Search

A number of challenges hinder artificial intelligence (AI) models from effective clinical translation. Foremost among these challenges is the lack of generalizability, which is defined as the ability of a model to perform well on datasets that have different characteristics from the training data. We recently investigated the development of an AI pipeline on digital images of the cervix, utilizing a multi-heterogeneous dataset of 9,462 women (17,013 images) and a multi-stage model selection and optimization approach, to generate a diagnostic classifier able to classify images of the cervix into "normal", "indeterminate" and "precancer/cancer" (denoted as "precancer+") categories. In this work, we investigate the performance of this multiclass classifier on external data not utilized in training and internal validation, to assess the generalizability of the classifier when moving to new settings. We assessed both the classification performance and repeatability of our classifier model across the two axes of heterogeneity present in our dataset: image capture device and geography, utilizing both out-of-the-box inference and retraining with external data. Our results demonstrate that device-level heterogeneity affects our model performance more than geography-level heterogeneity. Classification performance of our model is strong on images from a new geography without retraining, while incremental retraining with inclusion of images from a new device progressively improves classification performance on that device up to a point of saturation. Repeatability of our model is relatively unaffected by data heterogeneity and remains strong throughout. Our work supports the need for optimized retraining approaches that address data heterogeneity (e.g., when moving to a new device) to facilitate effective use of AI models in new settings.

Background: Women treated for cervical precancerous lesions have up to a 30 % increased risk of developing cervical cancer compared to women in the general population. The aim of this study was to identify predictors for adherence to follow-up among women treated for precancerous lesions of the cervix in Cameroon. Materials and Methods: The study design was a retrospective cohort analysis of a five-year follow-up for women in Cameroon who were initially treated for cervical precancer lesions in 2013. Logistic regression models were used to determine factors associated with adherence to post-treatment follow-up. Statistical significance was set at p

Cameroon is a low-and-middle income country (LMIC) with one of the highest incidence and mortality from cervical cancer in Africa. In this Central African country where the prevalence of human immunodeficiency virus (HIV) is high and the screening coverage is low, cervical cancer is the most deadly and the second most common cancer among women. Notwithstanding the growing burden of cervical cancer in Cameroon, most patients - often of lower socioeconomic status - continue to encounter multi-level barriers to timely and adequate care. These include the lack of physical and financial access to healthcare facilities, limited quality pathology, imaging and treatment services, ignorance of disease by the population, shortage of a well-trained oncology workfroce, which result in significant delays in gaining access to screening, diagnosis, treatment and care. This paper presents 3 cases of patients with advanced cervical cancer who had surgery (hysterectomy) as primary treatment, without appropriate post-surgical investigation to further specify disease stage, persistence of residual disease, and need for adjuvant chemoradiation. Pathology services and diagnostic imaging procedures remain scarce and underused in LMIC countries like Cameroon. Healthcare professionals involved in patient care lack adequate knowledge, skills and collaborative strategy to properly navigate these patients. To address these challenges, the health system should be reinforced with adequate infrastructures, sustainable funding should be secured to enhance universal health coverage and promote cancer prevention and control programs, multidisciplinary teams and coordination of care among providers should be improved, and relevant health indicators should be put in place to better monitor the quality of care delivered to patients who are mostly vulnerable and uninformed.

IntroductionEvidence suggests that advice from health care professionals (HCP) increases the likelihood of quit attempts and successful quitting of tobacco use. However, previous studies primarily focussed on cigarette smoking and did not include all forms of tobacco products. This study aimed to investigate the prevalence and disparities in receiving HCP’s advice to quit tobacco use (combustible or noncombustible) in the US adult population.MethodsUsing the 2022 National Health Information Survey (NHIS) data, we examined 4,424 adults who reported (i) any tobacco product use within the past 12 months and (ii) having seen an HCP within the past 12 months. The outcome variable included the receipt of advice to quit tobacco use from an HCP, and predictors included sociodemographic variables. Weighted prevalence estimates were calculated, and multivariable regression analyses were conducted.ResultsOver 38% of tobacco users who visited an HCP were advised to quit. The odds of receiving such advice were lower among Hispanics (AOR: 0.625; 95% confidence interval (CI) [0.464–0.843];p = 0.002), males (AOR: 0.767; 95% CI [0.659–0.893], p = 0.001), those above the poverty level (AOR: 0.795; 95% CI [0.641–0.987];p = 0.037), foreign-born (AOR: 0.664; 95% CI [0.496–0.888]; p = 0.006), those with a bachelor’s degree or higher educational level (AOR: 0.477; 95% CI [0.349–0.653]; p < 0.001) and those aged less than 45 years (AOR: 0.404; 95% CI: [0.344–0.473]; p

