Results - Part 6

Vectorborne diseases

Vectorborne diseases are infections transmitted by arthropods such as mosquitoes and ticks. A vectorborne disease may involve a simple transfer via the arthropod, or may involve replication of the disease-causing organism in the vector.²² Vectorborne diseases of public health importance in Australia listed in this chapter are: arbovirus not elsewhere classified (NEC); Barmah Forest virus (BFV) infection; dengue virus (DENV) infection; Japanese encephalitis virus (JEV) infection; infections with the Kunjin lineage of West Nile virus (KUNV, which is probably limited to the Australian mainland or possibly Papua New Guinea), and other lineages of West Nile virus (WNV), malaria, Murray Valley encephalitis virus (MVEV) infection and Ross River virus (RRV) infection. Some vectorborne diseases, including yellow fever virus infection, plague and certain viral haemorrhagic fevers, are listed under quarantinable diseases. The National Arbovirus and Malaria Advisory Committee (NAMAC) provides expert technical advice on vectorborne diseases to the Australian Health Principal Protection Committee through CDNA.

Alphaviruses

Viruses in the genus Alphavirus that are notifiable in Australia are BFV and RRV. These viruses are unique to the Australasian region.¹⁰⁷ Infection can cause a clinical illness, which is characterised by fever, rash and polyarthritis. The viruses are transmitted by numerous species of mosquito that breed in diverse environments.¹⁰⁸ The alphavirus chikungunya was not nationally notifiable in 2014, and thus not included in this annual report. However, it is notifiable in all states and territories except the Australian Capital Territory, and states and territories send information about cases to the Australian Government for national collation and analysis.^109,110 Chikungunya virus infection was made nationally notifiable in January 2015.

The national case definitions for RRV and BFV currently require only a single IgM positive test to 1 virus, in the absence of IgM to the other.¹¹¹ False positive IgM diagnoses for BFV in particular are a known issue, thus it is unclear what proportion of notifications represent true cases. Importantly, the case definitions were reviewed by the CDWG and endorsed by CDNA. The revised case definitions were implemented on 1 January 2016.

Top of page

Barmah Forest virus infection

Epidemiological situation in 2014

In 2014, there were 741 notifications of BFV infection, representing a rate of 3.2 per 100,000. This compared with a 5-year mean of 2,155 notifications and a 5-year mean rate of 9.6 per 100,000. The number of notifications of BFV was dramatically decreased compared with 2013, when there were 4,239 notifications, representing a rate of 18.3 per 100,000 (Figure 84). This previous increase in 2013 was considered likely to have been due to a high rate of false positive IgM test results produced by a commercial test kit in private laboratories, and which resulted in a recall of the affected kits in September 2013.¹¹²

Figure 84: Notifications of Barmah Forest virus infection, Australia, 2009 to 2014, by month and year and state or territory

Top of page

Geographic distribution

Comparisons between regions are likely to be influenced by the accuracy of case-ascertainment, which may vary between jurisdictions due to differences in reporting criteria and diagnostic tests used and seasonal trends vary between states and territories. More details are reported in the NAMAC annual reports.¹¹⁰ More than half of all BFV notifications in 2014 were from Queensland (64%, 473) and population rates were highest in the Northern Territory (12.3 per 100,000) and Queensland (10.0 per 100,000). These rates were less than half the 5-year mean, with rate ratios of 0.2 and 0.5 respectively for the Northern Territory and Queensland for 2014 compared with the 5-year mean rate, noting that the 5-year mean rate is strongly affected by the increase between late 2012 and late 2013.

Age and sex distribution

In 2014, BFV infection was most frequently notified in people aged between 40 and 54 years (median 47 years, range 3 to 92 years). Age and sex specific rates were highest among females in the 40–44 and 50–54 years age groups (Figure 85). In 2014, 54% (403) of cases were female and rates were higher in females overall than in males (3.4 and 2.9 per 100,000 respectively).

Figure 85: Notification rates for Barmah Forest virus, 2014 and 2009 to 2013, by age group and sex

Top of page

Seasonality

Peak incidence of BFV could be expected to occur during the warmer months (or during wetter months in northern areas of Australia) when mosquito numbers are high. However, seasonality of notifications is less marked than expected (Figure 84), and a high proportion of interseasonal notifications are thought to be due to false positive diagnoses. Peak notification of BFV in 2014 was between January and April, with 57% (422) of notifications being during this period, a stronger seasonality than observed between 2009 and 2012 (45%, 4,898/10,775) and 2013 (47% 1,998/4,239).

