Table 1
PECOS framing of the scoping review on heat and health in Gauteng.
| Population | Studies conducted in Gauteng province (including Johannesburg) |
| Exposure | Studies that assessed environmental factors, agents, or exposures related to heat, heatwaves, and high temperature |
| Concept/context | Literature focused on heat in Gauteng province (including Johannesburg) |
| Outcome | Outcomes related to climate change, including climate change‑related morbidities or mortalities |
| Study design | Primary studies/original research (qualitative and quantitative) that used or investigated the impact of heat on health outcomes |
Table 2
Inclusion and exclusion criteria for the scoping review on heat and health in Gauteng.
| INCLUSION | EXCLUSION |
|---|---|
| People of all ages | Animals |
| Weather‑related elevated temperatures, including heatwaves and hot days | Non‑weather‑related heat such as body temperature, incubator temperature |
| Short‑term and long‑term exposure | No reference to heat exposure |
| Outdoor ambient temperature | Indoor temperature |
| Reference to air pollution/air quality that exacerbates heat | Air pollution/air quality but no reference to heat |
| Temperature, relative humidity, and other heat metrics | Exclusively studies with relative humidity or precipitation |
| Physical health: respiratory illnesses, cardiovascular issues, kidney conditions, and other illnesses; birth‑related outcomes from pregnancy such as preterm birth, stillbirth, low birth weight, and birth defects | Studies on heat but no health‑related outcomes |
Figure 1
Methodological framework for the benchmarking exercise.
Figure 2
PRISMA diagram for this study.
Table 3
Summary of descriptive findings of the 36 included studies in this scoping review.
| FIRST AUTHOR, YEAR OF PUBLICATION | STUDY DESIGN, PLACE | METHODS APPLIED IN THE STUDY | FINDINGS |
|---|---|---|---|
| Brimicombe, 2024 [22] | Observational, Africa | Pooled time‑series analysis Risk ratio for mortality rates | The influence of extreme heat on mortality risk in children under five years varies by age group, region, and season. There was a significant increase in mortality risk for children in eastern Africa, Senegal, and The Gambia. The study highlights the importance of considering seasonal and regional variations when assessing heat exposure effects. |
| Buhler, 2022 [23] | Observational, Limpopo | Distributed lag, non‑linear model | Both warm and cold apparent temperatures are associated with an increase in cardiovascular disease hospital admissions, with a cumulative effect over 21 days. The effect of warm temperatures is immediate and short‑lived, lasting two to four days, while the effect of cold temperatures is more inconsistent. Cold temperatures account for a larger fraction (8.5%) of CVD admissions compared to warm temperatures (1.1%). |
| Burkart, 2021 [24] | Observational, global | Estimated population attributable fractions for high and low temperatures | The study estimated that 1.69 million deaths were attributable to non‑optimal temperature globally in 2019. Cold‑attributable mortality generally exceeded heat‑attributable mortality, with regional variations in burden. High heat‑attributable burdens were observed in South and Southeast Asia, sub‑Saharan Africa and North Africa, and the Middle East, while high cold‑attributable burdens were seen in Eastern and Central Europe and Central Asia. |
| Gates, 2019 [25] | Observational, South Africa | Time‑stratified case‑crossover | A 1°C increase in same‑day maximum temperature was associated with a 1.5% increase in definite homicides and a 1.2% increase in total homicides. Significant positive associations were observed across different temperature metrics and lags. The temperature–homicide association was linear, with no clear non‑linearities. |
| Hugo, 2023 [19] | Observational, mixed methods, Pretoria | Indoor monitoring Local weather data Surveys, reflexive diaries | Informal dwellings in Melusi, Tshwane, South Africa, perform poorly due to endogenous factors, leading to extreme heat stress conditions for 6 to 10 h daily during peak summer. The dwellings’ spatial and material conditions are similar, with limited thermal control strategies, resulting in high indoor temperatures. The study highlights the need for significant adaptations to mitigate heat stress but notes that such adaptations may be socially and economically unfeasible. |