BackgroundCigarette smoking is the major preventable cause of premature deaths in the United States. Attempting to quit smoking is an important step toward smoking cessation. Although it has been studied extensively, limited information on the association between attempts to quit smoking and neighborhood air quality problems is available. Therefore, we examined the association between attempts to quit smoking in the past year and perceived neighborhood air quality problems among adult Texans who smoke.MethodsIn 2018, a cross-sectional multistage area probability design-based survey was administered to collect sociodemographic, behavioral, and health-related information from a representative sample of 2050 Texas residents. The current study included 486 adult respondents who reported smoking within the past 12 months. The association between attempts to quit smoking and perceived neighborhood air quality (measured by self-reported problems with neighborhood air quality) was examined using a population-weighted multivariable logistic regression analysis.ResultsOverall, 60.7% of the 486 respondents attempted to quit cigarette smoking. The prevalence of attempting to quit was 74.6% for those reporting perceived neighborhood air quality problems. In the multivariable analysis, a higher likelihood of attempting to quit smoking was found among individuals with perceived neighborhood air quality problems (AOR: 1.906 [1.104–3.289]) and those who were married or living as married (AOR: 1.876 [1.161–3.033]). The likelihood of attempts to quit smoking was lower among males (AOR: 0.629 [0.397–0.995]) and decreased with age (AOR: 0.968 [0.951–0.984]).DiscussionThe perceived neighborhood air quality problems were found to independently predict attempts to quit cigarette smoking in Texas. To encourage quitting smoking among individuals living in neighborhoods with poor air quality, such neighborhoods should receive tailored and evidence-based interventions to improve community education, social support, and healthcare professionals’ assistance to quit smoking.

Background: Cancer is a rapidly rising cause of morbidity and mortality in sub-Saharan Africa. Cervical cancer, in particular, is still one of the leading causes of mortality for women in this setting. The uptake of healthcare services is in part influenced by patients’ belief systems. We sought to better understand the perception of cancer in the Kom tribe of Northwest Cameroon. Methods: A qualitative research study was completed using a semi-structured interview guide and one-on-one interviews with 45 parents of girls aged 9–14 years. These girls were candidates for free HPV vaccination to prevent cervical cancer. The interviews were recorded, transcribed, and analyzed using ATLAS.ti 9. Results: Thirty-five mothers and ten fathers with a median age of 42 yo were interviewed from Mbingo, Belo, Njinikom, and Fundong. Half of the parents were farmers, with three being herbalists or traditional medicine doctors. Seventy-seven percent had either no or only primary school education. None had had cancer. All knew at least one person with cancer. The most common word for cancer in the Kom language is “ngoissu”. It can refer to a bad infection or cancer. The occurrence of ngoissu is the result of either a curse placed on you, ancestral retribution, or transgressing the ngoulatta (snail shell spoken over and usually placed in a garden). The implications are that treatment of ngoissu must involve the traditional doctor who determines the spiritual issue and prescribes a remedy (like a herb or tea) and/or an incantation. Within the context of cancer, this can lead to a delay in diagnosis until the disease is no longer curable by conventional therapies. Conclusion: Ways to bridge biomedical healthcare services and traditional medicine are needed, especially in tribal contexts where the latter is an integral part of daily life.

