Spatial distribution and determinants of HIV prevalence among adults in urban Ethiopia: Findings from the Ethiopia Population-based HIV Impact Assessment Survey (2017-2018).

The design and evaluation of national HIV programs often rely on aggregated national data, which may obscure localized HIV epidemics. In Ethiopia, even though the national adult HIV prevalence has decreased, little information is available about local areas and subpopulations. To inform HIV prevention efforts for specific populations, we identified geographic locations and drivers of HIV transmission. We used data from adults aged 15-64 years who participated in the Ethiopian Population-based HIV Impact Assessment survey (October 2017-April 2018). Location-related information for the survey clusters was obtained from the 2007 Ethiopia population census. Spatial autocorrelation of HIV prevalence data were analyzed via a Global Moran's I test. Geographically weighted regression analysis was used to show the relationship of covariates. The finding indicated that uncircumcised men in certain hotspot towns and divorced or widowed individuals in hotspot woredas/towns might have contributed to the average increase in HIV prevalence in the hotspot areas. Hotspot analysis findings indicated that, localized, context-specific intervention efforts tailored to at-risk populations, such as divorced or widowed women or uncircumcised men, could decrease HIV transmission and prevalence in urban Ethiopia.

Introduction

In resource-limited settings, such as sub-Saharan Africa, sustained HIV interventions targeting all populations are not feasible [1, 2]. Understanding the HIV epidemic at the local level and reallocating resources for its control in specific areas in countries with a generalized epidemic has been recommended to increase cost-effectiveness [2, 3]. Identifying HIV clusters can inform tailored HIV interventions [3].

Ethiopian Population-based HIV Impact Assessment (EPHIA) survey data suggest that the incidence of HIV in urban Ethiopia is around 0.5% [4], much lower than the Joint United Nations Programme on HIV/AIDS benchmark (3%) [5]. Even though adult HIV prevalence, or the proportion of persons in a population who are living with HIV at a specific point in time, has declined at the national level, little information is available about sub-geographic areas and certain subpopulations in urban Ethiopia. Analysis of data from the 2011–2016 demographic and health survey in Ethiopia indicated that the distribution of HIV infection in Ethiopia is not random [6-8]. A study in one region in Ethiopia revealed that low educational status and migration status are determinants of HIV infection in Ethiopia [9]. Moreover, HIV service coverage varies among subpopulations and locations [10]. The HIV epidemic in certain localities could emerge or re-emerge if not addressed in these sub-regions and subpopulations [6].

Spatial analysis in epidemiology links spatial data (either from an absolute location and/or relative spatial arrangement of the data) to disease spread or high-risk populations [11] and has been applied to HIV intervention research in Africa [12]. Geospatial analysis of epidemiological data can generate precise maps of hotspot locations where HIV prevalence is concentrated [13]. Local spatial analyses can show geographic variation of the HIV epidemic and its drivers and inform targeted interventions; however, few geospatial analyses use data from sub-populations in Ethiopia [6-9].

With a heterogeneous HIV burden and decreasing national HIV prevalence, Ethiopia’s epidemic is primarily concentrated in certain populations [7]. Low national HIV prevalence may obscure localized epidemics in urban Ethiopia, and identifying hotspot areas could help target these high-risk populations. The purpose of this study is to identify geospatial clustering of HIV infections and hotspot areas by subpopulation groups to inform targeted interventions with the limited resources available.

Methods

Study design and population

Our geospatial analyses used data from adults who participated in the EPHIA survey (October 2017–April 2018) [14]. EPHIA was a nationally and regionally representative, cross-sectional, household-based sero-survey conducted among 19,136 adults aged 15–64 years and 4,729 children aged 0–14 years. The reference population comprised individuals in urban areas who were present in households (i.e., “slept in the household”) on the night before the interview. All individuals in either the de facto (i.e., those who slept in the household the night before the survey) or de jure (i.e., those individuals who are usual residents of the household regardless of whether they were present in the household during the previous night) populations were included in the rosters compiled for sampling purposes, although our analysis was limited to the de facto population only. According to the population projection made from 2007 to 2037, currently 20.4% of Ethiopia’s population lives in urban areas [15].

