Understating the Barriers to Achievement of the UNAIDS 90–90–90 Goal in Tanzania Using a Population-based HIV Impact Assessment Survey 2016–2017

Abstract

The Joint United Nations Programme on HIV/AIDS (UNAIDS) and partners launched the 90–90–90 targets. The three targets provide that by 2020, 90% of all people living with HIV (PLWH) will know their HIV status, 90% of all people with diagnosed HIV infection will receive sustained antiretroviral therapy (ART), and 90% of all people receiving ART will have viral load (VL) suppression (Abuogi et al., 2018; Granich et al., 2017; Levi et al., 2016). We use Tanzania HIV Impact Survey (THIS) data to study the barriers to achieve these targets.

THIS 2017 was a population-based with a stratified multistage stage survey sampling design. We used weighted logistic regression to associate three targets with socio-demographics, HIV-related discrimination, fear and shame. We defined HIV awareness by combination of self-reported of HIV status positive and antiretrovirals (ARVs) detected in blood among PLWH. On ART was defined as those who self-reported among awareness. VL suppression was defined as 400 copies/ml or less in the blood sample.

The three targets were estimated at 61–90-85 in Tanzania from the weighted analysis. The first target was far from being achieved. The weighted regression showed that being female, having attained higher education, married, having insurance, and living in urban areas were associated with a high likelihood of having ever tested for HIV. Yet, we found HIV related discrimination and fear were the barrier of HIV awareness and hence lower rates of being on ART and VL suppression.

The second and third targets are derivatives of the first one. The results indicated that intervention programs in Tanzania should focus on the first target. To achieve the first target, population-based massive screen program is necessary. Intervention programs should be designed for each target separately. Integrated strategies in context of low and middle income countries are needed to achieve these targets.

Introduction

In 2014, the Joint United Nations Programme on HIV/AIDS (UNAIDS) with efforts of 11 UN organizations launched the 90–90–90 targets (Levi et al., 2016). UNAIDS leads the global effort to end AIDS as a public health threat by 2030. The three 90–90–90 targets provide that by the year 2020, 90% of all people living with HIV (PLWH) will know their HIV status, 90% of all people with diagnosed HIV infection will receive sustained antiretroviral therapy (ART), 90% of all people receiving ART will have viral load (VL) suppression. We use Tanzania HIV Impact Survey (THIS) data to study the barriers to achieve these targets.

To measure the national and regional progress of UNAID’s 90–90–90 targets and to provide guidance for policy and funding priorities, a nationally representative, household-based, cross-sectional survey was conducted in 15 countries sponsored by U.S. President’s Emergency plan for AIDS Relief (PEPFAR). It is the population-based HIV impact assessment (PHIA) (Cervantes et al.) survey led by ICAP at Columbia University in partnership with the US Centers for Diseases Control and Prevention (CDC), and the Ministry of Health in each country. PHIA showed that HIV efforts made critical progress toward achieving an AIDS-free generation. It measured national HIV incidence and HIV VL suppression in population. The surveys were designed and implemented by experts in epidemiology, laboratory science, program services, and workforce training. PHIA in Tanzania recently released the survey conducted between October 2016 and August 2017. This is the first survey that measured national HIV incidence and VL suppression in population in Tanzania. THIS has transformed our understanding of the HIV epidemic in Tanzania and identified some programmatic gaps related to control of the HIV epidemic in in achieving the targets.

In 2017, Mozambique, South Africa and Tanzania accounted for more than half of new infections in SSA (UNAIDS, 2017). In 2018 the number of people living with HIV was estimated at 1.6 million in Tanzania (UNAIDS, 2018). In Tanzania, although new infections have declined, more needs to be done to scale-up HIV testing and treatment to avert new infections. In Tanzania young people 15–24 years particularly young women are disproportionately affected by the HIV epidemic (Plotkin et al., 2018; Sidibé, 2018). Over the last decade, Tanzania made significant progress in the scale up of ART following the national HIV/AIDS programme launched in 2004 (Somi et al., 2009). Recent research done in Tanzania show barriers to linkage in care include lack of disclosure, stigma, spiritual beliefs, poor provider relationship and attitudes, delays at clinic and lack of human resource (Abuogi et al., 2018; Kayabu, Ngocho, & Mmbaga, 2018; Nyato et al., 2019). The barriers to achieving viral suppression included ART toxicity, simplified regimens, stigma, discrimination, lack of data for monitoring, and the availability of supportive services (Abuogi et al., 2018). While there is evidence for some barriers in the context of pregnant mothers, adolescents and key populations, less is known about barriers in achieving the targets to Tanzania’s national HIV response of ending the HIV epidemic (Gourlay, Birdthistle, Mburu, Iorpenda, & Wringe, 2013).

