*4.1. Data Source*

We analyzed PDR and sociodemographic data from 1823 persons that participated in a large nationally representative HIV PDR survey carried out in Mexico from 09/2017 to 03/2018 who entered to care in Ministry of Health clinics [7], and had follow-up data in Mexico's national Ministry of Health ART database (SALVAR). The SALVAR database comprises information regarding mortality, retention in HIV care, ART regimen history, and CD4+ T cell count, and viral load follow-up for persons without social security in Mexico. The administrative closure date for the dataset used in this study was 12/2019. The last status reported in the database was used to classify the final outcome of the participants.

## *4.2. Sample Selection*

Persons recorded as alive and in follow-up in the SALVAR database were classified as remaining in care. Persons with a non-active status due to ART abandonment, migration to other healthcare systems [18], unknown status, as well as lack of viral load follow-up for more than 6 months at the dataset closure date, were classified as LTFU. Persons who died were not included in the LTFU group. Participants were stratified into four groups according to prior ART exposure and presence of NNRTI-PDR: experienced + resistant, experienced + non-resistant, naïve + resistant, and naïve + non-resistant. Retention in care, LTFU and death were compared between groups as final outcomes.

## *4.3. Statistical Analysis*

We estimated the percentage of persons with viral suppression (last viral load < 200 copies/mL) at the end of follow-up by group, among those persons with viral load information available (within 6 months prior to the last database entry for each participant). A logistic regression model was developed to assess the relationship between ART exposure, NNRTI resistance, and viral suppression, adjusting for sociodemographic characteristics in the main analysis. A model including change in ART status after an HIV drug resistance test was performed to see the potential impact of this variable on the probability of reaching viral suppression. A Fine-Gray model was used to compare retention in care and viral suppression at the end of follow-up, with LTFU and death as final outcomes and competing events. In this model, we censored persons who ended in care but did not achieve viral suppression or did not have viral load information available. We included ART exposure drug resistance status groups, age, sex, mode of transmission, and educational level

as co-variables. Due to the small size of groups, we combined sex and mode of HIV transmission to include them in the models. The variable "sex and mode of transmission" was categorized as: MSM, heterosexual cisgender men and heterosexual cisgender women. People who inject drugs were excluded from the models. Additional analyses using multiple imputations with 10 replications were conducted for viral suppression since we observed that 10% of the records had missing information for the last viral load. In particular, 30% (6/20) of the experienced + resistant group had missing viral load data. Additionally, to evaluate potential biases and the possible generalization of our results, we described and compared the characteristics of HIV drug resistance testing between persons with and without information in SALVAR. All the analyses were performed using R Version 1.2.5019.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/pathogens10121569/s1, Figure S1: Estimated probability for remaining in care under viral suppression in competing risk with death and LTFU. Table S1: Clinical outcomes according to availability of viral load information in Mexican persons living with HIV, 2017–2019.

**Author Contributions:** Conceptualization, Y.C.-V., C.G.-M., G.R.-T., J.G.S.-M. and S.Á.-R.; data curation, C.C., C.G.-M. and S.Á.-R.; formal analysis, Y.C.-V., F.A.-E., S.S.-R. and C.C.; funding acquisition, F.A.-E. and E.A.E.; investigation, A.P.-M. and C.G.-M.; methodology, Y.C.-V., F.A.-E., E.A.E., S.S.-R. and C.C.; resources, A.P.-M., C.G.-M. and S.Á.-R.; supervision, G.R.-T., J.G.S.-M. and S.Á.-R.; validation, C.G.-M. and S.Á.-R.; writing—original draft, Y.C.-V., A.P.-M. and S.Á.-R.; writing— review and editing, Y.C.-V., F.A.-E., E.A.E., C.C., A.P.-M., C.G.-M., G.R.-T., J.G.S.-M. and S.Á.-R. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by Consejo Nacional de Ciencia y Tecnología (CONACyT SALUD-2017-01-289725), the Mexican Government (Programa Presupuestal P016; Anexo 13 del Decreto del Presupuesto de Egresos de la Federación), and the Canadian Institutes of Health Research (grants PJT-148621 and PJT-159625). E.A.E., F.A.-E. and C.C. had funding provided by the Center for Global Health and Social Responsibility at the University of Minnesota. August 2018–July 2019.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán (protocol code INF-2795-18-19-1, date of approval 19 October 2018).

**Informed Consent Statement:** Patient consent was waived because we are using secondary information from datasets previously recorded; the written informed consent was obtained from the patient(s) to analyze the original information.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. Databases with patient sequences and follow up data are not public due to national regulations.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
