Changes in HIV Clinical Visit Frequency and Viral Suppression During and After COVID-19 Restrictions: A Nested Study Within NA-ACCORD in the US, 2018–2021 Authors: Yue Zhang, Brenna C Hogan, Kelly A Gebo, Peter F Rebeiro, Milton L Wainberg, Catherine R Lesko, Vincent C Marconi, N Sarita Shah, Raynell Lang, Nel Jason L Haw, Yawei Cheng, Michael A Horberg, Jennifer O Lam, Michael J Silverberg, George Yendewa, Joseph J Eron, Richard D Moore, Angela M Parcesepe, Keri N Althoff, for the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) Published: 2026 ABSTRACT AbstractBackgroundTemporary COVID-19-related restrictions disrupted routine HIV care and viral load (VL) testing for people with HIV (PWH). Whether and when encounters and VL suppression returned to prepandemic levels remains unclear. We aimed to evaluate changes in HIV clinical encounter frequency and VL suppression from 2018 through 2021, and assess the impact of COVID-19 restriction policy.MethodsAdults (≥18 years) with HIV from 12 US clinical cohorts with ≥1 VL test from June 2020 to December 2021 were categorized into mutually exclusive HIV care groups: (1) in-care before restrictions, (2) new-to-care, and (3) re-engaging during restrictions. Logistic regression with generalized estimating equations was used to estimate adjusted odds ratios (aOR) of unsuppressed VL (>200 copies/mL) for HIV care groups from June 2020 to December 2021. A linear mixed model evaluated associations between state-level COVID-19 policy strength (mild, moderate, or strong) and unsuppressed VL.ResultsAmong PWH in-care (n = 48 605), encounter frequency decreased sharply during early COVID-19 restrictions but returned to prepandemic levels by December 2020, and suppressed VL rate remained above 91%. New-to-care (n = 105) or re-engaging (n = 483) individuals did not increase odds of unsuppressed VL (vs in-care, aOR = 0.91 [95% CI: 0.73, 1.13]). Those living in states with strong (vs mild, aOR = 1.31 [0.93, 1.86]) COVID-19 restriction policies had increased odds of unsuppressed VL, though not significant.ConclusionsHIV care encounters returned to prepandemic levels by late 2020, with sustained viral suppression. Prolonged COVID-19 restrictions were associated with higher risk of unsuppressed VL. HIV patients, providers, and clinics showed resilience during COVID-19 disruptions, offering lessons for future emergencies.Keywords: COVID-19 pandemic, COVID-19 restrictions, HIV, HIV care encounters, viral load suppression Human immunodeficiency virus type 1 (HIV) remains a significant public health challenge in the United States (U.S.), with an estimated 1.2 million people living with HIV and nearly 35 000 new diagnoses each year [1]. With timely access to antiretroviral therapy (ART), HIV is a chronic condition that requires regular clinical follow-up and routine viral load (VL) monitoring to sustain viral suppression and reduce transmission [2–4]. HIV clinical guidelines recommend VL testing every 4–8 weeks for individuals initiating ART, and every 3–6 months depending on suppression status [5, 6]. Nationwide public health measures implemented in 2020–2021 to reduce COVID transmission disrupted outpatient HIV care, particularly in the first 3 months of the pandemic. The COVID-19 U.S. State Policy (CUSP) database shows restrictions were rapidly implemented beginning in March 2020, with most states adopting strong policy responses such as stay-at-home orders and social distancing mandates by May 2020 [7, 8]. Such policies and overwhelmed healthcare systems limited access to routine outpatient HIV clinical encounters and delayed VL monitoring [9–12]. Studies showed a greater impact of restricted access to care among people with HIV (PWH) compared with people without HIV in some U.S. states [13, 14]. A study with the Johns Hopkins HIV Clinical Cohort (Baltimore, Maryland) demonstrated that early pandemic restrictions and temporary onsite laboratory closures significantly delayed VL monitoring [15]. Despite evidence of HIV care disruptions during the early months of the COVID pandemic, little is known about whether and when HIV clinical encounters and VL monitoring returned to prepandemic levels and if state-level COVID-related restrictions impacted VL suppression. This study aimed to assess: a) the frequency of HIV clinical encounters and VL monitoring during the first 2 years of the COVID pandemic in the U.S.; and b) the association of state-level COVID-19 restriction policies and VL suppression. Findings are intended to inform strategies to minimize healthcare disruptions and maintain monitoring of a transmissible viral infection during future public health emergencies. METHODS