In this study we aim to study barriers that are associated with achieving the three targets among adults (>15 years) in Tanzania. We give some recommendations about the next step for ending the HIV epidemic in Tanzania. The lessons learned from ending AIDS threat in Tanzania can be carried forward to other countries and other infectious diseases.

Methods

We used Tanzania HIV Impact Survey 2016–2017 (THIS 2016–2017), which is the Population-based HIV Impact Assessment (PHIA) Surveys conducted between October 2016 and August 2017 in Tanzania (Tanzania, 2017). The consent form of THIS is completed by the household head or designee. This data set was released to public in 2019. THIS applied a stratified multistage stage probability sample design to represent the population. The strata was defined by 31 regions of the country. First-stage sampling units (primary sampling units (PSU)) were defined by enumeration areas (EAs) within strata. Second-stage sampling units were the households within EAs. The PSUs were selected with probability proportional to number of households. The sampling design achieved precision levels of national estimates of HIV incidence rates and regional estimates of viral load suppression (Tanzania, 2017). The sampling weights were calculated to compensate for the probability of selected variables, nonresponse rate of subsamples, and under coverage of certain population groups.

All analysis performed in the paper involving human participants were in accordance with ethical standards of the institutional research committee. The ethical approval is not required because the data is publicly available.

THIS offered HIV counseling and testing, and collected information about household and individual characteristics, HIV awareness and uptake of HIV care and treatment services. HIV prevalence testing was conducted in each household using a serological rapid diagnostic testing algorithm based on Tanzania’s national guidelines. Viral load testing was conducted on all HIV positive samples.

In this paper, we included PLWH in THIS adult’s sample who were aged 15 and above and completed both individual interview and biomarker blood test. Biomarker testing was offered to all adults who completed an individual interview and consented or assented to provide blood samples. We used biomarker sampling weights in the proceeding analysis. The blood sample tested three ARVs, Efavirenz, Lopinavir, and Nevirapine.

The outcomes of this paper were three 90–90–90 targets that included HIV awareness among people living with HIV (PLWH), on ART among those aware of their HIV status, viral load (VL) suppression among those on ART. HIV awareness was defined as a two-level variable namely (1) those who reported during the interview they that had taken the HIV test and knew their positive status; (2) PLWH but were unaware of their status. The second target was defined as (1) among those who were aware of their HIV status and if during the interview, the participant reported they are on ART; (2) Aware of status but not on ART. The third target was VL suppression among those who were on ART. We used VL 400 copies per ml or less in the blood sample as a cutoff value (Geretti et al., 2008).

In the analysis, we included individual-level socio-demographic variables of age, gender, education, marital status, health insurance status, area of residence (urban or rural), and HIV related discrimination, fear, and feeling shame if family members are HIV positive. Age is defined as a categorical variable with groups as 15–24, 25–34, 35–44, 45–54 and 55 and above. We defined HIV related discrimination by assigning a 1 id a responded a “yes” to either of the two questions: “Would you buy fresh vegetables from a shopkeeper or vendor if you knew that this person had HIV?” and “Do you think that children living with HIV should be able to attend school with children who are HIV negative?”. The HIV fear and shame were defined by answering yes to two questions respectively. The questions asked “Do you think people hesitate to take an HIV test because they are afraid of how other people will react if the test results is positive for HIV?” and “Do you agree or disagree with the following statement: I would be ashamed if someone in my family had HIV?”.

In this paper, we used weighted logistic regression (Barone, Lombardo, & Tarantino, 2007) to assess the association between three targets with the background socio-demographic and behavioral variables. The socio-demographics were also assessed for associations with having ever tested for HIV in the population. All analysis was performed using SAS 9.4 (Institute, 2015). We used univariate, bivariate and multivariate analysis to study the barriers associated with the three targets. At the end of the methods session, we assessed the variable indicator for those who reported had ever tested for HIV in population with the same set of covariates. We finally extend the analyses to barriers of population screening for HIV.

