HIV risk behavior and associated factors among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis.

BACKGROUND: HIV risk behavior among people living with HIV/AIDS (PLWHA) is a major public health concern as it increases HIV transmission. In Ethiopia, findings regarding HIV risk behavior have been inconsistent and inconclusive. Therefore, this meta-analysis aimed to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia.
METHODS: International databases, including Google Scholar, Cochrane library, HINARI, Pub Med, CINAHL, and Global Health were systematically searched to identify articles reporting the prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia. The data were analyzed using STATA/SE version-14. The random-effects model was used to estimate the pooled effects. I-squared statistics and Egger's test were used to assess the heterogeneity and publication bias respectively.
RESULTS: A total of 4,137 articles were reviewed and fourteen articles fulfilling the inclusion criteria were included in this meta-analysis. The pooled prevalence of HIV risk behavior in Ethiopia was 34.3%% (95% CI: 28.2, 40.3). Severe heterogeneity was observed between the included research articles (I2 = 96.6, p = 0.000). Alcohol use (OR = 1.9, 95%, CI: [1.6, 2.3]), HIV status non-disclosure (OR = 2.3, 95% CI: [1.3, 4.0]) and perceived stigma (OR = 2.3, 95% CI: [1.3, 4.1]) had a significant association with HIV risk behavior.
CONCLUSION: The prevalence of HIV risk behavior among PLWHA in Ethiopia was high. Alcohol use, HIV status non-disclosure, and perceived stigma had a significant association with HIV risk behavior. In addition to promoting access to Antiretroviral Therapy (ART) treatment and improving medication adherence among PLWHA, various intervention programs focusing on the associated factors have to be implemented to tackle high-risk sexual behavior and go forward toward ending the HIV/AIDS pandemic.

Background

HIV/AIDS remains a major public health challenge to the world population. In 2019, an estimated 37.9 million people were infected with HIV globally and 68% of patients were from sub-Saharan African (SSA) countries [1]. In Ethiopia, an estimated 722,248 people were living with HIV, and 14,872 people died from AIDS-related illnesses in 2017 [2].

The United States Centers for Disease Control and Prevention defines sexual risk behavior as a behavior that increases a person’s risk of contracting sexually transmitted infections (STI) including HIV and/or experiencing unintended pregnancy [3]. It includes one or more of the following characteristics: sex without the use of condoms (condomless sex) [4-7], inconsistent condom use [5, 6, 8, 9], having multiple sexual partners [5, 6], sex with the influence of alcohol [5, 6], casual sex [5-7] and sexual exchange (exchange of money for sexual intercourse) [5, 6].

HIV risk behavior is frequently practiced by PLWHA. The result of the meta-analytic review showed that between 10% and 60% of PLWHA practiced condomless sex [10]. Reports suggested a high incidence of STI [11-13] and continued fertility desire among PLWHA on ART [14, 15], which indicate the practice of (or in the latter case, the need for) sexual intercourse unprotected by a condom. In SSA, more than 1 in 3 PLWHA practiced HIV risk behaviors [16-18]. In Ethiopia, the prevalence of sexual risk behaviors among PLWHA ranged from 15.8% to 56% [7, 19].

PLWHA following ART treatment show general improvements in their overall physical and clinical status and many of them believe that they are no longer infectious since they are on ART [20-22]. There is robust evidence that PLWHA with undetectable viral load cannot transmit HIV through sexual intercourse [23]. However, PLWHA may have challenges in accessing and adhering to ART, which can result in detectable viral loads and the potential for HIV transmission [24-27].

Various studies have found that HIV risk behavior among PLWHA was determined by patient age and sex [28, 29], educational status [18, 30], marital status [31], alcohol consumption [18, 28], duration on ART [31], HIV serostatus disclosure [32-34] and perceived stigma [35, 36].

HIV prevention policies and programs have concentrated heavily on HIV-negative individuals.

However, strategies to decrease the infectiousness of PLWHA to prevent secondary HIV transmission should be an integral part of the prevention policy to achieve the United Nations Sustainable Development Goal (UN SDG) 3.3 which aims to end the epidemics of AIDS by 2030 [37, 38]. Bearing in mind the high rate of HIV infection in low-income countries including Ethiopia, various strategies have been tried so far to decrease high-risk sexual practice among PLWHA, in particular, consistent condom use to prevent further transmission of HIV/AIDS among PLWHA and in the general population [39, 40].

Various studies have been conducted to assess HIV risk behavior and associated factors among PLWHA in Ethiopia [4–9, 19, 41–46]. However, studies had inconsistent findings with the prevalence ranging from 15.8% in the Amhara region in the last twelve months before the survey [7] to 56% in Felege Hiwot Referral Hospital in the last three months preceding the survey [19]. Alcohol use, HIV status non-disclosure, and perceived stigma were factors that merit consideration as ones that could impact sexual behaviors and might be amenable to interventions.

Thus, this meta-analysis aimed to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia. The study generates crucial evidence that will be an input for program planners, policymakers, and health service providers to design and implement evidence-based interventions to reduce the transmission of HIV.

Materials and method

Study design and searching strategies

This systematic review and meta-analysis was conducted to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia. This meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist [47] (S1 Table). Six international databases—Google Scholar, Pub Med, Cochrane Library, HINARI, CINAHL, and Global Health—were searched systematically to find all relevant articles. A snowball search for the references of relevant articles was also performed. In addition, digital libraries were searched to identify grey literature. The search of research articles was carried out from June 1 to September 30, 2020, by three authors (YD, MY, and BK), and articles published from 2000 up to September 30, 2020, were included in the review.

Endnote software was used to collect, organize, and remove the duplications of search outcomes. All potential articles were accessed by using following keywords: “HIV risk behavior", "high risk behavior", "risky sexual practice", "risky behavior", "sexual behavior", "unprotected sexual practice","condom use", "inconsistent condom use", "consistent condom use", "casual sex", "multiple sexual partner", "alcohol use", "alcohol consumption", "HIV status disclosure", "perceived stigma", "associated factors”, "HIV patients", "AIDS patients", "people living with HIV/AIDS", "ART attendees", "HIV positive adults", "Ethiopia" independently and in combination using Boolean operators "OR" or "AND". The search strategy was formulated by using Medical Subject Headings (MESH), adding terms and keywords from a primary search (S1 File).

Inclusion criteria

Population

This meta-analysis included studies conducted among male, female, or both male and female adults living with HIV/AIDS in Ethiopia.

Exposure

PLWHA who drank alcohol, patients who did not disclose HIV serostatus, and those who experienced perceived stigma.

Comparison

PLWHA who didn’t drink alcohol, who disclosed their HIV serostatus, and those who did not experience perceived stigma.

Outcome

Studies assessed HIV risk behavior as a primary outcome, whether same-sex or different-sex risk behavior.

Study design

All types of observational studies (cross-sectional, case-control, and cohort) were included.

Study setting

All facility-based, as well as community-based studies

Time frame

Articles published from the beginning of 2000 up to September 30, 2020

Country

Studies conducted only in Ethiopia

Language

Articles written in the English language

Publication

Published articles were included in this study.

Outcome measurement

The primary outcome of this study was the pooled prevalence of HIV risk behavior among PLWHA in Ethiopia, which was computed by dividing the number of PLWHA practicing HIV risk behavior by the total sample size multiplied by 100.

The second outcome was the association between alcohol use, HIV status non-disclosure, and perceived stigma and HIV risk behavior in the form of the log odds ratio.

Studies measured history of alcohol use differently; three studies measured it as consuming any alcoholic beverage in the last three months [4, 42, 46], one study measured it as ever used alcohol [7], and the remaining study measured it as alcohol consumption in the last 12 months [41].

Primary studies measured HIV serostatus disclosure as "YES" if respondents had disclosed their HIV positive status to their recent sexual partner and "NO" if they had not [4, 19, 41–44, 46].