Study variables and measurement

The primary outcome variable for the analysis was geo-linked HIV prevalence. HIV prevalence in each cluster was calculated from the HIV test results obtained in the survey (the denominator is the total population tested for HIV, and the numerator is the number of HIV-positive persons identified in this study). HIV testing was conducted in each household using the Ethiopian National HIV testing algorithm in accordance with national guidelines. Individuals with a nonreactive result on the screening test were reported as HIV negative. Individuals with a reactive screening test underwent confirmatory testing. Those with reactive results on both the screening and confirmatory tests were classified as HIV positive. [4]. The HIV prevalence in each town was the outcome variable. HIV prevalence was defined as HIV infection on the day the blood sample was taken and HIV testing was conducted. The independent variables included survey-weighted proportions of women, mean age of all participants, being divorced or widowed, being uncircumcised, being unaware of HIV-positive status during EPHIA, having no regular sexual partner, not using condoms, having had sex before age 15 years, and using injected drugs. These socioeconomic, demographic, and biological variables were selected for this analysis because these factors have been associated with the spatial distribution of HIV and risk of HIV infection in other studies [16-24]. Data were aggregated at the town level.

Sample size and sampling procedures

The EPHIA methodology, including data collection procedures, has been previously reported [14]. Briefly, the survey used a two-stage, stratified sampling design to provide national estimates from 11 regions across the country. In the first stage, 395 enumeration areas (EAs) were selected, and 393 EAs were included in the EPHIA survey (Fig 1).

Fig 1

Distribution of enumeration areas (N = 393) selected for Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

The first-stage or primary sampling units for EPHIA are defined as EAs created for the 2007 Ethiopia Population and Housing Census. The 2007 sampling frame consisted of slightly over 17,000 urban EAs containing an estimated 3.0 million households [15]. EAs in six of the nine zones in the Somali region (Wogob, Jarar, Shebelle, Afder, Korahe, and Doolo) were excluded from the sampling frame for security reasons in the EPHIA survey. From each EA, a random sample of 30 households, on average, was selected for a total of 11,810 households. At each stage of the process, consent was indicated by signing or making a mark on the consent form on a tablet and on a printed copy, which was retained by the participant. After a designated head of household provided written consent for household members to participate in the survey, individual members were rostered during a household interview. Of the 20,170 adults who participated in the survey, 19,136 provided consent to be tested for HIV. Participants aged 15–64 years and emancipated minors ages 13–17 years provided written consent on a tablet for an interview and for participation in the biomarker component of the survey, including home-based testing and counselling, with return of HIV-test results as well as participation future research.

Data collection

EPHIA staff used a questionnaire prepared for the survey to collect data about household and individual characteristics. Initial household-based HIV testing was performed with the national HIV-testing algorithm using three HIV rapid tests. Individuals with a reactive screening test underwent confirmatory testing using the Uni-Gold HIV-1/2 (Trinity Biotech, Bray, Ireland). Individuals with discordant results were administered a tiebreaker test (Vikia HIV-1/2, bioMérieux, Marcy-l’Étoile, France). All HIV-positive individuals identified in the field received confirmatory testing in a satellite laboratory using the Geenius HIV 1/2 Supplemental Assay (Bio-Rad, Hercules, CA USA) [4]. Both the questionnaire and field laboratory data were collected on mobile tablet devices using the application Open Data Kit, an open-source mobile data collection application. The Global Positioning System was used to identify and record the geographical coordinates of each EPHIA sample location. Cluster geolocation data were loaded onto encrypted and passcode-protected tablet computers in keyhole markup language format using the MAPS.ME mobile app (https://maps.me), which includes OpenStreetMap (.

