COVID deaths in South Africa: 99 days since South Africa's first death

Abstract

Background:
Understanding the pattern of deaths from COVID-19 in South Africa (SA) is critical to identifying individuals at high risk of dying from the disease. The Minister of Health set up a daily reporting mechanism to obtain timeous details of COVID-19 deaths from the provinces to track mortality patterns.

Objectives:
To provide an epidemiological analysis of the first COVID-19 deaths in SA.

Methods:
Provincial deaths data from 28 March to 3 July 2020 were cleaned, information on comorbidities was standardised, and data were aggregated into a single data set. Analysis was performed by age, sex, province, date of death and comorbidities.

Results:
SA reported 3 088 deaths from COVID-19, i.e. an age-standardised death rate of 64.5 (95% confidence interval (CI) 62.3 - 66.8) deaths per million population. Most deaths occurred in Western Cape (65.5%) followed by Eastern Cape (16.8%) and Gauteng (11.3%). The median age of death was 61 years (interquartile range 52 - 71). Males had a 1.5 times higher death rate compared with females. Individuals with two or more comorbidities accounted for 58.6% (95% CI 56.6 - 60.5) of deaths. Hypertension and diabetes were the most common comorbidities reported, and HIV and tuberculosis were more common in individuals aged <50 years.

Conclusions:
Data collection for COVID-19 deaths in provinces must be standardised. Even though the data had limitations, these findings can be used by the SA government to manage the pandemic and identify individuals who are at high risk of dying from COVID-19.

Full text

1 Published online ahead of print
RESEARCH
This open-access article is distributed under
Creative Commons licence CC-BY-NC 4.0.
COVID deaths in South Africa:
99 days since South Africa’s rst death
V Pillay-van Wyk,1 PhD, Co-Director; D Bradshaw,1 DPhil, Chief Specialist Scientist; P Groenewald,1 MB ChB, MPH, Specialist Scientist;
I Seocharan,2 BTech, IT Specialist; S Manda,3 PhD, Director; R A Roomaney,1 MPH, Senior Scientist; O Awotiwon,1 MSc, Senior Scientist;
T Nkwenika,3 MSc, Junior Statistician; G Gra y,4 MB BCh, FC Paed (SA), President and CEO; S S Buthelezi,5 MB ChB, Director-General of
Health; Z L Mkhize,5 MB ChB, Minister of Health
1 Burden of Disease Research Unit, South African Medical Research Council, Cape Town, South Africa
2 Biostatistics Unit, South African Medical Research Council, Durban, South Africa
3 Biostatistics Unit, South African Medical Research Council, Pretoria, South Africa
4 South African Medical Research Council, Cape Town, South Africa
5 National Department of Health, Pretoria, South Africa
Corresponding author: V Pillay-van Wyk (victoria.pillayvanwyk@mrc.ac.za)
Background. Understanding the pattern of deaths from COVID-19 in South Africa (SA) is critical to identifying individuals at high risk
of dying from the disease. The Minister of Health set up a daily reporting mechanism to obtain timeous details of COVID-19 deaths from
the provinces to track mortality patterns.
Objectives. To provide an epidemiological analysis of the first COVID-19 deaths in SA.
Methods. Provincial deaths data from 28 March to 3 July 2020 were cleaned, information on comorbidities was standardised, and data were
aggregated into a single data set. Analysis was performed by age, sex, province, date of death and comorbidities.
Results. SA reported 3 088 deaths from COVID-19, i.e. an age-standardised death rate of 64.5 (95% confidence interval (CI) 62.3 - 66.8)
deaths per million population. Most deaths occurred in Western Cape (65.5%) followed by Eastern Cape (16.8%) and Gauteng (11.3%). The
median age of death was 61 years (interquartile range 52 - 71). Males had a 1.5 times higher death rate compared with females. Individuals
with two or more comorbidities accounted for 58.6% (95% CI 56.6 - 60.5) of deaths. Hypertension and diabetes were the most common
comorbidities reported, and HIV and tuberculosis were more common in individuals aged <50 years.
Conclusions. Data collection for COVID-19 deaths in provinces must be standardised. Even though the data had limitations, these findings
can be used by the SA government to manage the pandemic and identify individuals who are at high risk of dying from COVID-19.
S Afr Med J. Published online 30 September 2020. https://doi.org/10.7196/SAMJ.2020.v110i11.15249
e article in context
South Africa (SA) has a quadruple burden of disease, which includes a high burden of HIV and tuberculosis. The impact of COVID-19 was
unclear with our disease burden. We searched PubMed on 5 August 2020 for articles published from December 2019 until that date, using the
following search string: ((“coronavirus”[MeSH Terms] OR “coronavirus”[Title] OR “2019-nCoV”[Title] OR “2019nCoV”[Title] OR “COVID-
19”[Title] OR “SARS-CoV-2”[Title]) AND (“Africa”[MeSH Terms]) AND “humans”[MeSH Terms] AND (“2019/12”[Date - Publication]:
“3000”[Date - Publication])). Filters used: Books and Documents, Meta-Analysis, Review, Systematic Review. We excluded clinical trials
and randomised controlled trials. No language restrictions were applied. Twelve studies were identified. An additional evidence-based brief
on COVID-19 and non-communicable diseases was identified; however, none of the studies reported on COVID-19 age-specific and age-
standardised death rates, sex mortality ratios and distribution of comorbidities among individuals who died from COVID-19 for SA.
Our study reports the pattern of COVID-19 deaths by age, sex, province and comorbidities for SA for the period 28 March 2020 to 3 July
2020. Even though our finding of males having higher death rates than females has been seen in other countries, the male/female ratio is not as
high in SA as in some countries; the continued increase in the male-female mortality ratio in individuals aged ≥70 years was also not observed
in other countries. At the current stage of the pandemic, the COVID-19 male/female mortality ratio had a similar pattern to the all-cause
male/female mortality ratio for SA, with the exception of individuals aged ≥70 years, who experienced an even higher COVID-19 mortality
differential than the usual all-cause. Most of our deaths occurred in individuals with comorbidities, and 58% had two or more comorbidities.
Hypertension and diabetes were the most common comorbidities reported.
The pattern of COVID-19 deaths in SA may be different from patterns in other countries. However, this may change as the pandemic
progresses. Our findings have health systems and public health implications, as they identify individuals at high risk of dying from COVID-
19, and this information can be used to manage the pandemic. Even though we observed differential patterns of COVID-19 deaths by sex,
age, comorbidities and province, individuals with hypertension and diabetes are at high risk of dying from COVID-19 in SA and should be
managed carefully when they test positive for COVID-19. Furthermore, more stringent social distancing measures should be promoted to
reduce transmission of COVID-19 among those who have hypertension and diabetes, particularly those who have challenges with controlling
their levels of blood pressure and glucose.
Our study has identified areas in the data collection process that should be improved, and challenges with the reporting of cause of death
information in SA. Addressing these shortcomings will benefit the health system beyond the pandemic.