Primary studies calculated perceived stigma from respondents’ responses to stigma-specific questions. It was measured as "having perceived stigma" if respondents scored greater than or equal to the mean value and “have no perceived stigma” if respondents scored below the mean value [4, 19, 42, 43].

Operational definitions

HIV risk behavior

Is defined as practicing one or more of the following: sex without the use of condoms, inconsistent condom use, having multiple sexual partners, casual sex, sex with the influence of alcohol, and sexual exchange (exchange of money for sexual intercourse).

Inconsistent condom use

Responding "never", "sometimes", and "usually" to questions regarding the frequency of condom use.

Condom less sex

Not using a condom all the time for every act of sexual intercourse.

Multiple sexual partners

Having more than one sexual partner.

Casual sex

Sex with a non-regular sexual partner.

Alcohol use

Alcohol use in this study was measured as ever consuming any alcoholic beverage.

Data extraction

Six reviewers (YD, BA, MA, AM, MG, and WM) independently extracted all the necessary data from the relevant studies using a standardized and a pretested data extraction format prepared in a Microsoft™ Excel spreadsheet. The data extraction format includes the name of the first author followed by initials, region, publication year, study area, study setting (whether it is institution- or community-based), study design, sample size, response rate, the residence of PLWHA, number of PLWHA with HIV risk behavior, prevalence of HIV risk behavior, data collection method, the definition of outcome variable across the included studies, the definition of alcohol use across the included studies and frequencies of alcohol use, HIV status non-disclosure and perceived stigma in the form of two-by-two tables. Disagreement raised at the time of data extraction was resolved through discussion. Whenever possible, the corresponding authors of the research articles were contacted for clarification and additional information.

Quality assessment

After duplicate files were removed, the relevant articles identified from all databases were screened for inclusion by three reviewers (YD, BK, and MY). Then the Joana Brigg Institute (JBI) critical appraisal checklist for prevalence studies was used to assess the quality of each article (S2 File).

The checklist has nine questions with four response options (yes, no, unclear, and not applicable). These nine questions are: Was the sample frame appropriate to address the target population?, Were study participants sampled appropriately?, Was the sample size adequate?, Were the study subjects and the setting described in detail?, Was the data analysis conducted with sufficient coverage of the identified sample?, Were valid methods used for the identification of the condition?, Was the condition measured in a standard, reliable way for all participants?, Was there appropriate statistical analysis? And was the response rate adequate, and if not, was the low response rate managed appropriately?

Three authors (YD, BA, and MA) independently assessed the quality of each study out of 100%. In the JBI quality assessment tool, 50% was used as a cutoff point for the inclusion of research articles [48, 49]. In this study, since all of the included articles scored 50% or more, all are included in the review. The difference in the result at the time of quality assessment was settled by taking the mean score of the results of all reviewers.

Data analysis

The relevant data extracted using a Microsoft Excel spreadsheets were exported into STATA/SE version-14 statistical software for analysis. The heterogeneity among the included studies was statistically estimated by using the I2 test, with a variation in outcomes greater than 75% being taken as high heterogeneity.

Due to the presence of severe heterogeneity between the studies, DerSimonian and Liard’s method of random effect model at a P-value less than 0.05 with a 95% confidence interval was used to estimate the pooled prevalence of HIV risk behavior in Ethiopia [50]. Subgroup analysis was performed by various study level characteristics such as region (Amhara, Addis Ababa, Oromia, and Southern Nation Nationalities and Peoples’ Region [SNNPR]), residence (urban versus urban and rural), sample size (<400 versus ≥400) and the type of outcome assessed among the included articles. Moreover, univariate meta-regression analysis was done using sample size and publication year as factors to identify the possible source of heterogeneity and to decrease the random variations among the point estimates of original articles.

Moreover, a forest plot was used to present the point estimates with their 95% confidence interval, and a log odds ratio was used to determine the association between alcohol use, HIV status non-disclosure, and perceived stigma with HIV risk behavior among PLWHA. Finally, the presence of publication bias was objectively assessed using Egger’s regression test at a p-value < 0.05, which was considered a statistically significant publication bias [51].

Results

Study selection

A total of four thousand one hundred thirty-seven (4,137) published and unpublished studies were identified through electronic databases (Google Scholar, Cochrane Library, Pub Med, CINAHL, HINARI, and Global Health) and a digital library search. Among these, 4,123 articles were excluded as a result of duplication, due to their titles and abstracts, and as a result of not fulfilling the inclusion criteria. Finally, 14 eligible articles were included for analysis (Fig 1).

Fig 1

PRISMA flow diagram describing the selection of studies for systematic review and meta-analysis.

Characteristics of the included studies

All the fourteen studies were institution-based cross-sectional studies published between 2008 and 2020 [4–9, 19, 41–46]. A total of 6,179 PLWHA were involved to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia. The sample size of the included studies ranged from a minimum of 109 samples in a study conducted at Felege Hiwot Referral Hospital, Amhara region [19] to a maximum of 745 samples in a study conducted in the same region [7].

The lowest prevalence of HIV risk behavior (15.8%) was reported by a study done in the Amhara region [7] whereas the highest prevalence of HIV risk behavior (56%) was observed by a study done at Felege Hiwot Referral Hospital, Amhara region [19]. The meta-analysis included 7studies from Amhara region [6–9, 19, 44], 3 studies from Oromia region [41, 43, 45], 3 studies from Addis Ababa [4, 5, 42]and 1 study from Southern Nation Nationalities and Peoples Region (SNNPR) [46] (Table 1).

Table 1

Descriptive summary of fourteen studies included in the meta-analysis of HIV risk behavior and associated factors among PLWHA in Ethiopia, 2020.

Authors	Publication year	Region	Study area	Study design	Sex	Outcome assessed
Dessie et al.[4]	2011	Addis Ababa	Addis Ababa	Cross-sectional	Both	Condom-unprotected sex
Tadesse and Gelagay [5]	2019	Addis Ababa	Addis Ababa	Cross-sectional	Both	Sex without condoms and/or, inconsistent condoms use and/or multiple sexual partners and/or casual sex and/or sex with the influence of alcohol and/or sexual exchange for money
Alene et al. [8]	2014	Amhara	Gondar, Dessie and DebreMarkos	Cross-sectional	Female	Inconsistent condom use
Engedashet et al. [41]	2014	Oromia	Deberezeit	Cross-sectional	Both	Condom-unprotected sex
Demissie et al.[42]	2015	Addis Ababa	Addis Ababa	Cross-sectional	Both	Inconsistent condom use and/or unprotected sex
Molla and Gelagay [6]	2017	Amhara	Gondar	Cross-sectional	Both	Multiple sexual partners and/or casual sex and/or sex without condom and/or inconsistent condom use and/or sex with the influence of alcohol.
Mosisa et al. [43]	2018	Oromia	Nekemte	Cross-sectional	Both	Inconsistent use of condoms and/or no condom-protected sex with HIV-negative
Ali et al. [44]	2019	Amhara	KollaDiba	Cross-sectional	Both	Inconsistent use of condoms
Shewamene et al. [9]	2015	Amhara	Gondar	Cross-sectional	Both	Inconsistent use of condoms
Yalew et al.[19]	2012	Amhara	FelegeHiwot	Cross-sectional	Both	Inconsistent use of condoms
Yalew et al. [19]	2012	Amhara	FelegeHiwot	Cross-sectional	Both	Inconsistent use of condoms
Deribe et al. [45]	2008	Oromia	Jimma	Cross-sectional	Both	Unprotected sex
Moges et al. [7]	2020	Amhara	Amhara region	Cross-sectional	Both	inconsistent condom use and/or not using a condom
Abebo et al. [46]	2019	SNNPR	ArbaMinich	Cross-sectional	Both	Unprotected sexual practice
Definition of alcohol use	Data collection method	Sample size	Response rate	Prevalence (%)	Quality score
Alcohol use in the last three months	Face-to-face interview	601	100	36.9	85.2%
	Face-to-face interview	562	100	39.1	81.5%
	Face-to-face interview	351	98	43.3	63%
Alcohol use in the past one year	Face-to-face interview	667	100	22.2	85.2%
Alcohol use in the last three months	Face-to-face interview	376	100	30.4	66.7%
	Face-to-face interview	513	99	38	77.8%
	Face-to-face interview	337	100	32.9	66.7%
	Face-to-face interview	358	91	43.5	59.2%
	Face-to-face interview	317	100	21.1	66.7%
	Face-to-face interview	113	100	44.2	55.6%
	Face-to-face interview	109	100	56	55.6%
	Face-to-face interview	705	100	24	88.9%
Ever used alcohol	Face-to-face interview	745	98	15.8	88.9%
Alcohol use in the last three months	Face-to-face interview	513	100	52	81.5%