Discussion

The previously reported dramatic increase due to false positive IgM diagnoses declined in late 2013. In addition to the withdrawal of some affected batches of the test kit, the widely acknowledged unreliability of the IgM test kit may have led to a sustained decline in testing for BFV infection in private laboratories, which may not have an alternative diagnostic method to the IgM test kits. On recommendation from NAMAC, the CDWG undertook a review of the surveillance case definitions for BFV infection and for RRV infection. The CDNA surveillance case definition for BFV in 2014 allowed for confirmation based on a single positive IgM, in the absence of IgM to other alphaviruses. Under the revised case definition, a single IgM positive result will no longer constitute laboratory evidence for infection, and where a single result is IgM and IgG positive, it may be notified as a probable case. A confirmed case will require IgG seroconversion or a significant increase in IgG antibody level (e.g. 4-fold or greater rise in titre). There is currently no plan to undertake a retrospective revision of notifications to apply the new case definitions because there is insufficient information on the diagnosis method information available in NNDSS. Therefore, the historical data prior to the upcoming change of case definition will continue to be considered unreliable. The new case definition was implemented from 1 January 2016.¹¹¹

Ross River virus infection

Epidemiological situation in 2014

In 2014, there were 5,316 notifications of RRV, representing a rate of 22.6 per 100,000. This compares with a 5-year mean of 4801.0 cases and a 5-year mean rate of 21.5 per 100,000.

Geographic distribution

In 2014, similar to previous years, nearly half of all RRV infections were from Queensland (44%, 2,344), representing a rate of 49.6 per 100,000), but population rates were highest in the Northern Territory (168.4 per 100,000).

Age and sex distribution

RRV was most frequently reported in adults aged in their 30s or 40s (median 44 years, range 1 to 98 years), similar to previous years. Rates were similar in females and males (rates of 24.9 and 20.4 per 100,000 respectively for a rate ratio of 1.2), similar to previous years. In 2014, age specific rates were highest among females in the 35–49 years age group, and males in the 35–44 years age group (Figure 86).

Figure 86: Notification rates for Ross River virus, 2014, by age group and sex

Top of page

Seasonality

The peak notification for RRV in 2014 was between January and April (Figure 87), and 39% (2,077) of cases were diagnosed during these months. Between 2009 and 2012, 58% (11,358/19,689) of notifications were between January and April, indicating that in 2014, similar to 2013,³⁷ the proportion of inter-seasonal notifications was higher than in previous years. It is important to note that seasonal trends vary between and within states and territories according to differences in mosquito vectors, hosts and climate. In addition, comparisons between regions are likely to be influenced by the accuracy of case-ascertainment, which may vary between jurisdictions because of some differences in reporting criteria and also quality of diagnostic tests used, with false positive IgMs a long term issue.^113,114

Figure 87: Notifications of Ross River virus, Australia, 2009 to 2014, by month and year and state or territory

Top of page

Discussion

The CDNA surveillance case definition for RRV in 2014 allowed for confirmation based on a single positive IgM, in the absence of IgM to other alphaviruses, and there have been differences in laboratory and notification practices. Similar to BFV, there has been a widely acknowledged unreliability of diagnosis based on IgM-only, particularly during the off-season.¹¹⁴ As mentioned under BFV, the case definition was reviewed by the CDWG. Under a revised definition a confirmed case requires IgG seroconversion or a significant increase in IgG antibody level (e.g. 4-fold or greater rise in titre). The revised definition was implemented on 1 January 2016.

Flaviviruses

No specific treatment is available for infections with the flaviviruses DENV, WNV/KUNV, MVEV or JEV and care is largely supportive. A vaccine is available to prevent JEV infection³² and YFV infection, but there are no vaccines currently for DENV, MVEV or KUNV infection.

Infection with MVEV, KUNV or JEV is usually asymptomatic or produces a non-specific illness, but a small percentage of cases progress to encephalomyelitis of variable severity. Culex annulirostris is the major vector of MVEV, JEV and KUNV while Aedes aegypti is the major vector of DENV in Australia.

