HIV/AIDS reporting system

1. The Department of Health has implemented a voluntary anonymous HIV/AIDS reporting system since 1985. The system received reports from doctors and laboratories. Medical doctors report newly diagnosed positive cases by a standard form (DH2293). (http://www.info.gov.hk/aids/english/surveillance/form.pdf) In the past, only cases with Western Blot confirmed HIV antibody positive laboratory result were counted as HIV infection for cases aged above 18 months. Since the 4th quarter of 2006, cases with a PCR positive result and clinical or laboratory indication of recent infections were also counted as HIV infection in the reporting system, in view of the increasing regular detection of such cases.

HIV Surveillance at a glance (2008) 435 HIV reports and 96 AIDS reports Gender: 80.2% male Ethnicity: 60.9% Chinese Age: Median 36 Risks: 30.1% Heterosexual contact 33.3% Homo/bisexual contact 9.2% Injecting drug use 0.7% Blood contact 26.7% undetermined CD4 at reporting: Median 190 Subtypes: commonest are CRF01_AE and B Primary AIDS defining illness: Commonest PCP Seroprevalences Blood donors: < 0.01% Antenatal women: < 0.01% STI clinic attendees: 0.23% Methadone clinics attendees: 0.47% Men who have sex with men: 4.31%

2. In 2008, the department received 435 HIV reported cases and 96 AIDS reports, which increased respectively by 5.1% in HIV cases and 21.5% in AIDS cases as compared with 2007. It made the cumulative totals reached 4047 and 1030 for HIV and AIDS reports respectively. Under the revised definition, 13 cases of PCR positive with clinical or laboratory indication of recent infections were included as HIV infection in 2008. Public hospitals/clinics/laboratories were still the commonest source of HIV reports in 2008, which accounted for 39.5% of the reports. Private hospitals/clinics/laboratories were another common source of HIV reports (23.2%). Notably, the AIDS service organisations played a more significant role in HIV reporting in 2008 (8.3%). The number of reports from other sources has remained stable.

3. Around eighty percent of reported HIV cases were male. The male-to-female ratio was 4.1:1 in 2008, slightly lower than that in 2007 of 4.8:1. About 60% of reported cases were Chinese. Asian accounted for 17.2% of reports. The median age of reported HIV cases was 36. Over 63% of reported cases were believed to acquire the virus through sexual transmission in 2008. Injecting drug use accounted for 9.2% of HIV infections in 2008. There was no report of HIV transmission through perinatal contact in 2008. The suspected routes of transmission were not reported in about a quarter of cases. This means that sexual transmission has accounted for about 87% of HIV reports with defined risks.

Rising trend in men who have sex with men persisted

4. Sexual contact remained the commonest route of HIV transmission in Hong Kong. Both heterosexual and homosexual/bisexual contacts were important risk factors. In 1980s and early 1990s, the early years of AIDS epidemic in Hong Kong, it used to report more cases from men who have sex with men, including both homosexual and bisexual contacts. The trend then reversed with more heterosexual transmission reported since 1993. A rising trend in MSM has been observed since 2004 and the situation was consistent this year with 145 MSM cases identified out of 319 cases with defined risks. (Table 1).

Table 1. Distribution of reported HIV cases by exposure category (1984 - 2008)