SSNPR: Southern Nations, Nationalities, and Peoples Region

The prevalence of HIV risk behavior in Ethiopia

Analysis of the results of 14 studies showed that the pooled prevalence of HIV risk behavior among PLWHA in Ethiopia was 34.3% (95% CI: 28.2, 40.3). Severe heterogeneity was observed among the included research articles (I2 = 96.6, p = 0.000). As a result, the DerSimonian and Laird random-effects meta-analysis model was used to estimating the pooled prevalence of HIV risk behavior ().

The presence of publication bias was subjectively assessed using a funnel plot. The plot showed an asymmetrical distribution of the effect estimates indicating the presence of publication bias (Fig 3). However, Eggers test statistics confirm the absence of significant publication bias (P = 0.107). Furthermore, a sensitivity analysis was done to identify a single study’s effect on the overall pooled estimate. As a result, no evidence of a single study effect on the overall pooled prevalence was found.

Fig 3

Funnel plot of the pooled prevalence of HIV risk behavior in Ethiopia, 2020.

Subgroup analysis

To identify the variation among the individual studies, subgroup analysis was conducted based on the region where the studies were done, residence, sample size, and the type of outcome assessed across studies. Even though heterogeneity still existed in the subgroup analysis of all parameters mentioned above, the result indicated that the lowest pooled prevalence of HIV risk behavior was observed by studies conducted in Addis Ababa as compared to studies conducted in the SNNPR, Oromia, and Amhara regions [35.7% (95% CI: 30.9, 40.4)] (Table 2).

Table 2

Subgroup prevalence of HIV risk behavior in Ethiopia, 2020 (n = 14).

Variables	Characteristics	Included studies	Sample size	Prevalence (95% CI)	I2	P-value
Region	Amhara	7	2,506	34.8 (24.7, 44.9)	96.9%	<0.001
	Addis Ababa	3	1,539	35.7 (30.9, 40.4)	74.2%	0.021
	Others	4	2,222	32.7 (19.9, 45.6)	97.8%	<0.001
Residence	Rural and urban	10	4061	35.2 (26.9, 43.5)	97.2%	<0.001
	Urban	4	2,206	32.2 (23.9, 40.4)	94.5%	<0.001
Sample size	≤400	7	1,961	35.9 (28.5, 43.4)	92.1%	<0.001
	>400	7	4,306	32.5(23.3, 41.7)	97.9%	<0.001
Type of outcome assessed	Unprotected sex	4	2,486	33.7 (21.1, 46.3)	98.0%	<0.001
	Practicing one or more of the following:✓ Sex without condoms✓ Inconsistent condoms use✓ Multiple sexual partners✓ Casual sex✓ Sex with the influence of alcohol✓ Sexual exchange for money	2	1,075	38.6 (35.7, 41.5)	94.2%	<0.001
	Inconsistent condom use	5	1,248	38.1 (26.6, 49.6)	94.7%	<0.001
	Inconsistent condom use or unprotected sex	3	1,458	26.3 (14.4, 38.2)	96.2%	<0.001

SNNPR-Southern Nation Nationalities and Peoples Region; Others-Oromia and SNNPR.

In addition, the highest prevalence of HIV risk behavior was observed by studies conducted in both urban and rural settings [35.2% (95% CI: 26.9, 43.5)] and studies with a sample size of ≤400 [35.9% (95% CI: 28.5, 43.4)] as compared to studies conducted only in an urban setting and those with a sample size of >400 (Table 2).

In addition, a univariate meta-regression was conducted using sample size and publication year as factors. However, neither of them was found to be statistically significant sources of heterogeneity. Although not significant sources of variability, as the sample size and publication year increased, the prevalence of HIV risk behavior decreased (Table 3).

Table 3

Univariate meta-regression analysis to determine factors related to the heterogeneity of the prevalence of HIV risk behavior in Ethiopia, 2020.

Variables	Coefficient	P-value
Sample size	-0.0336989	0.053
Year of publication	0.2425176	0.792

Factors associated with HIV risk behavior

The association between alcohol use and HIV risk behavior was examined based on the result of five studies [4, 7, 41, 42, 46]. The association was positively significant in three studies [4, 41, 46] and not significant in the other two studies [7, 42]. In this meta-analysis, patients who consumed alcohol were 1.9 times more likely to be engaged in HIV risk behavior as compared to their non-alcohol-consuming counterparts (OR = 1.9, 95%, CI: [1.6, 2.3]). A fixed-effect meta-analysis model was used to examine the association between alcohol use and HIV risk behavior due to the absence of significant heterogeneity (I2 = 23.7%, p = 0.264) (Fig 4). Publication bias was also assessed using Egger’s tests, the result indicating the absence of publication bias (P = 0.439).

Fig 4

The pooled odds ratio of alcohol use among PLWHA in Ethiopia, 2020.

A total of seven studies [4, 19, 41–44, 46] were included to identify the association between HIV status non-disclosure and HIV risk behavior. The association was significant in one study [43] and non-significant in the other studies [4, 19, 41, 42, 44, 46]. The result of the random-effect meta-analysis showed that patients who did not disclose their HIV status to their sexual partner were 2.3 times more likely to be engaged in HIV risk behavior as compared to their counterparts who did disclose their HIV status (OR = 2.3, 95%CI: [1.3, 4.0]) (Fig 5). High heterogeneity was exhibited among the included studies (I2 = 80.4%, p = 0.000) and there was a low possibility of publication bias as indicated by Egger’s tests (P = 0.786).

Fig 5

The pooled odds ratio of HIV status non-disclosure among PLWHA in Ethiopia, 2020.

The effect of perceived stigma on HIV risk behavior was assessed based on the results of four research articles [4, 19, 42, 43]. The effect was positively significant in two studies [19, 43] and non-significant in the other two studies [4, 42]. In this study, patients who experienced perceived stigma were 2.3 more likely to practice HIV risk behavior as compared to patients without perceived stigma (OR = 2.3, 95%CI: [1.3, 4.1]) (Fig 6). The random effect meta-analysis model was used to estimate the pooled effect due to the presence of severe heterogeneity among studies (I2 = 79.2%, p = 0.002) and Egger’s tests showed the absence of publication bias (P = 0.682).

Fig 6

The pooled odds ratio of perceived stigma among PLWHA in Ethiopia, 2020.

Discussion

In this meta-analysis, the pooled prevalence of HIV risk behavior is congruent with studies conducted in Sao Paulo, Brazil (28.7%) [28], South Africa (30%) [52], Togo (34.6%) [18], Tanzania (40%) [53], and another study conducted in Italy (40%) [54]. However, it is higher than the results found by studies conducted in India (13%) [55], Kenya (28%) [17], Croatia (23%) [56], and South Africa (24.2%) [30].

Similarly, the finding is higher than that of other studies conducted in the United States (23%) [57] and Jamaica (25%) [32]. But the finding was lower than studies conducted in Nigeria (56%) [58], Kumasi, Ghana (51%) [59], and Uganda (58.7%) [16] and another study conducted in Nigeria (42.4%) [29]. The variation could be due to differences in the study settings, the differences in their definition used to define HIV risk behavior, time variation, and the difference in socio-economic status, educational status, and cultural and contextual factors.

