Abstract
Background
The COVID-19 pandemic underscored the relationship between infectious diseases and non-communicable diseases (NCDs), highlighting increased risks for people with pre-existing chronic conditions. This study piloted an integrated care model for NCD and infectious disease screening in healthcare settings. The aim of this study is to show that integrating NCD screening into existing HIV clinics and COVID-19 vaccine delivery clinics leads to better pandemic preparedness.
Methods
From October 2023 to December 2023, we conducted a cross sectional prospective, low interventional study involving 461 participants from two health care clinics in Johannesburg, South Africa using convenience sampling. Screening included blood pressure, body mass index (BMI), HbA1c, lipid profile, glucose, and kidney function assessments. Data were analyzed to identify NCD prevalence and key risk factors.
Results
Newly diagnosed NCDs were found in 73.9% of participants. The most common were kidney disease (31.7%), obesity (28.9%), and dyslipidaemia (25.8%). Modifiable cardiovascular disease (CVD) risk factors, including obesity (63.9%) and alcohol use (50%), were widespread.
Conclusion
Integrating NCD screening into HIV and COVID-19 services is achievable, offers a valuable opportunity for early detection and intervention. Scaling such models could strengthen health systems in resource-limited settings.
Keywords: health care, HIV, integrated care, NCDS, South Africa
Introduction
The COVID-19 pandemic brought renewed attention to the often-overlooked link between infectious diseases and NCDs (1). Individuals with NCDs faced higher risks of severe outcomes and mortality from SARS-CoV-2, underscoring the biological and health system vulnerabilities created by multimorbidity (2–4). These risks were most acute in low- and middle-income countries (LMICs), where weak primary care systems and siloed disease programmes failed to meet the needs of patients with complex chronic conditions (1, 5, 6).
Globally, NCDs account for 41 million deaths annually, with more than 80% occurring in LMICs (7). In South Africa, NCDs now account for over half of all deaths (57.4%), with CVD, T2D, and chronic respiratory conditions leading the burden (8). Despite global momentum behind cost-effective “best buy” interventions to help countries meet NCD targets under Sustainable Development Goal (SDG.) 3.4, implementation has been uneven (9). Progress has been made, particularly in taxation of tobacco and alcohol, however, efforts to improve diets through salt and sugar intake reduction or elimination of trans-fats remain slow. As a result, current efforts fall short of achieving SDG goal of reducing premature deaths by 33% by 2030 (9).
South Africa has a quadruple burden of disease namely communicable diseases such as HIV and tuberculosis (TB), maternal and child morbidity mortality, violent crimes as well as a rising NCD prevalence (10). Chronic kidney disease (CKD), often linked to unmanaged hypertension (HPT) and diabetes, represents a growing but under-recognised threat.
The pandemic not only amplified clinical risks for individuals with NCDs but also disrupted essential health services. World Health Organization’s (2) national pulse survey reported that 90% of countries experienced interruptions to NCD care, with the largest gaps in HPT, diabetes, and mental health services. In South Africa, excess mortality during the pandemic was disproportionately driven by people with poorly managed chronic diseases. These disruptions exposed structural weaknesses in health systems but also created new opportunities for innovation.
Integrated service delivery models, leveraging existing platforms such as HIV and COVID-19 programmes, offer a pragmatic and cost-efficient pathway to strengthen chronic care. Evidence from trials such as INTE-AFRICA and recent work by the NCD Alliance demonstrate that integrated care not only improves health outcomes but also reduces patient costs and improves access (10–13). However, real-world implementation remains limited, and few studies have examined feasibility within urban primary care settings in southern Africa.
In this pilot study, we evaluate the implementation of an integrated NCD screening model within two healthcare clinics in Johannesburg, South Africa. Building on COVID-19 and HIV infrastructure we aimed to identify the prevalence of HPT, obesity, T2D, dyslipidaemia, and CKD, while also assessing cardiovascular risk profiles and training primary healthcare staff in referral protocols. Our findings aim to inform scalable models for integrated chronic care in urban African settings.
The overall objective was to leverage COVID-19 and HIV pandemic response infrastructure for screening and monitoring of NCDs using NCD risk assessment tools.
The primary objective of the study was to assess the prevalence, incidence, and impact of CVD risk factors in HPT, obesity, T2D, dyslipidemia patients receiving COVID-19 vaccines or ART.
The secondary objective was:
Integrated care: training and supervision improvement of nurses to establish a primary health care (PHC) model of screening, and management, referral, and follow-up for HPT, T2D, dyslipidemia patients receiving COVID-19 vaccines or ART. Thus, facilitating integration of NCDs and HIV management.
Methods
Study design and sites
This cross sectional prospective, low interventional, pilot feasibility study was conducted, from October 2023 to December 2023 at two health care clinics in Johannesburg, South Africa. The first site, Riverpark Clinic in Alexandra Township is a large site, that provides nurse-led chronic disease management, COVID-19 vaccination and other primary care services. The second site, Ezintsha, Clinical Trials Unit, is a research unit of the University of the Witwatersrand that focuses on HIV care and clinical research, with advanced diagnostic capabilities for urban populations. Both sites provide services to underserved populations, living in urban, crowded and poorer communities of Johannesburg. The study was approved by Human Resource Ethics Committee.
Participants
Adults aged 18 and older attending either clinic for antiretroviral therapy (ART) or COVID-19 vaccinations were eligible. Informed consent was obtained from all participants. Participants were approached while queuing and invited to join the study without losing their place in line. Convenience sampling was used. As a pilot feasibility study, a formal prior sample size or power calculation was not performed. The sample size was determined by the duration of the study period and practical clinic constraint. Informed consent was obtained before any procedures. Participants were excluded if they were unable or unwilling to provide consent, or unable to complete all study procedures. Additionally, individuals judged by investigators to be at significant risk of protocol noncompliance were excluded.