Results

There were 38,680 observations in the adult individual interview data. Among them, 34,060 observations had biomarker data available. There were 1,895 PLWH representing a total of 1,532,483 population with HIV prevalence of 4.6% in Tanzania. VL results were available for 1,894 observations. The participants who did not have VL available were not included in further analyses (12%).

Table 1 shows the weighted Univariate analysis results. In the population, there were 939,903 (61%) PLWH who were aware of their HIV status. Among those who were aware of their status, about 842,365 (90%) were on ART. Among those who were on ART, about 716,993 (85%) were suppressed people whose viral load were under 400 copies/ml. The 90–90–90 target was achieved by 61–90-85 in Tanzania. In the population, there were 224,616 (15%) individuals with CD4 count less than 200 cells/mm³, who were at stage III of HIV or were diagnosed with acquired immunodeficiency syndrome (AIDS) (Organization, 2007). The mean age of PLWH was 39.7 years (SD=12, range from 15 to 80). Young adults aged 15 to 24 were 10% . The other age groups were 26% aged 25 to 34 years, 32% aged 35 to 44 years, 19% aged 45 to 54 years and 12% aged 55 years and above. Most PLWH were female (66%) especially young women. About 19% of the population did not attend school at all. There were about 69% with education primary or less. The rest 12% had education with secondary and above. About 56% of the population were married or lived with a partner. Most of the population (85%) did not have health insurance. There were about 46% of the population that lived in the urban area. There were 95.7% individuals reporting that they would not buy fresh vegetables from a person who has HIV, and about 10% who believed that children with and without HIV should not go to the same school together. There were 20% individuals who felt hesitant to take an HIV test because of the reaction from other people and 10% felt ashamed if someone in the family had HIV.

Table 1.

Social demographics of Tanzania population

Variables	Weighted Frequency (percent)
HIV awareness
Not aware	592,580 (38.67%)
Aware	939,903 (61.33%)
HIV treatment
Not on ARV	690,118 (45.03%)
On ARV	842,365 (54.97%)
Viral Load Suppression
> 400	774,214 (50.58%)
Under 400	756,440 (49.42%)
Age group
Age 15–24	153,400 (10.01%)
Age 25–34	401,208 (26.18%)
Age 35–44	495,163 (32.31%)
Age 45–54	297,554 (19.42%)
Age 55+	185,157 (12.08%)
Gender
Female	1,013,819 (66.16%)
Male	518,664 (33.84%)
Education
No school at all	292,665 (19.10%)
Primary or less	1,052,599 (68.69%)
Secondary and above	187,219 (12.22%)
Marital status
Never married	166,668 (10.88%)
Married/living together	859,338 (56.08%)
Divorced/Widowed	506,338 (33.04%)
Insurance
No insurance	1,295,839 (84.56%)
Have insurance	236,644 (15.44%)
Residential area
Urban	697,524 (45.52%)
Rural	834,959 (54.48%)
HIV discrimination
No	66,499 (4.34%)
Yes	1,465,984 (95.66%)
HIV fear
No	318,949 (22.05%)
Yes	1,127,511 (77.95%)
HIV shame
No	1,369,658 (90.18%)
Yes	149,209 (9.82%)

In the weighted bivariate analysis (Table 2), we examined the association between each covariate variable with the three targets respectively. Age was significantly associated with both HIV awareness and ART, but not with VL suppression. Gender was significantly associated with all three variables. Females were more likely to be aware of their HIV status, being on ART and VL suppression. Those who did not participate in schooling were less likely to be aware of their positive status. Living in urban area was associated with higher HIV awareness. Among PLWH, those who did not have HIV related fear and HIV related discrimination were less likely to be aware of their status. However, those who did not have HIV related shame were more likely to know their status. However, all these factors were not associated with two other targets in the subset analysis. Those who had insurance were more likely to be on ART (p-value <0.0001) and with VL under 400 copies/ml.

Table 2.