The other possible causes for the variation are the differences in the length of time used to measure the prevalence of HIV risk behavior, methodological variation (data collection method used and sampling of study participants), and the difference in the concern of different governmental and non-governmental organizations in HIV prevention and control across countries. For example, all studies included in this analysis used face-to-face interviews to collect the data and most of them assessed inconsistent condom use as HIV risk behavior [8, 9, 19, 42–44].

The high rate of high-risk sexual behavior among PLWHA in this meta-analysis has implications for the continued risk of HIV transmission in the country despite the increasing access to ART treatment. Although ART treatment reduces the risk of HIV transmission [60, 61], the finding of this study calls for designing behavioral intervention programs while at the same time scaling up access to Highly Active Antiretroviral Treatment (HAART) to address high-risk sexual behavior and to reduce HIV transmission.

The result of subgroup analysis indicated that the lowest pooled prevalence of HIV risk behavior was observed by studies conducted in Addis Ababa as compared to studies conducted in SNNPR, Oromia, and Amhara regions. This could be due to differences in educational status, media exposure, and difference in knowledge of HIV prevention methods and comprehensive knowledge about HIV/AIDS. For instance, according to Ethiopian Demographic and Health Survey (EDHS) 2016, respondents who lived in Addis Ababa had better knowledge of HIV prevention methods and comprehensive knowledge about HIV/AIDS than respondents who lived in SNNPR, Oromia, and Amhara regions [62].

The likelihood of engagement in HIV risk behavior was higher among PLWHA who consumed alcohol as compared to their non-drinking counterparts. This is in agreement with studies conducted in South Western Uganda [63], South Africa [52], and New Guinea [64]. The results of studies conducted in Sao Paulo, Brazil [28], Togo [18], and another study conducted in Uganda [16] also showed that alcohol consumption had a positive association with HIV risk behavior.

The finding of a systematic review and meta-analysis conducted in the United States also witnessed this association [65]. This association may be because alcohol use can inhibit an individual’s perception of the risk of HIV transmission and hinder thinking and decision-making ability about safe sex as a result of alcohol’s restricting effect on cognitive capacity causing someone to focus only on impelling immediate cues [66]. Bearing in mind the strong association between alcohol use and HIV risk behavior, clinicians need to integrate alcohol counseling into routine HIV care for PLWHA. It is also necessary to implement policies and strategies focusing on decreasing alcohol consumption and HIV risk behavior among PLWHA.

Non-disclosure of HIV serostatus to sexual partners had a significant association with HIV risk behavior. This is consistent with studies conducted in Jamaica [32], Cape Town, South Africa [33], and Cameroon [34] where HIV risk behavior was reported mostly among patients who did not disclose their HIV status to sex partners. A similar finding was also documented in a study conducted in Johannesburg, South Africa [30]. This could be due to the possibility that PLWHA who did not know the serostatus of their sexual partners’ might not be driven to use a condom at the time of sexual intercourse.

Disclosing HIV status is the key component of HIV prevention as it encourages partners to know each other’s HIV status, increases adherence to therapy, and improves the rate of CD4 recovery following ART treatment. Studies showed that HIV status disclosure decreases the risk of HIV transmission by 17.9% to 40.6% [67, 68]. However, a recent systematic review and meta-analysis study conducted in Ethiopia showed that 25.6% of HIV patients did not disclose their HIV status to their sexual partners [69]. More health education interventions including extensive adherence counseling of all PLWHA about the benefit of the disclosure are required to increase the HIV status disclosure rate and to decrease high-risk sexual behavior among PLWHA.

Experiencing perceived stigma also had a significant effect on HIV risk behavior. The finding is in line with studies conducted in India and the United States [35, 36]. This might be due to the possibility that individuals who have experienced perceived stigmatization could be afraid to disclose their HIV serostatus to their sexual partners. In Ethiopia, HIV-related stigma is the most common problem affecting 42–72.2% of HIV patients despite the effort made by NGOs and the government [70, 71]. Stigma not only affects the sexual behavior of HIV patients but also hinders ART adherence and thereby hastens disease progression. Thus, the government and all other concerned bodies should be a voice for PLWHA and do more to confront and reduce HIV-related stigma.

Like other studies, the results of this meta-analysis must be interpreted carefully in accordance with the strengths and limitations of the included studies. As far as the authors know, this systematic review and meta-analysis was the first study that attempts to assess the pooled prevalence of HIV risk behavior among PLWHA in the Ethiopian context, although it has some limitations. As a limitation, it is restricted to articles published in the English language and it may not represent research published in other languages.

The approaches used to measure independent and outcome variables differ across the included studies. Besides, all of the articles included in the analysis were cross-sectional and had a small sample size and this may have affected the pooled estimate. In addition, heterogeneity was detected across all analyses even if we conducted subgroup analyses and meta-regression. Moreover, the study may not be representative of all regions since the included articles were only from one city administration and three regions of Ethiopia.

Conclusions

The pooled prevalence of HIV risk behavior among PLWHA in Ethiopia was high. Alcohol use, HIV status non-disclosure, and perceived stigma were factors significantly associated with HIV risk behavior. Besides increasing access and adherence to ART treatment, the national program against HIV/AIDS should enhance health promotion activities including consistent condom use and the creation of a positive social environment to mitigate high-risk sexual practices among PLWHA. Screening and treatment of individuals with alcohol use disorder should be an integral part of the routine HIV treatment, care, and support package and health care providers should offer risk reduction counseling, especially for clients who use alcohol and/or experience perceived stigma, including encouraging disclosure of HIV serostatus among patients whose sexual partners are of unknown status.

Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist.

(DOC)

Click here for additional data file.

Search strategy used to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia.

(DOCX)

Click here for additional data file.

JBI critical appraisal checklist for prevalence studies.

(DOCX)

Click here for additional data file.

The data set used to estimate the pooled prevalence of HIV risk behavior and associated factors among PLWHA in Ethiopia.

(XLSX)

Click here for additional data file.

6 Jun 2021

Submitted filename: Response to reviewer.docx

Click here for additional data file.

17 Nov 2021

Risky sexual practice and its association with alcohol intake, HIV status disclosure, and perceived stigma among adult HIV infected patients in Ethiopia. A systematic review and meta-analysis

PLOS ONE

Dear Dr. Adane,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I agree with the Reviewers' comments and addressing these will be a requirement for acceptance. In addition, the authors must revisit the concept of "risky" sexual behavior prior to acceptance. People living with HIV who use antiretroviral therapy and attain a suppressed viral load cannot transmit HIV through sexual contact, and this is a critical concept that is not highlighted in the article, and which must be integrated into any study on sexual behaviors among people living with HIV (or those who have sex with members of this population). Further, the term "risky" is stigmatizing, and so it is no longer appropriate for describing sexual behaviors. The term "HIV risk behaviors" is more acceptable. The authors have also not adequately discussed the reasons why people engage in what they define as "risky" behaviors; is it for sexual pleasure, because of alcohol use disorders, or other reasons that can help inform ways to improve HIV prevention intervention in Ethiopia? Finally, I agree that the services of a copy editing expert and more detailed proofreading will be essential to ensure that the paper is easy to read and suitable for publication in this journal.

Please submit your revised manuscript by Dec 30 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Douglas S. Krakower, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: I Don't Know

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Title: change “adult HIV infected patients” to “adults living with HIV”

Key words: revise to remove standard terms i.e. prevalence.

Terminology: use correct terminology for “HIV infected patients”; use consistently: PLWHA

Abstract

Line 23: risky sexual practice does not increase HIV incidence; it increases the risk of HIV transmission.

Background

Line 53: CDC only used once so no need for abbreviation.