2 Published online ahead of print
RESEARCH
South Africa (SA) is one of the first countries on the African
continent to tackle COVID-19 head-on. Our country is different
from the developed countries that have already faced the full force
of the pandemic, due to its quadruple burden of disease of HIV/
AIDS and tuberculosis (TB); other communicable diseases, perinatal
conditions, maternal causes and nutritional deficiencies; non-
communicable diseases; and injuries.[1,2] These factors have left much
uncertainty about how the pandemic would progress in SA. COVID-
19 reached SA about 6 weeks after the outbreak in Europe[3,4] and on
11 March 2020 the World Health Organization (WHO) characterised
COVID-19 as a pandemic.[5] The SA government, guided by the
WHO and the progress of the outbreak in other parts of the world,
locked down the country to give provinces time to prepare for the
onslaught of COVID-19. On 27 March 2020, we heard the sad news
that SA had experienced its first death from COVID-19.[6] By the end
of April there had been 103 deaths,[7] which increased to 683 deaths
by the end of May[8] and to over 2 657 deaths by the end of June.
[9]
Due to the time lag for the national cause of death data based on
death notifications to be processed, the Minister of Health has set up
a daily reporting mechanism to obtain timeous details of COVID-19
deaths from the provinces to track the progression of the pandemic.
Objectives
This article reports on COVID-19 deaths for SA and its nine
provinces for the first 99 days since our first COVID-19 death.
Methods
The COVID-19 death reports provided to the National Department of
Health (NDoH) by each province between 28 March and 3July 2020
were used in this analysis. The data were provided in an unstandardised
format from each province in either PDF documents or Microsoft Excel
workbooks. The variables and clinical information provided by each
province varied. In data sets with identifiers, duplicate records were
identified and removed. Twenty-four duplicates were identified and
removed from the data, 23 from Gauteng and one from Western Cape.
It was not possible to identify whether there were any duplicate cases in
Eastern Cape, as no identifiers were provided. The clinical data provided
were interrogated in order to identify comorbidities, which were then
recorded in a standard format for each individual (AnnexureA, http://
samj.org.za/public/sup/15249-1.pdf). Comorbidities were reported as a
mix of full disease names and abbreviated names, sometimes including
conditions that were a consequence of COVID-19 itself rather than
comorbidities that the individual had prior to becoming ill with
COVID-19. Variables that were common across the provinces were
collated into a single data set and included date of death, age, province
and comorbidities. The data set was checked for missing data (Annexure
B, http://samj.org.za/public/sup/15249-2.pdf). Age was categorised into
5-year age bands, 10-year age bands and broad age bands <50 years,
50- 69 years and ≥70 years.
Statistical analysis was conducted using Microsoft Excel (Microsoft
Corp., USA) and Stata 15 (StataCorp, USA).
Summary statistics involved expressing continuous data as
medians and interquartile ranges (IQRs) for age distribution across
sex, province and comorbidities. Discrete or categorical data such as
comorbidities were summarised using frequencies and percentages
with the associated 95% confidence intervals (CIs) by age group,
sex, province and comorbidities. Associations between presence or
absence of a specific comorbidity and age group, sex and province
were quantified by χ2 tests and p-values, using a cut-off value of 0.05
for a statistically significant association. Death rates and associated
95% CIs[10] were calculated using mid-year population estimates
from the Thembisa model[11] and standardised using the WHO world
standard population.[12]
Results
A total of 3 088 deaths were provided from the nine provinces for
the period 28 March - 3 July 2020. Table 1 shows the characteristics
of individuals who were reported to have died from COVID-19. The
median (IQR) age was 61 (52 - 71) years. The age was similar for
both sexes. Only 20.3% of the deaths were in individuals aged <50
years, and 29.4% of those who died were aged ≥70 years. There were
six deaths reported in children aged <10 years, accounting for 0.2%
of the total deaths.
Table 1. Characteristics of reported COVID-19 deaths in
South Africa, 3 July 2020
Characteristics n%(95% CI)
Age (years),
median (IQR)*
Sex
Male 1602 52.0 (50.1 - 53.6) 61 (52 - 71)
Female 1486 48.0 (46.4 - 49.9) 62 (52 - 72)
Tot al 3088 100 - 61 (52 - 71)
Age categories
(years), 10-year
bands
0 - 9 6 0.2 (0.1 - 0.4) -
10 - 19 7 0.2 (0.1 - 0.5) -
20 - 29 48 1.6 (1.2 - 2.1) -
30 - 39 189 6.2 (5.4 - 7.1) -
40 - 49 372 12.1 (11.0 - 13.3) -
50 - 59 744 24.3 (22.8 - 25.8) -
60 - 69 799 26.0 (24.5 - 27.6) -
70 - 79 562 18.3 (17.0 - 19.7) -
≥80 341 11.1 (10.0 - 12.3) -
Tot a l 3068 100 -
Age categories
(years), broad
age bands
<50 622 20.3 (18.9 - 21.7) -
50 - 69 1543 50.3 (48.5 - 52.1) -
≥70 903 29.4 (27.8 - 31.1) -
Tot al 3068 100 -
Province
Eastern Cape 519 16.8 (15.5 - 18.2) 60 (52 - 68)
Free State 19 0.6 (0.4 - 1.0) 68 (55 - 76)
Gauteng 348 11.3 (10.2 - 12.4) 63 (51 - 73)
KwaZulu-Natal 98 3.2 (2.6 - 3.9) 63 (52 - 73)
Limpopo 31 1.0 (0.7 - 1.4) 56 (42 - 66)
Mpumalanga 7 0.2 (0.1 - 0.5) 45 (39 - 52)
North West 36 1.2 (0.8 - 1.6) 58 (51 - 67)
Northern Cape 6 0.2 (0.1 - 0.4) 59.5 (55 - 71)
Western Cape 2024 65.5 (63.8 - 67.2) 62 (52 - 72)
Tot al 3088 100 - 61 (52 - 71)
Comorbidity
None 102 4.2 (3.4 - 5.0) 56 (45 - 65)
≥1 2355 95.8 (95.0 - 96.6) 62 (53 - 71)
Tot a l 2457
(100)
-61 (52 - 71)
CI = confidence interval; IQR = interquartile range.
*Median age and IQR not calculated for all groups.