Measures
We used standardised protocols to collect anthropometric and clinical data. These included height, weight, and body mass index (BMI), calculated as weight (kg) divided by height squared (m2). Blood pressure and pulse were measured to screen for HPT, defined according to South African HPT Society guidelines. Point-of-care (POC) tests assessed HbA1c, total cholesterol, HDL, random plasma glucose, serum creatinine, estimated glomerular filtration (eGFR) and urine microalbumin levels. Definitions and clinical thresholds are detailed in Appendix 1.
Point-of-care testing equipment
POC testing was conducted using the Cobas b 101 system (Roche Diagnostics), a compact, portable analyser suitable for multi-analyte testing from capillary blood samples. This device allowed rapid results for HbA1c, lipids, and glucose making it ideal for resource-limited settings. Urine microalbumin was measured using the Clinitek Microalbumin Reagent Strips with a handheld analyser.
Statistical analysis
Data from 460 participants with complete records were exported from REDCap and analysed using SPSS Version 29. Bivariate analyses were conducted to identify associations at a 5% significance level. Primary outcomes included the prevalence of NCDs (HPT, T2D, obesity, dyslipidaemia, and kidney impairment). Cardiovascular risk was estimated using the Framingham risk score. Kidney disease classification followed KDIGO CKD 2024 guidelines. Variables with p-values ≤ 0.05 were included in logistic regression models to assess predictors of kidney impairment.
Sociodemographic and clinical characteristics
A total of 460 participants had complete information and were included in the analysis. Table 1 shows the male to female comparison of participants’ sociodemographic and clinical characteristics. Most of the study participants were female (74.8%), black/African (99.1%) and 50.5% were unemployed. The mean BMI was significantly higher at 34.56 ± 7.67 among the female compared with 27.02 ± 5.92 among the male participants (p < 0.001). There was also significant male to female differences in the family history of NCDs and substance use among the study participants. The mean systolic blood pressure was 135.64 ± 21.38 mmHg while the mean diastolic blood pressure was 85.01 ± 13.45 mmHg. About one-third (31.3%) of the participants had blood pressure measurements consistent with a diagnosis of HPT. Of these, 5.6% had isolated systolic HPT. Only 25 (5.4%) of the participants had random plasma glucose >11.0 mmol/L while 26.1% were classified as diabetic based on the HbA1c results. Of the 120 participants with HbA1c ≥ 6.5%, 93 had a pre-existing diagnosis of diabetes mellitus and were already on treatment. The remaining 27 participants represented newly detected cases with no prior diagnosis.
Table 1.
Sociodemographic and clinical characteristics.
| Variable | Female | Male | Total$ | Chi square | p value |
|---|---|---|---|---|---|
| Study location | |||||
| Ezintsha | 154 (70.0) | 66 (30.0) | 220 (100) | 5.114 | 0.025* |
| Riverpark | 190 (79.2) | 50 (20.8) | 240 (100) | ||
| Age | |||||
| 19 to 29 | 38 (76.0) | 12 (24.0) | 50 (100) | 6.203 | 0.184 |
| 30 to 39 | 106 (81.5) | 24 (18.5) | 130 (100) | ||
| 40 to 49 | 100 (69.9) | 43 (30.1) | 143 (100) | ||
| 50 to 59 | 58 (76.3) | 18 (23.7) | 76 (100) | ||
| ≥ 60 | 42 (68.9) | 19 (31.1) | 61 (100) | ||
| Mean age | 43.90 ± 12.22 | 45.39 ± 12.69 | |||
| Race | |||||
| Black/African | 340 (74.6) | 116 (25.4) | 456 (100) | 1.361# | 0.715 |
| White | 2 (100) | 0 (0.0) | 2 (100) | ||
| Coloured | 1 (100) | 0 (0.0) | 1 (100) | ||
| Indian/Asian | 1 (100) | 0 (0.0) | 1 (100) | ||
| Employment status$ | |||||
| Employed | 162 (73.6) | 58 (26.4) | 220 (100) | 0.326 | 0.589 |
| Unemployed | 177 (76.0) | 56 (24.0) | 233 (100) | ||
| Covid-19 vaccination$ | |||||
| Yes | 275 (77.9) | 78 (22.1) | 353 (100) | 7.213 | 0.008* |
| No | 69 (65.1) | 37 (34.9) | 106 (100) | ||
| BMI | |||||
| Normal weight (18.5–24.9) | 33 (38.8) | 52 (61.2) | 85 (100) | 89.926 | <0.001* |
| Pre-obesity (25.0–29.9) | 56 (69.1) | 25 (30.9) | 81 (100) | ||
| Obesity class I (30.0–34.9) | 116 (81.1) | 27 (18.9) | 143 (100) | ||
| Obesity class II (35.0–39.9) | 66 (85.7) | 11 (14.3) | 77 (100) | ||
| Obesity class III (≥ 40) | 73 (98.6) | 1 (1.4) | 74 (100) | ||
| Mean BMI | 34.56 ± 7.67 | 27.02 ± 5.92 | |||
| Family history of NCDs$ | |||||
| Yes | 187 (82.7) | 39 (17.3) | 226 (100) | 16.200 | <0.001* |
| No | 149 (66.2) | 76 (33.8) | 225 (100) | ||
| Tobacco use | |||||
| Yes | 50 (56.2) | 39 (43.8) | 89 (100) | 20.250 | <0.001* |
| No | 294 (79.2) | 77 (20.8) | 371 (100) | ||
| Smoking status$ | |||||