Exposure Category (%)	Year
	1984	1985	1986	1987	1988	1989	1990	1991	1992	1993	1994	1995	1996	1997	1998	1999	2000	2001	2002	2003	2004	2005	2006	2007	2008	Total
Heterosexual	1	0	0	3	6	11	12	29	32	47	73	81	94	117	133	127	115	127	147	116	112	113	130	109	131	1866
	(14.3)	(0.0)	(0.0)	(9.1)	(21.4)	(28.9)	(35.3)	(48.3)	(45.1)	(59.5)	(70.2)	(66.4)	(70.1)	(64.6)	(70.4)	(59.6)	(62.8)	(59.6)	(56.5)	(50.7)	(41.8)	(36.1)	(34.9)	(26.3)	(30.1)	(46.1)
Homosexual	1	10	6	12	12	15	8	18	27	20	22	26	20	33	16	33	21	37	47	45	62	86	108	156	132	973
	(14.3)	(21.7)	(30.0)	(36.4)	(42.9)	(39.5)	(23.5)	(30.0)	(38.0)	(25.3)	(21.2)	(21.3)	(14.9)	(18.2)	(8.5)	(15.5)	(11.5)	(17.4)	(18.1)	(19.7)	(23.1)	(27.5)	(29.0)	(37.7)	(30.3)	(24.0)
Bisexual	0	1	2	7	2	6	5	8	2	2	4	4	3	10	6	10	7	7	9	5	6	10	15	18	13	162
	(0.0)	(2.2)	(10.0)	(21.2)	(7.1)	(15.8)	(14.7)	(13.3)	(2.8)	(2.5)	(3.8)	(3.3)	(2.2)	(5.5)	(3.2)	(4.7)	(3.8)	(3.3)	(3.5)	(2.2)	(2.2)	(3.2)	(4.0)	(4.3)	(3.0)	(4.0)
Injecting drug use	0	1	0	0	2	2	0	0	3	1	2	2	1	2	1	6	10	11	10	11	22	31	57	42	40	257
	(0.0)	(2.2)	(0.0)	(0.0)	(7.1)	(5.3)	(0.0)	(0.0)	(4.2)	(1.3)	(1.9)	(1.6)	(0.7)	(1.1)	(0.5)	(2.8)	(5.5)	(5.2)	(3.8)	(4.8)	(8.2)	(9.9)	(15.3)	(10.1)	(9.2)	(6.4)
Blood contact	5	32	10	7	2	2	5	0	1	1	1	0	0	1	0	2	0	0	0	0	0	4	0	2	3	78
	(71.4)	(69.6)	(50.0)	(21.2)	(7.1)	(5.3)	(14.7)	(0.0)	(1.4)	(1.3)	(1.0)	(0.0)	(0.0)	(0.6)	(0.0)	(0.9)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(1.3)	(0.0)	(0.5)	(0.7)	(1.9)
Perinatal	0	0	0	0	0	0	0	0	0	0	1	2	1	0	2	4	2	2	1	0	0	2	2	1	0	20
	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(0.0)	(1.0)	(1.6)	(0.7)	(0.0)	(1.1)	(1.9)	(1.1)	(0.9)	(0.4)	(0.0)	(0.0)	(0.6)	(0.5)	(0.2)	(0.0)	(0.5)
Undetermined	0	2	2	4	4	2	4	5	6	8	1	7	15	18	31	31	28	29	46	52	66	67	61	86	116	691
	(0.0)	(4.3)	(10.0)	(12.1)	(14.3)	(5.3)	(11.8)	(8.3)	(8.5)	(10.1)	(1.0)	(5.7)	(11.2)	(9.9)	(16.4)	(14.6)	(15.3)	(13.6)	(17.7)	(22.7)	(24.6)	(21.4)	(16.4)	(20.8)	(26.7)	(17.1)
Total	7	46	20	33	28	38	34	60	71	79	104	122	134	181	189	213	183	213	260	229	268	313	373	414	435	4047
	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)	(100)

5. A high weighting of MSM in HIV reports continued in 2008 despite being a drop compared to that of 2007. Over 40% of male HIV reports in 2008 contracted the virus through homosexual or bisexual contact. Heterosexual contact in male cases accounted for about 24%, whereas the routes of transmission were not reported in the rest of the 23% male cases. The ratio of heterosexual men against MSM dropped from its peak of 4.2:1 in 1998 to 0.6:1 in 2008 with the lowest ratio recorded at 0.4:1 in 2007 during the period. (Figure 1) That still indicated more men continued to be infected through homosexual/bisexual contact than heterosexual contact.

Figure 1. The number of MSM cases is taking over heterosexual men cases in the reporting system again.

