A Temporal Analysis of the Impact of COVID-19 on Diabetes Epidemiology in KwaZulu-Natal, South Africa | ||
Afro-Egyptian Journal of Infectious and Endemic Diseases | ||
Articles in Press, Accepted Manuscript, Available Online from 21 October 2025 PDF (1.3 M) | ||
Document Type: Original Article | ||
DOI: 10.21608/aeji.2025.389237.1482 | ||
Authors | ||
Nikita Sahadew* 1; Somasundram Pillay2; Veena Singaram1 | ||
1Clinical and Professional Practice, School of Clinical Medicine, University of KwaZulu-Natal, Durban, South Africa. | ||
2Clinical and Professional Practice, School of Clinical Medicine, University of KwaZulu-Natal, Durban, South Africa. Department of Internal Medicine, Victoria Mxenge Hospital, Durban, South Africa. | ||
Abstract | ||
Background and study aim: The COVID-19 pandemic disrupted healthcare systems globally, raising concerns about its impact on chronic disease management, including diabetes. This study assessed the pandemic’s effects on diabetes incidence, healthcare utilization, glycaemic control, and diabetes-related amputations in KwaZulu-Natal (KZN), South Africa. We aimed to analyse trends in diabetes incidence, screening, healthcare engagement, glycaemic control, and amputations across pre-pandemic, pandemic, and post-pandemic periods in KZN, while identifying gaps and challenges in diabetes care recovery. Patients and Methods: A retrospective analysis of provincial health data from public sector facilities was conducted, comparing pre-pandemic (2018–2019), pandemic (2020–2021), and post-pandemic (2022–2023) periods. Indicators included diabetes incidence (per 100,000 population), screening volume and yield, clinic visits, HbA1c <7% rates, and amputation rates. Paired t-tests and Wilcoxon signed-rank tests assessed significance. Results: Diabetes incidence rose from 526.9 (pre) to 570.0 during COVID (+8.2%) and surged to 1104.2 post-pandemic (+93.7% vs COVID). Reported incidence underestimated true rates, with undercounts rising from 46.9% to 74.6%. Screening volumes declined by 1.7% during COVID but rose by 24.3% post-pandemic, with yields <0.003%. Treatment visits declined during COVID and partially recovered, with Amajuba exceeding pre-COVID engagement. Glycaemic control improved post-pandemic (p = 0.041; p = 0.00098). Amputation rates declined overall, though some districts remained high. Conclusion: COVID-19 disrupted diabetes care in KZN, contributing to underdiagnosis and service reductions, with a sharp post-pandemic rise in cases. Strengthening chronic care systems is essential for resilience during health crises. | ||
Highlights | ||
| ||
Keywords | ||
COVID-19; Diabetes Mellitus; Epidemiology; KwaZulu-Natal; South Africa | ||
Full Text | ||
INTRODUCTION The prevalence of type 2 diabetes mellitus (T2DM) in sub-Saharan Africa has escalated significantly, transitioning from a once-rare condition to a prominent public health concern. This surge is largely attributed to rapid urbanization and lifestyle changes, leading to a shift from communicable to non-communicable diseases in the region [1, 2, 3]. Despite this growing burden, there remains a scarcity of high-quality, population-representative data to inform effective interventions [1]. In South Africa, the diabetes epidemic poses a significant challenge when considering the high rates of urbanisation and resource limited climate [4, 5,6]. Factors such as limited access to healthcare, economic constraints, and a high proportion of undiagnosed cases exacerbate the situation [2]. The COVID-19 pandemic has further complicated diabetes management, as healthcare resources were reallocated to address the immediate crisis, potentially disrupting routine care for chronic conditions like diabetes. The interplay between COVID-19 and diabetes is multifaceted. Individuals with preexisting diabetes are at a heightened risk of severe COVID-19 outcomes, including increased mortality rates [7, 4-5]. Conversely, emerging evidence suggests that COVID-19 may precipitate new-onset diabetes in some patients, possibly due to the virus's impact on pancreatic beta cells and the body's inflammatory response [7]. In South Africa, the onset of the COVID-19 pandemic prompted a rapid and stringent national response, including international travel bans, one of the strictest lockdowns globally, and wide-ranging public health measures such as school closures, social distancing, and hygiene protocols [6]. These interventions were intended to curb viral spread and preserve health system capacity, but they also carried significant socioeconomic consequences and disrupted routine chronic disease services [4,8]. Although South Africa reported comparatively lower mortality rates than several similarly affected countries, excess deaths and health service interruptions highlighted the indirect impact of the pandemic on non-communicable disease care [3,5,7]. The unique trajectory of the