| Non-smoker | 294 (79.2) | 77 (20.8) | 317 (100) | 40.865 | <0.001* |
| Current smoker | 31 (44.9) | 38 (55.1) | 69 (100) | ||
| Other tobacco use | 19 (95.0) | 1 (5.0) | 20 (100) | ||
| Illicit drug use | |||||
| Yes | 3 (37.5) | 5 (62.5) | 8 (100) | 6.001 | 0.027* |
| No | 341 (75.4) | 111 (24.6) | 452 (100) | ||
| Drinks alcohol | |||||
| Yes | 158 (66.9) | 78 (33.1) | 236 (100) | 15.770 | <0.001* |
| No | 186 (83.0) | 38 (17.0) | 224 (100) | ||
| Haemoglobin A1c (%)$ | |||||
| < 6.5 (inconclusive) | 248 (74.7) | 84 (25.3) | 332 (100) | 0.183 | 0.713 |
| ≥ 6.5 (diabetic) | 92 (76.7) | 28 (23.3) | 120(100) | ||
| Estimated GFR category | |||||
| G1: Normal or high | 240 (78.7) | 65 (21.3) | 305 (100) | 14.686# | 0.005* |
| G2: Mildly decreased | 92 (70.8) | 38 (29.2) | 130 (100) | ||
| G3a: Mildly to moderately decreased | 8 (50.0) | 8 (50.0) | 16 (100) | ||
| G3b: Moderately to severely decreased | 4 (57.1) | 3 (42.9) | 7 (100) | ||
| G4: Severely decreased | 0 (0.0) | 2 (100) | 2 (100) | ||
| Estimated GFR in two categories | |||||
| Normal or high GFR (G1) | 240 (78.7) | 65 (21.3) | 305 (100) | 7.323 | 0.007* |
| Decreased GFR (G2 to G4) | 104 (67.1) | 51 (32.9) | 155 (100) | ||
| Albumin-creatinine ratio$ | |||||
| < 3 mg/mmol | 330 (74.3) | 114 (25.7) | 444 (100) | 1.170 | 0.279 |
| 3–30 mg/mmol | 13 (86.7) | 2 (13.3) | 15 (100) | ||
| Random plasma glucose (mmol/L) | |||||
| < 5.6 (diabetes excluded) | 199 (72.6) | 75 (27.4) | 274 (100) | 2.212 | 0.331 |
| 5.6–11.0 (inconclusive) | 127 (78.9) | 34 (21.1) | 161 (100) | ||
| ≥ 11.1 (diabetic) | 18 (72.0) | 7 (28.0) | 25 (100) | ||
| Total cholesterol$ | |||||
| < 200 mg/dL (desirable) | 339 (74.8) | 114 (25.2) | 453 (100) | 0.579# | 0.447 |
| 200–239 mg/dL (borderline) | 1 (100) | 0 (0.0) | 1 (100) | ||
| HDL | |||||
| < 40 mg/dL (low – increased risk) | 327 (75.3) | 107 (24.7) | 434 (100) | 7.832 | 0.020* |
| 40–59 mg/dL (borderline) | 9 (56.3) | 7 (43.8) | 16 (100) | ||
| ≥ 60 mg/dL (high/protective) | 9 (100) | 0 (0.0) | 9 (100) | ||
| Blood pressure categories | |||||
| Normal | 74 (80.4) | 18 (19.6) | 92 (100) | 9.554# | 0.145 |
| Optimal blood pressure | 28 (80.0) | 7 (20.0) | 35 (100) | ||
| High normal blood pressure | 138 (73.0) | 51 (27.0) | 189 (100) | ||
| Grade 1 HPT | 70 (79.5) | 18 (20.5) | 88 (100) | ||
| Grade 2 HPT | 20 (64.5) | 11 (35.5) | 31 (100) | ||
| Grade 3 HPT | 9 (52.9) | 8 (47.1) | 17 (100) | ||
| Isolated systolic HPT | 5 (62.5) | 3 (37.5) | 8 (100) | ||
| Mean systolic blood pressure | 134.32 ± 20.76 | 139.57 ± 22.74 | |||
| Mean diastolic blood pressure | 82.51 ± 13.78 | 84.73 ± 12.35 | |||
| HPT | |||||
| Yes | 104 (72.2) | 40 (27.8) | 144 (100) | 0.729 | 0.418 |
| No | 240 (75.9) | 76 (24.1) | 316 (100) | ||
| Newly diagnosed NCD | |||||
| Yes | 251 (73.8) | 89 (26.2) | 340 (100) | 0.636 | 0.465 |
| No | 93 (77.5) | 27 (22.5) | 120 (100) | ||
| Newly diagnosed diabetes mellitus | |||||
| Yes | 21 (77.8) | 6 (22.2) | 27 (100) | 0.220 | 0.820 |
| No | 232 (73.7) | 83 (26.3) | 315 (100) | ||
| Newly diagnosed HPT | |||||
| Yes | 32 (61.5) | 20 (38.5) | 52 (100) | 4.794 | 0.039* |
| No | 219 (76.0) | 69 (24.0) | 288 (100) | ||
| Newly diagnosed dyslipidaemia | |||||
| Yes | 96 (80.7) | 23 (19.3) | 119 (100) | 4.195 | 0.041* |
| No | 155 (70.5) | 65 (29.5) | 220 (100) | ||
| Newly diagnosed kidney disease | |||||
| Yes | 105 (71.9) | 41 (28.1) | 146 (100) | 0.520 | 0.533 |
| No | 147 (75.4) | 48 (24.6) | 195 (100) | ||
| Framingham 10-year risk category | |||||
| Low risk | 159 (88.8) | 20 (11.2) | 179 (100) | 54.123 | <0.001* |
| Medium risk | 155 (70.8) | 64 (29.2) | 219 (100) | ||
| High risk | 26 (55.3) | 21 (44.7) | 47 (100) | ||
| Very high risk | 0 (0.0) | 8 (100) | 8 (100) | ||
| Referred following screening | |||||
| Yes | 265 (74.0) | 93 (26.0) | 358 (100) | 0.123 | 0.726 |
| No | 72 (75.8) | 23 (24.2) | 95 (100) | ||
#Likelihood ratio reported because of cells with expected counts <5 *statistically significant. $ Totals vary across variables due to missing data. Percentages are calculated using the valid N for each variable. There were some missing clinical data (e.g., employment status, family history, smoking history, newly diagnosed NCD’s) which were due to subjects “not knowing” or “refusing” to answer. For instance, many subjects are not formally employed, worked intermittently, rely on family support and/or social grants or a combination of these. The POC (HbA1c, Cholesterol, and Albuminuria) discrepancies were related to “device days off” when consumable shortages were experienced (wastage, re-calibration needs). We elected to include these patients in the analysis in line with the goal of “real world” challenges in this pilot study.