Study Population This study was nested in the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD) [16]. Eligibility criteria for adults (≥18 years) in NA-ACCORD contributing clinical cohorts include successful linkage to HIV care, defined as having ≥ 2 HIV care encounters within a 12-month period. Briefly, individual-level longitudinal data from PWH across the U.S. are harmonized, pooled, and quality-assured [16]. All contributing cohorts have received ethics approval from their respective local institutional review boards, and the NA-ACCORD collaboration has been approved by the Johns Hopkins Bloomberg School of Public Health. Our nested study population included adults from 12 U.S. clinical cohorts who had ≥1 VL test between June 1, 2020 and December 31, 2021 (Supplementary Figure 1). Participants without any VL measurement during this period (ie, those who did not return for VL monitoring after the initial COVID-19-related restrictions) were excluded. Time within our study period is categorized as before (September 1, 2018 to February 29, 2020), during (restriction period: March 1, 2020 to May 31, 2020), and after (June 1, 2020 to December 31, 2021) the initial phase of COVID-19 pandemic-related restrictions were implemented. We categorized the study population into 3 HIV care groups: In care before and returning after restrictions (in-care): PWH with ≥1 CD4, VL, or HIV encounter before the pandemic. Baseline defined as the date of the last CD4 or VL measurement or encounter before the pandemic. In care before and returning after restrictions (in-care): PWH with ≥1 CD4, VL, or HIV encounter before the pandemic. Baseline defined as the date of the last CD4 or VL measurement or encounter before the pandemic. New-to-care during restrictions (new-to-care): (a) No CD4, VL, ART use, or HIV encounters before March 1, 2020, and (b) ≥ 1 CD4, VL, or HIV encounter during the restriction period with unsuppressed VL (>200 cp/mL at, or within the 30 days of baseline). Baseline defined as the first CD4, VL, or HIV encounter during the restriction period. New-to-care during restrictions (new-to-care): (a) No CD4, VL, ART use, or HIV encounters before March 1, 2020, and (b) ≥ 1 CD4, VL, or HIV encounter during the restriction period with unsuppressed VL (>200 cp/mL at, or within the 30 days of baseline). Baseline defined as the first CD4, VL, or HIV encounter during the restriction period. Re-engaging or transferring during restrictions (re-engaging): (a) No evidence of being in HIV care (ie, no CD4, VL, ART use or HIV encounter while enrolled in the NA-ACCORD contributing cohort) before the pandemic, and (b) evidence of care prior to September 2018, and (c) at least 1 CD4, VL, or encounter during the restriction period. Baseline defined as the first CD4, VL or HIV encounter during the restriction period. Re-engaging or transferring during restrictions (re-engaging): (a) No evidence of being in HIV care (ie, no CD4, VL, ART use or HIV encounter while enrolled in the NA-ACCORD contributing cohort) before the pandemic, and (b) evidence of care prior to September 2018, and (c) at least 1 CD4, VL, or encounter during the restriction period. Baseline defined as the first CD4, VL or HIV encounter during the restriction period. To assess whether the proportion of participants excluded from the cohort was higher during the pandemic than during the prepandemic period, we constructed a comparison population of PWH who had ≥1 VL test between June 1, 2018 and December 31, 2019. For the in-care group, participants were also required to have ≥1 HIV encounter, VL, or CD4 measurement between September 1, 2016 and February 28, 2018. This comparison population represents the corresponding period 2 years before the primary study population (Supplementary Figure 2). Using this comparison population, we assessed whether the proportion of participants who did not return to care during the pandemic differed from that during the comparable prepandemic period. Outcome The outcome of our study was unsuppressed VL defined as an HIV RNA >200 copies/mL, consistent with public health definitions of nonsuppression and potential transmissibility [17]. State-level COVID-Related Restriction Policy Strength State-level COVID-19 restriction policy strength was categorized per month as strong, moderate, or mild based on closures, shelter-in-place orders, and reopening measures, as classified in the COVID-19 U.S. State Policies database [8] (Supplementary Method 1). All states except South Dakota implemented strong restriction policies in April 2020; no states had strong restriction policies after April 2021 (Supplementary Figure 3). Covariates Covariates