South African response provides important context for interpreting the disruption and rebound observed in diabetes incidence and management across KwaZulu-Natal [5,6,8]. Given these complexities, it is imperative to understand how the COVID-19 pandemic has influenced diabetes epidemiology in a developing country such as South Africa. This study aims to analyse epidemiological data from 2018 to 2023 within one South African province, encompassing periods before, during, and after the pandemic, to assess changes in diabetes incidence, management, and outcomes. By identifying trends and disruptions, the research seeks to inform strategies for enhancing healthcare resilience and ensuring continuity of care for chronic disease patients during global health crises. PATIENTS AND METHODS Study Design and Setting: This retrospective, observational study was conducted to assess the impact of the COVID-19 pandemic on diabetes epidemiology in KwaZulu-Natal (KZN), South Africa. The study utilised data from public healthcare facilities across all 11 districts of KZN, encompassing both urban and rural settings and uses the district health information system (DHIS) as the collection tool of choice. KwaZulu-Natal Province comprises eleven health districts: Amajuba, eThekwini, Harry Gwala, iLembe, King Cetshwayo, Ugu, uMgungundlovu, uMzinyathi, uThukela, uThungulu, and Zululand. Further detail on these regions of interest can be found in Sahadew et al. [4]. Study Periods: The analysis covered six consecutive years, categorized into three distinct periods: Pre-pandemic period: 2018–2019 Pandemic period: 2020–2021 Post-pandemic period: 2022-2023 Data Sources and Collection: Data were extracted from the DHIS, a centralized database that compiles routine health service data from public health facilities. The dataset included aggregated monthly reports on diabetes-related indicators, such as:
Inclusion and Exclusion Criteria: Inclusion: All public health facilities in KZN reporting complete monthly data on diabetes indicators for the study period. Exclusion: None Data Analysis: Descriptive statistics were used to summarise the data, presenting frequencies and percentages for categorical variables. Comparative analyses were conducted to assess changes in diabetes indicators across the three defined periods. Chi-square tests were employed for categorical variables, and a p-value of <0.05 was considered statistically significant. The primary diabetes-related indicators collected from the DHIS for this study included:
Incidence measures Two incidence measures were available for analysis. First, we calculated incidence rates by dividing the number of newly diagnosed diabetes cases (from DHIS new patient entries) by the corresponding mid-year district population estimates, expressed per 100 000 population. Second, we extracted the DHIS field labelled “Incidence – all ages,” which represents the routinely reported incidence generated within the health information system. For consistency and transparency, both calculated and reported incidence were included in analyses, and discrepancies between the two are described in the Results. RESULTS Provincial incidence of diabetes - Calculated As seen in figure 1, the incidence of diabetes in KwaZulu-Natal increased steadily across the three periods. A slight rise was observed during the COVID period (570.0 per 100,000) compared to the pre-COVID years (526.9 per 100,000), representing an 8.2% increase. Notably, the post-COVID period showed a marked escalation to 1104.2 per 100,000—a 93.7% increase compared to the COVID period. These findings suggest a significant rebound in diagnoses or a possible surge in new cases following the pandemic, highlighting the long-term epidemiological impact of COVID-19 on non-communicable diseases like diabetes. Provincial incidence of diabetes - Calculated Vs Reported The above data in Figure 1 were calculated using the number of new patients reported and the mid-year estimates as published by the South African government for the respective years. Upon examination of the raw data, it was established that there did exist a field called ‘Incidence-all ages’. The DHIS includes a precomputed indicator labelled ‘Incidence – all ages’; we refer to this as the reported incidence and present it in Figure 2.. As seen from Figure 2, the comparison of calculated and reported diabetes incidence per 100,000 population in KwaZulu-Natal during the pre-COVID (2018–2019), COVID (2020–2021), and post-COVID (2022–2023) periods. Calculated incidence, derived from raw new case data and population estimates, was consistently higher than the corresponding reported values. The gap widened over time—from a 46.9% difference pre-COVID to a 74.6% difference post-COVID—indicating increasing underreporting or missed case detection. These discrepancies raise important concerns about the reliability of routine health