study, data were collected from the medical record at the time of discharge from November 2021 to December 2023. Outcomes were recorded in the Tsepamo Birth Outcomes Surveillance and Safe Birth studies at Princess Marina Hospital in Botswana. We evaluated maternal mortality and obstetric morbidities by HIV status, including preeclampsia, eclampsia, hemorrhage, infection, and acute pulmonary or cardiac conditions at the time of hospital discharge.

RESULTS: We included 11,754 participants; 2,201 (18.7%) were pregnant people with HIV infection. Ninety-seven percent (2,135) were on antiretroviral therapy (ART) at time of delivery; 1,996 (93.5%) of those with a known ART regimen were on dolutegravir, tenofovir disoproxil fumarate, and lamivudine. Of the 1,090 people with HIV infection with known CD4 counts, 757 (69.4%) had more than 500 cells/microliter, and only 42 (3.9%) had fewer than 200 cells/microliter. Of 1,524 people with HIV infection with known viral loads, 1,436 (94.2%) were undetectable on initial testing. There were no statistically significant differences in incidence of hemorrhage (90 [4.1%] vs 370 [3.9%], adjusted risk ratio [RR] 0.93, 95% CI, 0.73-1.17), infection (38 [1.7%] vs 126 [1.3%], adjusted RR 1.56, 95% CI, 0.97-2.51), eclampsia (6 [0.3%] vs 28 [0.3%], adjusted RR 1.12, 95% CI, 0.50-2.53), acute pulmonary or cardiac conditions (15 [0.7%] vs 43 [0.4%], adjusted RR 1.22, 95% CI, 0.65-2.27), transfusion of 2 or more units of packed red blood cells (33 [36.7%] vs 110 [29.8%], P= .21), additional uterotonics (48 [53.3%] vs 173 [47.1%], P= .29), use of tranexamic acid (31 [ 34.4%] vs 106 [29.0%], P= .31), intensive care unit admission (4 [0.2%] vs 10 [0.1%], P= .31), mechanical ventilation (3 [0.1%] vs 6 [0.1%], P= .38), pressor support (2 [0.1%] vs 2 [0.0%], P= .16), or mortality (5 [0.2%] vs 11 [0.1%], adjusted RR 1.44, 95% CI, 0.46-4.57) in people with HIV infection compared with those without HIV infection. There were few notable differences, including a slightly reduced risk of preeclampsia (184 [8.4%] vs 818 [8.6%], adjusted RR 0.84, 95% CI, 0.71-0.98) and, although rare, an increased risk of uterine rupture (12 [0.5%] vs 8 [0.1%], adjusted RR 6.54, 95% CI, 2.33-18.33) in people with HIV infection compared with those without HIV infection.

CONCLUSION: There was little difference in adverse maternal obstetric outcomes between people with and those without HIV infection in the treat-all era with integrase strand inhibitors (primarily dolutegravir); notable exceptions included a slightly reduced risk of preeclampsia and, although rare, an increased risk of uterine rupture in those with HIV infection.