Data extraction

The 2017 EPHIA geospatial data sets were downloaded in Stata format with permission from the PHIA project website (at https://phia-data.icap.columbia.edu/datasets?country_id=12) [26]. After understanding the detailed data sets and coding, further data recoding was carried out. Data sets of Global Positioning System (GPS) coordinates of EPHIA clusters were merged for this analysis as appropriate per the Reference Guide for Using Geospatial Data from the Population-based HIV Impact Assessments [25].

Analysis

All households and individuals within a dataset were assigned the location of the centroid of their respective survey EA, and then the centroid was randomly displaced within an area around the actual center of the EA [25].

Interpolation maps of HIV prevalence were created through ordinary kriging‥ Kriging is a statistical model that infers the dependence between pairs of points by examining all similarly distant pairs in the data [27]. A hotspot analysis was conducted using Getis-Ord statistic. ArcGIS (v.10.2, Esri, Redlands, CA USA) was used to detect statistically significant clusters of high or low HIV prevalence.

The spatial autocorrelation of HIV prevalence data was analyzed by performing a global Moran’s I statistic to test the null hypothesis that observed data at one location are independent of data at other locations. Moran I is a measure of spatial autocorrelation in area data, explaining the degree of dependence between values of a variable at different geographic locations [28].

Because the Moran I statistic showed significant spatial autocorrelation of our HIV prevalence data (Moran I, 0.057; p = 0.00014) within a distance of 150 km., we subsequently analyzed the data using kriging [27]. The enumeration area (EA) in Ethiopia were used as the spatial mapping unit at the woreda/town level in this study which comprises about 249 woreda in this analysis, since the woreda is the basic decentralized administrative unit [29], and the woreda health system is the primary level of health care provided in Ethiopia [30]. These basic units were used based on previous studies [31], which suggested that spatial analysis results should be aggregated to the lower levels of services or programs relevant for decision-making, budgeting, and monitoring (including the district or facility levels).

To explore the correlations between covariates and HIV prevalence, we used Pearson correlation coefficient and geographically weighted regression (GWR) analysis. GWR) is a widely used tool for exploring spatial heterogeneity of conditions over geographic space [32]. Covariates that had low correlation (with correlation coefficient [r] <0.2) with HIV prevalence were not included in the models. Highly correlated covariates were not simultaneously included in the regression models to avoid model redundancy and multicollinearity. We used an explanatory approach and used the maps of hotspots of HIV prevalence to select covariates of interest (p<0.05). The prevalence of HIV in the hotspot towns was significantly correlated with the percentage of women who were not virally suppressed; individuals who were widowed, divorced, or separated; uncircumcised men; and individuals who lived in women-headed households during the survey period.

Following correlation analysis, we fitted an ordinary least squares (OLS) model to explain the global relationship between HIV prevalence and the covariates using a sample of 249 towns (observations). The outcome variable for the regression model (OLS) was the weighted HIV prevalence in each town. Where HIVi is the estimated prevalence, (i = 1, 2…249), β0 is the global intercept, βs are the regression coefficients and the covariates, and εi is the error term:

β1i = weighted proportion of females

β2i: weighted average age

β3i: weighted proportions of widowed or divorced individuals

β4i: weighted proportion of men who are not circumcised

β5i: weighted proportion of women-headed households

β6i: weighted proportion of food-insecure households

β7i: weighted proportion of individuals who had first sexual encounter before age 15 years

OLS results are unreliable when two or more variables exhibit multicollinearity, but GWR builds a local regression equation for each feature in the dataset [33]. The OLS regression model estimates a parameter of interest independent of the location of the particular observation, whereas the GWR model estimates the parameter of interest under more localized conditions by considering the location of that observation [34]. Therefore, the GWR model is advantageous over OLS when dealing with such spatial datasets, so we used the GWspatial lag regression model to show how the covariate relationships changed in each woreda/town. For all the statistical analysis, statistical significance was decided at p<0.05. Multicollinearity was assessed through variance inflation factor (VIF) and condition number [35], where VIF values greater than 10 indicated multicollinearity [35].