While 66% of the study population had normal to high glomerular filtration rates (GFR), the proportion who had impaired kidney function was significantly higher among the females than the male participants. The cardiovascular risk measured with the Framingham 10-year risk classification showed that 38.9% of the study population had low risk, 12% had high to very high risk while eight (8) men, i.e., 1.7% of the population had very high risk.
HIV status and multimorbidity prevalence
While 17 (3.7%) of the population had no NCDs reported or diagnosed in the study, 11 (2.4%) of the participants had four NCDS while 69 (15.0%) had three (Table 2). The prevalence of multimorbidity in the study population was 17.4%. The proportion who had co-existing HIV and multimorbidity was 12.8%.
Table 2.
HIV and co-existing NCD.
| Variable | HIV positive | HIV negative | Total | Chi square | p value |
|---|---|---|---|---|---|
| Multimorbidity | |||||
| Yes | 23 (12.8) | 57 (21.3) | 80 | 5.375 | 0.020 |
| No | 157 (87.2) | 210 (78.7) | 367 | ||
| Newly diagnosed NCDs | |||||
| Yes | 163 (85.8) | 177 (65.6) | 340 | 23.680 | <0.001 |
| No | 27 (14.2) | 93 (34.4) | 120 | ||
| Newly diagnosed diabetes | |||||
| Yes | 15 (9.1) | 12 (6.7) | 27 | 0.679 | 0.410 |
| No | 149 (90.9) | 166 (93.3) | 315 | ||
| Newly diagnosed HPT | |||||
| Yes | 16 (9.8) | 147 (90.2) | 52 | 7.253 | 0.007 |
| No | 36 (20.3) | 141 (79.7) | 288 | ||
| Newly diagnosed dyslipidaemia | |||||
| Yes | 81 (50.0) | 38 (21.5) | 119 | 30.224 | <0.001 |
| No | 81 (50.0) | 139 (78.5) | 177 | ||
| Newly diagnosed renal disease | |||||
| Yes | 109 (66.9) | 37 (20.8) | 146 | 73.805 | <0.001 |
| No | 54 (33.1) | 141 (79.2) | 178 | ||
Predictors of kidney impairment
We examined the factors associated with renal impairment assessed with the GFR (Table 3). Factors that were significant on bivariate analysis were fitted into a logistics regression model. We performed a binary logistic regression using the Enter method to assess the impact of study location, sex, age, BMI, family history of NCDs, smoking status, tobacco use, HIV status, blood glucose, use of antiretroviral and diabetes medications and the cardiovascular risk measure with the Framingham risk category on the likelihood of kidney impairment measured with the GFR. The Hosmer Lemeshow goodness of fit test showed that the model was a good fit (χ2 = 8.261 at 8 degrees of freedom with p value = 0.408). Our findings showed that females had a higher risk of kidney impairment than males. The risk of kidney impairment was also higher among people who were 45 years and over compared with those who were less than 35 years old. Other significant predictors of kidney impairment in this population were BMI, random plasma glucose in the diabetic range, having a newly diagnosed NCD and cardiovascular risk were significant predictors of kidney impairment in this study population.
Table 3.
Predictors of kidney impairment.