Weighted bivariate analysis between targets and socio-demographics

	Target I		Target II		Target III
	HIV awareness		On ART		VL suppressed
	N= 939,903		N=842,365		N= 716,993
Age groups
Age 15–24	79,495 (67.49%)	<.0001	68,942 (86.72%)	0.0001	56,379 (81.78%)	NS
Age 25–34	207,869 (67.48%)		171,416 (82.46%)		150,255 (87.66%)
Age 35–44	326,057 (74.54%)		291,673 (89.45%)		246,577 (84.54%)
Age 45–54	212,874 (77.85%)		200,440 (94.16%)		170,365 (85.00%)
Age 55+	113,608 (72.13%)		109,895 (96.73%)		93,417 (85.01%)
Gender
Female	664,794 (65.57%)	<.0001	604,706 (90.96%)	0.0075	524,593 (86.75%)	0.0261
Male	275,109 (53.04%)		237,659 (86.39%)		192,400 (80.96%)
Education
No school at all	160,932 (54.99%)	0.01	144,173 (89.59%)	NS	124,986 (86.69%)	NS
Primary or less	674,309 (64.06%)		603,895 (89.56%)		510,550 (84.54%)
Secondary and above	104,662 (55.90%)		94,298 (90.10%)		81,457 (86.38%)
Marital status
Never married	97,093 (58.26%)	NS	84,480 (87.01%)	NS	69,925 (82.77%)	NS
Married/living together	518,065 (60.29%)		460,795 (88.95%)		400,408 (86.90%)
Divorced/Widowed	324,606 (64.11%)		296,952 (91.48%)		246,522 (83.02%)
Insurance
No insurance	775,934 (59.88%)	0.022	683,460 (88.08%)	<.0001	582,110 (85.17%)	NS
Have insurance	163,969 (69.29%)		158,906 (96.91%)		134,883 (84.88%)
Residential area
Urban	451,756 (64.77%)	0.038	404,476 (89.53%)	NS	346,763 (85.73%)	NS
Rural	488,147 (58.46%)		437,889 (89.70%)		370,230 (84.55%)
HIV discrimination
No	18,186 (27.35%)	<.0001	17,535 (96.42%)	NS	13,784 (78.61%)	NS
Yes	921,717 (62.87%)		824,831 (89.49%)		703,209 (85.26%)
HIV fear
No	173,447 (54.38%)	0.0021	153,969 (88.77%)	NS	134,510 (87.36%)	NS
Yes	716,074 (63.51%)		644,391 (89.99%)		550,385 (85.41%)
HIV shame
No	869,032 (63.45%)	0.0001	779,251 (89.67%)	NS	661,906 (84.94%)	NS
Yes	66,172 (44.35%)		58,960 (89.10%)		52,386 (88.85%)

Weighted logistic regression models were performed on the HIV testing in the population and on the three targets with same set of socio-demographic variables, shown in Table 3. The major issue for achieving three targets by 2020 was to increase the HIV awareness in the population. The socio-demographic variables selected in the analysis were all significantly associated with having HIV testing in the population. Among the three models, each analysis was performed on a subsample of previous analysis.

Table 3.

Weighted Logistic Regression for the three WHO goals

	Target I			Target II			Target III
	N=1,425,972			N=870,992			N=783,712
	OR	95% CI		OR	95% CI		OR	95% CI
Age 25–34 vs 15–24	0.89	0.56	1.42	1.42	0.64	3.15	0.58	0.29	1.15
Age 35–44 vs 15–24	0.47**	0.30	0.73	0.67	0.30	1.49	0.64	0.33	1.26
Age 45–54 vs 15–24	0.36***	0.23	0.56	0.34*	0.13	0.88	0.48	0.23	0.99
Age 55+ vs 15–24	0.54*	0.32	0.93	0.15**	0.04	0.51	0.38*	0.16	0.92
Male vs Female	2.00***	1.57	2.55	2.23***	1.51	3.29	1.96**	1.20	3.20
No school vs secondary+	1.05	0.67	1.67	1.22	0.60	2.51	1.12	0.61	2.08
<primary vs secondary+	0.74	0.50	1.08	1.31	0.65	2.63	2.00*	1.18	3.41
Married vs never married	1.17	0.72	1.88	0.78	0.36	1.69	0.75	0.42	1.35
Divorced vs never married	1.30	0.79	2.14	0.80	0.36	1.81	1.21	0.62	2.36
Have insurance vs not	0.73	0.50	1.06	0.35**	0.17	0.70	1.35	0.84	2.17
Living in urban vs rural	1.09	0.82	1.44	0.89	0.53	1.51	1.01	0.65	1.56
HIV discrimination vs not	0.24***	0.13	0.43	2.16	0.40	11.70	0.64	0.17	2.41
HIV fear vs not	0.80	0.63	1.03	0.87	0.55	1.36	1.10	0.69	1.73
HIV shame vs not	2.03*	1.29	3.21	0.86	0.41	1.81	0.65	0.27	1.59