The clinical illness for DENV infection illness is characterised by mild to severe febrile illness with fever, headache, muscle/joint pain and sometimes a rash. A minority of cases progress to severe dengue with haemorrhage and shock. DENV has 4 serotypes, each containing numerous genotypes. The serotypes isolated from returning travellers (and thus involved in local outbreaks) vary by year and geographical region. Infection with 1 serotype probably confers lifelong immunity to that serotype,²² but subsequent infection with a different serotype is 1 factor thought to increase the risk of severe outcomes, along with the infecting serotype and genotype, and host factors.^22,115–117

Top of page

Arbovirus NEC /Flavivirus unspecified

Epidemiological situation in 2014

In 2014, there were 28 notifications of arbovirus NEC, compared with an average of 12.0 during the previous 5 years. Most notifications were from Queensland (22), with 5 from New South Wales and 1 from Victoria. These notifications comprised Kokobera (1 cases) and Zika virus (ZIKV) infection and probable ZIKV infection (16 cases) and 11 other notifications for which the infecting flavivirus could not be determined or was not supplied (Table 19). Of particular note were the 11 ZIKV and 1 probable ZIKV infections in March and April acquired in the Cook Islands and 1 in February acquired in Samoa. The first report of an outbreak of ZIKV in the South Pacific was through PacNet¹¹⁸ in February 2013, and reported in the Cook Islands. Outbreaks were later reported in New Caledonia and French Polynesia and these continued to mid 2014.

Table 19: Notifications of arbovirus NEC, Australia, 2014, by country of acquisition and infecting species

Country of acquisition	Virus species				Total
	Kokobera	Zika	Zika (probable)*	Unspecified
nfd Not further defined. * This case was a confirmed flavivirus infection, and there was some laboratory evidence of the infecting species being Zika virus, but this was not conclusive.
Australia	1	0	0	0	1
Central and West Africa, nfd	0	0	0	1	1
Cook Islands	0	10	1	1	12
Fiji	0	0	0	1	1
French Polynesia	0	1	0	0	1
Indonesia	0	0	0	2	2
Maldives	0	1	0	0	1
Papua New Guinea	0	0	0	2	2
Philippines	0	0	0	1	1
Samoa	0	1	0	0	1
Sub-Saharan Africa, nfd	0	0	0	1	1
Thailand	0	1	0	0	1
Vanuatu	0	0	0	1	1
Unknown	0	1	0	1	2
Total	1	15	1	11	28

Top of page

Information about the country of acquisition was available for 93% of cases (26) and all of these were acquired overseas (Table 19).

The median age of cases was 39 years (range 18 to 79 years) and 46% of cases (13).

Dengue virus infection

Local transmission of DENV in Australia is normally restricted to areas of northern Queensland where the key mosquito vector, Ae. aegypti is present.¹¹⁹ DENV is not endemic in North Queensland, but local transmission can occur upon introduction of the virus to the mosquito vector by a viraemic tourist or a resident returning from a dengue-affected area overseas.¹²⁰

Epidemiological situation in 2014

There were 1,716 notifications of DENV infection in 2014, which was 1.3 times the 5-year mean of 1,366.4 notifications. Most infections were known to have been acquired overseas (n=1,520) (Figure 88). There were 186 infections acquired in Australia. For 10 cases, no information was supplied on the place of acquisition.

Figure 88: Notifications of dengue virus infection, Australia, 2009 to 2014, by month and year and place of acquisition

Top of page

Geographic distribution

More than 99% (1,706) of notifications in 2014 contained complete information on the place of acquisition. Overseas acquired infections comprised 89% of notifications (1,520) (Table 20). The number of overseas-acquired infections was slightly lower than in 2013 (n=1,596), which was the largest number ever reported (Table 21).

Table 20: Notified cases of dengue virus, Australia, 2014, by serotype and place of acquisition