6. The major attributes of the rise in MSM were Chinese and of age group 20-39. Over 86% of MSM cases in 2008 were Chinese. Caucasians accounted for only 11%. A rising trend in the number of reported Chinese MSM cases was observed in recent years despite a modest drop this year. (Figure 2) The median age of MSM cases at report was 36, as compared to 43 of heterosexual man cases. Moreover, the median age of HIV infected MSM population has been relatively stable in the last decade whereas that of heterosexual men was on a rising trend. (Figure 3) Age group 30-39 remained the commonest age of reporting in MSM, which accounted for 38% in 2008, followed by 26% in the age group 20-29. (Figure 4)

Figure 2. Ethnicity Breakdown of HIV-infected MSM cases (1984-2008)

Figure 3. Median age of HIV-infected MSM cases, heterosexual man and heterosexual women (1984-2008)

Figure 4. Age breakdown of HIV-infected MSM cases (1984 - 2008)

7. Efforts have been made to track the HIV prevalence and risk behaviours among MSM in Hong Kong, including the second community-based survey (PRiSM) in gay saunas, bars and clubs which was conducted in 2008 and revealed a HIV prevalence of 4.31% among MSM attending these venues. The level of consistent condom use with regular sex partners and non-regular sex partners were 45% and 75% respectively, which were very similar to those figures of 2006 survey. On the other hand, rate of HIV testing within the last one year increased from 24% in 2006 to 36% in 2008. AIDS Concern's voluntary HIV testing service targeting MSM was another source to estimate the prevalence in MSM, although the data was affected by participant bias to a larger extent. A rising trend in prevalence was observed since 2004 but it seemed to be stable in recent years. (Table 2) This may be affected by the expansion of AIDS Concern's service from higher-risk MSM settings to general-risk MSM populations in recent years.

Table 2. HIV seroprevalence in AIDS Concern's MSM voluntary HIV testing service

Year	No. of blood samples	No. of samples tested anti-HIV+	Prevalence (%)	95% C.I. for prevalence (%)
2000	38	0	0	(--- - ---)
2001	107	1	0.93	(0.024-5.207)
2002	130	1	0.77	(0.019-4.286)
2003	223	2	0.90	(0.109-3.240)
2004	332	6	1.81	(0.663-3.934)
2005	483	12	2.48	(1.284-4.340)
2006	610	10	1.64	(0.786 -3.015 )
2007*	723	17	2.35	(1.370 -3.765 )
2008	905	15	1.66	(0.928 -2.734 )

* figures in 2007 revised

8. The regular condom use rate of MSM attending AIDS Counselling and Testing Service was observed to be higher in recent two years for both regular partners and casual partners, although the trends observed for condom use in last anal sex among MSM were relatively less obvious. On the other hand, the trends derived from MSM attending AIDS Concern's testing service remained static for both regular condom use and condom use for last anal sex. (Figure 5)

Figure 5. Regular condom use* among adult Men who have Sex with Men (MSM)

*	Regular condom use is defined as always or usually using a condom on a 4-level scale time period: ACTS: past one year / AC: past 3 months
**	Regular sex partners used to refer to long-term sex partners including spouse, mistress, and steady boy/girl friends for at least one year, or if less than one year, one with whom is expected to continue sexual relationship. This definition of regular sex partners in 2008 has been futher refined to include (other than the long-term sex partners) sex buddy that refers to regular sex only partner for at least 6 months, or if less than 6 months, one with whom is expected to continue sexual relationship
***	Casual sex partners, the two do not have steady relationship.
Remarks :	ACTS - AIDS Counselling and Testing Service; AC - AIDS Concern

The number of heterosexual contact cases increased in 2008

9. The number of heterosexual cases, after an obvious drop from 130 in 2006 to 109 in 2007, appeared to rebound to the similar level in 2006, back up to 131 this year. Because of this rebound in the number of heterosexual contact, the proportion accounted by heterosexual contact jumped back from 26.3% in 2007 to 30.1% in 2008. (Table 1) The male to female ratio for heterosexual cases was 1.8:1. The median age of heterosexual cases in 2008 was 40. Heterosexual male cases were mainly (75% in year 2008) Chinese whereas Chinese accounted for less than half (46% in year 2008) of female cases.

10. A majority of Social Hygiene Clinics attendees reported unprotected heterosexual contact. The seroprevalence of Social Hygiene Clinic attendees remained stable at below 0.3% (0.23% in 2008). On the other hand, the trend of sexually transmitted infections (STI) provides information for the understanding of risk of HIV infection in the community. Although it was estimated that Social Hygiene Clinics took care of only 20% of STI cases in the territory, it was still a very important sentinel site. It continued to record a decrease in the total number of STI cases in Social Hygiene Clinics, an aggregate of 13,867 in 2008 as compared with 14,305 cases in 2007. A 3% drop was observed in all the common STI diagnosis. The decrease of cases was more obvious in non-specific genital infection / non-gonococcal urethritis from 6,761 cases in 2007 to 6,518 cases in 2008.

