Cancer diagnostic service use in people living with HIV in South Africa: A cross-sectional study.

OBJECTIVE: The objective of this study was to map place of cancer diagnosis in relation to Human Immunodeficiency Virus (HIV) care centre among people living with HIV (PLHIV) within South Africa (SA) using national laboratory database. DESIGN: We linked HIV and cancer laboratory data from 2004-2014 using supervised machine-learning algorithms. We performed a cross-sectional analysis comparing province where individuals accessed their HIV care versus where they had their cancer diagnosis. SETTING: We used laboratory test records related to HIV diagnostics and care, such as CD4 cell counts and percentages, rapid tests, qualitative Polymerase Chain Reaction (PCR), antibody and antigen tests for HIV data that was documented as HIV positive and laboratory diagnosed cancer records from SA. STUDY POPULATION: Our study population consisted of HIV records from the National Health Laboratory Service (NHLS) that linked to cancer record at the National Cancer Registry (NCR) between 2004-2014. PRIMARY AND SECONDARY OUTCOMES: We linked HIV records from NHLS to cancer records at NCR in order to study the inherent characteristics of the population with both HIV and cancer. RESULTS: The study population was 68,284 individuals with cancer and documented HIV related laboratory test. The median age at cancer diagnosis was 40 [IQR, 33-48] years for the study population with most cancers in PLHIV diagnosed in females 70.9% [n = 46,313]. Of all the PLHIV and cancer, 25% (n = 16,364 p < 0.001) sought treatment outside their province of residence with 60.7% (n = 10,235) travelling to Gauteng. KZN had 46.6% (n = 4,107) of its PLHIV getting cancer diagnosis in Gauteng. Western Cape had 95% (n = 6,200) of PLHIV getting cancer diagnosis within the province. CONCLUSIONS: Our results showed health systems inequalities across provinces in SA with respect to cancer diagnosis. KZN for example had nearly half of the PLHIV getting cancer diagnosis outside the province while Western Cape is able to offer cancer diagnostic services to most of the PLHIV in the province. Gauteng is getting over burdened with referral for cancer diagnosis from other provinces. More effort is required to ensure equitable access to cancer diagnostic services within the country.

Cancer diagnostic service use in people with HIV in South Africa: a cross-sectional study.

Objective: The objective of this study was to map place of cancer diagnosis in relation to Human Immunodeficiency Virus (HIV) care centre among people with HIV (PWH) within South Africa (SA) using national laboratory database. Design: We linked HIV and cancer laboratory data from 2004-2014 using supervised machine-learning algorithms. We performed a cross-sectional analysis comparing province where individuals accessed their HIV care versus where they had their cancer diagnosis. Setting: We used laboratory test records related to HIV diagnostics and care, such as CD4 cell counts and percentages, rapid tests, qualitative Polymerase Chain Reaction (PCR), antibody and antigen tests for HIV data that was documented as HIV positive and laboratory diagnosed cancer records from SA. Study population: Our study population consisted of HIV records from the National Health Laboratory Service (NHLS) that linked to cancer record at the National Cancer Registry (NCR) between 2004- 2014. Primary and secondary outcomes: We linked HIV records from NHLS to cancer records at NCR in order to study the inherent characteristics of the population with both HIV and cancer. Results: The study population was 68,284 individuals with cancer and documented HIV related laboratory test. The median age at cancer diagnosis was 40 [IQR, 33-48] years for the study population with most cancers in PWH diagnosed in females 70.9% [n=46,313]. Of all the PWH and cancer, 25% (n=16,364 p < 0.001) sought treatment outside their province of residence with 60.7% (n=10,235) travelling to Gauteng. KZN had 46.6% (n=4,107) of its PWH getting cancer diagnosis in Gauteng. Western Cape had 95% (n=6,200) of PWH getting cancer diagnosis within the province. Conclusions: Our results showed health systems inequalities across provinces in South Africa with respect to cancer diagnosis. KZN for example had nearly half of the PWH getting cancer diagnosis outside the province while Western Cape is able to offer cancer diagnostic services to most of the PWH in the province. Gauteng is getting over burdened with referral for cancer diagnosis from other provinces. More effort is required to ensure equitable access to cancer diagnostic services within the country.

Rule-Based Information Extraction from Free-Text Pathology Reports Reveals Trends in South African Female Breast Cancer Molecular Subtypes and Ki67 Expression.