Ethical considerations

The survey protocol was approved by the Institutional Review Boards of the Ethiopian Public Health Institute (EPHI, Ethiopia), Centers for Disease Control and Prevention (Atlanta, GA USA), and Columbia University (New York, NY USA). The EPHIA Data Analysis Advisory Committee at the EPHI approved the analysis of the data.

Results

Characteristics of the study population

Table 1 provides the characteristics of the study population.

Table 1

Demographic, socioeconomic, and behavioral characteristics of adults aged 15–64 years by HIV status in urban Ethiopia (N = 19,136), Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

Background characteristics	N (Weighted %)	HIV status		P-value
		HIV negative	HIV positive
		Weighted % (95% CI)	Weighted % (95% CI)
Women	11,599 (50.5)	95.9 (95.5–96.3)	4.1 (3.7–4.5)	<0.0001*
Men	7,537 (49.5)	98.1 (97.7–98.4)	1.9 (1.6–2.3)
Age, years
15–24	7,547 (34.9)	99.3 (99.0–99.5)	0.7 (0.5–1.0)	<0.0001*
25–34	5,664 (30.3)	97.4 (96.9–97.8)	2.6 (2.2–3.1)
35–44	3,136 (18.9)	93.8 (92.8–94.6)	6.2 (5.4–7.2)
45–54	1,651 (10.1)	93.9 (92.5–95.1)	6.1 (4.9–7.5)
55–64	1,138 (5.8)	96.6 (95.3–97.6)	3.4 (2.4–4.7)
Man-headed HH	9,343 (53.7)	97.8 (97.4–98.1)	2.2 (1.9–2.6)	<0.0001*
Woman-headed HH	9,793 (46.3)	96.0 (95.5–96.4)	3.6 (3.6–4.5)
Marital status
Never married	7,103 (35.5)	99.0 (98.7–99.3)	1.0 (0.7–1.3)	<0.0001*
Married or living together	9,418 (52.1)	97.2 (96.8–97.6)	2.8 (2.4–3.2)
Divorced or separated	1,723 (8.5)	92.3 (90.7–93.5)	7.7 (6.5–9.3)
Widowed	7,723 (0.9)	85.3 (82.3–87.9)	14.7 (12.1–17.7)
Education level
No education	200 (11.9)	94.8 (93.7–95.8)	5.2 (4.2–6.3)	<0.0001*
Primary	6,803 (35.5)	95.8 (95.2–96.3)	4.2 (3.7–4.8)
Secondary	5,488 (28.7)	97.6 (97.1–98.0)	2.4 (2.0–2.9)
More than secondary	4,376 (24.0)	99.0 (98.6–99.3)	1.0 (0.7–1.4)
Food insecurity in the past 4 weeks
No	18,223 (95.5)	97.1 (96.8–97.3)	2.9 (2.7–3.2)	<0.0001*
Yes	808 (4.5)	95.0 (93.2–96.3)	5.0 (3.7–6.8)
Age at first sex
First sex at age ≥15years	17,735 (95)	97.1 (96.9–97.4)	2.9 (2.6–3.1)	<0.0001*
First sex at age <15 years	1,014 (5.0)	93.2 (91.3–94.7)	6.8 (5.3–8.7)
Total	19,136	97.0 (96.7–97.2)	3.0 (2.8–3.3)

Abbreviations: X2, chi-square statistics; CI, confidence interval; HH, household.

*P-values of <0.0001 indicate, statistically significant.