11. The regular condom use rate with commercial partners remained stably high among adult heterosexual men attending Social Hygiene Clinics and AIDS Counselling and Testing Service. The level was over 80% among those attending AIDS Counselling and Testing Service and about 70% in Social Hygiene Clinic attendees.

Small but significant numbers of infection in injecting drug users reported

12. In 2008, the reporting system recorded 40 cases of HIV transmission through injecting drug use. The number was similar to that of 2007, but still remaining at a relatively higher level as compared with several years ago. Most of the cases were Asian, non-Chinese. The median age was 30. 10% of injecting drug user cases was reported from methadone clinics.

13. The Universal HIV Antibody (Urine) Testing Programme (MUT) replaced the unlinked anonymous screening (UAS) in methadone clinic as the seroprevalence study in 2004. 7723 attendees participated in the programme in 2008 with a coverage rate of 83%, a better coverage than in 2007 of 80%. The programme tested 7942 urine samples, 24 positive cases in 2008 and with the 13 previously known positive cases still attending methadone clinics, totally there were 37 HIV positive drug users attending methadone clinic this year. The seroprevalence over the year was stable at below 1%. The seroprevalence of methadone clinic attendees in 2008 was 0.47%, which remained at a similar level as in previous years.

14. Although a significant proportion of drug users were injectors, various surveys revealed that the proportion of needle sharing was relatively low. The trend remained stable over the years, with sharing rate revealed from street survey generally higher than from treatment and rehabilitation services.

No case of perinatal transmission recorded

15. In 2008, three cases reported blood contact transmission. Actually no HIV infection from local contaminated blood or blood product was found in the recent several years. The seroprevalence of new blood donors at Hong Kong Red Cross Blood Transfusion Service was at a low level of 0.012% in 2008.

16. In 2008, there was no perinatal HIV infections reported. The Universal Antenatal HIV Testing was implemented in September 2001. Over 40,000 pregnant women attending public antenatal services were tested every year and the coverage of the programme reached 98.2% in 2008 and revealed the seroprevalence of HIV infection in pregnant women to be 0.004%, which remained at a low level as in previous years. Two pregnant women were tested positive in the programme this year. One woman terminated her pregnancy and one woman delivered her baby by Caesarean Section.

Cases with undetermined risk factor was significant

17. The information of voluntary reporting was becoming incomplete as there are an increasing proportion of cases reported without a risk factor. Slightly more than past years, around a quarter of cases reported without a suspected route of transmission. These cases were usually those without a physician's reporting. Undetermined risk is commoner in cases reported by private hospitals/clinics/laboratories. Although this was a voluntary reporting system, physicians were strongly encouraged to make the system comprehensive. The data is of paramount importance for understanding the local HIV epidemiology and its evolution.

Pneumocystis Pneumonia and Tuberculosis remained the commonest Primary AIDS Defining Illness

18. Hong Kong adopts a CDC-modified system of HIV classification and AIDS definition system laid down by the then Scientific Committee on AIDS in 1995. (http://www.info.gov.hk/aids/pdf/g40.pdf) The annual number of reported AIDS cases was dropping since 1997, the year of introducing highly active antiretroviral therapy (HAART) in Hong Kong but a slow increasing trend was observed since 2005. Ninety-six AIDS cases, the highest annual number, were reported in 2008 as compared with 79 cases in 2007. The increase was compatible with increase in newly reported and cumulative HIV cases. 90 cases (94%) of the AIDS reports this year has their AIDS reported within 3 months of HIV reporting.

19. The primary AIDS defining illness (ADI) pattern of the reported cases also changed slightly in recent years. Pneumocystis jirovechi pneumonia (previously named Pneumocystis carinii) was the commonest ADI in Hong Kong in 2008 which accounted for 37 cases (38.5%), an increase of over 3% in terms of the proportion of ADI as compared with 2007. This year, 32 cases (33.3%) reported Mycobacterium tuberculosis as the primary ADI which was following right after Pneumocystis jirovechi pneumonia as the second commonest ADI. They were followed by Penicilliosis (6, 6.3%), and Cytomegalovirus diseases (6, 6.3%). On the other hand, unlinked anonymous testing in tuberculosis patients demonstrated a HIV seroprevalence of 0.51% in 2008 while universal voluntary testing showed a seroprevalence of 1.17%. HIV positivity rate in TB patients was consistently higher than many at-risk populations over the years.