of interest included baseline age, race and ethnicity, sex, clinical cohorts, and HIV acquisition group, which was categorized using a mutually exclusive hierarchy: (1) injection drug use (IDU), (2) men who have sex with men (MSM), (3) heterosexual contact, and (4) other or unknown risk. Baseline CD4 count (+/− 90 days of baseline) was categorized as <200, 200–499, ≥500 cells/mm3, or not assessed. Statistical Analysis We assessed differences in the distribution of characteristics among the 3 HIV care groups. Care patterns were examined by month from September 2018 to December 2021 for the in-care group, and March 2020 to December 2021 for the new-to-care and re-engaging groups: (1) number of HIV care encounters (categorized as in-person or telemedicine); (2) proportion with unsuppressed VL (defined as the number of unsuppressed tests divided by the total number of tests in that month); (3) number of days between HIV care encounters; (4) number of days between VL measurements. To estimate the crude (OR) and adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) of unsuppressed VL after the restrictions lifted (June 1, 2020 to December 31, 2021) by HIV care group (and accounting for covariates), logistic regression models with general estimating equations (GEE) and an exchangeable working correlation structure were used. To test the robustness of the results, subgroup analyses were conducted by excluding participants from the 3 cohorts with different barriers to care compared with community- or academic-based clinics. To estimate the relationship between state-level COVID-19 restriction policy strength (categorized monthly) and unsuppressed VL from May 2020 to April 2021 (the calendar months with heterogeneity in the strength of monthly COVID-19-related restriction policies), a generalized linear mixed model was used to account for random effects for each PWH. An interaction term for policy strength and calendar month tested the hypothesis that the relationship between policy strength and unsuppressed VL changed during and after the COVID-19 pandemic-related restrictions (Supplementary Method 2). All analyses were conducted using in R software V.4.5 (R Foundation for Statistical Computing, Vienna, Austria) and SAS Software V.9.4 (SAS Institute). P-value < .05 guided statistical interpretation. RESULTS Characteristics of the Study Population The study population included 49 193 people. The in-care group (N = 48 605) had a median age of 55 [interquartile range [IQR] 43, 63] years, while those new-to-care (N = 105) and re-engaging (N = 483) were younger, with median ages of 38 [IQR 29, 52] years and 42 [IQR 32, 55] years, respectively (Table 1). Most individuals were male (87% in-care, 77% new-to-care, 89% re-engaging), and non-Hispanic Black was the largest racial group (43% of those in-care, 50% of those new-to-care and 42% re-engaging in care). Compared with our study population (which is predominantly those in-care), the 12 783 PWH excluded from our study because they did not have a VL measurement after pandemic restrictions (between June 1, 2020 and December 1, 2021) were older (median age 56 vs 54 years), had lower median CD4 counts (538 vs 627), and a lower proportion with suppressed VL at baseline (34% vs 61%) (Supplementary Table 1). Of the 49 193 individuals in the study population, 1307 died during the study period (September 1, 2018 to December 31, 2021). In the comparison population, among PWH with ≥1 HIV encounter, CD4, or VL measurement between September 1, 2016 and February 29, 2018 (N = 57 073), 9201 (16.1%) did not return to care (Supplementary Figure 2). In the primary study population, among PWH with ≥1 HIV encounter, CD4, or VL measurement between September 1, 2018 and February 29, 2020 (N = 59 022), 10 417 (17.6%) did not return to care after the pandemic-related restrictions (Supplementary Figure 1). The proportion of participants who did not return to care was similar during the pandemic and the comparable prepandemic period. Table 1.Baseline Characteristics, by HIV Care Group Engaging Prior to (In-care) and During (New-to-care and re-engaging) the Pandemica,b,cIn-careNew-to-careRe-engagingTotalN = 48 605N = 105N = 483N = 49 193Median age [interquartile range]55 [44, 63]38 [29, 52]42 [32, 55]54 [44, 62]Sex………… Male42 936 (87%)81 (77%)430 (89%)43 447 (88%) Female5669 (12%)24 (23%)53 (11%)5746 (12%)Race and ethnicity………… Non-Hispanic White18 973 (39%)28 (27%)177 (37%)19 178 (39%) Non-Hispanic Black20 802 (43%)53 (50%)204 (42%)21 059 (43%) Hispanic5731 (12%)13 (12%)69 (14%)5813 (12%) Other/unknown3099 (6%)11 (11%)33 (7%)3143 (6%)HIV acquisition group………… Injection drug use2592 (5%)3 (3%)27 (6%)2622 (5%) Men who have sex with men15 545 (32%)37 (35%)196 (41%)15 778 (32%) Heterosexual contact7526 (16%)29 (28%)83 (17%)7638 (16%) Other/unknown22 942 (47%)36 (34%)177 (37%)23 155 (47%)Median CD4 count [interquartile range]629 [424, 860]382 [190, 610]513 [268, 782]627 [422, 859]CD4 (count +/− 90 d of baseline)32 944 (68%)86 (82%)394 (82%)33 424 (68%)HIV RNA level (measured +/− 30 d of baseline)………… HIV RNA ≤ 200 copies/mL29 972 (62%)0 (0%)231 (48%)30 203 (61%) HIV RNA > 200 copies/mL2177 (4%)67 (64%)141 (29%)2385 (5%) HIV RNA not assessed16 456 (34%)38 (36%)111 (23%)16 605 (34%)ARV regimend………… INSTI30 875 (64%)70 (67%)343 (71%)31 288 (64%) NNRTI5916 (12%)3 (3%)28 (6%)5947 (12%) PI2768 (6%)2 (2%)22 (5%)2792 (6%) Dual agent7498 (15%)3 (3%)39 (8%)7540 (15%) Other/unknown851 (2%)0 (0%)9 (2%)860 (2%) None697 (1%)27 (26%)42 (9%)766 (2%) aBaseline: measured at the date of the last CD4 or VL measurement or encounter prior to March 1, 2020 among those in-care, and at the first visit on or after March 1, 2020 among those new-to-care and re-engaging. bPercentages may not sum to 100% due to rounding. c P-values calculated using χ2 tests or Fisher's exact tests (if cell counts are <5) for all categorical variables and ANOVA tests for all continuous variables. P < .001 for differences in all baseline characteristics by group, except for race (P = .07). dARV regimen recorded +/− 30 d of baseline.Abbreviations: ART, antiretroviral therapy; INSTI, integrase strand transfer inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor. Baseline Characteristics, by HIV Care Group Engaging Prior to (In-care) and During (New-to-care and re-engaging) the Pandemica,b,c aBaseline: measured at the date of the last CD4 or VL measurement or encounter prior to March 1, 2020 among those in-care, and at the first visit on or after March 1, 2020 among those new-to-care and re-engaging. bPercentages may not sum to 100% due to rounding. c P-values calculated using χ2 tests or Fisher's exact tests (if cell counts are <5) for all categorical variables and ANOVA tests for all continuous variables. P < .001 for differences in all baseline characteristics by group, except for race (P = .07). dARV regimen recorded +/− 30 d of baseline. Abbreviations: ART, antiretroviral therapy; INSTI, integrase strand transfer inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor. Trends in the Number of HIV Care Encounters and Proportion With Suppressed VL Among the in-care group, the number of monthly encounters ranged from 19 680 to 23 889 in 2019 (Figure 1A). Monthly encounters dropped sharply from March 2020 (n = 21 009) to May (n = 14 747), then began recovering in June (n = 18 679) and July (n = 19 413). Telemedicine use rose dramatically from 4% of encounters in January 2020 to 33% in April 2020. By the end of 2021, total encounters exceeded prepandemic levels and the proportion of telemedicine encounters declined and plateaued above prepandemic levels. The proportion with suppressed VL remained stable and >90% in all months from 2019 to 2021, ranging from 91% (February 2019) to 94% (July 2020). In the new-to-care group, the number of encounters declined slightly from n = 63 in March 2020 to n = 42 in December 2021 and suppressed VL increased to 87% (June 2020) remaining high through December 2021 (Figure 1B). In the re-engaging group, the number of encounters declined from n = 280 in March 2020 to n = 138 in December 2021 (Figure 1C). The proportion with suppressed VL increased from 61% (March 2020) to 85% (July 2020) and remained between 85% and 95% through December 2021. Number of HIV care encounters and HIV RNA measurements from March 2020 to December 2021 among those (A) in-care prior to the pandemic, (B) new-to-care, and (C) re-engaging in HIV care during pandemic restrictions. For those new-to-care or re-engaging in care during the pandemic, the x-axis begins in March 2020 and extends to December 2021 due to the definitions of these groups. The Days Between Encounters and VL Measurements Before, During, and After Pandemic-related Restrictions Among the in-care group, the days between HIV care encounters remained relatively stable, with median of 70 [IQR: 21, 150] days between encounters before restrictions and a median of 73 [IQR: 19, 154] days between encounters during restrictions (Figure 2A). After restrictions, the interval decreased to a median of 55 [IQR: 15, 141] days. Both the new-to-care and re-engaging groups had shorter intervals between encounters during restrictions, with medians of 28 [IQR: 7, 70] and 49 [IQR: 14, 133] days, respectively. After restrictions were lifted, these intervals increased to a median of 59 [IRQ: 