| Kapwata, 2022 [26] | Observational, South Africa | Diurnal temperature range threshold defined heatwave Calculated hot days per district | The highest frequency of heatwaves in South Africa occurred during the austral summer, with 150 events out of 270 from 2014 to 2019. Future climate projections indicate an 80%–87% increase in heatwave events during summer months from 2020 to 2039. The study highlights the need for targeted heat‑health adaptation strategies by identifying provinces and towns with intense, long‑lasting heatwaves. |
| Kapwata, 2021 [27] | Observational, Limpopo | Wavelet transform analysis | The study found that changes in air quality precondition the prevalence of pneumonia, with a time delay of 10 to 15 days. Malaria in South Africa is a multivariate event initiated by the co‑occurrence of high temperature and rainfall, with a time delay of 30 days. The findings have relevance for early warning system development and climate change adaptation planning to protect human health and well‑being. |
| Kapwata, 2024 [28] | Observational, South Africa | Distributed lag, non‑linear modeling | Maximum temperature was identified as the most statistically significant predictor of all‑cause mortality in 40 out of 52 districts in South Africa. The spatial distribution of maximum temperature thresholds varied by climate zone, with higher thresholds in hotter regions indicating population adaptation to local climate. The study recommends using maximum temperature as a key metric for heat‑health warning systems due to its strong association with mortality and adaptability to local climate conditions. |
| Lokotola, 2020 [29] | Case‑crossover, Cape Town | Time‑stratified case‑crossover Distributed lag, non‑linear modeling | The 15–64‑year‑old group was more at risk for CVD hospitalization with increasing air pollution levels on warm and cold days compared to other age groups. Females were more at risk from PM10, while males were more vulnerable to NO2 and SO2. Temperature was identified as a modifier of the effects of air pollution on CVD hospital admissions, with varying impacts across different population subgroups. |
| Mathee, 2009 [30] | Qualitative, Johannesburg and Upington | Focus group discussions Interviews Thematic analysis | Workers in Upington experienced heat‑related health effects such as sunburn, sleeplessness, irritability, and exhaustion, which affected their work levels and output. Few measures were taken by employers to protect workers’ health or improve their comfort. The study highlights the need for further research to quantify the effects of outdoor work on workers due to its public health importance. |
| McElroy, 2021 [31] | Pooled time‑stratified case‑crossover, global | Pooled time‑stratified case‑crossover design DLNM | Higher maximum temperatures and smaller diurnal temperature ranges in the week before birth increased the risk of preterm birth and stillbirth. The risk of preterm birth increased with exposure to extreme heat within seven days before birth, indicating an immediate effect. A temperature range of 20°C–30°C was identified where the risk of stillbirth is heightened, indicating a threshold beyond which the risk increases. |
| McMichael, 2025 [32] | Observational, global | Modeling | Temperature is associated with daily mortality in cities across different climates, with increased risk at both high‑ and low‑temperature extremes. Cities with hotter summers have higher heat thresholds, reflecting population adaptation to local climate conditions. Populations in low‑ and middle‑income countries are adversely affected by high temperatures and may be especially vulnerable to future climate change impacts. |
| Mfamadi, 2023 [33] | Cross‑sectional, Johannesburg | Temperature measurements Questionnaire | Median indoor apparent temperature in shipping container units ranged from 6°C to 42°C, indicating potential heat‑health risks. All units experienced heat‑health risks, with some exceeding danger thresholds, and participants reported symptoms like headaches and fatigue. Insulation may not be adequate for maintaining thermal comfort, suggesting a need for alternative solutions to mitigate heat‑related health impacts. |
| Moodley, 2024 [34] | Observational, KwaZulu‑Natal | Population‑based cohort analysis | Each additional hot day during the month preceding conception was associated with 26% higher odds of miscarriage. No significant association was found between heat exposure during the week preceding the outcome and miscarriage. The relationship between heat exposure and miscarriage was J‑shaped, with most miscarriages occurring within the first four weeks post‑conception. |