Line 54: lower case for “Sexually Transmitted Infection”. One does not contract AIDS but HIV, please edit the sentence.

Line 60: Use abbreviation PLWHA

Line 80-83: Consider revising or add to methods or discussion.

Methodology

Line 118-124: What was the inclusion criteria? The section on inclusion criteria is not clear. Were studies included limited to those conducted in Ethiopia?

Line 131-133: revise sentence e.g. Studies measured alcohol intake as “taking any alcoholic beverage in the last three months” [24], one study used “ever used alcohol” [35] and the remaining one study measured it as “alcohol consumption in the last 12 months” [27]

Line 135: Risky sexual practice?

Line 140: Six reviewers….

Line 141: reference the pretested data extraction format.

Line 147-149: Were there any corresponding authors that were contacted for clarification?

Line 192-194: Were there any regional differences worth discussing?

Results

Line 197: typo: Severe heterogeneity….

Line 209-210: Revise the first two sentences. Was this meant to be a standalone paragraph?

Line 216-217: Please add a comment in your discussion regarding regions.

Line 222: Use scientific writing… What does “a little bit high” mean? Please revise.

Line 229: For OR, please use 1 decimal point throughout and be consistent.

Discussion

Do not repeat your results in your discussion. You can still highlight specifics but limit repetition of results.

Line 253-254: Is 39% prevalence high? Significant?

Reviewer #2: This a meta-analysis of the prevalence of sexual risk behavior among people living with HIV in Ethiopia. The researchers also examined the relationship between sexual risk behavior and alcohol use, stigma, and HIV disclosure. Although the study covers an important topic, it has some conceptual and methodological problems.

The researchers justify the study by indicating that it will solve controversies regarding the factors that affect sexual risk behaviors in Ethiopia. However, there is no discussion of the differences the study intends to address. Moreover, although alcohol use and stigma may affect risk behaviors, moderators such as such as age, gender, access to treatment, and type of partner are likely to explain the heterogeneity in the reviewed studies. In fact, the characteristics and behaviors of the sample included in the review are not described.

The description of the outcome and predictor variables is incomplete. Risk behaviors include number of sexual partners, inconsistent condom use, and having had casual partners. How were these standardized? What happened with studies that included multiple risk-behavior measures or reported continuous outcomes (e.g., average number of sexual partners)? Given the high heterogeneity in the studies and the diverse indicators of risk behaviors, it would be important to conduct moderator analyses in terms of type of outcome measure, in addition to key demographic characteristics aside from the study setting.

How were stigma and HIV status disclosure measured in the primary studies? Were the variables dichotomous?

The description of the quality assessment is incomplete, how many studies were excluded because of the quality assessment?

In the discussion, the extensive descriptions of the levels of risk behavior in different regions needs to be streamlined so it is easier to compare and interpret findings.

It is difficult to conclude that alcohol use affects sexual risk behaviors, given the measures that are included in the studies (e.g., lifetime alcohol use, alcohol use last 12 months).

I would recommend having the paper revised by a native English speaker or professional translator. There are lines that sound awkward or are unclear, and/or need to be revised in terms of word choice and punctuation. There is also some inconsistency in the use of terms. In addition, some paragraphs would benefit from reorganization.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Tafireyi Marukutira

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

29 Dec 2021

Douglas S. Krakower, MD

Academic Editor, PLOS ONE

RE: Manuscript ID: PONE-D-21-17808 (HIV risk behavior and its association with alcohol intake, HIV status non-disclosure, and perceived stigma among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis.)

Dear Dr. Douglas S. Krakower,

Thank you very much for your email and the comments/suggestions of the reviewers and academic editor. We have looked at the comments and have revised our paper accordingly. We hope our paper improved as a result of incorporating the reviewers' and academic editor's comments and suggestions.

Please find for your kind consideration the following:

�  A rebuttal letter that responds to each point raised by the academic editor and reviewer labeled 'Response to Reviewers'. The point-by-point responses are written in italic font style.

�  A revised manuscript with track changes labeled 'Revised Manuscript with Track Changes'.

�  A revised paper without tracked changes labeled 'Manuscript'

While hoping that these changes would meet with your favourable consideration, we are happy to hear if there are more comments and suggestions. Please do not hesitate to let us know if you have any questions.

Yours Sincerely,

Metadel Adane

Department of Environmental Health, Wollo University

Dessie, Ethiopia

Tel:+251910336962

E-mail: metadel.adane2@gmail.com

Point by point response

Editor Comments:

I agree with the Reviewers' comments and addressing these will be a requirement for acceptance. In addition, the authors must revisit the concept of "risky" sexual behavior prior to acceptance. People living with HIV who use antiretroviral therapy and attain a suppressed viral load cannot transmit HIV through sexual contact, and this is a critical concept that is not highlighted in the article, and which must be integrated into any study on sexual behaviors among people living with HIV (or those who have sex with members of this population). Further, the term "risky" is stigmatizing, and so it is no longer appropriate for describing sexual behaviors. The term "HIV risk behaviors" is more acceptable. The authors have also not adequately discussed the reasons why people engage in what they define as "risky" behaviors; is it for sexual pleasure, because of alcohol use disorders, or other reasons that can help inform ways to improve HIV prevention intervention in Ethiopia? Finally, I agree that the services of a copy editing expert and more detailed proofreading will be essential to ensure that the paper is easy to read and suitable for publication in this journal.

Response: Thank you for your constructive comment. We have tried to highlight this critical concept in the introduction section line 58-67. In addition, we have tried to use the term “HIV risk behavior” throughout the document. Furthermore, efforts were made to revise our document by language expert.

Reviewer #1:

1. Title: change “adult HIV infected patients” to “adults living with HIV”

Response: Thank you dear reviewer. We have amended it.

2. Key words: revise to remove standard terms i.e. prevalence.

Response: Thank you. We have removed it.

3. Terminology: use correct terminology for “HIV infected patients”; use consistently: PLWHA

Response: Thank you for your comment. We have tried to use the term “PLWHA” throughout our document accordingly.

Abstract

4. Line 23: risky sexual practice does not increase HIV incidence; it increases the risk of HIV transmission.

Response: Thank you very much for your constructive comments. We have tried to modify it accordingly.

Background

5. Line 53: CDC only used once so no need for abbreviation.

Response: Thank you for your suggestion. We have amended it.

6. Line 54: lower case for “Sexually Transmitted Infection”. One does not contract AIDS but HIV, please edit the sentence.

Response: Thank you. We have modified it accordingly.

7. Line 60: Use abbreviation PLWHA

Response: Thank you. The comment is accepted and addressed accordingly.

8. Line 80-83: Consider revising or add to methods or discussion.

Response: Thank you. We have revised it.

Methodology

9. Line 118-124: What was the inclusion criteria? The section on inclusion criteria is not clear. Were studies included limited to those conducted in Ethiopia?

Response: Many thanks for your comment. It is editorial problem. Studies included in this meta-analysis are limited to those conducted in Ethiopia. These are additional inclusion criteria’s.

Population: This meta-analysis include studies conducted among male, female or both male and female adults living with HIV in Ethiopia.

Exposure: PLWHA who took alcohol, disclosed their HIV sero-status and those who experienced perceived stigma.

Comparison: PLWHA who didn’t take alcohol, patients with HIV sero-status non-disclosure and those without perceived stigma.

Outcome: Studies assessed risky sexual practice as a primary outcome; whether it is same-sex or different-sex risk behavior.

Study design: all types of observational studies (Cross-sectional, case-control, and cohort) were included.

Study setting: all facility-based, as well as community-based studies

Time frame: articles published from the beginning of 2000 up to September 30, 2020

Country: studies conducted only in Ethiopia

Language: articles written in the English language

Publication: both published and unpublished articles were included in this study.