IntroductionAlthough the burden of cervical cancer in Africa is highest, HPV vaccination coverage remains alarmingly low in this region. Providers’ knowledge and recommendation are key drivers of HPV vaccination uptake. Yet, evidence about providers’ knowledge and recommendation practices about the HPV vaccine against a backdrop of emerging vaccine hesitancy fueled by the COVID-19 pandemic is lacking in Africa.MethodsA cross-sectional study was conducted in 2021–2022 among healthcare providers involved in cervical cancer prevention activities in Africa. They were invited to report prior training, the availability of the HPV vaccine in their practice, whether they recommended the HPV vaccine, and, if not, the reasons for not recommending it. Their knowledge about the HPV vaccine was assessed through self-reporting (perceived knowledge) and with three pre-tested knowledge questions (measured knowledge).ResultsOf the 153 providers from 23 African countries who responded to the survey (mean age: 38.5 years, SD: 10.1), 75 (54.0%) were female and 97 (63.4%) were based In countries with national HPV immunization programs. Overall, 57 (43.8%) reported having received prior training on HPV vaccine education/counseling, and 40 (37.4%) indicated that the HPV vaccine was available at the facility where they work. Most respondents (109, 83.2%) reported recommending the HPV vaccine in their practice. Vaccine unavailability (57.1%), lack of effective communication tools and informational material (28.6%), and need for adequate training (28.6%) were the most commonly reported reasons for not recommending the HPV vaccine. While 63 providers (52.9%) reported that their knowledge about HPV vaccination was adequate for their practice, only 9.9% responded correctly to the 3 knowledge questions.ConclusionTo increase HPV vaccination coverage and counter misinformation about this vaccine in Africa, adequate training of providers and culturally appropriate educational materials are needed to improve their knowledge of the HPV vaccine and to facilitate effective communication with their patients and the community.

Introduction Despite human papilloma virus (HPV) vaccine being incorporated in the National Immunisation Programme in Cameroon in 2019, HPV vaccine uptake among eligible girls has been poor (5%). The barriers to HPV vaccination in this high-burden setting have not been previously studied, especially in rural areas. We propose to evaluate the knowledge, attitudes and beliefs of parents of girls aged 9–14 years regarding HPV vaccine.Methods and analysis Study design: a qualitative descriptive study will be conducted using one-on-one semi-structured interviews with parents of girls aged 9–14 years from 3 health districts in Cameroon (Mbingo, Njinikom and Fundong) who can converse in English or Pidgin English. Enrolment will occur until thematic saturation—approximately 40 participants. Analysis: quantitative methods will be used to describe the interviewees. All interviews will be audio recorded, transcribed and loaded into a tool to facilitate analysis (ATLAS.ti). Transcripts will be coded and thematic analysis will be conducted. Analysis will occur concurrent with interviews.Ethics and dissemination Ethics: institutional review board approval will be obtained from the Cameroon Baptist Convention Health Board (CBCHB), Cameroon and McMaster University, Hamilton, Canada. Dissemination: study findings will be presented via a report and webinar to the Ministry of Health, the funders, the CBCHB and in person to healthcare providers and interested members of the general population in the study region. Plans are to share findings internationally through peer-reviewed publication(s) and presentation(s).Trial registration number NCT05325138.

Background: Implementation of onchocerciasis elimination programmes has been delayed in Central Africa because of the risk of ivermectin-related serious adverse events (SAEs) in individuals with high Loa loa microfilarial densities (MFD). We developed the first statistical models enabling prediction of SAE risk in individuals with a given MFD. Methods: We used individual participant data from two trials conducted in loiasis-onchocerciasis co-endemic areas in Cameroon. among the 10 506 ivermectin-treated subjects included in the analysis, 38 (0·36%) developed an ivermectin-related SAE. To predict individual-level risk of SAE, we developed mixed multivariate logistic models including subjects’ sex, age, pre-treatment L loa and Mansonella perstans MFDs, and study region. Findings: The models predicted that regardless of sex, about 1% of people with 20 000 L loa microfilariae per millilitre of blood (mf/mL), 10% of people with 50 000 mf/mL and about one third of those with 100 000 mf/mL will develop an SAE. For a given MFD, males have a three-fold higher risk of developing an SAE than females. Interpretation: By enabling the prediction of post-ivermectin SAE risk in communities with known distribution of L loa MFDs, our results can guide decisions on the choice of ivermectin-based treatment strategies. They also predict that 37 SAEs were prevented in 2015 by using a Test-and-Treat strategy in the Okola District of Cameroon. Funding: UNDP/World Bank/WHO Special Programme for Research and Training in Tropical Diseases; Institut de Recherche pour le Développement; Mectizan Donation Program; Bill & Melinda Gates Foundation.