| Manyuchi, 2022 [10] | Observational, Agincourt | Interviews | Heat exposure has serious effects on morbidity, particularly affecting vulnerable populations such as the elderly and children. Future health awareness campaigns are essential to reduce vulnerability, morbidity, and mortality. The study provides location‑specific data for an area expected to experience significant heat stress, underscoring the need for heat‑health interventions. |
| Mwase, 2022 [35] | Case‑crossover, Vaal Triangle Air Pollution Priority Area | Time‑stratified case‑crossover | The association between respiratory disease hospitalizations and air pollution in the Vaal Triangle Air Pollution Priority Area is modified by apparent temperature (Tapp). Moderate Tapp levels worsen the effects of PM2.5, PM10, SO2, and BC, while high Tapp levels are more associated with increased effects of NO2 and O3. The elderly and females are more vulnerable to air pollution, especially on days with moderate Tapp levels. |
| Naicker, 2017 [36] | Cross‑sectional, Johannesburg | Indoor monitoring Questionnaire | There was a significant difference in monthly mean indoor temperature between different housing types, with low‑cost government‑built houses and informal settlement houses experiencing the greatest variation. These housing types were warmer than outdoor temperatures by 4°C–5°C, indicating a higher risk for indoor temperature extremes. The presence of ceilings significantly affected indoor temperature stability, with dwellings without ceilings being more sensitive to ambient temperature changes. |
| Ncongwane, 2021 [37] | Observational, multiple | Heat stress indicators using apparent temperature | The study found positive trends in heat stress in 88% of the weather stations, with approximately 47% being statistically significant at a 5% level. High annual climatological median values (>32°C) were observed at 42 stations, mostly in low‑altitude regions along coastlines. The study identified regions in South Africa susceptible to heat stress, particularly low‑altitude areas in provinces like MP, KZN, LP, WC, NC, and EC. |
| Nhamo, 2025 [38] | Observational, South Africa | In‑depth interviews Thematic analysis | Droughts, extreme heat, wildfires, and floods are the most prevalent climate‑induced weather extremes in South African national parks. These weather extremes lead to loss of biodiversity, wildlife stress, damage to infrastructure, reduced tourist arrivals, and loss of revenues. Park managers adopt strategies such as building back better infrastructure, creating fire breaks, and implementing water conservation measures to combat these impacts. |
| Olutola, 2023 [39] | Observational, Cape Town | Time‑stratified case‑crossover Effect modification by temperature | The elderly and females are more vulnerable to air pollutants, especially at high and moderate apparent temperature levels. NO2 has stronger effects on CVD mortality compared to PM10 and SO2. Harvesting effects were observed at longer lags, indicating decreased risk over time. The results can be used to develop an early warning system for CVD mortality in Cape Town. |
| Orlov, 2024 [40] | Observational, global | Model simulations | The study found that under the RCP1.9 scenario, there is a reduction in total excess mortality by the end of the century compared to 1980–1989, primarily due to a decrease in cold‑related deaths. The results indicate limited net negative impacts of temperature‑related mortality under RCP1.9, with some regions experiencing a marginal net positive benefit. The study highlights substantial differences in temperature responses across the three Earth system models used. |
| Part, 2022 [41] | Observational, Johannesburg | Time‑to‑event | High temperatures in early pregnancy, particularly during the third and fourth weeks, increase the risk of pre‑eclampsia, eclampsia, and HELLP syndrome. Low temperatures during the third trimester are associated with an increased risk of high blood pressure, hypertension, and gestational hypertension. High temperatures in early pregnancy likely affect placental development, increasing the risk of severe hypertensive disorders. |
| Rother, 2019 [42] | Observational, Western Cape | Focus group discussions | Forestry workers in South Africa experience health impacts from heat and sun exposure, using local coping mechanisms like wearing ochre for protection. The study identified gaps in current protective measures and suggested improvements such as providing individual water bottles, sunscreen, and flexible work hours. Workers initially did not differentiate between sun and heat exposure, but through discussions, they began to understand the differences. |