| Variable | Decreased GFR n (%) | Adjusted odds | 95% CI |
|---|---|---|---|
| Study location | |||
| Ezintsha | 104 (47.3) | Ref | |
| Riverpark | 51 (21.3) | 0.852 | 0.422–1.720 |
| Sex | |||
| Female | 104 (30.2) | 2.855 | 1.392–5.855* |
| Male | 51 (44.0) | Ref | |
| Age | |||
| <35 | 21 (13.5) | Ref | |
| 35–44 | 49 (31.6) | 1.493 | 0.594–3.750 |
| 45–64 | 64 (41.3) | 4.866 | 1.584–14.946* |
| ≥ 65 | 21 (13.5) | 19.011 | 3.152–114.667* |
| BMI | |||
| Normal weight (18.5–24.9) | 60 (70.6) | Ref | |
| Pre-obesity (25.0–29.9) | 46 (56.8) | 0.403 | 0.182–0.892* |
| Obesity class I (30.0–34.9) | 34 (23.8) | 0.093 | 0.040–0.211* |
| Obesity class II (35.0–39.9) | 8 (10.4) | 0.023 | 0.007–0.074* |
| Obesity class III (≥ 40) | 7 (9.5) | 0.023 | 0.007–0.074* |
| Family history of NCDs | |||
| Yes | 62 (27.4) | 0.990 | 0.578–1.696 |
| No | 89 (39.6) | Ref | |
| Tobacco use | |||
| Yes | 40 (44.9) | 0.777 | 0.215–2.814 |
| No | 115 (31.0) | Ref | |
| Smoking status | |||
| Non-smoker | 115 (31.0) | Ref | |
| Current smoker | 35 (50.7) | 0.815 | 0.196–3.393 |
| Other tobacco use | 5 (25.0) | ||
| HIV | |||
| Yes | 91 (47.9) | 0.845 | 0.309–2.310 |
| No | 64 (23.7) | Ref | |
| Haemoglobin A1c (%) | |||
| < 6.5 (inconclusive) | 117 (35.2) | Ref | |
| ≥ 6.5 (diabetic) | 33 (27.5) | 1.050 | 0.548–2.012 |
| Random plasma glucose (mmol/L) | |||
| < 5.6 (diabetes excluded) | 110 (40.1) | Ref | |
| 5.6–11.0 (inconclusive) | 42 (26.1) | 0.907 | 0.495–1.661 |
| ≥ 11.1 (diabetic) | 3 (12.0) | 0.102 | 0.017–0.604* |
| Newly diagnosed NCD | |||
| Yes | 129 (37.9) | 3.193 | 1.620–6.293* |
| No | 26 (21.7) | Ref | |
| On diabetes medication | |||
| Yes | 147 (35.1) | 1.219 | 0.398–3.739 |
| No | 8 (19.5) | Ref | |
| On antiretroviral medication | |||
| Yes | 59 (23.7) | 2.037 | 0.818–5.069 |
| No | 96 (45.5) | Ref | |
| Framingham 10-year risk category | |||
| Low risk | 46 (25.7) | Ref | |
| Medium risk | 81 (37.0) | 2.683 | 1.474–4.881* |
| High risk | 18 (38.3) | 5.088 | 1.887–13.718* |
| Very high risk | 5 (62.5) | 19.909 | 2.253–175.902* |
#Likelihood ratio reported because of cells with expected counts <5*cells with expected counts <5 statistically significant.
Factors associated with referral for further care
Examining the pattern of referral for further care among the study population (Table 4), we found that individuals were more likely to be referred if they had been newly diagnosed with NCDs including diabetes, dyslipidaemia and kidney disease with impaired estimated GFR.
Table 4.
Factors associated with referral for further care.
| Referred for further care (n = 453) | Chi square | p value | |||
|---|---|---|---|---|---|
| Variable | Yes | No | Total | ||
| Sex | |||||
| Female | 265 (78.6) | 72 (21.4) | 337 (100) | 0.123 | 0.792 |
| Male | 93 (80.2) | 23 (19.8) | 116 (100) | ||
| Age | |||||
| 19 to 29 | 35 (70.0) | 15 (30.0) | 50 (100) | 9.326 | 0.053 |
| 30 to 39 | 98 (76.0) | 31 (24.0) | 129 (100) | ||
| 40 to 49 | 117 (83.0) | 24 (17.0) | 141 (100) | ||
| 50 to 59 | 65 (87.8) | 9 (12.2) | 74 (100) | ||
| ≥ 60 | 43 (72.9) | 16 (27.1) | 59 (100) | ||
| BMI | |||||
| Normal weight (18.5–24.9) | 71 (83.5) | 14 (16.5) | 85 (100) | 2.699 | 0.609 |
| Pre-obesity (25.0–29.9) | 64 (79.0) | 17 (21.0) | 81 (100) | ||
| Obesity class I (30.0–34.9) | 110 (79.1) | 29 (20.9) | 139 (100) | ||
| Obesity class II (35.0–39.9) | 54 (73.0) | 20 (27.0) | 74 (100) | ||
| Obesity class III (≥ 40) | 59 (79.7) | 15 (20.3) | 74 (100) | ||
| Haemoglobin A1c (%) | |||||
| < 6.5 (inconclusive) | 262 (79.2) | 69 (20.8) | 331 (100) | 0.060 | 0.807 |
| ≥ 6.5 (diabetic) | 89 (78.1) | 25 (21.9) | 114 (100) | ||
| Random plasma glucose (mmol/L) | |||||
| < 5.6 (diabetes excluded) | 213 (78.6) | 58 (21.4) | 271 (100) | 2.476 | 0.290 |
| 5.6–11.0 (inconclusive) | 123 (77.8) | 35 (22.2) | 158 (100) | ||
| ≥ 11.1 (diabetic) | 22 (91.7) | 2 (8.3) | 24 (100) | ||
| HPT | |||||
| Yes | 112 (78.9) | 30 (21.1) | 142 (100) | 0.003 | 1.000 |
| No | 246 (79.1) | 65 (20.9) | 311 (100) | ||
| Newly diagnosed NCD | |||||
| Yes | 291 (86.9) | 44 (13.1) | 335 (100) | 47.659 | <0.001* |
| No | 67 (56.8) | 51 (43.2) | 118 (100) | ||
| Newly diagnosed diabetes mellitus | |||||
| Yes | 24 (100) | 0 (0.0) | 24 (100) | 6.986+ | 0.008* |
| No | 269 (85.9) | 44 (14.1) | 313 (100) | ||
| Newly diagnosed HPT | |||||
| Yes | 48 (92.3) | 4 (7.7) | 52 (100) | 1.598 | 0.266 |
| No | 243 (85.9) | 40 (14.1) | 283 (100) | ||
| Newly diagnosed dyslipidaemia | |||||
| Yes | 114 (98.3) | 2 (1.7) | 116 (100) | 20.369 | <0.001* |
| No | 176 (80.7) | 42 (19.3) | 218 (100) | ||
| Newly diagnosed kidney disease | |||||
| Yes | 146 (100) | 0 (0.0) | 146 (0.0) | 38.905 | <0.001* |
| No | 146 (76.8) | 44 (23.2) | 190 (100) | ||
| Estimated GFR category | |||||
| G1: Normal or high | 214 (71.8) | 84 (28.2) | 298 (100) | 34.596# | <0.001* |
| G2: Mildly decreased | 120 (92.3) | 10 (7.7) | 130 (100) | ||
| G3a: Mildly to moderately decreased | 16 (100) | 0 (0.0) | 16 (100) | ||
| G3b: Moderately to severely decreased | 6 (85.7) | 1 (14.3) | 7 (100) | ||
| G4: Severely decreased | 2 (100) | 0 (0.0) | 2 (100) | ||
| Estimated GFR in two categories | |||||
| Normal or high GFR (G1) | 214 (71.8) | 84 (28.2) | 298 (100) | 27.368 | <0.001* |
| Decreased GFR (G2 to G4) | 144 (92.9) | 11 (7.1) | 155 (100) | ||
| Albumin-creatinine ratio | |||||
| <3 mg/mmol | 345 (78.9) | 92 (21.1) | 437 (100) | 0.161 | 0.689 |
| 3 – 30 mg/mmol | 13 (86.7) | 2 (13.3) | 15 (100) | ||
| Framingham 10-year risk category | |||||
| Low risk | 132 (75.0) | 44 (25.0) | 176 (100) | 2.934 | 0.403 |
| Medium risk | 177 (81.6) | 40 (18.4) | 217 (100) | ||
| High risk | 36 (80.0) | 9 (20.0) | 45 (100) | ||
| Very high risk | 7 (87.5) | 1 (12.5) | 8 (100) | ||
#Likelihood ratio reported because of cells with expected counts <5 *statistically significant.