Place of acquisition	Serotype						Unknown/untyped	Total
	DENV 1	DENV 1 and 3	DENV1 And 4	DENV 2	DENV 3	DENV 4
Locally-acquired
Australia	130	0	0	0	1	0	55	186
Overseas-acquired
Indonesia	267	2	0	70	52	21	399	811
Thailand	16	0	0	19	1	7	96	139
Fiji	2	0	0	9	25	0	70	106
Malaysia	8	0	0	16	2	2	53	81
Philippines	5	0	0	11	2	6	42	66
Sri Lanka	7	0	0	1	0	2	30	40
India	2	0	0	5	1	1	29	38
Timor-Leste	12	0	0	0	4	0	20	36
Vanuatu	1	0	1	0	12	1	18	33
Singapore	8	0	0	0	0	0	7	15
Tonga	0	0	0	0	3	0	12	15
Vietnam	1	0	0	4	0	1	9	15
Nauru	0	0	2	0	5	0	4	11
Papua New Guinea	0	0	0	3	4	0	4	11
Cambodia	0	0	0	1	0	1	8	10
French Polynesia	2	0	0	0	0	0	8	10
Other countries	6	0	2	9	2	0	61	80
Overseas-country unknown	0	0	0	0	0	0	3	3
Total overseas-acquired	337	2	5	148	113	42	873	1,520
Unknown
Place of acquisition unknown	4	0	0	1	0	0	5	10
Total	471	2	5	149	114	42	933	1,716

Top of page

 

Table 21: Notifications of dengue virus infection acquired overseas, Australia, 2009 to 2014, by selected countries of acquisition

Country of acquisition	Year of diagnosis
	2009	2010	2011	2012	2013	2014	Total
Indonesia	172	717	461	804	801	811	3,766
Thailand	25	124	84	279	269	139	920
Fiji	8	1	6	32	14	106	167
Malaysia	16	17	21	20	53	81	208
Philippines	9	42	24	55	63	66	259
Sri Lanka	0	4	12	26	28	40	110
India	15	43	31	60	58	38	245
Timor-Leste	25	37	12	52	49	36	211
Vanuatu	10	4	0	0	5	33	52
Singapore	1	4	5	3	18	15	46
Tonga	15	1	0	0	0	15	31
Vietnam	20	34	14	21	25	15	129
Papua New Guinea	13	21	15	16	35	11	111
Nauru	0	0	0	0	0	11	11
French Polynesia	3	0	0	0	5	10	18
Cambodia	5	11	6	31	31	10	94
Other/unknown countries	137	80	42	76	142	83	560
Total	474	1,140	733	1,475	1,596	1,520	6,939

Top of page

Cases acquired in Indonesia continue to account for the largest number and proportion of all notifications, accounting for 53% (811/1,520) of all overseas-acquired cases in 2014 (Table 21). This was similar to the percentage in 2012 and 2013. The serotype of DENV infections acquired in Indonesia, where known, was frequently serotype 2 or 3 although data completeness for serotype was very low. Other frequently reported source countries in 2014 included Thailand, Fiji and Malaysia.

In Queensland, a single case of locally acquired DENV infection is considered an outbreak. All of the of the 186 locally acquired cases in 2014 were reported in NNDSS to have been associated with 1 of the outbreaks of locally acquired infection in Queensland in 2014 and/or acquired in North Queensland and reported by other states or territories.

Age and sex distribution

DENV infections acquired overseas in 2014 were most commonly reported among younger and middle aged adults (median 38 years, range 0 to 83 years), with a slight peak of notifications among females aged 25–34 years and males aged 25–29 years (Figure 89). Females comprised 50% (757) of overseas acquired cases.

Figure 89: Notifications of overseas-acquired dengue virus infection, 2014, by age group and sex (n=1,520)

Top of page

Locally acquired cases peaked in several adult age groups, but was less common among people aged less than 15 years or more than 64 years (Figure 90). The median age of locally-acquired cases was 38 years (range 1 to 75 years). Females comprised 45% (83/186) of locally-acquired cases.

Figure 90: Notifications of dengue virus infection acquired in Australia, 2014, by age group and sex (n=186)

Top of page

Seasonality

In 2014, the largest number of overseas-acquired cases were diagnosed in January (193 cases) and the lowest number in September and October (63 and 68 cases respectively), but there is no consistent pattern of seasonality from year to year (Figure 88). For locally acquired cases, only 11 cases were reported between June and November demonstrating that there is no on-going local transmission of dengue during the cooler months and that DENV is not endemic in North Queensland.

Microbiological trends

In 2014, serotype information was available for 46% of notifications (783/1,716), which was a decrease compared with the 5-year mean of 53% (Table 22). In 2014, 60% (471/783) of cases with a known serotype were due to DENV 1, similar to 2013, but in contrast to 2012 when DENV2 was more frequently reported, noting the low completeness (Table 22).