| Scovronick, 2012 [43] | Observational, Western and Eastern Cape | Distributed lag, non‑linear model | Future mortality burdens would be lower under a policy scenario prioritizing the replacement of informal housing over traditional dwellings. A maximum protection scenario with formal wealthy housing could reduce temperature‑related mortality by over 50%, approximately 5000 deaths annually. Well‑planned housing policies can reduce the future burden of temperature‑related mortality, emphasizing the need to improve low‑cost housing thermal performance. |
| Scovronick, 2018 [44] | Observational, South Africa | Distributed lag, non‑linear model | The study found an association between daily maximum temperature and mortality, with increased risks at both cold and hot temperatures. Overall, 3.4% of deaths in South Africa were attributed to non‑optimum temperatures, with the majority due to cold temperatures. The strongest associations were observed in the youngest and oldest age groups, as well as for cardiorespiratory causes. |
| Teare, 2020 [45] | Observational, Port Elizabeth, and Johannesburg | Temperature measurements Questionnaire | Upper respiratory tract symptoms were significantly associated with reports of high levels of indoor dust. Lower respiratory tract symptoms were significantly associated with low income, overcrowding, and having a young child in the household. |
| Wen, 2024 [46] | Observational, global | Quasi‑Poisson time‑series Multi‑level analysis | Intra‑day temperature variability (TV) is associated with a higher mortality risk than inter‑day TV for all‑cause, cardiovascular, and respiratory mortality. The mortality risk increases by 0.59% for all‑cause, 0.64% for cardiovascular, and 0.65% for respiratory mortality per IQR increase in intra‑day TV 0–7. Intra‑day TV accounts for 1.45% of all‑cause deaths, while inter‑day TV accounts for 0.35%. |
| Wichmann, 2017 [47] | Case‑crossover, Cape Town, Durban, and Johannesburg | Time‑stratified Meta‑analysis | A significant increase in mortality was observed when apparent temperature exceeded city‑specific thresholds. The elderly (≥ 65 years) were more at risk, with a significant increase in mortality per IQR increase in Tapp. The observed risks are similar to those in other countries, indicating that South Africa’s population is not more vulnerable to heat‑related mortality. |
| Wright, 2017 [48] | Observational, Limpopo | Indoor monitoring Questionnaire | Mean monthly indoor temperatures in clinics were higher during summer months, with the highest mean temperature of 31.4 ± 2.7°C recorded in February 2016. Indoor temperatures were 2°C–4°C higher than outdoor temperatures on average, with apparent temperatures suggesting discomfort and potential health risks. The study supports the need for controlled temperatures, reduced waiting times, and other measures to mitigate heat‑related health impacts in clinics. |
| Wright, 2017 [49] | Observational, Durban | Cross‑sectional questionnaire | Glare from the sun and excessive sweating were commonly reported sun‑related health symptoms among informal workers. The use of protective clothing was more prevalent among those who perceived sun exposure as a hazard. Informal workers implemented their own protective measures, such as portable shade and protective clothing, due to high sun exposure. |
| Wright, 2019 [50] | Observational, Johannesburg | Cross‑sectional questionnaire | Individuals with hypertension and those over 60 years old are more likely to experience heat‑related health effects. Living in government‑sponsored detached housing and houses with asbestos roofs increases the risk of heat‑related health effects. Targeted awareness campaigns for individuals with pre‑existing diseases and the elderly, and improvements in housing infrastructure, are recommended to mitigate heat‑related health impacts. |
| Wright, 2025 [51] | Observational, Durban | Mixed methods Temperature monitoring Questionnaire | Temperatures inside minibus taxis reached up to 39°C, significantly warmer than outdoors, and were above 27°C for extended periods, posing a threat to human health. Taxi drivers reported feeling hot and used water to cool down, indicating subjective heat stress. The study suggests that high temperatures could lead to dehydration and heat‑related illnesses among drivers, necessitating mitigation strategies due to climate change. |