16, 147] days for the new-to-care group and 85 [IQR: 23, 181] days for the re-engaging group. Distribution of days (A) between HIV care encounters and (B) between HIV RNA tests, HIV care group (in-care, new-to-care, and re-engaging in care) and calendar time (before, during, and after COVID-19 restrictions). Before restrictions: The number of days between an individual's last 2 encounters or viral load (VL) tests before March 1, 2020 within the calendar time range of September 1, 2018 to February 29, 2020. During restrictions: The number of days between an individual's first encounter occurring between March 1, 2020 and May 31, 2020 and a subsequent encounter before December 31, 2021. After Restrictions: The number of days between an individual's last 2 encounters or VL tests occurring between June 1, 2020 and December 31, 2021. % Definition: Represents the proportion of individuals included in the boxplot relative to the total number of individuals in that group with recorded data for the corresponding time periods. Total number of in care group: N = 48 605; New-to-care group: N = 105; Re-engaging group: N = 483. Similar patterns were observed in the median days between VL tests across HIV care groups and calendar time (in-care group: median number of days between VL tests was 168 [IQR: 105, 199] before, 174 [IQR: 112, 217] during, and 176 [IQR: 117, 211] after restrictions; new-to-care: median number of days between VL tests was 51 [IQR: 35, 91] during and 179 [IQR: 98, 196] after restrictions; re-engaging group: median number of days between VL tests was 121 [IQR: 62, 206] during and 169 [IQR: 105, 210] after restrictions (Figure 2B). Factors Associated With Unsuppressed VL Compared with the in-care group, the combined group of those new-to-care or re-engaging in care had a 70% increase in the crude odds of unsuppressed VL, which attenuated to the null in multivariable analyses (Table 2). The findings remained robust after excluding the 3 cohorts with lower barriers to accessing care. There was a decrease in the odds of unsuppressed VL with increasing age, and an increase among non-Hispanic Black (vs white) individuals, those in the heterosexual or IDU acquisition risk group (vs men who have sex with men), and those with lower CD4 counts (vs ≥ 500 cells/mm3); findings were robust in subgroup analyses. Table 2.Factors Associated With Unsuppressed Viral Load (>200 Copies/mL) After the Restriction (June 2020–December 2021)aPrimary AnalysisbSubgroup AnalysiscFactorsCrude OR (95% CI)Adjusted OR (95% CI)Crude OR (95% CI)Adjusted OR (95% CI)HIV Care Groups………… In care prior to the pandemicReferenceReferenceReferenceReferenceNew-to-care or re-engaging in care during pandemic restrictions 1.70 (1.40, 2.07) 0.91 (.73, 1.13) 1.68 (1.27, 2.23)0.83 (.60, 1.14)Age (in decades)d 0.82 (.80, .84) 0.76 (.74, .77) 0.75 (.72, .78) 0.71 (.68, .74) Sex………… MaleReferenceReferenceReferenceReference Female 1.30 (1.20, 1.41) 1.05 (.95, 1.16) 1.28 (1.15, 1.42)1.10 (.96, 1.26)Race and ethnicity … ……… Non-Hispanic WhiteReferenceReferenceReferenceReference Non-Hispanic Black 1.75 (1.65, 1.87) 1.51 (1.42, 1.61) 1.95 (1.76, 2.16) 1.52 (1.36, 1.69) Hispanic1.07 (.97, 1.18)0.95 (.86, 1.05) 1.28 (1.09, 1.51)1.03 (.88, 1.22) Other/unknown1.01 (.89, 1.16)0.88 (.77, 1.01)1.26 (.99, 1.65)0.98 (.76, 1.26)HIV acquisition risk group………… Men who have sex with menReferenceReferenceReferenceReference Heterosexual contact 1.57 (1.44, 1.71) 1.32 (1.19, 1.47) 1.49 (1.34, 1.65) 1.27 (1.11, 1.46) Injection drug use 1.81 (1.61, 2.04) 1.81 (1.60, 2.05) 1.87 (1.62, 2.15) 1.87 (1.60, 2.19) Other/unknown 1.34 (1.25, 1.44) 1.51 (1.41, 1.62) 1.28 (1.06, 1.55)1.02 (.84, 1.24)CD4 countse (cells/mm3)………… CD4 ≥ 500ReferenceReferenceReferenceReference CD4 200-499 2.24 (2.11, 2.38) 2.39 (2.25, 2.55) 2.58 (2.33, 2.85) 2.66 (2.40, 2.95) CD4 < 200 8.58 (7.92, 9.29) 8.65 (7.97, 9.38) 9.74 (8.58, 11.04) 9.41 (8.27, 10.71) Not assessed1.53 (.75, 3.14)1.47 (.68, 3.16)1.21 (.32, 4.52)1.55 (.40, 5.93) aLogistic regression models with general estimating equations (GEE) were used to account for the correlation of repeated HIV RNA measurements for each participant and to estimate crude and adjusted odds ratios (aOR). In addition to the covariates in the table, the adjusted models also accounted for differences by cohort. bPrimary analysis: With participants from all 12 clinical cohorts (N = 49 193). cSubgroup analysis: Exclude participants from 3 cohorts with different barriers to care (N = 17 164). dAge modeled per 10-y (decade) increase. eCD4 count is defined as the measurement closest to the start of the postpandemic restriction era. For the in-care group, the range within