Discussion
This study serves as a pilot to demonstrate that integrating screening for NCDs including HPT, T2D, obesity, dyslipidaemia, and kidney disease, into health care services is achievable. Future studies will be needed to establish cost effectiveness and implementation.
Using existing infrastructure and staff, a nurse-led approach identified a high burden of undiagnosed NCDs among adults attending PHC clinic and HIV treatment clinic in Johannesburg.
The WHO advocates POC testing for NCDs as part of its package for essential intervention in PHC especially for LMICs (14). There are advantages and barriers to POC testing that should be considered when scalling a screening program. A review study by Tenorio-Mucha et al. (14) revealed advantages to POC such as convenience, simple tests, real time results facilitating medical decision making, as well as barriers such as accuracy of results when starting patients on treatment for life. Szablowski et al. (15) found POC HbA1c tests to be as accurate as lab based. Mutepfa et al. (16) evaluation of POC cholesterol machines can be used in PHC as long as they are validated. We opted to use POC testing due to convenience for the participants and resourse constraints and had referral systems to a medical doctor for those with abnormalities.
Our findings align with previous studies showing the utility of POC screening in PHC settings.
For example, Brennan et al. (17) integrated POC screening for HPT and T2D into South Africa’s COVID-19 vaccine programme and identified similarly high levels of undiagnosed disease. However, our study expands this approach by incorporating kidney function assessment and a broader range of cardiometabolic risk factors. Compared to Brennan et al.’s cohort, we found higher prevalence of HPT (31.3% vs. 19%) and a slightly lower rate of newly diagnosed T2D (5.9% vs. 7.8%), likely reflecting differences in demographic and clinical profiles.
Around half of the participants drank alcoholic beverages, the most common form being beer (24.1%) on a weekly basis. Alcohol consumption has been linked globally to 474000 CVD in 2019 (18). The level of risk of CVD from alcohol is linked to the volume of alcohol consumed at a time. Unfortunately, we did not measure this in the study and are unable to define if this is high risk, we can only assume that it has an effect due to the high percentage reported. To our surprise 80.7% of the participants were non-smokers. We suspect this is in keeping with patients receiving health education while taking meds for HIV and previous CVD. Almost 50 percent of the population were unemployed and from poorer communities in Johannesburg which can also be considered a risk factor for CVD.
A key secondary finding was the high proportion of participants with obesity (63.9%), particularly among women. This is consistent with national data showing higher obesity rates among South African women and aligns with global concerns about rising obesity-related cardiovascular risk (19). In 2019 five million deaths from NCDs were attributed to high BMI (20). We also observed a high burden of kidney disease (31.7%), identified using eGFR thresholds from the 2024 KDIGO CKD guidelines. Predictors of kidney impairment included female sex, older age, obesity, HIV status, family history of NCDs, and both newly diagnosed and treated diabetes. Interestingly, HPT did not emerge as a significant predictor, which may be explained by the relatively young mean age of our cohort (44.4 years), where end-organ damage may not yet be clinically apparent. End organ complications usually take years to be detectable, hence, the high prevalence of HPT in this population implies the potential for end organ damage including renal impairment a few years down the line. This is therefore a population to watch and screen at regular intervals, so changes are detected early.
The observed clustering of metabolic, renal, and cardiovascular risk factors supports the emerging concept of cardiovascular-kidney-metabolic (CKM) syndrome (21). This syndemic model recognises the shared biological pathways and social determinants that drive overlapping chronic conditions. Each of these 3 conditions can lead to or worsen the other leading to complications such as heart or kidney failure (21). Screening for this syndrome involves looking at basic biological assessments, POC tests as well social determinants of health such as education, unemployment and mental health issues (21). Monitoring BP, lipid profiles, blood glucose, BMI, and kidney function is imperative for the identification and management of CKM as the syndrome is divided into stages that determine the medical intervention.
The age profile of the cohort explains the absence of significant associations between BMI, hypertension, and kidney impairment in the multivariable model. Age was the dominant predictor of kidney impairment (AOR 4.866 for ages 45–64; AOR 19.011 for ≥65), yet the mean age was 44.4 years with 39.1% of participants under 40—below the threshold at which obesity-related nephropathy and hypertensive nephrosclerosis typically become clinically manifest. The high burden of both risk factors in this young cohort reflects a population at an early but critical stage of chronic disease progression, where timely intervention and long-term surveillance offer the greatest opportunity to prevent future CKD.