3 Published online ahead of print
RESEARCH
Provinces with <100 deaths were combined due to small numbers
and presented as ‘other provinces’. Western Cape reported the highest
number of deaths for SA (65.5%), followed by Eastern Cape (16.8%)
and Gauteng (11.3%). The six other provinces collectively reported
6.4% of the deaths for SA (Fig. 1A).
Fig. 1B shows the increase in the number of deaths over time.
March and April were combined due to small numbers, and June
and July were combined as deaths for July only included deaths from
1 July to 3 July. There was an almost four-fold increase in deaths
between May and June-July. Fig. 1C shows the age-standardised
death rate; this measure removes the effect of the age structure
of the population and allows for direct comparison of death rates
across provinces. Western Cape had the highest age-standardised
death rate, followed by Eastern Cape and Gauteng.
The crude death rate for Western Cape was 288.9 deaths per
million population (95% CI 276.3 - 301.5), that for Eastern Cape was
77.1 deaths per million population (95% CI 70.4 - 83.7) and that for
Gauteng was 23.8 deaths per million population (95% CI 21.3 - 26.2).
SA had a crude death rate of 52.1 deaths per million population (95%
CI50.3 - 53.9) (data not reported).
Fig. 2A shows the age pattern of deaths for the provinces. Even
though the age pattern appears similar across the provinces, there
were some differences, i.e. the highest number of deaths in Western
Cape was in 60 - 64-year-olds, in Eastern Cape it was in 55 - 59-year-
olds, and in Gauteng the majority of deaths were in 70 - 74-year-olds.
Fig. 2B shows age-specific death rates; this measure accounts for the
age structure and population size for each province. Age-specific
death rates increased dramatically with age for Western Cape from
≥35 years and for Gauteng from ≥70 years, with Eastern Cape
showing a different pattern (Fig. 2B).
Fig. 3 shows the age distribution of deaths for males and females
in SA. The highest percentage of deaths occurred in the 55 - 59- and
60 - 64-year age groups for both males and females (Fig. 3A). The
percentage of deaths of individuals aged <30 years was very low for
both males and females, i.e. 2.0% (61deaths). Age-specific death rates
increased as age increased for both males and females (Fig.3B). The
death rate was was higher in males compared with females in those aged
A. Provincial distribution of deaths (N=3 088)
B. Number of deaths over time (N=3 088)
2 500
2 000
1 500
1 000
500
0
Month
400
350
300
250
200
150
100
50
0
64.5
C. Age-standardised death rate by province and national (N=3 088)
South Africa
Western Cape
65.5%
Gauteng
11.3%
Eastern Cape
16.8%
Other provinces
6.4%
Deaths, n
March - April May June - July
Western Cape Eastern Cape Gauteng Other provinces
Death rate, per million population
87.0
30.3
324.4
8.5
Eastern Cape
Gauteng
Western Cape
Other provinces
Fig. 1. Provincial distribution and death rates of COVID-19 deaths, as at
3July 2020.
Eastern Cape Gauteng Western Cape Other provinces
Eastern Cape Gauteng Western Cape Other provinces
25
20
15
10
5
0
Age group (years)
3 500
3 000
2 500
2 000
1 500
1 000
500
0
A. Percentage of deaths by age group and province (N=3 068)
Deaths, %
Age group (years)
B. Age-specic death rates by province (N=3 068)
Death rate, per million population
0 - 4
5 - 9
10 - 14
15 - 19
20 - 24
25 - 29
30 - 34
35 - 39
40 - 44
45 - 49
50 - 54
55 - 59
60 - 64
65 - 69
70 - 74
75 - 79
80 - 84
85 - 89
≥90
0 - 4
5 - 9
10 - 14
15 - 19
20 - 24
25 - 29
30 - 34
35 - 39
40 - 44
45 - 49
50 - 54
55 - 59
60 - 64
65 - 69
70 - 74
75 - 79
≥80
Fig. 2. Age distribution and death rates of COVID-19 deaths by province, as
at 3 July 2020. For B, mid-year population estimates only available for age
≥80 years for the calculation of age-specic death rates.