In the multivariate analysis, adjusting for gender, education, marital status, insurance, living area, HIV related discrimination, fear and shame, age group was significantly associated with all four dependent variables. In population, young adults (age 15–24) were less likely to take the HIV test compared with age groups 25–34 and 35–44. However, young adults were more likely to take the HIV test compared with older adults aged 45–54 and 55+. Among PLWH, young adults were less likely to know their HIV positive status compared with age groups 35–44 and 45–54. Among those who aware of positive status, age groups 45 and older (both 45–54 and 55+) were less likely to report on ART. Among those who reported on ART, older adults (55+) were less likely to have VL suppression compared with young adults (15–24).

In the population, males were less likely to report having taken an HIV test compared to females. However, when reviewed all covariates in three weighted multivariate models together, being male was the very important factor that was associated with 1.9 times the likelihood of knowing their HIV positive status, 2.2 times likely to be on ART, and 2.0 times likely to have VL suppressed when adjusted for age, education, marital status, insurance, living area, and HIV related discrimination, fear and shame. Being female, they were less likely to achieve any of the three targets when adjusted for the other socio-demographic factors.

In THIS data, education was significantly associated with having an HIV test in the population (Table 4). Those who had reported education as secondary and above were more likely to have HIV test in the population. However, education was not significantly associated with HIV positive status awareness and being on ART. Those who had some education but less than primary education were twice likely to have VL suppressed compared with those with secondary education and above. In population, marital status was significantly associated with having an HIV test. Those who were married or living with a partner, and those who were divorced or widowed, were more likely to have an HIV test compared with those who had never been married. However marital status was not significantly associated with all three targets. Those who had insurance were more likely to have an HIV test in the population. Among PLWH, those who had insurance were less likely to know their positive status but associations were non-significant. Hence, among those who knew their status, those who had insurance were less likely to be on ART. Among those who were on ART, having insurance was significantly associated with VL suppression though non-significant. Those who lived in the urban area were less likely to have an HIV test compared to those who lived in the rural area in the population. Although among PLWH, those living in the urban area were more likely to know their status, this association was non-significant.

Table 4.

Weighted Logistic Regression for the HIV test in population

	HIV test in population
	N=28,595,966
	Odds Ratio	95% CI
Age 25–34 vs 15–24	2.90***	2.56	3.29
Age 35–44 vs 15–24	1.58***	1.38	1.82
Age 45–54 vs 15–24	0.74***	0.63	0.87
Age 55+ vs 15–24	0.31***	0.27	0.37
Male vs Female	0.61***	0.57	0.66
No school vs secondary+	0.37***	0.32	0.42
<primary vs secondary+	0.61***	0.55	0.67
Married vs never married	7.10***	6.27	8.03
Divorced vs never married	5.56***	4.77	6.48
Have insurance vs not	1.39***	1.22	1.57
Living in urban vs rural	0.66***	0.59	0.74
HIV discrimination vs not	1.52***	1.34	1.72
HIV fear vs not	1.26***	1.16	1.38
HIV shame vs not	0.74***	0.68	0.81

In the population, people with HIV related discrimination and HIV related fear were more likely to have an HIV test. However, among PLWH, people with HIV related discrimination and HIV related fear were less likely to be aware of their positive status. In the population, those had HIV related shame were less likely to have an HIV test. However, those who felt HIV related shame were more likely to know their positive status and less likely to be on ART and have VL suppressed.

Discussion

We aimed to explore the barriers for achieving 90–90–90 targets in 2020. We used weighted logistic regression analysis to investigate the associations of socio-demographic factors and all targets. The first target provided the foundation for the next two targets. The availability of test in population made it feasible for HIV awareness among PLWH. The accessibility and utility to HIV clinic for PLWH made it possible for those who knew their status to be on ART. Those who were on ART with good adherence to ARVs were more likely to have their VL suppressed. The first goal provides the foundation for the rest two goals. The first 90-goal was merely achieved by 61. When HIV awareness was low among PLWH, the gap to achieving the rest of the goals, that is on ART and viral suppression were wider. It should be misleading to claim the second goal was completed if the first goal was not reached.