The majority of our cohort came from low-income communities, and nearly half were unemployed, factors known to increase vulnerability to NCDs through limited access to healthy food, physical activity, and healthcare.
Despite strengths, this study has several limitations. Convenience sampling used can lead to selection bias there is a potential for single time point measurement bias. The study attempted to mitigate this bias by ensuring instruments across both sites were identical, validated with suppliers has having high sensitivity and specificity and approved protocol were followed by both sites. However, we did not do multiple sample measurements or multivariate analysis which opened to this bias. Single-occasion assessments are subject to biological and measurement variability and may over- or underestimate true disease prevalence. All participants with abnormal screening results were referred to a clinician for confirmatory assessment. Prevalence estimates from this study should therefore be interpreted as screening yields rather than confirmed disease prevalence.
We were unable to distinguish between type 1 diabetes and T2D among participants already on treatment, which may have introduced misclassification. While referrals were made for participants with abnormal findings, we did not track whether they accessed follow-up care, limiting our ability to assess downstream clinical impact. Alcohol use was reported but not quantified, preventing an assessment of dose-dependent cardiovascular risk. Finally, the cross-sectional design limits causal inference.
Data in Table 2 demonstrates higher rates of newly diagnosed NCDs among HIV-positive participants, these comparisons are unadjusted and may reflect confounding by ART use, immune status, and clinical characteristics. A fully adjusted analysis of NCD outcomes stratified by HIV status is warranted in future, larger studies, and should account for ART regimen type, CD4 count, viral load, and duration of infection. Our findings have several implications for practice and policy. They support expanding routine NCD screening within PHC using existing personnel, particularly nurses, with minimal disruption to services. The high burden of undiagnosed and overlapping conditions underscores the need for integrated, person-centered care. In settings with a high prevalence of HIV and constrained health resources, CKM screening may offer a cost-effective approach to identifying individuals at risk for multiple chronic conditions. Additionally, the disproportionate burden among women highlights the importance of gender-responsive strategies.
Future research should evaluate the long-term outcomes of integrated NCD screening, including linkage to care, treatment uptake, and clinical endpoints such as cardiovascular events or renal decline. Implementation studies should also explore how best to operationalise CKM screening within PHC, including task-shifting models, digital decision support, and integration with existing HIV and maternal health programmes.
Acknowledgments
Medical writing and editorial services have been provided by Shantha Kumar V and Shruthi VB.
Glossary List of commonly used abbreviationsAbbreviationDescription
- ACC
American college of cardiology
- AHC
American heart association
- ART
Antiretroviral therapy
- BP
Blood pressure
- CKD
Chronic kidney disease
- COVID-19
Coronavirus disease-19
- Crr
Creatinine
- CKM
Cardiovascular-Kidney -Metabolic
- CVD
Cardiovascular disease
- e-GFR
Estimated glomerular filtration rate
- HbA1C
Haemoglobin A1C
- HDL
High density lipoprotein
- HPT
Hypertension
- HREC
Human research ethics committee
- ICF
Informed consent form
- LMIC
Lower- and middle-income countries
- NCDs
Non-communicable diseases
- POC
Point of care
- PHC
Primary health care
- SAHPRA
South african health products regulatory authority
- SEMDSA
Society of endocrine, metabolism, and diabetes of South Africa
- SDG
Sustainable development goal
- SA
South africa
- T2D
Type 2 diabetes mellitus
- TB
Tuberculosis
- WHO
World health organisation
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. The study was funded by Viatris.
Footnotes
Edited by: Gregory Wayne Heath, University of Tennessee at Chattanooga, United States
Reviewed by: Damaris Ochanda, Masinde Muliro University of Science and Technology, Kenya
Richard Osei-Yeboah, University of Edinburgh, United Kingdom
Author contributions
NJ: Methodology, Formal analysis, Resources, Data curation, Visualization, Investigation, Validation, Funding acquisition, Software, Conceptualization, Writing – review & editing, Project administration, Supervision, Writing – original draft. NC: Project administration, Formal analysis, Visualization, Supervision, Methodology, Writing – original draft, Writing – review & editing, Software, Data curation, Funding acquisition, Resources, Conceptualization, Validation, Investigation. OO: Resources, Validation, Conceptualization, Investigation, Visualization, Project administration, Writing – review & editing, Funding acquisition, Methodology, Formal analysis, Data curation, Writing – original draft, Supervision, Software. JW: Formal analysis, Project administration, Supervision, Data curation, Writing – review & editing, Methodology, Validation, Writing – original draft, Visualization, Investigation, Funding acquisition, Conceptualization, Software, Resources.
Conflict of interest
NC has received research funding from Viatris Inc via the institute, she is associated with and has received honoraria from Novo Nordisk. JW has received research grant from Viatris Inc. Author NJ was employed by company Viatris.
The authors declare that this study received funding from Viatris Inc. However, the funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.