4 Published online ahead of print
RESEARCH
≥40 years (Fig. 3C). As age increased, the COVID-19 mortality ratio
increased. Comparison of the COVID-19 mortality ratio with the all-
cause mortality ratio for the year 2012 from the Second National Burden
of Disease Study[2] for SA revealed that the COVID-19 mortality ratio
was higher in individuals aged ≥70 years (Fig.3C).
Fig. 4 shows the sex distribution of deaths for province and
national. The pattern of male and female deaths across all provinces
followed the national pattern, with more male deaths occurring
compared with females, except for Eastern Cape, which reported
more female deaths (Fig. 4A). However, once the effect of the
population structure was removed, the age-standardised death rates
showed that the male/female mortality ratio for Eastern Cape was
similar to the national pattern (Fig. 4B and C). The age-standardised
COVID-19 male/female mortality ratio was 1.5 compared with the
all-cause mortality ratio of 1.4 (data not reported).
Table 2 shows the distribution of co morbidities. In a relatively
high propor tion of the deaths, information about co morbidities was
missing or unknown (18.3% and 2.2%, respectively) (Annexure B,
http://samj.org.za/public/sup/15249-2.pdf). At least one co morbidity
was reported for 2 355 (95.8%) of individuals who died (Table 1).
Hypertension and diabetes were by far the most common
comorbidities among both males and females, occurring in more than
half of the national deaths (60.5%; 95% CI 58.5 - 62.4 and 54.9%; 95%
CI 52.9 - 56.9, respectively). Hypertension was significantly more
common in females (66.2%; 95% CI63.5 - 68.8) than males (55.0%;
95% CI52.3 - 57.8), and diabetes was more common in males (56.4%;
95% CI 53.6 - 59.1) than females (53.3%; 95% CI50.5 - 56.2). These
were followed by HIV (13.9%; 95% CI12.6 - 15.3), chronic respira-
tory conditions, i.e. asthma and chronic obstructive pulmonary
A. Percentage of deaths by sex for provinces and national (N=3 088)
48.1
70
60
50
40
30
20
10
0
Males Females
Deaths, %
45.9
54.1 54.9
45.1 52.1 47.9
59.9
40.1
51.9
Males Females
Eastern
Cape
Gauteng Western
Cape
Other
provinces
South Africa
Eastern
Cape
Gauteng Western
Cape
Other
provinces
South Africa
Eastern
Cape
Gauteng Western
Cape
Other
provinces
South Africa
500
400
300
200
100
0
Death rate, per million population
94.4 73.8 35.0 23.6
372.0
270.1
12.5 5.5
79.2 52.6
B. Age-standardised death rates by sex for provinces and national (N=3 088)
C. Ratio of male/female age-standardised death rates
2.5
2.0
1.5
1.0
0.5
0.0
1.5
Male/female mortality ratio
1.3 1.5 1.4
2.3
Fig. 4. Sex distribution of COVID-19 deaths by province and national, as
at 3 July 2020.
A. Percentage of deaths by age group and sex (N=3 068)
COVID-19 mortality ratio All-cause mortality ratio
25
20
15
10
5
0
Age group (years)
Deaths, %
Age group (years)
B. Age-specic death rates by sex (N=3 068)
Males Females
Males Females
1 000
800
600
400
200
0
Death rate, per million population
C. Ratio of male/female age-specic death rates
Male/female mortality ratio
Age group (years)
2.0
1.5
1.0
0.5
0.0
1.9
1.3
1.6
1.8 1.7
1.4 1.3
0.8 0.9
1.2
1.2
1.5 1.6
1.6
0 - 4
5 - 9
10 - 14
15 - 19
20 - 24
25 - 29
30 - 34
35 - 39
40 - 44
45 - 49
50 - 54
55 - 59
60 - 64
65 - 69
70 - 74
75 - 79
80 - 84
85 - 89
≥90
0 - 4
5 - 9
10 - 14
15 - 19
20 - 24
25 - 29
30 - 34
35 - 39
40 - 44
45 - 49
50 - 54
55 - 59
60 - 64
65 - 69
70 - 74
75 - 79
≥80
<30
30 - 39
40 - 49
50 - 59
60 - 69
70 - 79
≥80
Fig. 3. Age distribution of COVID-19 deaths by sex, as at 3 July 2020. For B
and C, mid-year population estimates only available for age ≥80 years for the
calculation of age-specic death rates. All-cause age-specic death rates were
calculated using deaths from the Second National Burden of Disease Study
for South Africa for 2012[2] and Dorrington[13] mid-year population estimates.