| Wu, 2022 [52] | Two‑stage time series, observational, global | Temperature variability calculation Attributable fraction | The study found that increased temperature variability (TV) is associated with higher heat‑related mortality risks, with attributable fractions ranging from 0.70% to 2.73% for different quartiles of TV. TV has a significant modification effect on the heat‑mortality association, leading to a higher mortality burden with increased TV. The combination of heat exposure and high TV significantly increases mortality risk during the warm season, with disparate geographical variations in the impact of TV. |
| Wu, 2024 [53] | Observational, global | Two‑stage time series Multilevel meta‑analysis | The study observed a consistent decrease in mortality risk as the frequency of temperature increases, indicating adaptation to local climate through frequent exposure. The average increase in mortality risk from the 10th to the 100th percentile of normalized frequency was 13%, with significant regional differences. The increase in mortality risk varied by temperature component, with extreme cold showing the highest increase, followed by extreme hot, moderate cold, and moderate hot. |
| Zhu, 2023 [54] | Cross‑sectional, sub‑Saharan Africa | Predictive modeling | The study found that each 1°C increment in annual temperature was associated with increased odds of anemia in children. The burden of childhood anemia attributable to climate warming is projected to increase significantly under a high‑emission scenario by the end of the 21st century. Regional disparities were observed, with Central Africa expected to bear the highest burden of childhood anemia due to global warming. |
| Zhao, 2024 [55] | Observational, global | Modeling framework, distributed lag non‑linear model | Heatwave‑related excess deaths accounted for 0.94% of global deaths during the warm seasons of 1990 to 2019, equating to 236 deaths per 10 million residents. The global heatwave‑related excess death ratio remained relatively constant, while the excess death rate declined by 7.2% per decade. The findings indicate the potential benefit of governmental actions to enhance health sector adaptation and resilience, accounting for inequalities across communities. |
Table 4
Benchmark matrix of HAP elements that Gauteng Province should consider adopting for their own HAP.
| REVIEWED LOCATIONS | ||||
|---|---|---|---|---|
| DIMENSION AND LOCATION INFORMATION | MAHARASHTRA | VICTORIA | KHYBER PAKHTUNKHWA | IMPLICATION FOR GAUTENG |
| Lead and roles | State Disaster Management Plan with line‑department playbooks | Health‑led and emergency management integration | Provincial Department Management Authority declaration and Emergency Operations Centre protocols | Gauteng Department of Health lead; Provincial Disaster Management Centre co‑lead; formal memorandum of understandings with metros and SAWS |
| Triggers | Indian Meteorological Department anomaly‑based (graded) | District‑specific forecast mean thresholds | Heat index and declaration steps | District‑specific thresholds with SAWS, published in advance |
| Early warning | Statewide cascade; media/toolkits | Pre‑scripted alerts to sectors and public | Provincial Department Management Authority alerts; inter‑agency activation | WhatsApp/SMS and radio cascade; signage at clinics/taxi ranks |
| Health system | Heat illness surveillance and facility actions | Surge plans; checklists for services | Standard Operating Procedures across health and rescue | Emergency Department triage, IV fluids, electrolytes stock, staff rostering for hot days |
| Community measures | Public advice, hydration, shade | Local government activation and outreach | Cooling camps during heat events | Pop‑up cooling points/water points |
| Equity | Vulnerable district targeting | Targeted messaging to services | Community‑level access points | Prioritize informal settlements, elderly, outdoor workers |
| Monitoring and evaluation | State Guidance via National Centre for Disease Control | Annual plan refresh | Post‑season reviews implied | After‑action review each season; track heat‑related illnesses, mortality, alert reach |
| Long‑term | Heat‑resilient infrastructure and urban planning linkages (per global guidance) | Integrated with broader risk management | Standard Operating Procedures and risk analysis baseline | Trees, cool roofs, sharing at clinics and schools, aligning with the Sustainable Development Goals and Sendai Framework |