information systems in capturing the true burden of diabetes, particularly during and after public health emergencies. District-Level Incidence Rates – as calculated Figure 3 presents the calculated incidence of diabetes (per 100,000 population) across KwaZulu-Natal's districts, grouped by pre-COVID (2018–2019), COVID (2020–2021), and post-COVID (2022–2023) periods. A consistent pattern emerged across most districts, with a decline or stagnation in incidence during the COVID period, followed by a marked increase post-COVID. The uMgungundlovu and eThekwini districts demonstrated the most substantial post-pandemic surges, while Amajuba and iLembe reflected more modest trends. These shifts suggest delayed diagnoses and possible backlogs in care during COVID-19 restrictions, followed by increased case capture and reporting in the recovery period. Calculated vs reported district-level incidence rates – ‘Incidence – all ages’ Figure 4 illustrates the comparison between calculated and reported diabetes incidence per 100,000 population across KwaZulu-Natal’s districts for the pre-COVID, COVID, and post-COVID periods. Across all districts and timeframes, calculated values significantly exceeded the reported incidence, with discrepancies ranging from approximately 90% to 97%. The post-COVID period shows the largest divergence, particularly in districts like uMgungundlovu, where the calculated incidence approached 1,300 per 100,000, compared to a reported figure near 50 per 100,000. These discrepancies suggest substantial underreporting or systemic gaps in surveillance and data capture, emphasizing the need to strengthen health information systems and improve case detection mechanisms across the province. Provincial screening for diabetes as reported Figure 5 illustrates the cumulative number of diabetes screenings conducted provincially across KwaZulu-Natal during three distinct periods: pre-COVID (2018–2019), COVID-period (2020–2021), and post-COVID (2022–2023). A black trend line connects the tops of each bar to highlight directional shifts across the timeframes. Compared to the pre-COVID baseline, screenings decreased slightly during the COVID period (−1.7%), reflecting service disruptions and reduced access to routine care. However, a substantial recovery is evident post-COVID, with a 24.3% increase in total screenings compared to the COVID period. The annotations and arrows emphasise these percentage changes, underscoring the rebound in public health service delivery in the post-pandemic phase. District-level screening for diabetes The bar graph in figure 6 displays the total number of diabetes screenings conducted across 11 districts in KwaZulu-Natal over three distinct periods: pre-COVID (2018–2019), COVID-period (2020–2021), and post-COVID (2022–2023). A dashed horizontal line represents the average screening volume during the pre-COVID period, serving as a reference benchmark. Most districts experienced a decline or stagnation in screenings during the COVID period, reflecting service disruptions. However, all districts showed a recovery post-COVID, with iLembe, Zululand, and Umzinyathi recording the highest relative increases compared to the pre-COVID baseline. Percentage change annotations above each post-COVID bar highlight the extent of this rebound in screening activity. The yield of diabetic screening To gain deeper insight into the effectiveness of diabetes case detection efforts across KwaZulu-Natal, we conducted a comparative analysis of diabetes screening volumes, and the number of newly diagnosed diabetes cases reported during the same periods. By examining the yield of diabetes screening—defined as the proportion of individuals screened who were subsequently diagnosed with diabetes—we aimed to evaluate the efficiency of screening strategies over time and across districts. This analysis also helps assess the extent to which fluctuations in screening activity, particularly during the COVID-19 period, may have influenced the observed incidence of newly diagnosed diabetes cases. Identifying discrepancies between screening coverage and incidence trends may reveal missed opportunities for early diagnosis or highlight districts where targeted interventions are needed. Figure 7 illustrates the percentage of individuals screened who were newly diagnosed with diabetes across KwaZulu-Natal districts over three periods. Overall, the yield remained extremely low in all districts (typically under 0.003%), indicating a low rate of case detection despite high volumes of screening. This trend suggests that many individuals screened were not diagnosed with diabetes, which could reflect inefficient or poorly targeted screening, underreporting of new cases, or gaps in follow-up and diagnostic confirmation. Notably, yields declined further during and after the COVID-19 period, despite a rebound in screening numbers, highlighting potential missed