Table 22: Serotype of dengue virus infections, Australia, 2009 to 2014

Serotype	2009	2010	2011	2012	2013	2014
DENV1	82	190	140	138	507	471
DENV 1 and DENV 3	0	0	0	0	0	2
DENV 1 and DENV 4	0	0	0	1	0	5
DENV 2	54	255	159	477	179	149
DENV 3	771	106	78	113	142	114
DENV 4	43	47	43	16	55	42
Untyped/unknown	452	630	401	796	957	933
Total	1,402	1,228	821	1,541	1,840	1,716
% with a serotype supplied	68	49	51	48	48	46

Top of page

Discussion

The number of overseas-acquired cases reported in Australia has tended to increase each year, although numbers in 2014 were lower than the previous year. In recent years, improved diagnostic techniques, in particular the availability of the rapid non-structural protein 1 (NS1) antigen detection kit, have improved detection and would have contributed to the observed increase in reported numbers of overseas-acquired dengue in Australia.¹²¹ The dramatic re-emergence and geographical expansion of DENV overseas over the past 50 years with explosive outbreaks,¹¹⁷ as well as increases in the number of Australians travelling overseas each year to DENV endemic countries, particularly tourist destinations such as Bali, Indonesia would also have contributed to this increase.

While local outbreaks of DENV infection occur during the warmer months in North Queensland, each outbreak since 2010 has been relatively small, and prompt and effective responses by public health authorities in Queensland have ensured that the disease has not become endemic.

The number of DENV infections that are serotyped continues to decline. The decreased reporting of a serotype may reflect the increasing use of NS1 antigen detection and/or other diagnostic methods, which do not provide a serotype.

Top of page

Japanese encephalitis virus infection

Epidemiological situation in 2014

There was 1 notification of JEV infection in 2014, which was reported to have been acquired in Indonesia. There were 4 notifications in 2013, all of which were acquired overseas. The last locally acquired case was in 1998.¹²²

West Nile virus/Kunjin virus infection

Epidemiological situation in 2014

There was 1 notification of WNV/KUNV infection in 2014, which was reported to have been acquired in Indonesia. This overseas-acquired case was likely to be from a WNV lineage other than the KUNV lineage due to the limited range of KUNV. There were 2 notifications of WNV/KUNV infection in 2013 (previously published as 3 notifications³⁷ but this number has since been updated).

Murray Valley encephalitis virus infection

Epidemiological situation in 2014

There were no notifications of MVEV infection in 2014. MVEV is a rare disease in Australia with a 5-year mean of 4.4 cases.

The largest number of cases in recent years was in 2011, when 16 cases were reported, including an outbreak in south-east Australia, and these have been described elsewhere.^123–126

Top of page

Malaria

Malaria is caused by a protozoan parasite in the genus Plasmodium, and 5 species are known to infect humans; Plasmodium vivax, P. falciparum, P. malariae, P. ovale and P. knowlesi.^22,127 Malaria is a serious acute febrile illness that is transmitted from person-to-person via the bite of an infected mosquito of the genus Anopheles. Australia is free of endemic malaria as declared in 1981,¹²⁸ but suitable vectors are present in northern Australia, and the area remains malaria-receptive. Malaria is the most frequently reported cause of fever in returned travellers worldwide.¹²⁹ A case series in the Northern Territory showed that malaria cases were reported in travellers returning from endemic areas, but also reflected current events such as military operations and increased refugee arrivals from malaria endemic areas.¹³⁰ Malaria cases in Australia can be found either through testing of symptomatic persons with a compatible travel history, or through screening of refugees who may be asymptomatic, and people may be tested for other reasons.

Epidemiological situation in 2014

There were 322 cases of malaria notified in Australia in 2014, a 23% decrease compared with a 5-year mean of 417.4 cases, and continuing the trend of gradually decreasing notifications since 2005 (Figure 91). The largest number of cases was reported by Queensland (86 cases).

Figure 91: Notifications of malaria, Australia 2000 to 2014, by year

Top of page

Geographic distribution

Malaria in Australia is a disease associated with residing or travelling overseas in areas with endemic transmission. The last cases acquired on mainland Australia were during an outbreak in North Queensland in 2002.¹³¹ Limited transmission occurs occasionally in the Torres Strait following importation, with the most recent being a single case in 2013 acquired on Saibai Island in the Torres Strait and 7 locally acquired cases in the Torres Strait in 2011.