5 Published online ahead of print
RESEARCH
disease (13.3%; 95% CI 12.0 - 14.7) and chronic kidney disease
(13.3%; 95% CI 12.0 - 14.7). Cardiovascular diseases were reported
for 6.3% (95% CI 5.4 - 7.3), obesity for 5.5% (95% CI 4.7 - 6.5) and
cancers for 3.3% (95% CI 2.7 - 4.1) of the deaths. Other diseases
combined were reported for 9.1% (95% CI8.0- 10.3) of the deaths,
and eight females were pregnant (data not reported). Current TB
was reported in 3.3% (95% CI 2.6 - 4.0) of the deaths and differed
significantly by sex, with 4.0% (95% CI 3.0- 5.2) in males and 2.5%
(95% CI 1.8 - 3.6) in females.
There is a strong age pattern in the reported comorbidities. The
ranking of the five most common comorbidities reported for all ages
differed when disaggregated by broad age group. HIV and TB were
much higher among individuals aged <50 years (33.2%; 95% CI 29.1-
37.6 and 7.8%; 95% CI5.7 - 10.6, respectively), and chronic kidney
disease was much higher in individuals aged ≥70 years (19.2%;
95% CI16.4 - 22.3). Diabetes ranked highest among <50-year-olds
(41.2%; 95% CI 36.9- 45.7), followed by hypertension (34.2%; 95%
CI 30.1 - 38.7) and HIV (33.2%; 95% CI29.1- 37.6). Hypertension
ranked highest in both 50 - 69-year-olds (64.3%; 95% CI61.7- 66.9)
and ≥70-year-olds (71.2%; 95% CI 67.7 - 74.4), followed by diabetes
(62.4%; 95% CI 59.7 - 65.1 in 50 - 69-year-olds and 50.4%; 95%
CI46.7 - 54.1 in ≥70-year-olds).
Hypertension was the most common co- morbidity reported in
WesternCape(63.8%; 95% CI 61.4 - 66.2), Eastern Cape (58.6%; 95%
CI 53.8 - 63.3) and Gauteng (48.1%; 95% CI 42.6 - 53.6), followed
by diabetes (Table 2). The comorbidity propor tions were statistically
significantly different across provinces. Differences were observed
in the top five comorbidities across provinces; for example, Gauteng
and Eastern Cape reported cardiovascular disease in their top five
comorbidities. TB was more common in Western Cape than in the
other provinces (3.9%; 95% CI 3.1 - 5.0). Of the 167 deaths in which
previous TB was reported, 159 were from Western Cape (data not
reported). The findings on comorbidities must be interpreted with
caution, as the proportions could be influenced by the low proportion
of deaths with no comorbidities, which in turn could be related to the
high proportion of unknowns in the data.
Fig. 5 shows the number of comorbidities per individual where
comorbidity infor mation was available. Most individuals had either
one comorbidity (37.3%; 95% CI35.4 - 39.2) or two comorbidities
reported (36.3%; 95% CI 34.4 - 38.2). Only 4.2% (95% CI 3.4 - 5.0)
had no comorbidities. More than half (58.6%; 95% CI 56.6 - 60.5) of
the individuals who died had two or more comorbidities.
Fig. 6 shows the overall pattern of comorbidities, showing
how often each condition was reported. This includes multiple
comorbidities for individuals with more than one comorbidity.
Hypertension and diabetes accounted for >50% of the comorbidities
reported for most provinces except North West and Mpumalanga.
Table 2. Median age by comorbidity and prevalence of comorbidities by sex, age and province for COVID-19 deaths, as at 3 July 2020
Comorbidities*
Total, %
(N=2 457)
Age (years), median
(IQR) (N=2 447)
Sex, % Broad age groups (years), % Province, %
Male
(N=1 259)
Female
(N=1 198) p-value
<50
(N=476)
50 - 69
(N=1 270)
≥70
(N=704) p-value
Western
Cape
(N=1 587)
Eastern
Cape
(N=406)
Gauteng
(N=312)
Other
provinces
(N=152) p-value
Hypertension 60.5 64 (56 - 73) 55.0 66.2 <0.001 34.2 64.3 71.2 <0.001 63.8 58.6 48.1 55.9 <0.001
Diabetes 54.9 61 (54 - 70) 56.4 53.3 0.128 41.2 62.4 50.4 <0.001 60.3 51.0 36.9 46.1 <0.001
HIV 13.9 51 (41 - 59) 12.8 15.0 0.109 33.2 12.7 3.1 <0.001 16.3 6.9 11.9 11.2 <0.001
Asthma/COPD 13.3 65 (52 - 71) 12.7 13.9 0.369 9.7 12.8 16.8 0.002 16.1 8.1 9.0 7.2 <0.001
Chronic kidney disease 13.3 66.5 (57 - 75) 12.3 14.3 0.152 6.6 12.6 19.2 <0.001 16.5 4.7 9.0 11.2 <0.001
Cardiovascular disease 6.3 68.5 (56 - 76) 6.2 6.3 0.879 4.4 5.0 9.9 <0.001 4.0 7.4 11.2 16.4 <0.001
Obesity 5.5 56.5 (46.5 - 63) 5.2 5.9 0.408 9.5 5.8 2.4 <0.001 5.9 3.4 5.1 8.6 0.092
Cancer 3.3 68.5 (55 - 78) 3.8 2.8 0.142 2.7 2.5 5.0 0.010 2.1 3.2 7.7 6.6 <0.001
TB current 3.3 51 (39 - 61.5) 4.0 2.5 0.041 7.8 2.6 1.4 <0.001 3.9 2.0 1.9 2.6 0.102
Autoimmune disease 0.1 59 (51 - 76) 0.1 0.2 0.535 0 0.2 0.1 0.695 0.2 0 0 0 0.649
Other disease† 9.1 67 (56 - 78) 8.7 9.5 0.459 7.6 6.8 14.3 <0.001 7.3 9.4 15.4 13.8 <0.001
No comorbidity 4.2 56 (45 - 65) 5.2 3.1 0.010 7.6 3.7 2.6 <0.001 0.7 10.8 11.5 7.2 <0.001
IQR = interquartile range; COPD = chronic obstructive pulmonary disease; TB = tuberculosis.
*One individual can have more than one comorbidity.
†Other disease refers to external causes and medical conditions including mental illness (see Annexure A, http://samj.org.za/public/sup/15249-1.pdf).
45
40
35
30
25
20
15
10
5
0
5.5
Number of comorbidities reported
Deaths, %
4.2
37.3 36.3
16.8
None One Two Three Four or more
Fig. 5. Percentage of COVID-19 deaths by number of comorbidities
reported, as at 3 July 2020 (N=2457).