opportunities for earlier diagnosis. Glycaemic control trends during and post covid periods Glycaemic control trends among diabetes clients with HbA1c <7% from 2020 to 2023 were examined across the eleven health districts of KwaZulu-Natal Province. The HbA1c <7% threshold is internationally recognized as a marker of optimal glycaemic control and is associated with reduced risk of microvascular and macrovascular complications. The World Health Organization and various national diabetes guidelines recommend this target for the majority of adults with type 2 diabetes, unless individualized targets are indicated (10). Monitoring the percentage of patients achieving this target provides a valuable proxy for the effectiveness and continuity of diabetes management services. By tracking these figures through the COVID-19 pandemic and into the post-pandemic recovery period, we gain insight into the resilience and adaptation of public health systems in maintaining chronic care. Figure 8 illustrates the total number of diabetes clients with HbA1c levels <7% during the COVID-19 period (2020–2021) compared to the post-COVID period (2022–2023) across the eleven districts of KwaZulu-Natal Province. Across all districts, there was a substantial increase in the number of patients achieving optimal glycaemic control in the post-COVID period. The most notable improvements were observed in eThekwini, uMgungundlovu, and Zululand, where client numbers more than doubled. These findings suggest a significant rebound in diabetes service delivery following pandemic-related disruptions. The overall upward trend may reflect improved health system responsiveness, enhance chronic disease follow-up, or expanded access to care as facilities adapted to post-pandemic conditions. To assess whether the observed improvements in glycaemic control were statistically significant, a paired analysis was conducted comparing the number of diabetes clients with HbA1c <7% during the COVID-19 period (2020–2021) and the post-COVID period (2022–2023) across all 11 districts. Both the paired t-test (p = 0.041) and the Wilcoxon signed-rank test (p = 0.00098) indicated statistically significant increases in the number of clients achieving optimal HbA1c levels post-COVID. These findings confirm that the improvements illustrated in Figure 8 were unlikely due to random variation and instead reflect meaningful gains in diabetes management and service recovery following pandemic-related disruptions. Diabetes treatment visits and visit-to-incidence ratios Figure 9 displays a dual-axis comparison of diabetes treatment visits and Visit-to-Incidence Ratios across the eleven districts in KwaZulu-Natal, segmented by three time periods: pre-COVID (2018–2019), COVID (2020–2021), and post-COVID (2022–2023). As illustrated by the bar plots, all districts experienced a notable decline in treatment visits during the COVID period, with partial recovery observed post-COVID. For instance, eThekwini, the most populous district, experienced a drop in visits from 907,772 pre-COVID to 613,179 during COVID, with a rebound to 734,798 post-COVID. Similar trends were observed across most districts, although Uthukela and Ugu did not return to pre-COVID treatment levels even in the post-pandemic period. The Visit-to-Incidence Ratio represents the number of treatment visits relative to the number of new diabetes cases in each district, calculated as: Visit-to-Incidence Ratio = Treatment Visits ÷ New Diabetes Cases This ratio serves as a proxy for health system responsiveness and care engagement, indicating how well districts maintained or adapted care delivery in proportion to the growing diabetes burden. A higher ratio suggests that more treatment visits were being made per case of diabetes, which reflects better follow-up, access to care, or retention in the health system. Conversely, a lower ratio may indicate strained systems, disrupted care pathways, or declining engagement, despite rising incidence. The line plots in Figure 9 show that: Pre-COVID, most districts had relatively high Visit-to-Incidence Ratios, suggesting that the care system was adequately servicing known diabetic patients relative to incidence. During COVID, ratios declined sharply across all districts. This reflects both the direct disruption of routine healthcare services and the reduced ability to sustain ongoing care per diagnosed case. For example, Zululand’s ratio dropped by 66.6%, and eThekwini experienced a 49.3% decline. Post-COVID, only Amajuba demonstrated a positive net recovery in the Visit-to-Incidence Ratio (+73.3% compared to pre-COVID). All other districts had ratios that remained below baseline, despite increasing absolute numbers of treatment visits, indicating a mismatch between growing disease burden and service availability. eThekwini and King Cetshwayo, while maintaining the highest volume of visits, exhibited large