Of 11,581 eligible households, 90.9% completed the household interview. We found significantly higher HIV prevalence among women (4.1%) than men (1.9%; p<0.0001), among all participants aged 35–44 years (6.2%) than those aged 15–24 years (0.7%; p<0.0001), and among those with no education (5.2%) than those with more than secondary education (1.0%; p<0.0001). Similarly, women-headed households (4.0%) had higher HIV prevalence than men-headed households (2.2%; p<0.0001). Widowed individuals (14.7%) had significantly higher HIV prevalence than individuals who never married (1.0%; p<0.0001). Participants who had sex before age 15 years (6.8%) had a significantly higher prevalence of HIV than those who had sex at age >15 years (2.9%; p<0.0001 compared with their counterparts (Table 1).

Geographic distribution of HIV in Ethiopia

HIV prevalence among adults in Ethiopia was 3.0%, which significantly varied by administrative regions (p<0.0001) and ranged from 0.8% in Somali to 5.7% in Gambela (Fig 2).

Fig 2

Weighted distribution of HIV prevalence (%) by region, Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

*Area boundaries indicate administrative regions in Ethiopia.

Of the 249 woredas/towns included in this analysis, 167 (67.1%) had at least one HIV-positive resident. The crude estimates of HIV prevalence among these 167 woredas/towns ranged from 0.6% in Hosana to 25.1% in Meki (Fig 3).

Fig 3

Distribution of HIV positive cases by enumeration areas, Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

Spatial heterogeneity of HIV prevalence in urban Ethiopia

HIV prevalence was spatially auto correlated (Moran I, 0.057; p = 0.00014). The interpolated data show the spatial heterogeneity of HIV prevalence (range, 0%–25%), independent of regional boundaries. A higher HIV prevalence was observed in Gambela region, Center-North, and some parts of Eastern Ethiopia. In contrast, HIV prevalence was spatially low in the Center-West and Southern regions (Fig 4). Because EPHIA did not include the eastern parts of the Somali administrative region, the HIV prevalence for this region could not be interpolated in this analysis.

Fig 4

Predicted HIV prevalence among adults aged 15–64 years in urban Ethiopia, Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

Spatial determinants of HIV infection in urban Ethiopia

Table 2 describes the spatial determinants of HIV infection in urban Ethiopia.

Table 2

Spatial determinants of HIV infection using ordinary least square among adults aged 15–64 years in urban Ethiopia, Ethiopia Population-based HIV Impact Assessment Survey (2017–2018).

Variable	Coefficient	Standard Error	Robust P-value	VIF**
Intercept	-4.10	2.653	0.1238	--------
Women	0.01	0.033	0.7873	1.73
Age	0.08	0.105	0.4438	1.29
Divorced/widowed	0.16	0.047	0.0007*	1.48
Not circumcised (only for men)	0.10	0.027	0.0005*	1.11
Women-headed households	0.02	0.016	0.1522	1.50
Food-insecure household	0.06	0.040	0.1590	1.07
Age at first sexual encounter (<15 years)	0.09	0.059	0.1329	1.16

Abbreviations: VIF, variance inflation factor.

* P-values <0.05 were considered statistically significant.

** Large VIF values (>7.5) indicate redundancy among explanatory variables.

Compared with the woredas/towns with low HIV prevalence, there was a disproportionately higher rate of men who were not circumcised in Gambela; high HIV burden woredas/towns also had more participants who were not virally suppressed and who were divorced/separated/widowed. In hotspot woredas/towns, the average increase in HIV prevalence among participants who were divorced/separated/widowed was 0.16 and among uncircumcised men was 0.10 (Table 2).

The GWR model is a better fit because it explains 33.4% of total variation in HIV prevalence for the seven variables compared with the OLS model, which explains 25% of the total variation.