6 Published online ahead of print
RESEARCH
Table 3 reports on the top ten combina tions of comorbidities in
individuals. For those reported to have a comorbidity, the combination
of diabetes and hypertension (19.9%; 95% CI 18.3 - 21.6) was most
common, followed by hypertension only (13.4%; 95% CI 12.1 - 14.8)
and diabetes only (12.7%; 95% CI 11.5 - 14.1).
Discussion
Even though SA is still in the early stages of the COVID-19 pandemic,
it is important to track how the pandemic is progressing and
understand the pattern of deaths as they unfold. The SA government’s
swift actions to set up a daily reporting of deaths to the NDoH has
provided some understanding of the nature of the pandemic in SA.
Our analysis found that the various provinces in SA are at different
stages of the COVID-19 pandemic, with Western Cape being at a
much later stage than its counterparts, contributing to 65% of total
deaths with a death rate of 324.4 deaths per million population, a rate
that is five times higher than the national death rate of 64.5 deaths
per million population. The number of deaths peaked at 55 - 59 years
in Eastern Cape, 60 - 64 years in Western Cape and 75 - 79 years
in Gauteng. Most deaths occurred in individuals aged ≥30 years,
starting slightly younger than in other countries, which could be due
to SA having a younger population; SA’s median age is 27.6 years,
compared with Italy with 47.3 years and Spain with 44.9 years.[14] As
age increased, death rates increased. The age pattern observed for
Eastern Cape is different where their number of deaths peaked at a
younger age and no observable increase in death rates as age increases
is difficult to explain and requires further investigation.
Our finding that males had higher death rates than females is
similar to other countries.[15] However, the divergence of death rates
between males and females as age increased was not observed in
other countries. This could be indicative of the stage of the pandemic
in SA. We found that the COVID-19 male/female mortality ratio of
1.5 for SA is similar to the all-cause male/female mortality ratio of 1.4
for our country for 2012.[2] The age pattern of SA’s all-cause mortality
ratio was similar to that observed in other countries where the male/
female gap for death rates becomes smaller after 69years of age. The
male-female gap for COVID-19 death rates became bigger after 69
years of age. The COVID-19 male/female mortality ratio observed
for the provinces was similar to the national ratio.
For individuals where information was available, at least one
comorbidity was repor ted for 2 355/2 457 (95.8%) of the individuals
who died. More than half (58.6%) of the deaths occurred in individuals
with two or more comorbidities. Hypertension and diabetes were the
most common comorbidities in individuals who were reported to
have died from COVID-19; this finding is similar to what was found
in Europe.
[16] Hypertension and diabetes were the most common
comorbidities in individuals across provinces, and across the broad
age bands <50 years, 50 - 59 years, 60 - 69 years and ≥70 years, but the
prevalence of these conditions increased significantly with age. Our
findings that HIV and current TB were more common in <50-year-
olds[17] and chronic kidney disease in ≥70-year-olds have been reported
previously.
[17,18] The top five comorbidities did differ by province. At the
current stage of the COVID-19 pandemic, very few deaths have been
reported in children.
From a health services perspective, our analysis showed that most
individuals who died of COVID-19 presented with both hypertension
and diabetes, followed by hypertension only and then diabetes only.
This information is critical to identifying individuals who are at high
risk of dying from COVID-19 and may indicate the importance of
screening for and managing these conditions among people infected
with SARS-CoV-2.
While a large proportion of South Africans have HIV, TB or both
diseases, our analysis shows that these conditions were not the most
common comorbidities in individuals who died from COVID-19.
Table 3. Combinations of comorbidities for COVID-19 deaths (N=2 355), as at 3 July 2020
Comorbidities* n %(95% CI)
Diabetes and hypertension 469 19.9 (18.3 - 21.6)
Hypertension only 315 13.4 (12.1 - 14.8)
Diabetes only 300 12.7 (11.5 - 14.1)
HIV only 100 4.2 (3.5 - 5.1)
Diabetes and CKD and hypertension 100 4.2 (3.5 - 5.1)
Asthma/COPD and diabetes and hypertension 58 2.5 (1.9 - 3.2)
Asthma/COPD only 52 2.2 (1.7 - 2.9)
CKD and hypertension 49 2.1 (1.6 - 2.7)
Other disease†44 1.9 (1.4 - 2.5)
Asthma/COPD and hypertension 44 1.9 (1.4 - 2.5)
Diabetes and HIV 38 1.6 (1.2 - 2.2)
Hypertension and other disease†37 1.6 (1.1 - 2.2)
Diabetes and HIV and hypertension 35 1.5 (1.1 - 2.1)
CI = confidence interval; COPD = chronic obstructive pulmonary disease; CKD = chronic kidney disease.
*One individual can have more than one comorbidity.
†Other disease refers to external causes and medical conditions including mental illness (see Annexure A, http://samj.org.za/public/sup/15249-1.pdf).
100
80
60
40
20
0
Western Cape
Comorbidities, %
Eastern Cape
Free State
Gauteng
KwaZulu-Natal
Limpopo
Mpumalanga
North West
Northern Cape
Hypertension
Diabetes
Cardiovascular disease
Obesity
Asthma/COPD
Autoimmune disease
HIV
Chronic kidney disease
Cancer
TB current
Other
Fig. 6. Percentage of comorbidities by province for COVID-19 deaths, as at
3 July 2020.