This study assessed the potential barriers using interview data from a survey. The association in the weighted logistic regression analysis indicated that HIV related discrimination and HIV related fear were still the leading factors for increasing the HIV awareness rate. We found it consistent with other studies that young women were the majority that were not aware of the HIV status in Tanzania (Fonner, Mbwambo, Kennedy, & Sweat, 2019; Parcesepe et al., 2020). This study suggested that HIV intervention programs focused on scaling-up the HIV testing rate and hence increase the rate of HIV awareness among PLWH were needed. In order to complete the 90–90–90 goal, it is particularly important that intervention programs in Tanzania should focus on HIV testing and that intervention programs should be customized for each goal. Current programs scaling up testing could identify interventions that would then increase linkage to care after diagnosis and monitoring viral suppression. Several interventions were implemented in SSA to scale up testing, treatment and suppression in Tanzania from 2005 to 2016 (Kalinjuma et al., 2020). Additionally, the intervention programs should be designed for each target individually to target different groups. Different targets might require different strategies and there should be no program that could solve all problems.

The analysis indicated the programs targeted to increase the HIV test rate may not be extended automatically to increase the rate of being on ART and viral suppression. Interventions or treatment programs should be designed individually for each target. The intervention programs that have worked in developed countries may not work in developing countries, especially low-income countries in Africa (Caldwell, 2000). For example, in western countries, the current trend is “U=U”, that is undetectable equals untransmittable (Eisinger, Dieffenbach, & Fauci, 2019). The target moved to have VL undetectable (<200 copies/mL) and hence zero transmission. However, though PLWH may live with this virus lifelong, they do not transmit the virus to others if they maintain VL suppression. This is promising and encouraging for PLWH. The population-based survey is critical to understand the current situation in these countries in order to redesign the program. It indicated the road for Tanzania in achieving UNAIDS target by 2030 (ending HIV epidemic) is still long. The HIV prevention work should focus on increasing the HIV screening in population and increasing the proportion of PLWH aware of their HIV status.

One of the limitations of this study was that the blood test only evaluated three ARVs, Efavirenz, Lopinavir, and Nevirapine. Currently new ARVs were available in Tanzania by 2017, including Tenofovir, Lamivudine and Dolutegravir, that is a potent well-tolerated drug (Cahn et al., 2013). The drug concentration in the blood highly depended on the adherence in the past several days. The drug level in plasma might be associated with the adherence to medication but not treatment outcomes (Nwaiwu, Akindele, Adeyemi, Akinleye, & Akanmu, 2019). However, adherence to medication was associated with viral load suppression because suboptimal intake of ART may lead to virological failure (Ntamatungiro et al., 2017). Among those who were on ART, only about half (58%) participants reported they have optimal adherence and with ARV detected in the plasma. Adherence should be the key factor in order for the move from the second target to the third target, especially the keyword in second target was sustained ART.

Another limitation in the variable, HIV awareness, was to combine two levels of not aware together. One was those PLWH had never taken a test. In the population-based survey in Tanzania, there were 26% PLWH who tested but did not know their status. Those who self-reported tested but with unknown status were mostly because the clinic was too far or self-believed low risk for HIV. The easiness of test tools, e.g. home-based test, could increase the awareness rate for this group. Also, there were 13% PLWH had never taken HIV test among those who were not aware of the HIV status. Usually, noted disease symptoms served as a motivating factor for HIV testing (Plotkin et al., 2018) in Tanzania. The symptomatic HIV infection could be a factor associated with high rate of never testing. The population-based massive screening may be the solution of increasing the test rate and hence awareness.

Conclusion

In this paper, we reviewed the importance of HIV awareness, which is the foundation to achieve the goal of ending HIV epidemic in 2030. Though among those who were aware of their positive status, there were more than 90% on treatment. However, the three goals should be reviewed together. It can be extended to the senario of ending the COVID-19 pandemic. The first goal should be among those got infected, there should be at least 90% aware of the status. The second goal should be among those who were aware of their status, there should be at least 90% in the health system and be treated. The third goal should be among those at least 90% in hospital, at least 90% should be cured to stop further transmission. This paper reviewed the efforts in the past decades for ending the HIV epidemic. It sheds a light for current status and recommend the population-based screening should be the first step. From the current lesson, the massive testing and “brute-force strategies” implemented by South Korea (Walensky & del Rio, 2020) might provide an effective solution. Meanwhile, the massive testing manner in vulnerable communities that maximizes the benefits of the community but without increasing stigma and marginalization is crucial. To further investigate the barriers that vulnerable communities and low income families experience to avoid testing should be the next move.