declines in care per case, suggesting systemic pressure and possible gaps in follow-up. uMgungundlovu and Umkhanyakude also experienced steep declines in ratios post-COVID, despite relatively high incidence rates, indicating possible care fragmentation. Amajuba emerged as an outlier, with both incidence and treatment visits stabilizing, and an improved engagement ratio, pointing to potentially effective local adaptations or system resilience. Diabetes-related amputations Figure 10 presents the diabetes-related amputation rates per 100 new diabetes cases across all districts in KwaZulu-Natal, stratified by pre-COVID (2018–2019), COVID (2020–2021), and post-COVID (2022–2023) periods. The visualization highlights substantial inter-district variability and temporal trends, with most districts demonstrating a decline in amputation rates over time. Notably, King Cetshwayo, Amajuba, and uMgungundlovu recorded the highest relative rates during the pre-pandemic period, all exceeding 5%. A horizontal red reference line is included at the 5% threshold, which serves as a comparative benchmark for clinical concern. Rates above this level may signal systemic deficiencies in early detection, glycaemic control, foot care, or timely access to diabetic limb preservation services. Districts falling consistently above this line—particularly King Cetshwayo, which remained high even post-COVID—warrant further investigation and targeted intervention. It is important to note that use of a 5% reference line in Figure 10 serves as a visual aid to highlight districts with higher amputation rates, rather than reflecting a universally accepted clinical standard. Conversely, several districts including iLembe, Umzinyathi, and Zululand showed rates well below this threshold across all periods, though caution is warranted in interpreting these findings due to possible underreporting or limited access to surgical services. The overall post-COVID decline in relative amputation rates may reflect improved healthcare system recovery, renewed focus on diabetic foot care, or shifts in health-seeking behaviour. However, these findings should be interpreted within the context of the district-specific healthcare infrastructure and potential data quality limitations, as outlined in the limitations section. DISCUSSION The COVID-19 pandemic has reshaped the global health landscape, exposed systemic weaknesses while accelerated innovation and adaptation across healthcare systems. In South Africa, and specifically in the KwaZulu-Natal (KZN) province, the pandemic disrupted essential chronic disease services, including screening, diagnosis, and management of diabetes mellitus—a condition already posing a significant burden on the public sector. This study sought to assess temporal changes in diabetes-related indicators before (2018–2019), during (2020–2021), and after (2022–2023) the COVID-19 period, drawing on routinely collected health information system data across all districts in KZN. These results are discussed in the context of global literature examining the direct and indirect effects of COVID-19 on diabetes epidemiology, care continuity, health system resilience, and equity in service delivery. Existing literature highlights care deferrals, weakened surveillance, metabolic deterioration, and rising post-COVID chronic disease burdens as common themes internationally [3,2,7,8]. Understanding how these dynamics manifested locally allows for the identification of gaps, opportunities for system strengthening, and tailored responses to reduce long-term diabetes-related morbidity in the post-pandemic era. The findings from KZN clearly indicate a significant disruption in diabetes service delivery during the COVID-19 pandemic. There was a sharp decline in newly diagnosed diabetes cases, with calculated incidence only slightly increasing during the COVID period (570.0 per 100,000) compared to pre-COVID levels (526.9 per 100,000), followed by a near doubling post-COVID (1104.2 per 100,000). Screening volumes fell by 1.7% across the province, while treatment visits declined dramatically, as seen in eThekwini where they dropped from 907,772 pre-COVID to 613,179 during the pandemic. This downward trend in service utilization reflects the broader pattern of healthcare disruption noted globally. One study reported widespread deferral of care among patients with chronic conditions due to pandemic-related constraints and public fear of healthcare settings [8]. Another study similarly described how routine diabetes services were suspended or scaled back globally, leading to delayed diagnosis and suboptimal disease management [2]. The disruption in routine care likely allowed diabetes to go undiagnosed or unmanaged during this critical period, contributing to a potential worsening in glycaemic outcomes and increased long-term complications. The post-COVID period (2022–2023) revealed a sharp