Discussion

Prevalence of HIV infection among adults in Ethiopia was 3.0% (women, 4.1%; men, 1.9%). This corresponds to approximately 384,000 adults living with HIV in urban Ethiopia [4]. Our study estimated the clustering effect of HIV in urban Ethiopia. The crude estimates of HIV prevalence ranged from 0.6% to 25.1%, showing that the HIV prevalence data were spatially auto correlated (Moran I, 0.057; p = 0.00014). The data indicate that there is <1% likelihood that this cluster pattern of HIV prevalence could be the result of random chance. The spatial analysis helped to identify individuals at high risk of HIV infection due to their geographic location in Ethiopia. The interpolated data showed that there was spatial heterogeneity of HIV prevalence (range, 0%–25%) independent of regional boundaries, which is consistent with previous findings [1, 6, 16, 17, 20, 21, 28, 36, 37]he prevalence of HIV was 10%–21% in certain geographic clusters of Ethiopia [6], which is consistent with findings in other countries [1, 16, 17, 20, 21, 28, 36, 37] indicating a complex geographical variation of the HIV epidemic at the local level.

Our findings show the importance of considering geographic factors in addition to determining individual risk factors to HIV infection in Ethiopia. Previous studies have estimated HIV prevalence in Ethiopia at the national level, disaggregated by urban, rural, or regional levels [10, 38, 39], and most of these studies used blood samples drawn from pregnant women in antenatal facilities [40] and dried blood spot samples from adults [41]. These previous studies provided proximate estimates of prevalence in the overall urban population.

The HIV prevalence maps generated in this analysis describe the spatial disparities in the HIV epidemic within Ethiopia and identify areas with concentrated HIV burden and drivers of the HIV epidemic. Hotspot analysis results indicate that the current epidemic is concentrated in the Gambela region, in the Center-North region, and in some parts Eastern Ethiopia. The new hotspot in northern Ethiopia calls for further analysis to explore whether HIV prevalence is static or increasing in these areas.

Our findings suggest that localized and woreda/town-specific intervention strategies could help combat the spread of HIV/AIDS in Ethiopia. Prioritizing people at greatest risk of infection and locations with high HIV burden and adapting interventions to reflect the local epidemiological context could increase the efficiency and effectiveness of HIV prevention programs [42]. Our observation that urban individuals who are not virally suppressed, who are divorced/widowed/separated, or who are not circumcised had high HIV infection prevalence in the hotspot woredas/towns concurs with findings of other studies [20, 43–45].

District-based spatial clustering of the HIV epidemic can target interventions at populations with high transmission [46]. Spatial analysis enables policy makers to identify towns most affected and design effective and culturally acceptable preventive measures such as circumcision in the Gambela region, as suggested by other studies [42, 47]. Specific interventions targeted at woredas/towns are more appropriate than universal HIV reduction strategies, which may not be applicable to different cultural contexts in Ethiopia.

Our study also showed high HIV prevalence among women, which is consistent with other studies conducted in Ethiopia [6, 38] and with reports from other countries [1, 19, 23, 28, 45]. Our findings indicate that the widowed individual has high HIV prevalence, and 91.9% of this individuals are women. We found that the divorced/widowed individuals in hotspot woredas/towns had 0.16 average increase in HIV prevalence. The high percentage of widowed women in high HIV prevalence towns might be due to the increased migration of women from rural areas to woredas/towns for job opportunities.

We found that uncircumcised men in hotspot woredas/towns had 0.10 average increase in HIV prevalence. This finding highlights the need to target voluntary medical male circumcision programs in regions such as Gambela, where 61% of the men were circumcised (both medically and traditionally) [48], and less than one-fifth of the circumcisions (18.5%) were performed by health professionals [48].

Targeting regions that have limited capacity to diagnose and treat HIV also could improve outcomes [49]. Our findings contribute to the evidence in the literature and suggest that the national epidemics cannot continue to be assessed as a whole when there is clear evidence of substantial local heterogeneity in the HIV epidemic [50]. The literature from Ethiopia indicates that the overall capacity score (57.1%) for diagnosing and treating HIV in urban facilities was higher than that of the rural health facilities (38.2%) [49]. Our findings, which varied across woredas/towns, also suggest that context specific intervention efforts could more effectively reduce the burden of HIV in Ethiopia.