The remaining author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpubh.2026.1825868/full#supplementary-material
References
- 1.Onder G, Rezza G, Brusaferro S. Case-fatality rate and characteristics of patients dying in relation to COVID-19 in Italy. JAMA. (2020) 323:1775–6. doi: 10.1001/jama.2020.4683, [DOI] [PubMed] [Google Scholar]
- 2.World Health Organization . Second round of the national pulse survey on continuity of essential health services during the COVID-19 pandemic. Geneva: World Health Organization; (2021). Available online at: https://www.who.int/publications/i/item/WHO-2019-nCoV-EHS-continuity-survey-2021.1 (Accessed March 30, 2022). [Google Scholar]
- 3.Ennezat PV, Guerbaai RA. Fighting non-communicable diseases in the COVID-19 era: a catch 22? Acta Cardiol (2022) 77, 366–367. doi: 10.1080/00015385.2021.1887587, [DOI] [PubMed] [Google Scholar]
- 4.Olds PK, Nuwagira E, Obwoya PS, Kansiime G, Musinguzi N, Tumwekwatse L, et al. In-hospital trends of non-communicable disease mortality during the pandemic for patients without COVID-19 at a regional referral hospital in southwestern Uganda. BMC Glob Public Health. (2025) 3:37. doi: 10.1186/s44263-025-00155-9, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.World Health Organization . Mental health and COVID-19: early evidence of the pandemic’s impact: scientific brief. Geneva: World Health Organization. (2022) Available online at: https://www.who.int/publications/i/item/WHO-2019-nCoV-Sci_Brief-Mental_health-2022 (Accessed March 30, 2026). [Google Scholar]
- 6.Basa M, De Vries J, McDonagh D, Comiskey C. The impact of COVID-19 on non-communicable disease patients in sub-Saharan African countries: a systematic review. PLoS One. (2024) 19:e0293376. doi: 10.1371/journal.pone.0293376, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Martinez R, Lloyd-Sherlock P, Soliz P, Ebrahim S, Vega E, Ordunez P, et al. Trends in premature avertable mortality from non-communicable diseases for 195 countries and territories, 1990-2017: a population-based study. Lancet Glob Health. (2020) 8:e511–23. doi: 10.1016/S2214-109X(20)30035-8, [DOI] [PubMed] [Google Scholar]
- 8.Stats South Africa . Non-communicable Diseases in South Africa: Findings from Death Notification, 2008–2018. Pretoria: Statistics South Africa. (2023). Available online at: https://www.statssa.gov.za/publications/Report-03-08-01/Report-03-08-012018.pdf (Accessed March 28, 2026). [Google Scholar]
- 9.United Nations . Transforming our world: the 2030 agenda for sustainable development. New York, NY: United Nations. Available online at: https://sdgs.un.org/2030agenda (Accessed March 28, 2026) [Google Scholar]
- 10.McCombe G, Murtagh S, Lazarus JV, Van Hout MC, Bachmann M, Jaffar S, et al. Integrating diabetes, hypertension and HIV care in sub-Saharan Africa: a Delphi consensus study on international best practice. BMC Health Serv Res. (2021) 21:1235. doi: 10.1186/s12913-021-07073-0, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kivuyo S, Birungi J, Okebe J, Wang D, Ramaiya K, Ainan S, et al. Integrated management of HIV, diabetes, and hypertension in sub-Saharan Africa (INTE-AFRICA): a pragmatic cluster-randomised, controlled trial. Lancet. (2023). 402:1241–50. doi: 10.1016/S0140-6736(23)01573-8, [DOI] [PubMed] [Google Scholar]
- 12.Rocks S, Berntson D, Gil-Salmerón A, Kadu M, Ehrenberg N, Stein V, et al. Cost and effects of integrated care: a systematic literature review and meta-analysis. Eur J Health Econ. (2020) 21:1211–21. doi: 10.1007/s10198-020-01217-5, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.NCD Alliance . Policy research report: spending wisely: exploring the economic and societal benefits of integrating HIV/AIDS and NCDs service delivery. Geneva: NCD Alliance. (2023). Available online at: https://ncdalliance.org/sites/default/files/resource_files/spending_wisely-ncda_rti_report-gw4a23-en.pdf [Google Scholar]
- 14.Tenorio-Mucha J, Busta-Flores P, Lazo-Porras M, Vetter B, Safary E, Moran AE, et al. Facilitators and barriers of the implementation of point-of-care devices for cardiometabolic diseases: a scoping review. BMC Health Serv Res. (2023) 23:412. doi: 10.1186/s12913-023-09419-2, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Szablowski CJ, Suscha E, Davis K, Xie CZ, Moskowitz K, Anderson JH, et al. Point-of-care HbA1c—a case for diabetes screening and diagnosis. Diabetes. (2018) 67:1518. doi: 10.2337/db18-1518-P [DOI] [Google Scholar]
- 16.Mutepfa CC, Hicks TP, Winter A, Dickinson RE, Williams C, Harrison N, et al. Can we trust published evidence on point-of-care tests for cholesterol? A rapid review. BMJ Open. (2025) 15:e080726. doi: 10.1136/bmjopen-2023-080726, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Brennan AT, Vetter B, Masuku SD, Mtshazo B, Mashabane N, Sokhela S, et al. Integration of point-of-care screening for type 2 diabetes mellitus and hypertension into the COVID-19 vaccine programme in Johannesburg, South Africa. BMC Public Health. (2023) 23:2291. doi: 10.1186/s12889-023-17190-6, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.World Health Organization . Alcohol. Geneva: World Health Organization. (2024). Available online at: https://www.who.int/news-room/fact-sheets/detail/alcohol (Accessed December 30, 2024). [Google Scholar]
- 19.Govender I, Sunnasy A. The growing problem of obesity in South Africa. S Afr Fam Pract. (2025) 67:a6001. doi: 10.4102/safp.v67i1.6001, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.World Health Organization . Obesity and overweight. Geneva: World Health Organization. Available online at: https://www.who.int/news-room/fact-sheets/detail/obesity-and-overweight (Accessed December 25, 2025). [Google Scholar]
- 21.Ndumele CE, Rangaswami J, Chow SL, Neeland IJ, Tuttle KR, Khan SS, et al. Cardiovascular-kidney-metabolic health: a presidential advisory from the American Heart Association. Circulation. (2023) 148:1606–35. doi: 10.1161/CIR.0000000000001184, [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.