All cases of malaria notified in 2014 were known to have been acquired overseas; however, for 12 cases (4%) the country of acquisition information was incomplete or missing. The most frequent countries of acquisition for overseas acquired cases in 2014 were India (13%, 39/310 of cases with complete information) and Papua New Guinea (10%, 31/310) (Table 23) (Figure 92). Most cases acquired in Papua New Guinea were reported by Queensland (19/31 cases).

Figure 92: Notifications of malaria, Australia, 2009 to 2014, by month and year and place of acquisition

Top of page

Age and sex distribution

In 2014, malaria was most commonly reported in males (72%, 232 cases) with a peak of notifications in males in the 25–29 years age group (Figure 93). The median age of cases was 32 years (range 2 to 72 years).

Figure 93: Notified cases of malaria, Australia, 2014, by age group and sex (n=321)*

* Age was not reported for 1 case and this case is excluded.

Top of page

Seasonality

Increases in notifications or an observable pattern of seasonality in a predominantly overseas-acquired infection can relate to the seasonality of travel patterns, or to local disease epidemiology in the source countries. In 2014, while there were some fluctuations in monthly notifications, there was no clear pattern of seasonality, with notifications ranging between 19 and 37 per month.

Microbiological trends

The infecting species was supplied for 95% (307/322) of cases in 2014 (Table 23). The most frequent infecting species was P. falciparum (reported in 51% of cases with complete information). P. vivax was associated with Asia and the Pacific, whilst most cases acquired in African countries were P. falciparum.

Table 23: Notified cases of malaria, Australia 2014, by infecting species and region and country of acquisition

Country	P. falciparum	P. malariae	P. ovale	P. vivax	P. species	Total
nfd Not further defined.
Oceania
Papua New Guinea	7	1	0	22	1	31
Solomon Islands	0	0	0	4	1	5
North Africa and the Middle East
Egypt	0	1	0	0	0	1
Sudan	23		3	2	1	29
Western Sahara	1	0	0	0	0	1
South Sudan	3	0	0	0	0	3
United Arab Emirates	1	0	0	0	0	1
South-east Asia
Myanmar, The Republic of the Union of	0	0	0	1	0	1
Cambodia	0	0	0	4	0	4
Laos	0	0	0	1	0	1
Thailand	0	0	0	1	0	1
Brunei Darussalam	0	1				1
Indonesia	5	0	2	14	1	22
Malaysia	1	0	0	0	0	1
North-east Asia
China (excludes SARs and Taiwan)	0	0	0	1	0	1
Korea, Republic of (South)		0	0	1	0	1
Southern and Central Asia
India	0	0	0	35	4	39
Pakistan	0	0	0	12	2	14
Afghanistan	0	0	0	1	0	1
Peru	1	0	0	1	0	2
Sub-Saharan Africa
Sub-Saharan Africa, nfd	5	2	0	0	1	8
Central and West Africa, nfd	1	0	0	0	0	1
Burkina Faso	0	0	1	0	0	1
Cameroon	1	0	0	0	0	1
Chad	1	0	0	0	0	1
Congo, Republic of	5	0	0	0	0	5
Congo, Democratic Republic of	1	0	0	0	0	1
Cote d'Ivoire	1	0	1	0	0	2
Gabon	1	0	0	0	0	1
Ghana	13	0	1	1	0	15
Guinea	1	0	0	0	0	1
Guinea–Bissau	1	0	0	0	0	1
Liberia	3	1	0	0	0	4
Mali	3	0	0	0	0	3
Nigeria	5	0	0	0	0	5
Sierra Leone	12	0	3	0		15
Togo	2	0	0	0	0	2
Southern and East Africa
Southern and East Africa, nfd	1	0	0	0	0	1
Burundi	1	0	0	0	0	1
Eritrea	0	0	2	1	0	3
Ethiopia	2	0	0	3	0	5
Kenya	12	2	2	0	1	17
Malawi	1	0	0	0	0	1
Mozambique	2	0	1	0	0	3
Rwanda	1	0	0	0	0	1
South Africa	2	0	0	0	0	2
Tanzania	11	0	0	0	1	12
Uganda	10	3	2	4	0	19
Zambia	7	0	3	0	0	10
Zimbabwe	3	0	0	0	0	3
Southern and East Africa, nec	4	0	0	0	1	5
Overseas acquired – country and region not stated
Unknown/invalid	2	0	1	6	0	9
Overseas-country unknown	1	0	0	1	1	3
Total	158	11	22	116	15	322

Top of page