which CD4 was selected was January 1, 2016 to May 31, 2020. For the new/re-engaging group, the range within which CD4 was selected was January 1, 2016 to 90 d after the CD4, VL or encounter that occurred during the restriction period (March 1, 2020–May 31, 2020).Bold values indicate statistical significance at a P-value less than .05. Factors Associated With Unsuppressed Viral Load (>200 Copies/mL) After the Restriction (June 2020–December 2021)a aLogistic regression models with general estimating equations (GEE) were used to account for the correlation of repeated HIV RNA measurements for each participant and to estimate crude and adjusted odds ratios (aOR). In addition to the covariates in the table, the adjusted models also accounted for differences by cohort. bPrimary analysis: With participants from all 12 clinical cohorts (N = 49 193). cSubgroup analysis: Exclude participants from 3 cohorts with different barriers to care (N = 17 164). dAge modeled per 10-y (decade) increase. eCD4 count is defined as the measurement closest to the start of the postpandemic restriction era. For the in-care group, the range within which CD4 was selected was January 1, 2016 to May 31, 2020. For the new/re-engaging group, the range within which CD4 was selected was January 1, 2016 to 90 d after the CD4, VL or encounter that occurred during the restriction period (March 1, 2020–May 31, 2020). Bold values indicate statistical significance at a P-value less than .05. The Association of State-level COVID-19-related Restriction Policy Strength and Unsuppressed VL The relationship between state-level COVID-related restriction policy strength and unsuppressed VL was qualitatively different in May–June 2020 and July 2020–April 2021 (Figure 3), albeit the associations of policy strength and unsuppressed VL stratified by time were not statistically significant (Supplementary Table 2). From May to June 2020, those living in states with moderate (vs mild) and strong (vs mild) COVID-19 restriction policies had a 26% and 20% (respectively) decrease in the adjusted odds of unsuppressed VL. From July 2020 to April 2021, those in moderate (vs mild) restriction states had a similar adjusted odds of unsuppressed VL while those in strong (vs mild) restriction states had a 31% increase in the adjusted odds of unsuppressed VL compared with those in mild restriction states. Association between COVID-related restriction policy strength and unsuppressed viral load (>200 copies/mL), May 2020 to April 2021 (N = 44 807*). * Of the 48 605 participants from the in-care group, 44 807 had a recorded residence state and at least 1 viral load test between May 2020 and April 2021. A generalized linear mixed model with splined calendar month estimated adjusted ORs for the association between COVID-related restriction policy strength and unsuppressed VL, incorporating an interaction between policy strength and calendar month (with knot at June 2020). The model accounted for random effects per PWH and adjusted for age, sex, race, region, HIV acquisition risk, and baseline CD4 count. DISCUSSION Our findings suggest HIV care in the U.S. remained resilient during the first 2 years of the COVID-19 pandemic for those who had at least 1 VL measured after COVID-19 restrictions (June 1, 2020–December 31, 2021). The frequency of HIV care encounters and VL monitoring returned to prepandemic levels by December 2020 and were maintained in 2021. Viral suppression was largely maintained among PWH who were in care prior to the pandemic, as well as those new-to-care or re-engaging during the first months of the pandemic. HIV clinics, providers, and patients adapted quickly and effectively to the challenges caused by the disruptions to healthcare access—a success story during a public health emergency. Multiple factors likely contributed to reducing the impact of COVID-related disruptions on HIV care. The rapid scale-up of telehealth encounters provided patients with access to healthcare when in-person encounters were disrupted [18–21]. Our data support the assertion that HIV clinics may have prioritized those who were new-to-care or re-engaging in care when in-person access was limited, showing guideline-aligned intensive VL monitoring after ART initiation or restart likely occurred with shorter time intervals between initial encounters followed by longer intervals. At the local-level, programs targeting populations who are more vulnerable to adverse outcomes (eg, racial and ethnic minorities, sexual minorities, and persons who inject drugs) were launched to engage those during COVID-related healthcare disruptions, like the “Keep It Going If You Can” program in Seattle, Washington [22]. Although direct evidence of effectiveness is limited, the high rates of viral suppression achieved reflect multilevel efforts to minimize barriers to ART and viral suppression, and particularly those at higher risk for adverse outcomes. Ryan White—a federal program that supports PWH to achieve viral suppression—allowed flexible ART prescription refills, including 60- or 90-day supplies and home delivery [23]. The increase in viral suppression from March to June 2020 among those new-to-care and re-engaging during the early months of the pandemic was likely facilitated by the prioritization for care of these groups, with the median time between VL tests reflecting the ability of patients and providers to engage in routine clinical management despite the pandemic-related restrictions. Other studies have also demonstrated VL monitoring and suppression were similar to prepandemic levels by the end of 2020. El-Nahal et al., examined VL monitoring patterns among those in the Johns Hopkins HIV Clinical Cohort (a contributing cohort to NA-ACCORD) and reported that while laboratory closures (March 16, 2020–July 12, 2020) delayed VL testing, viral suppression remained high once the labs reopened (July 13, 2020–December 31, 2020) [15]. Factors not related to pandemic disruptions may have also influenced VL suppression, including the ability to maintain viral suppression despite gaps in care of up to 9 months, which is longer than the most restrictive COVID-related policies [24]. Importantly, patient-level factors independently associated with unsuppressed VL, were consistent with long-standing differences in HIV outcomes, including younger age, being non-Hispanic Black, having a history of IDU, and lower CD4 count [25–27]. These findings suggest that while the system showed resilience, the subgroups at greatest risk for unsuppressed VL continued to be at greatest risk during COVID-related disruptions and immediately after. Expanding existing efforts to reduce these differences across demographic and behavioral risk groups are necessary outside of public health emergencies to protect against exacerbation of these differences during emergencies. While we did not observe statistically significant associations between state-level COVID-related restriction policy strength and viral suppression, the magnitude of association suggests lower odds of unsuppressed VL in states with moderate or strong (vs mild) policies from May to June 2020, and a reversal of this relationship from July 2020 to 2021 among those in states with strong (vs mild) restriction policies from July 2020 to April 2021. These patterns suggest that prolonged strong restriction policies may challenge viral suppression for PWH, and early efforts to prioritize access to HIV care for subgroups with greater vulnerability to unsuppressed VL offset potential negative impacts of the restriction policies [28]. Our analysis was limited by the absence of local contextual data (eg, access to public transportation and city- or county-level COVID-restriction policies), and additional patient-level data (eg, employment status, educational background, income, and other barriers to healthcare). Future studies should assess how patient-level socioeconomic factors and local COVID-related restriction policies influenced HIV care encounters and viral suppression. Limitations The first and most important limitation of our study is that our study population was limited to individuals who had at least 1 VL test occurring June 1, 2020 and December 31, 2021, effectively restricting our study population to those who engaged in HIV care at least once following the period of strongest COVID-related restriction policies. Although this restriction was necessary to estimate the proportion with suppressed VL, we caution against generalizing our findings to those who did not engage in HIV care after initial COVID-related shut-downs (a population who may be more likely to experience disruptions to the continuity of HIV care and unsuppressed VL). Future studies are needed to collect data on PWH who did not re-engage in HIV care after the pandemic-related restrictions and to identify the factors that influenced their return to care. Second, we again urge caution regarding the generalizability of our findings to all PWH in the U.S. because many NA-ACCORD sites are highly resourced, with established systems to support patient retention and adherence (eg, specialized HIV pharmacies and mail-order ART delivery), which may result in a greater likelihood in continuity of HIV care and viral suppression. The sociodemographic characteristics of participants in the NA-ACCORD, however, are