rebound in diabetes indicators. The incidence rate nearly doubled, screening increased by 24.3%, and there was statistically significant improvement in glycaemic control (p = 0.041, Wilcoxon p = 0.00098). This rebound likely represents both the resumption of health services and a surge in diagnoses that were delayed during the pandemic. However, the rise in incidence may also reflect a true increase in diabetes burden. It has been reported that pandemic-related factors—such as increased sedentary behaviour, weight gain, mental health stressors, and disrupted routines—could contribute to higher diabetes incidence rates post-COVID [3]. The convergence of catch-up care and worsening risk factors underscores the importance of maintaining robust surveillance and continuity of chronic care. It also suggests that future planning for health emergencies must include strategies to preserve essential non-communicable disease services. An important and concerning finding of this study is the discrepancy between calculated and reported incidence of diabetes. Calculated figures consistently exceeded reported data from the District Health Information System, with the gap growing from 46.9% pre-COVID to 74.6% post-COVID. At the district level, some areas showed discrepancies greater than 90%. These variances suggest substantial underreporting or delays in capturing new diabetes cases. These discrepancies could be explained as overburdened health systems during crises often deprioritize non-urgent data entry and compromise surveillance infrastructure [7]. Inadequate training, staffing shortages, and reliance on manual data entry further compound inaccuracies. These findings call for investment in digital infrastructure, routine data audits, and integrated reporting systems to ensure that real-time data can accurately inform policy and resource allocation, especially in regions with high non-communicable disease burdens [1,2]. While screening activities recovered post-COVID, the effectiveness of these efforts, as measured by the yield (percentage of screened individuals who were diagnosed), remained low—below 0.003% across all districts. The post-pandemic increase in screening was not matched by a proportionate rise in new diagnoses, which may reflect poor targeting of high-risk populations or inefficiencies in diagnostic follow-up. These broad, non-targeted screening strategies may waste resources and fail to detect those most at risk unless accompanied by structured risk stratification tools [5]. In the KZN context, the low screening yield suggests a need to refine diabetes screening protocols by integrating demographic, behavioural, and clinical risk predictors. Moreover, underreporting of positive cases and inadequate linkages to diagnostic services may also contribute to the low yield, indicating a need for better case tracking and care coordination across levels of the health system. A notable finding from the study is the substantial post-COVID improvement in glycaemic control, as measured by the number of patients achieving HbA1c <7%. Districts like eThekwini, uMgungundlovu, and Zululand more than doubled their number of well-controlled patients. These improvements, confirmed statistically, indicate that health services were able to re-engage diabetes patients and re-establish chronic care routines. It has been reported that in sub-Saharan Africa, diabetes programs gradually resumed their activities with renewed focus on continuity and decentralization, which may explain this trend [1]. Telemedicine adaptations and community outreach likely helped bridge access gaps. However, maintaining these gains requires deliberate effort to sustain medication access, clinician follow-up, and health literacy support. The improvements seen post-pandemic are encouraging but highlight the importance of resilient systems that can rebound quickly after crises. Despite a rebound in absolute treatment visits, many districts failed to return to pre-COVID care engagement levels when adjusted for the growing incidence. The visit-to-incidence ratio remained depressed in all districts except Amajuba, suggesting that the growing burden of diabetes is outpacing the capacity of the health system to provide consistent care. Zululand and eThekwini had large declines in their ratios, while King Cetshwayo and Umkhanyakude also showed substantial care gaps. A possible solution to this challenge could be a deliberate policy shifts post-COVID to build chronic disease care capacity, including strategies like task-shifting to nurses, mobile outreach units, and better integration of community health workers [6]. Without alignment between incidence and care capacity, KZN risks increasing rates of uncontrolled diabetes, hospitalizations, and complications. The province must urgently address health system bottlenecks and prioritize district-level service expansion to meet rising