7 Published online ahead of print
RESEARCH
However, as the number of deaths increases in the country, this pattern
could change.
An analysis of Western Cape COVID-19 deaths by Boulle et
al.[19] showed similar findings to our analysis for the province. They
investigated 625 public sector COVID-19 deaths that occurred before
1 June 2020. Their study highlighted that HIV and TB should not
be forgotten in the COVID-19 pandemic, as there was a 3.3 times
increased risk of dying from COVID-19 in individuals with TB and a
2 times increased risk of dying in individuals with HIV, when adjusted
for age and sex.[19] That said, their study found that the risk of dying
from COVID-19 was also higher in individuals with hypertension (2.2
times) and diabetes (ranging from 6.1 to 12.9 times higher depending
on the level of glucose control), when adjusted for age and sex.[19]
Study limitations
The study findings reflect the data that were provided and have
limitations. Many provinces did not provide information on the dates
of COVID-19 tests and the dates on which the results were received.
It was therefore not possible to distinguish between a confirmed
and probable COVID-19 death.[20] It was also unclear whether the
data reported on comorbidities are complete, i.e. whether all the
comorbidities have been reported, and whether missing data means
no comorbidities in the individuals or this was just not reported.
Furthermore, the impact of the inconsistent reporting of information
across provinces and the completeness of reporting has not been
investigated.
Bradshaw et al.[21] have shown, using deaths data from the
Department of Home Affairs, that more deaths are being reported
during the period of the pandemic compared with previous years,
which could be attributed directly or indirectly to COVID-19. Their
numbers are much higher than those reported in this study. This could
be due to more individuals dying at home from COVID-19 and being
missed in our analysis, or dying from other conditions at home because
they are afraid to attend a health facility. The differences in the number
of deaths require further investigation.
Conclusions
Our study provides important epidemio logical information on the
pattern of the current state of COVID-19 mortality in SA that, even
though it is not perfect, can be used by the SA government to identify
high-risk individuals. Differential patterns of COVID-19 deaths by
sex, age, comorbidities and provinces point to the need for targeted
and localised interventions. That said, individuals with hypertension
and diabetes should be given careful attention during the COVID-19
pandemic across SA.
Furthermore, data collection for COVID-19 needs to be standardised
across provinces, and systems to verify the completeness and accuracy
of the data should be put in place to ensure better reporting of
information on COVID-19 deaths. The timeliness of the availability of
cause of death information needs to be addressed. Firstly, SA’s NDoH
should have access to the cause of death information at the time of
death registration in order to monitor epidemics more accurately.
Secondly, efforts should be made to expedite the flow and processing
of cause of death information between the SA’s National Department
of Home Affairs and the national statistics office, Statistics South
Africa. Doing this will have long-term benefits beyond the COVID-19
pandemic.
Declaration. None.
Acknowledgements. We acknowledge all the healthcare workers who
collated the deaths data from the various provinces.
Author contributions. All authors contributed to the analysis, interpretation
of ndings and nalising the manuscript.
Funding.is work was funded by the South African Medical Research
Council.
Conicts of interest.None.
1. Hofman K, Madhi S. e unanticipated costs of COVID-19 to South Africa’s quadruple disease
burden. S Afr Med J 2020;110(8):698-699. https://doi.org/10.7196/SAMJ.2020.v110i8.15125
2. Pillay-van Wyk V, Ms emburi W, Laubscher R, et al. Mortality trends and dierentials in South
Africa from 1997 to 2012: Second National Burden of Disease Study. L ancet Glob Health
2016;4(9):e642-e653. https://doi.org/10.1016/S2214-109X(16)30113-9
3. World Health Organization. Timeline of WHO’s response to COVID-19. https://www.who.int/