Our findings are subject to several limitations. We conducted the study in urban areas, so the interpolated data are limited to urban HIV prevalence in Ethiopia. Some regions such as Somali and Benishangul Gumuz had a relatively small number of HIV-positive individuals, which may raise questions related to accuracy of some estimates in these regions. The interpolation of HIV prevalence was not predicted for the six zones in the Somali region. The survey was not powered for analysis at the district level. However, we note that the observed geographical patterns of HIV prevalence parallel that observed in other studies [6], including national surveys such as the demographic and health surveys (DHS) [41] and key population surveillance reports from Ethiopia [40]. This study also has the inherent limitation of a cross-sectional study design, which does not allow for the examination of causal relationships.

The estimates of HIV prevalence across spaces provide an important tool for targeting the interventions that are necessary to bringing HIV infections under control in Ethiopia.

We found higher HIV prevalence in Gambela region, Center-North, and some parts of Eastern Ethiopia. Our findings suggest that uncircumcised men in the certain hotspot towns and divorced/widowed individuals in hotspot woredas/towns might have contributed to the average increase in HIV prevalence in these hotspot areas. Context-specific intervention efforts could decrease the burden of HIV in urban Ethiopia. Localized HIV prevention interventions tailored to at-risk individuals, such as divorced and widowed women and uncircumcised men in certain regions, could be essential to curbing HIV transmission in urban Ethiopia

(DOCX)

Click here for additional data file.

30 Dec 2021

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,

Mona Pathak, PhD

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

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“This project has been supported by the U.S. President’s Emergency Plan for AIDS Relief (PEPFAR) through the U.S. Centers for Disease Control and Prevention (CDC) under the terms of cooperative agreement #U2GGH001226. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the funding agencies. Conflict of interest.”

We note that you have provided additional information within the Acknowledgements Section that is not currently declared in your Funding Statement. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“The author(s) received no specific funding for this work.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

3. We note that you have stated that you will provide repository information for your data at acceptance. Should your manuscript be accepted for publication, we will hold it until you provide the relevant accession numbers or DOIs necessary to access your data. If you wish to make changes to your Data Availability statement, please describe these changes in your cover letter and we will update your Data Availability statement to reflect the information you provide.

4. We note that Figure 1, 2, 3 and 4 in your submission contain map images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

a. You may seek permission from the original copyright holder of Figure 1, 2, 3 and 4 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

b. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

5. Please ensure that you refer to Figure 1 and 2 in your text as, if accepted, production will need this reference to link the reader to the figure.

[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: No

**********

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

Reviewer #1: Yes

Reviewer #2: No

**********

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: No

**********

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: Yes

**********

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:

�  The scope of the study is well designed to delineate hotspot HIV transmission risk.

�  The study result cloud be used to control and management of the HIV infection in the vulnerable community through proper intervention measures target the gender and affected age groups with efforts of public health officials and national health programmers

�  Random Sampling methods for field data collection of HIV epidemiological information in the selected house hold, and geo-coordinates of the house hold location are depicted on map, using GPS.

�  Geo-Spatial analysis was applied significantly

�  Moran I geo-statistical analysis measure of spatial autocorrelation in area data, explaining the degree of dependence between values of a variable at different geographic locations.

�  Geographically weighted regression (GWR) analysis was used to establish the relationship of covariates.

�  The interpolation of spatial prediction of HIV infection using Kriging methods, and results reflect the real situation of the ground.

Reviewer #2: Please find my comments and suggestions below. I have mainly focused on the spatial analysis aspects of this study and please read my suggestions below.

Study variables and measurement (Lines 87-98)

Please clearly define your outcome variable. Please define the geo-linked HIV prevalence, geo-linked to what geographical level? It is not clearly described how the enumeration areas (EAs) are related to “towns” Does each enumeration area consist of one town? Or do you need to aggregate several EAs to the town level?