10. Line 131-133: revise sentence e.g. Studies measured alcohol intake as “taking any alcoholic beverage in the last three months” [24], one study used “ever used alcohol” [35] and the remaining one study measured it as “alcohol consumption in the last 12 months” [27]

Response: Thank you. We have tried to revise it accordingly.

11. Line 135: Risky sexual practice?

Response: Definitely, it is to mean risky sexual practice.

12. Line 140: Six reviewers….

Response: Thank you for your suggestion. We have corrected it.

13. Line 141: reference the pretested data extraction format.

Response: Thank you for your comment. But the comment is not clear for us. How can we reference it? If you want to see the format, the dataset is already uploaded as a supporting information so that you can see the data extraction form from the dataset. If this is not the case, make it clear for the next revision so that we will entertain it accordingly.

14. Line 147-149: Were there any corresponding authors that were contacted for clarification?

Response: Yes.

15. Line 192-194: Were there any regional differences worth discussing?

Response: Thank you for your valuable comment. Although it was not significant, the pooled prevalence of HIV risk behavior was relatively low among studies conducted in Addis Ababa as compared to studies conducted in SNNPR, Oromia and Amhara region.

Results

16. Line 197: typo: Severe heterogeneity….

Response: Thank you. We have amended it.

17. Line 209-210: Revise the first two sentences. Was this meant to be a standalone paragraph?

Response: Thank you for your suggestion. We have tried to revise it.

18. Line 216-217: Please add a comment in your discussion regarding regions.

Response: Thank you. We have tried to discuss the subgroup prevalence of HIV risk behavior across regions accordingly in line 235-237.

19. Line 222: Use scientific writing… What does “a little bit high” mean? Please revise.

Response: Thank you for your constructive comment. We have tried to revise it accordingly.

20. Line 229: For OR, please use 1 decimal point throughout and be consistent.

Response: Thank you. The comment is accepted and addressed accordingly.

Discussion

21. Do not repeat your results in your discussion. You can still highlight specifics but limit repetition of results.

Response: Thank you. We have amended it.

22. Line 253-254: Is 39% prevalence high? Significant?

Response: Yes. It is high as compared to different studies conducted before.

Reviewer #2:

1. This a meta-analysis of the prevalence of sexual risk behavior among people living with HIV in Ethiopia. The researchers also examined the relationship between sexual risk behavior and alcohol use, stigma, and HIV disclosure. Although the study covers an important topic, it has some conceptual and methodological problems.

Response: Thank you. We have tried to modify our document accordingly.

2. The researchers justify the study by indicating that it will solve controversies regarding the factors that affect sexual risk behaviors in Ethiopia. However, there is no discussion of the differences the study intends to address. Moreover, although alcohol use and stigma may affect risk behaviors, moderators such as such as age, gender, access to treatment, and type of partner are likely to explain the heterogeneity in the reviewed studies. In fact, the characteristics and behaviors of the sample included in the review are not described.

Response: Thank you for your important comment. We have tried to discuss the differences in the result of the primary studies regarding the factors that affect HIV risk behaviors in result section line 244-245, 253-254 and 262-263. Moreover, we have tried to conduct moderator analysis based on the sex of study participants. However, it is difficult to conduct moderator analysis based the patient age, access to treatment since all are on ART and type of sexual partner since the sexual partner is not specified on primary studies. However, the characteristics and behaviors of the samples included in the review are described on Table 1.

3. The description of the outcome and predictor variables is incomplete. Risk behaviors include number of sexual partners, inconsistent condom use, and having had casual partners. How were these standardized? What happened with studies that included multiple risk-behavior measures or reported continuous outcomes (e.g., average number of sexual partners)? Given the high heterogeneity in the studies and the diverse indicators of risk behaviors, it would be important to conduct moderator analyses in terms of type of outcome measure, in addition to key demographic characteristics aside from the study setting.

Response: Thank you for your important comment. We have tried to write the description of the outcome and predictor variables in method section line 135-152. In addition, we have conducted moderator analyses in terms of type of outcome measure and the sex of study participants as indicated in Table 3. But, we did not attempt to standardize definitions of number of sexual partners, inconsistent condom use, and having had casual partners rather we used them as used in each of the articles selected. None of the articles included in the review reported continuous outcomes.

4. How were stigma and HIV status disclosure measured in the primary studies? Were the variables dichotomous?

Response: Thank you for your constructive comment. Both stigma and HIV status disclosure were dichotomous variables. Primary studies measured HIV sero-status disclosure as “YES” if respondents told their HIV positive status to their recent sexual partner and “NO” if they didn’t. Primary studies calculated perceived stigma from respondents’ response to stigma-specific questions. It was measured as “having perceived stigma” if respondents scored greater than or equal to the mean value and “have no perceived stigma” if respondents scored below the mean value.

5. The description of the quality assessment is incomplete, how many studies were excluded because of the quality assessment?

Response: Thank you for your important comment. We have tried to revise the quality assessment section. In our study, all of the included articles scored 50% and more, thus all are included in the review.

6. In the discussion, the extensive descriptions of the levels of risk behavior in different regions needs to be streamlined so it is easier to compare and interpret findings.

Response: Thank you for your comment. We have tried to discuss the levels of risk behavior across regions in the discussion section line 286-292

7. It is difficult to conclude that alcohol use affects sexual risk behaviors, given the measures that are included in the studies (e.g., lifetime alcohol use, alcohol use last 12 months).

Response: Thank you for your constructive comment. It is the amount of alcohol used (rather than the timing of alcohol use) that largely determine the sexual behaviour of PLWHA. Since the amount of alcohol determine the HIV risk behaviour of PLWHA, It is also difficult to conclude that alcohol use in the last 3 months affects HIV risk behavior. In our opinion, it is PLWHA who have ever used alcohol and those used alcohol in the last 12 months who are likely to adapt and maintain the behaviour (alcohol use) and have a chance to be addicted with alcohol and hence practice sexual risk behaviour as compared to those who used alcohol recently. Due to these reason we can conclude that lifetime alcohol use, alcohol use last 12 months can affect sexual risk behaviors.

8. I would recommend having the paper revised by a native English speaker or professional translator. There are lines that sound awkward or are unclear, and/or need to be revised in terms of word choice and punctuation. There is also some inconsistency in the use of terms. In addition, some paragraphs would benefit from reorganization.

Response: The whole parts of the manuscript was revised by language expert for typing, formatting, word choice and punctuation issues. We also have tried to reorganize paragraphs and use terms consistently.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

1 Apr 2022

16 May 2022

Douglas S. Krakower, MD

Academic Editor, PLOS ONE

RE: Manuscript ID: PONE-D-21-17808R1 (HIV risk behavior and its association with alcohol intake, HIV status non-disclosure, and perceived stigma among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis.)

Dear Dr. Douglas S. Krakower,

Thank you very much for your email and the comments/suggestions of the reviewers and academic editor. We have looked at the comments and have revised our paper accordingly. We hope our paper improved as a result of incorporating the reviewers' and academic editor's comments and suggestions.

Please find for your kind consideration the following:

�  A rebuttal letter that responds to each point raised by the academic editor and reviewer is labeled 'Response to Reviewers'. The point-by-point responses are written in italic font style.

�  A revised manuscript with track changes labeled 'Revised Manuscript with Track Changes'.

�  A revised paper without tracked changes labeled 'Manuscript'

While hoping that these changes would meet with your favorable consideration, we are happy to hear if there are more comments and suggestions. Please do not hesitate to let us know if you have any questions.

Yours Sincerely,

Metadel Adane

Department of Environmental Health, Wollo University

Dessie, Ethiopia

Tel:+251910336962

E-mail:metadel.adane2@gmail.com

Point by point response

Editor’s comment

Comment 1: Addressing the most recent reviewers' comments that relate to the statistical methods used and their interpretation and discussion, to ensure methodological rigor in the final published manuscript.

Response: Thank you, dear editor. We have tried to address reviewers' comments accordingly.