news-room/detail/29-06-2020-covidtimeline (accessed 11 August 2020).
4. National Institute for Communicable Diseases. First case of COVID-19 announced – an update.
5 March 2020. https://www.nicd.ac.za/rst-case-of-covid-19-announced-an-update/ (accessed
19September 2020).
5. World Health Organization. WHO Director-General’s opening remarks at the media brieng
on COVID-19 – 11 March 2020. 11 March 2020. https://www.who.int/dg/speeches/detail/
who-director-general-s-opening-remarks-at-the-media-briefing-on-covid-19---11-march-2020
(accessed 11 August 2020).
6. National Department of Health, South Africa. Minister Zweli Mkhize conrms rst deaths due
to Coronavirus COVID-19. 27 March 2020. https://www.gov.za/speeches/minister-zweli-mkhize-
conrms-rst-deaths-due-coronavirus-covid-19-27-mar-2020-0000 (accessed 28 July 2020).
7. National Department of Health, South Africa. Minister Zweli Mkhize conrms total of 5 350 cases
of Coronavirus COVID-19. 29 April 2020. https://www.gov.za/speeches/minister-zweli-mkhize-
conrms-total-5-350-cases-coronavirus-covid-19-29-apr-2020-0000 (accessed 28 July 2020).
8. National Department of Health, South Africa. Minister Zweli Mkhize conrms total of 32 683 cases
of Coronavirus COVID-19. 31 May 2020. https://www.gov.za/speeches/minister-zweli-mkhizi-
conrms-total-32-683-cases-coronavirus-covid-19-31-may-2020-0000 (accessed 28 July 2020).
9. National Department of Health, South Africa. Minister Zweli Mkhize conrms total of 151 209 cases
of Coronavirus COVID-19. 30 June 2020. https://www.gov.za/speeches/minister-zweli-mkhize-
conrms-total-151-209-cases-coronavirus-covid-19-30-jun-2020-0000 (accessed 28 July 2020)
10. Keytz N. Sampling variance of standardized mortality rates. Hum Biol 1966;38(3):309-317.
11. Johnson LF, Dorrington RE. embisa version 4.2: A Model for Evaluating the Impact of HIV/AIDS
in South Africa. Cape Town: University of Cape Town, 2019. https://www.thembisa.org/content/
ledl/embisa4_2report (accessed 11 August 2020).
12. Ahmad OB, Boschi Pinto C, Lopez AD. Age standardization of rates: A new WHO standard. GPE
Discussion Paper Series: No 31. 2001:10-12. https://www.who.int/healthinfo/paper31.pdf (accessed
11August 2020).
13. Dorrington R. Alternative South African Mid-year Estimates, 2013. Cape Town: Centre for Actuarial
Research, University of Cape Town, 2013. https://www.commerce.uct.ac.za/Research_Units/CARE/
Monographs/Monographs/Mono13.pdf (accessed 11 August 2020).
14. United Nations, Department of Economic and Social Aairs. World Population Prospects 2019,
online edition. 2019. https://population.un.org/wpp/ (accessed 11 August 2020).
15. Ng J, Bakrania K, Russell R, Falkous C. COVID-19 mortality rates by age and gender: Why is the
disease killing more men than women? RGA, 3 July 2020. https://www.rgare.com/docs/default-source/
knowledge-center-articles/covid-19-case-fatality_agegender.p df?sfvrsn=5806a7ea_2 (accessed 11
August 2020).
16. Gold MS, Sehayek D, Gabrielli S, et al. COVID-19 and comorbidities: A systematic review and meta-
analysis. Postgrad Med 2020:1-7. https://doi.org/10.1080/00325481.2020.1786964
17. Western Cape Government. Analysis in comorbidities in adult COVID-19 deaths. 2020. https://www.
westerncape.gov.za/gc-news/147/54748 (accessed 28 July 2020).
18. Guan W-J, Liang W-H, Zhao Y, et al. Comorbidity and its impact on 1590 patients with
Covid-19 in China: A nationwide analysis. Eur Respir J 2020;55(5):2000547. https://doi.
org/10.1183/13993003.00547-2020
19. Boulle A, Davies MA, Hussey H, et al. Risk factors for COVID-19 death in a population cohort
study from the Western Cape Province, South Africa [published online ahead of print, 29 August
2020].Clin Infect Dis 2020;ciaa1198. https://doi.org/10.1093/cid/ciaa1198
20. Groenewald P, Awotiwon O, Hanmer L, Bradshaw D. Guideline for medical certication of death
in the COVID-19 era. S Afr Med J 2020;110(8):721-723. https://doi.org/10.7196/SAMJ.2020.
v110i8.15114
21. Bradshaw D, Laubscher R, Dorrington R, Groenewald P, Moultrie T. Report on weekly deaths in South
Africa: 1 January - 14 July 2020 (week 28). Cape Town: South African Medical Research Council,
2020. https://www.samrc.ac.za/sites/default/les/les/2020-07-22/WeeklyDeaths14July2020_0.pdf
(accessed 28July 2020).
Accepted 18 September 2020.

8 Published online ahead of print
RESEARCH