20. The median CD4 of newly reported HIV cases in 2008 was 190. Reporting of CD4 level was becoming a routine practice in physician. It provided useful information on the timing of diagnosis in the course of HIV infection. 61.4% of HIV cases in 2008 reported the CD4 level at diagnosis. (Table 3) The median CD4 for those aged less than 55 has been stable at around 200 (158 - 278) for the past 5 years. One the other hand, there was a continued decreasing trend in median CD4 count among those who are aged 55 and above. It suggested that more patients reported at age 55 or above were diagnosed at a late disease stage. (Table 4)

Table 3. Reported CD4 levels at HIV diagnosis

Year	No. of HIV reports	No. of CD4 reports (%)		Median CD4 (cell/ul)	CD4>=200 (cell/ul) (%)
2001	213	162	( 76.1% )	233.5	85	( 52.5% )
2002	260	201	( 77.3% )	197	100	( 49.8% )
2003	229	166	( 72.5% )	205	85	( 51.2% )
2004	268	179	( 66.8% )	208	95	( 53.1% )
2005	313	224	( 71.6% )	194	110	( 49.1% )
2006	373	278	( 74.5% )	224	150	( 54.0% )
2007	414	301	( 72.7% )	241	170	( 56.5% )
2008	435	267	( 61.4% )	190	128	( 47.9% )

Table 4. CD4 Reports by age group

Age	Year	No. of HIV reports	No. of CD4 reports (%)		Median CD4 (cell/ul)	% of CD4 >= 200 (cell/ul)
<55	2001	190	146	(76.8%)	258.5	54.1%
	2002	230	183	(79.6%)	196	49.7%
	2003	190	139	(73.2%)	228	52.5%
	2004	225	158	(70.2%)	220.5	55.7%
	2005	280	201	(71.8%)	192	48.8%
	2006	341	253	(74.2%)	241	56.9%
	2007	377	278	(73.7%)	254.5	57.9%
	2008	380	228	(60.0%)	217	51.8%
>=55	2001	22	16	(72.7%)	96	37.5%
	2002	24	18	(75.0%)	212.5	50.0%
	2003	32	27	(84.4%)	108	44.4%
	2004	32	21	(65.6%)	82	33.3%
	2005	29	23	(79.3%)	223	52.2%
	2006	28	25	(89.3%)	145	24.0%
	2007	33	23	(69.7%)	104	39.1%
	2008	53	39	(73.6%)	74	25.6%

The commonest HIV subtypes were CRF01_AE and B

21. In 2008, about 85% of HIV reports had their subtypes documented, at a comparable level as in the past years. CRF01_AE and Subtype B of HIV-1 strains were the most common subtypes identified in Hong Kong. They together accounted for 73% of all HIV cases. CRF_01AE was found to be commoner in female, Asian non-Chinese, heterosexuals and IDU. The subtype B was commoner in Caucasian, MSM and C subtypes in females, Asians and sexually transmitted cases. An increasing diversity of subtypes and its circulating recombinant forms was also noted. (Table 5)

Table 5. HIV-1 Subtypes in Hong Kong

	2003		2004		2005		2006		2007		2008
Annual HIV Reports	229		268		313		373		414		435
No of reports with subtypes (%)	204	(89%)	203	(76%)	258	(82%)	316	(85%)	354	(86%)	370	(85%)
Subtype (%)
CRF01_AE	99	(43%)	96	(36%)	125	(40%)	149	(40%)	158	(38%)	169	(39%)
B	60	(26%)	71	(26%)	101	(32%)	124	(33%)	155	(37%)	150	(34%)
C	21	(9%)	3	(1%)	2	(1%)	6	(2%)	2	(0%)	14	(3%)
CRF07_BC	9	(4%)	7	(3%)	6	(2%)	14	(4%)	10	(2%)	11	(3%)
CRF08_BC	4	(2%)	10	(4%)	6	(2%)	11	(3%)	14	(3%)	2	(0%)
Others	11	(5%)	16	(6%)	18	(6%)	12	(3%)	15	(4%)	24	(6%)