Clinical information on molecular subtypes and the Ki67 index is critical for breast cancer (BC) prognosis and personalised treatment plan. Extracting such information into structured data is essential for research, auditing, and cancer incidence reporting and underpins the potential for automated decision support. Herewith, we developed a rule-based natural language processing algorithm that retrieved and extracted important BC parameters from free-text pathology reports towards exploring molecular subtypes and Ki67-proliferation trends. We considered malignant BC pathology reports with different free-text narrative attributes from the South African National Health Laboratory Service. The reports were preprocessed and parsed through the algorithm. Parameters extracted by the algorithm were validated against manually extracted parameters. For all parameters extracted, we obtained accurate annotations of 83-100%, 93-100%, 91-100%, and 92-100% precision, recall, F 1-score, and kappa, respectively. There was a significant trend in the proportion of each molecular subtype by patient age, histologic type, grade, Ki67, and race. The findings also showed significant association in the Ki67 trend with hormone receptors, human epidermal growth factors, age, grade, and race. Our approach bridges the gap between data availability and actionable knowledge and provides a framework that could be adapted and reused in other cancers and beyond cancer studies. Information extracted from these reports showed interesting trends that may be exploited for BC screening and treatment resources in South Africa. Finally, this study strongly encourages the implementation of a synoptic style pathology report in South Africa.

Corrigendum: Predicting Colorectal Cancer Recurrence and Patient Survival Using Supervised Machine Learning Approach: A South African Population-Based Study.

[This corrects the article DOI: 10.3389/fpubh.2021.694306.].

Use of Machine Learning and Statistical Algorithms to Predict Hospital Length of Stay Following Colorectal Cancer Resection: A South African Pilot Study.

The aim of this pilot study was to develop logistic regression (LR) and support vector machine (SVM) models that differentiate low from high risk for prolonged hospital length of stay (LOS) in a South African cohort of 383 colorectal cancer patients who underwent surgical resection with curative intent. Additionally, the impact of 10-fold cross-validation (CV), Monte Carlo CV, and bootstrap internal validation methods on the performance of the two models was evaluated. The median LOS was 9 days, and prolonged LOS was defined as greater than 9 days post-operation. Preoperative factors associated with prolonged LOS were a prior history of hypertension and an Eastern Cooperative Oncology Group score between 2 and 4. Postoperative factors related to prolonged LOS were the need for a stoma as part of the surgical procedure and the development of post-surgical complications. The risk of prolonged LOS was higher in male patients and in any patient with lower preoperative hemoglobin. The highest area under the receiving operating characteristics (AU-ROC) was achieved using LR of 0.823 (CI = 0.798-0.849) and SVM of 0.821 (CI = 0.776-0.825), with each model using the Monte Carlo CV method for internal validation. However, bootstrapping resulted in models with slightly lower variability. We found no significant difference between the models across the three internal validation methods. The LR and SVM algorithms used in this study required incorporating important features for optimal hospital LOS predictions. The factors identified in this study, especially postoperative complications, can be employed as a simple and quick test clinicians may flag a patient at risk of prolonged LOS.

Improving long-lasting insecticidal nets use in Kayange community of north-western Burundi: a pilot study exploring the use of quality improvement methodologies in low-resource community settings.

Quality improvement (QI) approaches have demonstrated a lot of promise in improving clinical care processes, both in high-resource and low-resource settings. However, most examples of QI initiatives in healthcare in low-income countries are clinic-based. The objective of this study was to demonstrate feasibility of applying QI methods in low-resource community settings by applying them to the problem of correct utilisation of long-lasting insecticidal nets (LLINs) in a rural community in Burundi. Correct utilisation of LLINs had been shown to be a cost-effective approach to malaria prevention. In Burundi, LLINs utilisation is low. The Model for Improvement, a well-known QI approach, was used to increase LLINs utilisation in a rural community in Burundi. In the baseline, LLINs ownership and weekly utilisation together with factors affecting LLINs non-use were documented for a period of 4 weeks before intervention. Improvement ideas were collaboratively developed by a quality improvement team (QIT) and tested using Plan-Do-Study-Act (PDSA) cycles. The first PDSA cycle consisted of the demonstration of how to mount LLINs, the second was an implementation of reminders done by household 'watchdogs', the third cycle consisted of conducting two community reminders every week and the last cycle was a combination of the last two PDSA cycles. The intervention lasted 4 weeks and data were collected weekly. LLINs utilisation was calculated each week and plotted on a run chart to demonstrate improvement trends. LLINs utilisation data were collected for another 3 weeks postintervention. Of 96 households, 83 (87%) households owned at least one LLIN. After intervention, the number of LLINs used increased from 32% to 75% (134% increase) and the number of persons (general population) sleeping under LLINs from 35% to 73% (108% increase). The number of children under 5 years sleeping under LLINs increased from 31% to 76% (145% increase) and the number of pregnant women who slept under LLINs from 43% to 73% (69% increase). Also, the averages of the number of nights in each week that the general population slept under LLINs increased from 2.13 to 5.11 (140% increase), children under 5 years from 1.68 to 4.78 (184% increase) and pregnant women from 1.56 to 4.47 (186% increase). Each of the 4 PDSA cycles led to a significant increase in outcome indicators and the trends appear to persist even after the implementation was complete. While it is impossible to draw generalisable conclusions from a small pilot study, QI approaches appear to be feasible to implement in low-resource community setting and have promise in producing results. More research at larger scale should be encouraged to validate our initial findings.