When you define the prevalence of HIV, it is not clear what a cluster means. Does “cluster” refer to an EA or a town? When you are mentioning that you have calculated the weighted HIV prevalence in each town, but you did not describe the weighting method.

In Figure 2 you are presenting a prevalence map, but the area boundaries are not defined. Are those provinces in Ethiopia? Additionally, as you have mentioned just a small portion of EAs could be sampled from “Somali” district, and for that reason, you should change the map to include a small portion of that district. Figure 2 implies that HIV prevalence is low in Somali district, but we do not know because the population was not included in the EPHIA survey.

Line 91 – Please state that the outcome variable for your regression model (OLS) was the weighted HIV prevalence in each town.

Line 92 – “Study variables” is a confusing term, because study variables could include the outcome (e.g., dependent) variable and the predictor (e.g., independent) variables. You should use “predictor” or “independent” variables instead of study variables.

Analysis (Lines 139-190)

The hotspot analysis (Getis-Ord Gi*statistic) is not a spatial interpolation method. Moreover, you did not present any local clustering (hot spot) analysis. Please describe in detail your Kriging method, because there is more than one Kriging method that you can use. Kriging is a spatial interpolation method and you can not identify statistically significant clusters of high or low HIV prevalence based on Kriging analysis. You might confuse Kriging with the Getis-Ord Gi* statistic or the Local Moran’s I method where you can identify high or low (if you use Getis-Ord Gi*) and high-high, low-low, low-high, and high-low (if you use Local Moran’s I) HIV prevalence clusters, but you did not present any of those results.

You can read my previous papers where I have used and described various spatial epidemiology methods:

https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0235291

https://bmcinfectdis.biomedcentral.com/articles/10.1186/s12879-015-1106-6

Line 148 - Global Moran’s I statistic evaluates the extent of spatial heterogeneity. You should define at what distance band did you observe the highest Moran's I value (see my papers for the methods). Please also define what type of conceptualization parameter did you use, because you can use more than one (e.g., zone of indifference, inverse distance, etc.). I think that you should use and present local Moran’s I analysis to identify local HIV clusters.

Line 168 – Please define your GWR model. What type of spatial weights matrix did you use? What type of GRW did you use, spatial lag model or spatial error model?

You can read about them at:

https://s4.ad.brown.edu/resources/tutorial/modul2/geoda3final.pdf

Thank you!

**********

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: Masimalai Palaniyandi., M.Sc.,M.Tech.,Ph.D.,

Reviewer #2: Yes: Csaba Varga

[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.

Submitted filename: Reviewer Recommendation and Comments for Manuscript Number PONE.pdf

Click here for additional data file.

14 May 2022

The responses are incorporated in the manuscript. Double checked the link to data source and updated it as appropriate.

Submitted filename: A rebuttal letter .docx

Click here for additional data file.

27 Jun 2022

Spatial distribution and determinants of HIV prevalence among adults in urban Ethiopia: Findings from the Ethiopia Population-based HIV Impact Assessment Survey (2017–2018)

PONE-D-21-34281R1

Dear Dr. Argefa,

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.

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,

Mona Pathak, PhD

Academic Editor

PLOS ONE

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. 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: Yes

**********

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

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. 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

**********

5. 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: Yes

**********

6. 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: (No Response)

Reviewer #2: (No Response)

**********

7. 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: Dr.Palaniyandi Masimalai

Reviewer #2: Yes: Csaba Varga

**********

1 Jul 2022

PONE-D-21-34281R1

Spatial distribution and determinants of HIV prevalence among adults in urban Ethiopia: Findings from the Ethiopia Population-based HIV Impact Assessment Survey (2017–2018)

Dear Dr. Gelibo:

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. Mona Pathak

Academic Editor

PLOS ONE