Discussion

22. The number of HIV reports was persistently on a rise in 2008. The annual HIV reports used to be less than or around 300 before 2006. The total number of HIV reports in 2008 was 435, which was a 5% increase as compared to 2007. In the last few years, there was 10-20% increase in HIV reports every year except in 2003, when SARS outbreak occurred. The increasing reports from Men who have Sex with Men continued to contribute to the rise in HIV reports, although heterosexual contact appeared to be catching up in 2008. An increase in injecting drug users was observed but mainly in non-Chinese population, which suggested non-local infections.

23. The number of HIV reports among MSM continued to play a significant role and it accounted for consistently the largest proportion this year. The HIV situation in MSM was really worrisome because the increasing trend has persisted and in an escalating fashion. Data suggested that young MSM aged 20-29 was becoming more affected. The second community-based seroprevalence survey in 2008 revealed a slightly higher HIV prevalence of 4.31% when compared with the previous study in 2006. Both condom usage rates of MSM with casual and regular partners remained at a suboptimal level than that of heterosexual men visiting sex workers. Reporting data, prevalence data and behavioural data all suggested a persisting local HIV epidemic in MSM. The observation was in keeping with the regional picture of rising MSM HIV epidemic.

24. Heterosexual transmission appeared to be in a stable trend over the years although it appeared to increase in 2008. A significant proportion of non Chinese female cases might suggest infections outside Hong Kong. For heterosexual men, only 34% of their contacts were believed to have occurred in Hong Kong. The prevalence in social hygiene clinics attendees and antenatal women were all below 1%. The condom use rates of commercial sex were high as gauged from the reports of sex workers and their clients.

25. Although the number of HIV-infected injecting drug users was persistent at a higher level than a few years back, an escalating growth of HIV infections in injecting drug users was not expected at present. Same as last year, most who reported injecting drug use were Asian non-Chinese. It was believed that those non-Chinese acquired the infection outside Hong Kong. The number of HIV infections in drug users contributed by the local infections was not largely different from previous years.

26. In conclusion, the HIV infections in Hong Kong were increasing and mainly affecting MSM populations. The situation of heterosexual population and local injecting drug user population was relatively stable thus far. The HIV epidemiology in Hong Kong was also affected by the situation of neighbouring countries. A proportion of cases were infections which acquired outside Hong Kong. The number of people living with HIV was estimated to be 3600 as of 2007. Hence, HIV prevalence remained at <0.1% among the general population in Hong Kong.

27. The number of HIV reports among MSM continued to rise and it accounted for even a larger proportion this year. The HIV situation in MSM was really worrisome because the increasing trend has persisted and in an escalating fashion. The young MSM aged 20-29 was heavily affected. The community-based seroprevalence survey in 2006 revealed a high HIV prevalence. Both condom usage rates of MSM with casual and regular partners remained at a lower level than that of heterosexual men visiting sex workers. Reporting data, prevalence data and behavioural data all suggested a persisting local HIV epidemic in MSM.

28. Heterosexual transmission appeared to be in a stable trend over the years. A significant proportion of non Chinese cases suggested infections outside Hong Kong. For heterosexual men, only 33.3% of their contacts were believed to have occurred in Hong Kong. The prevalence in social hygiene clinics attendees and antenatal women were all below 1%. The condom use rates of commercial sex were high on both from the report of sex workers and their clients.

29. Although the number of HIV-infected injecting drug users was persistent at a high level, an escalating growth of HIV infections in injecting drug users was not expected at present. Same as last year, most reported injecting drug users were Asian non-Chinese. It was believed that those non-Chinese acquired the infection outside Hong Kong. The number of HIV infections in drug users contributed by the local infections was not largely different from previous years.

30. In conclusion, the HIV infections in Hong Kong were increasing and mainly affecting MSM populations. The situation of heterosexual population and local injecting drug user population was stable. The HIV epidemiology in Hong Kong was also affected the situation of neighbouring countries. A proportion of cases were infections which acquired outside Hong Kong.