Comment 2: The term "HIV risk behavior" is currently used in an imprecise manner in the paper. Please be specific in the behaviors that you are describing. If you mean condom-less sex, sex with multiple partners, or some other behavior, please mention these in specific. "Unprotected" sex is not specific, as PLWH who have a suppressed viral load cannot transmit HIV to their sexual partners, so sex without condoms in this scenario is "protected" from HIV transmission (even if not from other STIs or pregnancy); please change to "condomless sex" if that is the intended meaning.

Response: Thank you. Studies included in this meta-analysis did not address the same behavior. Some assessed risky sexual behavior, others assessed inconsistent condom use, condom unprotected sex, etc.….. So it is difficult to specify exactly those behaviors that are being addressed since the included studies assessed different behaviors. So the term that represents all these behaviors is risky sexual behavior. We used the term risky sexual behaviors to represent all these behaviors since risky sexual behavior according to the Center for Disease Control definition includes one or more of the following behaviors: sex without the use of condoms, inconsistent condom use, having multiple sexual partners, casual sex, sex with the influence of alcohol, and sexual exchange (exchange of money for sexual intercourse). Moreover, we have tried to use condomless sex to represent unprotected sex throughout the document.

Comment 3: In line 60, the term "premise" is not appropriate, as this suggests that the assertion that follows is uncertain or untested. Please change to: "There is robust evidence that PLWH who are undetectable on ART cannot transmit HIV to their sexual partners."

Response: Thank you, dear reviewer. We have amended it accordingly.

Comment 4: Lines 61 to 67 suggest that the acquisition of drug-resistant HIV strains among PLWH is a major factor in the need for second-line ART. The totality of the literature does not support this assertion. Further, the references cited are generally more than a decade old, which ignores a much more robust set of studies from 2010 onward that do not support the authors' statements. Without a more evidence-based and updated argument, the introduction is not appropriate for publication. It would be appropriate to state that "However, PLWH may have challenges in accessing and adhering to ART, which can result in detectable viral loads and the potential for HIV transmission" with the support of new references.

Response: Thank you, dear reviewer. The comment is accepted and addressed accordingly.

Comment 5: Lines 80-82: The authors do not justify why the factors listed are the "most fundamental to intervene," and this strong assertion needs to be moderated. It would be more appropriate to state that these factors merit consideration as ones that could impact sexual behaviors and might be amenable to interventions.

Response: Thank you for your constructive comment. We have amended it accordingly.

Comment 6: Line 129: the authors need to justify why unpublished studies were included in the analyses.

Response: Thank you dear reviewer for remembering an important issue. All the studies included in this meta-analysis were published articles. However, during the initial searching process, we found one unpublished study which has been posted in the research square. Due to this reason, we have said that “both published and unpublished studies were included in the review”. But finally, when we checked its publication status, it was already published in the journal. So in this study, only published studies were included. We have tried to revise the inclusion criteria in our revised manuscript.

Comment 7: Lines 136-139: The authors need to state clearly how they operationalized alcohol use given the heterogeneous definitions used in prior studies. Because the definition for this and other covariates of interest, such as HIV status disclosure, were heterogeneous, this needs to be discussed as a major limitation when considering the study findings in the Discussion.

Response: Thank you, dear reviewer. We have tried to revise the operational definition and discuss it as a limitation in the discussion section accordingly.

Comment 8: In the Discussion, as above, the authors need to be far more precise when discussing HIV risk behaviors, and must specify exactly those behaviors that are being addressed, as opposed to the general term of "HIV risk behaviors," which can mean many different things. Unless all of the studies in lines 278-285 addressed the same behaviors, then it is not appropriate to make quantitative comparisons across studies as the authors have done.

Response: Thank you for your constructive comment. As we have stated in the document, different literature defined risky sexual behavior as a behavior that includes one or more of the following behaviors: sex without the use of condoms, inconsistent condom use, having multiple sexual partners, casual sex, sex with the influence of alcohol, and sexual exchange (exchange of money for sexual intercourse). Studies included in this meta-analysis did not assess the same behaviors. Some assessed risky sexual behavior and others assessed inconsistent condom use or condom unprotected sex. So it is difficult to specify exactly those behaviors that are being addressed since they assessed different behaviors. Due to this reason, we used the term HIV risk behaviors to represent all these behaviors. We acknowledged it as a limitation in the revised manuscript.

Comment 9: Line 39 and elsewhere: the term "safe sex" is outdated, non-specific, and not appropriate. Please remove this and specify exactly what is meant, such as how alcohol could influence decisions around condom use, status disclosure, or other factors that can affect HIV transmission.

Response: Thank you, dear reviewer. We have amended it throughout the document.

Comment 10: Line 347: The term "alcohol abuse" is stigmatizing; please change it to "alcohol use" or "alcohol use disorders" or similar. As above, please also remove "unsafe" as a term.

Response: Thank you. The comment is accepted and addressed accordingly.

Comment 11: The Discussion and Conclusion are missing a critical emphasis on the importance of promoting access and adherence to ART as a way to decrease HIV transmission, as suppressed viral loads are the most effective way to prevent transmission - even more than condoms or any social-behavioral interventions. The authors can discuss how counseling about alcohol, disclosure, and stigma is important to decrease HIV transmission among those without the use of ART, but ART needs to be mentioned front and center given the immense strength of evidence behind this biomedical strategy.

Response: Thank you. We have tried to revise the discussion and conclusion section accordingly.

Reviewer #3

Comment 1: I read with great interest the Manuscript titled "Risky sexual practice and its association with alcohol intake, HIV status disclosure, and perceived stigma among adult HIV infected patients in Ethiopia. A systematic review and meta-analysis" which falls within the aim of PLOSE ONE Journal. In my honest opinion, the topic is interesting enough to attract the readers' attention. The methodology is accurate and conclusions are supported by the data analysis. Nevertheless, the authors should clarify some points and improve the operational definition of some variables.

Response: Thank you dear reviewer for your constructive comments. We have tried to revise the manuscript based on the comments of the reviewer accordingly.

Comment 2: The rationale that the authors stated is not consistent across the document. For instance, in the abstract, the authors said “ this meta-analysis aimed to estimate the pooled prevalence of the risky sexual practice and its association with……" but in the background section page 4, lines 80-81 the authors said, " However, we only examine the association between risky sexual behavior and alcohol intake, HIV status disclosure, and perceived stigma since they were the most frequently mentioned factors and have controversial findings among the included studies”.

Response: Thank you. We have amended it accordingly.

Comment 3: How the authors operationalized alcohol consumption, perceived stigma, and HIV status disclosure is not clear. Their definition across the studies might be different and the authors should add one table concerning the definition of these variables in each study included in this review.

Response: Thank you for your suggestion. The definition of perceived stigma and HIV status disclosure is similar across the included articles due for this reason, there is no need to put their definition in table 1. However, we have tried to put the definition of alcohol use across the studies in table 1 of our revised manuscript.

Comment 4:In the abstract section, the authors said “In Ethiopia, findings regarding risky sexual practice have been inconsistent and inconclusive". However, they said nothing about this in the background section. Please, show us the inconsistency you found by elaborating on this section.

Response: Thank you. We have tried to show the inconsistencies in the background section accordingly.

Reviewer #4:

Comment 1: This is a good manuscript, clearly written, with the correct methodology and correct statistics for meta-analysis. However, it needs revision to be improved.

Response: Thank you, dear reviewer. We have tried to revise our manuscript according to your comment.

Comment 2: Tile: It was too long! "It’s association with alcohol intake, HIV status non-disclosure, and perceived stigma" This listed associated factor can be written concisely.

Response: Thank you. We have modified the title to "HIV risk behavior and associated factors among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis.”