Objective: We assessed the global distribution and academic, administrative and research outcomes of international fellows (IFs) trained in Canadian gynecologic oncology (GO) programs. Methods: A web-based survey was sent to IFs who completed GO training in Canada. Using the Web of science database, we identified the publication list, citation record and H-index of IFs and classified them according to their region of practice: high-income countries (HIC), middle income countries (MIC), and low-income countries (LIC). Results: From 1996 to 2020, 81 IFs from 23 countries were trained in English-speaking (62,9%) and French-speaking Canadian universities (37,1%). Most IFs came from HIC (87,6%) and none from LIC. Only 12 IFs (14,8%) are now practicing in Canada. Of the 55 IFs who completed the survey (response rate: 67,9%), the majority (58,2%) reported working in an academic hospital and 29,1% were holding an executive position in a national scholar organization. IFs participated in mentoring residents (96.4 %) and medical students (83,6%) and 36,3% initiated a GO fellowship program in their home country. 67,3% of IFs were involved in international research collaboration and 52,7% participated in international clinical trials. The mean number of publications (22,36 vs 7,75, p = 0.007), citations (369,15 vs 45,12 p = 0.0006) and H-Index (6,88 vs 2,37 p = 0.0001) were significantly higher among IFs working in HIC compared to those in MIC. Most IFs (98,2%) recommended their Canadian GO fellowship program to a colleague from their home country. Conclusion: Most IFs trained in Canadian GO fellowship programs returned to their home countries and achieved important milestones in terms of academic, clinical and research accomplishments.

Background: In a context where specialist physicians are lacking and usually inaccessible, trained nurses can provide affordable cervical cancer screening and management services to women. Considering the rapid development of emerging cervical cancer prevention, diagnostic, and treatment procedures, nurses need to update their skills and knowledge regularly in the most appropriate way possible. Furthermore, the management of difficult cases in these settings is challenging, since many patients referred for specialty care cannot afford those services and must continue their care with local health providers. In this study, we applied and reviewed the telementoring scheme ECHO (Extension for Community Health Outcomes) designed to provide support for the Women's Health Program of the Cameroon Baptist Convention Health Services, a programme run by trained nurses that provides lower-cost cervical cancer screening and treatment to women in Cameroon. Methods: Since June 2018, the Women's Health Program of the Cameroon Baptist Convention Health Services has benefited from the expertise of specialist consultants in the USA, Canada, Europe, and Africa through monthly telementoring meetings using the ECHO model to review clinical cases and provide tailored didactic lectures. During each ECHO teleconference, clinicians from Cameroon and other African countries present two clinical cases to experts who provide recommendations for management, and experts give a didactic lecture on a topic relevant to providers, including cervical cancer prevention. In December, 2018, we administered a survey to Women's Health Program providers who had attended at least one ECHO session to collect data on baseline characteristics, previous training, and programme satisfaction. Findings: Between June, 2018, and August, 2019, 28 clinical cases were reviewed and 15 lectures presented. The number of participants in these ECHO sessions has increased from 5–10 to about 40–45. 16 nurses (median age 32 years [range 25–55 years]) completed the satisfaction survey; 15 were women (94%). Most participants had previously been trained to perform cervical cancer counselling (15 [94%]), to screen for cervical cancer with visual methods or to collect specimens for HPV testing (15 [94%]), to perform cervical biopsies (12 [75%]), and to treat with cryotherapy or thermal coagulation (13 [82%]). All respondents enjoyed the opportunity of learning with peers, 10 (63%) reported that ECHO sessions increased their ability to access specialty care, 14 (88%) reported that these sessions have improved the coordination of care, and 12 (75·0%) that these sessions have improved the quality of patients care. Most clinicians had adopted best-practice care following these sessions (12 [75%]), or developed additional clinical skills (10 [63%]). Interpretation: Within the Cameroon Baptist Convention Health Services Women's Health Program, results from Project ECHO have shown improved clinical skills and knowledge for providers, and enhanced quality of care for patients. This simple and inexpensive tool has the potential to foster quality improvement of cervical cancer prevention programmes in resource-limited countries. After the success of the programme in Cameroon, ten other African countries have joined the ECHO sessions. Funding: None