demand. The study found an encouraging overall decline in diabetes-related amputation rates across most districts, possibly reflecting improved foot care, earlier intervention, or better chronic disease management. However, King Cetshwayo remained above the 5% clinical concern threshold post-COVID, suggesting persistent challenges in this region. High amputation rates in uMgungundlovu and Amajuba pre-COVID also highlight longstanding system issues. It is well known that timely foot screening, multidisciplinary management, and access to limb preservation services are essential to reduce diabetes-related amputations [9]. In KZN, such services may be inconsistently available or underutilized. To address these disparities, the province should consider rolling out diabetic foot care bundles, expanding podiatry services, and ensuring timely referral pathways, especially in districts with historically high amputation burdens. This study is subject to several limitations. Firstly, the analysis relied on routinely collected health data from the District Health Information System (DHIS), which may be affected by underreporting, delayed reporting, or inconsistencies in data entry across districts and time periods. While every effort was made to validate and cross-check screening and incidence figures, some degree of data quality variability is inherent in large-scale administrative datasets. Although the DHIS includes an indicator labelled “detection rate,” this metric was excluded from comparative analysis due to the absence of a clearly defined calculation method or accompanying metadata. To maintain analytical transparency and internal consistency, we independently calculated screening yield using reported screening volumes, estimated district-level population figures, and incidence rates derived from the same dataset. HbA1c control data—specifically the proportion of diabetes clients with HbA1c <7%—was only available from 2020 onward in the DHIS dataset. As a result, this indicator could not be compared across all three periods (pre-, during, and post-COVID). While valuable for assessing quality of care among diagnosed patients, its limited temporal scope restricted our ability to evaluate long-term trends in glycaemic control or assess baseline management prior to the COVID-19 pandemic. A key limitation of this study is the presence of missing amputation data for certain districts and periods, most notably for Harry Gwala District (2018–2021) and Zululand District (2022–2023). These data gaps restrict the ability to conduct a complete temporal comparison across all districts and may obscure true trends in diabetes-related amputation rates. The absence of records may be attributable to inconsistent reporting systems, data capture errors, or service delivery interruptions during the COVID-19 pandemic. Consequently, interpretation of findings in these districts should be approached with caution, and future studies should aim to verify and supplement these data to ensure a more comprehensive provincial picture. Additionally, population estimates used to calculate incidence and screening yield were based on average values per district and period. While this approach provides a stable baseline, it does not account for intra-period population shifts such as migration, mortality, or demographic changes, which may have influenced screening coverage or detection outcomes. Finally, the use of aggregated data at the district level limits the ability to account for individual-level risk factors, comorbidities, or care-seeking behaviours that could influence diabetes diagnosis rates. Future studies incorporating disaggregated or patient-level data would be valuable for understanding these dynamics in more detail. CONCLUSION This study offers a comprehensive view of the evolving diabetes profile in KwaZulu-Natal across pre-, during-, and post-COVID periods. It highlights the significant disruptions caused by the pandemic, the resilience and recovery of the healthcare system, and the persisting gaps in surveillance, efficiency, and equitable access. To sustain recent improvements and address the growing burden of diabetes, policymakers must prioritize health information system strengthening, expand multi-disciplinary community-based chronic care, and invest in district-specific multi-disciplinary interventions tailored to local epidemiological trends. Ethical Considerations: The study was conducted in accordance with the ethical standards of the University of KwaZulu-Natal. No patient identifiers were used so anonymity was guaranteed (approval number HSS/1835/017D). Funding: The authors received no financial support for the research, authorship, and/or publication of this article. Conflict of Interest: The authors declare no conflicts of interest related to this study. Author Contributions:
| ||
References | ||
| ||
Statistics Article View: 2 PDF Download: 2 |