Abstract:

Comment 3:International databases such as Google Scholar, the Cochrane Library, HINARI, PubMed, CINAHL, and Global Health. Good to identify which one is a database and which one is a search engine? Furthermore, why are you using HINARI and Global Health? It is good to consider if you consider others, i.e., AJOL, WorldCat

Response: Thank you, dear reviewer. Except for google scholar; Cochrane Library, HINARI, PubMed, CINAHL, and Global Health are databases. We used HINARI and Global Health since they are easily accessible (Wollo university has the password for the HINARI database due to this we can easily access other databases via HINARI). We have already missed AJOL and WorldCat in the current meta-analysis. We will consider them for our future experience.

Comment 4:Why did you use the word "prevalence" for "HIV risk behavior"?

Response: Thank you, dear reviewer. Some studies included in this study used the word "prevalence". So, we used this term in order not to miss those studies.

Background

Comment 5: "There is also a premise that PLWHA who achieve and maintain viral load suppression cannot transmit HIV to their HIV-negative sexual partners [11]." I suggested instead of maintaining viral load suppression, changing to "undetectable viral load cannot pass HIV on through sex".

Response: Thank you for your suggestion. We have amended it accordingly.

Comment 6:There is previous SR and MA has already been performed (eg. Mekuriaw, B., Belayneh, Z., Molla, A. et al. A meta-analysis and systematic review20, 55 (in 2021).https://doi.org/10.1186/s12954-021-00503-6). What is the difference between this previous work (alcohol use and its determinants among adults living with HIV/AIDS) and the current work (alcohol use on HIV risk behavior)? Please provide information about the heterogeneity of this previous MA.

Response: Thank you. There is a big difference between the previous work and the current one. The outcome variable for the previous work was alcohol use among PLWHA and the outcome variable for the current study is HIV risk behavior. Alcohol use was used as an independent predictor of HIV risk behavior in the current study as opposed to the previous work which was used as an outcome variable.

Comment 7: More description of the "core problem" is required than justification.

Response: Thank you. The comment is accepted and addressed accordingly.

Comment 8: "Moreover, there is no single country-level figure estimating the pooled prevalence of HIV risk behavior among PLWHA in Ethiopia." Take a look at # 2 for an example.

Response: Thank you for your suggestion. We have amended it accordingly.

Methods:

Comment 9:Is there any amendments or changes? Did the authors write a protocol before doing this SR research (if any)? Additionally, there is a major difference in the description from the PROSPERO registration (CRD42020160018) e.g. The title at PROSPERO is "systematic review and meta-analysis on the magnitude and determinants of inconsistent condom use among adult HIV patients in Sub-Saharan Africa," which is very different from the title, which sets "HIV risk behavior and its association with alcohol intake, HIV status non-disclosure, and perceived stigma among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis." Please provide an explanation and revision at PROSPERO.

Response: Thank you for your constructive comment. Initially, we proposed to do SR research entitled "Systematic review and meta-analysis on the magnitude and determinants of inconsistent condom use among adult HIV patients in Sub-Saharan Africa," which was registered at PROSPERO with a registration number of CRD42020160018. However, due to the small number of articles, we have changed the title to "HIV risk behavior and its association with alcohol intake, HIV status non-disclosure, and perceived stigma among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis.” We used the registration number of the former title to the current one without amending the protocol which may mislead the reader. We acknowledged the problem and have removed the registration status of the title from our manuscript.

Comment 10: Search strategies: Why did the authors not use mesh terms in their search strategy? Moreover, the search strategies must be clearly stated in the annex (S2 File). You mention only PubMed, why not others? It should be clearly shown the search strategies by "PICO" or "PECO" or "CoCoP" and finally, use Boolean terms like "AND" or "OR." Then put in the exact date and time of the search with its findings. How do you get gray literature?

Response: Thank you for your comment. We have amended the search strategy and tried to put the findings from each searching database in S2 File accordingly. Gray literature were identified by traditional Google search and Addis Ababa digital library search.

Comment 11: Why did the authors collect studies only from 2000 to 2020? Shouldn't the result section have included data from 2008?

Response: Thank you. We have tried to collect articles published from the beginning of 2000 to September 30, 2020, to get more articles assessing HIV risk behavior. However, during the searching processes, we found only articles published from 2008-to 2020.

Comment 12:Authors should explain more data/information extraction selections in detail and place them in a table with the contents of the chosen article.

Response: Thank you, dear reviewer. We have tried to include more data in the data extraction section accordingly. However, we haven’t seen the advantage of putting the data extraction section in tablet form since it is already attached as supporting information.

Comment 13:More information about the quality assessment of selected articles is required.

Response: Thank you. We have modified the quality assessment section according to your comment.

Comment 14:Authors concluded heterogeneity in their MA by using the I2. The I2 is a value that quantifies heterogeneity, but the presence of significant heterogeneity should be done with the Q test and its respective p-value.

Response: Thank you for your constructive comment. Although Q test is the usual way of assessing whether a set of single studies are homogeneous in the meta-analysis, it only informs us of the presence versus the absence of heterogeneity, but it does not report on the extent of such heterogeneity. In addition, Q has low power as a comprehensive test of heterogeneity especially when the number of studies is small (Gavaghan et al, 2000). Recently, the I2 index has been proposed to quantify the degree of heterogeneity in a meta-analysis. A study which is highlighted in yellow color (Huedo-Medina TB, Sánchez-Meca J, Marin-Martinez F, Botella J. Assessing heterogeneity in meta-analysis: Q statistic or I² index?. Psychological methods. 2006 Jun; 11(2):193) showed that the I2 index as a complement to the Q test due to this reason we can use I2 and its respective p-value to indicate the presence of significant heterogeneity

Comment 15:The subgroup analysis, sensitivity analysis, and Meta-regression should be more explained in the data analysis method section as there is a finding based on this analysis in the result section.

Response: Thank you, dear reviewer. We have explained them in the data analysis section accordingly.

Comment 16:Authors are encouraged to test for group differences and within the group subgroups as well (a p-value for interaction between groups).

Response: Thank you. Have tried to test for group differences and within the group subgroups among the included studies. In addition, we have to put a p-value for interaction between groups in table 2 of our revised manuscript.

Results

Comment 17:The "systematic review" was really rushed. More explanation and description of SR is required.

Response: Thank you. The comment is not clear for as. As much as possible, we have tried to put the detail explanation and description of each content of the systematic review in our manuscript.

Comment 18: Please indicate the type of ES, e.g., figure 2.

Response: Thank you. The comment is accepted and addressed accordingly in fig 2.

Discussion

Comment 19: Please add more sentences on the implications of the main findings in this section.

Response: Thank you. We have tried our best to add the implications of the main findings in the discussion section accordingly.

Comment 20: Heterogeneity is very important across all analyses, even in the subgroup analyses. Authors are encouraged to discuss the manuscript's limit.

Response: Thank you. We have tried to discuss it as a limitation in the discussion section.

Comment 21:Is there any strength in the manuscript? Please explain this, e.g., adding a body of knowledge, and a methodological perspective.

Response: Thank you. We have tried to put the strength of our study in the discussion section.

Submitted filename: Response to Reveiwers.docx

Click here for additional data file.

19 May 2022

HIV risk behavior and associated factors among people living with HIV/AIDS in Ethiopia : A systematic review and meta-analysis.

PONE-D-21-17808R2

Dear Dr. Adane,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.  Please also address my comments to the authors before finalizing the manuscript.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Douglas S. Krakower, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

1. Please proofread carefully for typos, which occur in the abstract and manuscript.

2. Line 53: please specify that you are referring to the US Centers for Disease Control and Prevention.

3. In Lines 83 to 88, please specify what the percentages represent; is it the percent of PLWH who have engaged in an HIV risk behaviors ever, or during some particular time frame?

Reviewers' comments:

29 Jun 2022

PONE-D-21-17808R2

HIV risk behavior and associated factors among people living with HIV/AIDS in Ethiopia: A systematic review and meta-analysis

Dear Dr. Adane:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Douglas S. Krakower

Academic Editor

PLOS ONE