broadly similar to those of the national population of PWH [29]. Third, the definition of new-to-care is based on the absence of encounters, VL or CD4 tests, and ART prescriptions available in the NA-ACCORD data. It is possible that patients who transferred care from other clinical centers or had a prior period of HIV care engagement not observable within the NA-ACCORD were misclassified as new-to-care. Fourth, our study population only includes cisgender PWH; the findings may not be generalizable to transgender PWH. Finally, clinic-level spatial analysis was not possible because several NA-ACCORD cohorts are multisite cohorts, and participants are linked to cohorts, not specific clinics where care was received. We were, however, able to adjust for cohort-level variation, which helps to account for differences in clinical care policies. CONCLUSIONS Our study showed that the frequency of HIV care encounters declined during the initial months of the pandemic, rebounded to prepandemic levels by the end of 2020 and was maintained in 2021. Viral suppression was maintained during the pandemic among those who returned to care after the initial months of the pandemic (June 2020–December 2021). Those living in states with strong COVID-19-related restriction policies between July 2020 and April 2021 experienced a higher increase in the odds of viral suppression and additional studies investigating patient-, clinic-, and local-level drivers of this increase are needed. Our study highlights the resilience of the healthcare system in maintaining continuity of HIV care and viral suppression among PWH who remained engaged in care, despite widespread disruptions, and offers an example of how to prepare for, and respond to, future disruptions to care. Supplementary Material Notes Author Contributions. Concept and design: K.N.A., A.M.P., Y.Z., and B.C.H. Acquisition, analysis, or interpretation of data: B.C.H. and Y.Z. Drafting of the manuscript: Y.Z., K.N.A., and A.M.P. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: Y.Z. and B.C.H. Obtained funding: K.N.A. and A.M.P. Supervision: K.N.A. and A.M.P. Financial support. This nested study was supported in part by National Institute of Mental Health grant 1RF1MH133427. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The NA-ACCORD was supported by National Institutes of Health grants U01AI069918, F31AI124794, F31DA037788, G12MD007583, K01AI093197, K01AI131895, K01AI157849, K23EY013707, K24AI065298, K24AI118591, K24DA000432, KL2TR000421, N01CP01004, N02CP055504, N02CP91027, P30AI027757, P30AI027763, P30AI027767, P30AI036219, P30AI050409, P30AI050410, P30AI094189, P30AI110527, P30MH62246, R01AA016893, R01AG053100, R01AI177010, R01DA011602, R01DA012568, R24AI067039, R34DA045592, U01AA013566, U01AA020790, U01AI038855, U01AI038858, U01AI068634, U01AI068636, U01AI069432, U01AI069434, U01DA036297, U01DA036935, U10EY008057, U10EY008052, U10EY008067, U01HL146192, U01HL146193, U01HL146194, U01HL146201, U01HL146202, U01HL146203, U01HL146204, U01HL146205, U01HL146208, U01HL146240, U01HL146241, U01HL146242, U01HL146245, U01HL146333, U24AA020794, U54GM133807, UL1RR024131, UL1TR000004, UL1TR000083, UL1TR002378, Z01CP010214, and Z01CP010176; contracts CDC-200-2006-18797 and CDC-200-2015-63931 from the Centers for Disease Control and Prevention, USA; contract 90047713 from the Agency for Healthcare Research and Quality, USA; contract 90051652 from the Health Resources and Services Administration, USA; the Grady Health System; grants CBR-86906, CBR-94036, HCP-97105 and TGF-96118 from the Canadian Institutes of Health Research, Canada; Ontario Ministry of Health and Long Term Care, and the Government of Alberta, Canada. Additional support was provided by the National Institute Of Allergy And Infectious Diseases (NIAID), National Cancer Institute (NCI), National Heart, Lung, and Blood Institute (NHLBI), Eunice Kennedy Shriver National Institute Of Child Health & Human Development (NICHD), National Human Genome Research Institute (NHGRI), National Institute for Mental Health (NIMH) and National Institute on Drug Abuse (NIDA), National Institute On Aging (NIA), National Institute Of Dental & Craniofacial Research (NIDCR), National Institute Of Neurological Disorders And Stroke (NINDS), National Institute Of Nursing Research (NINR), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), and National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). This manuscript is the result of funding in whole or in part by the National Institutes of Health (NIH). It is subject to the NIH Public Access Policy. Through acceptance of this federal funding, NIH has been given a right to make this manuscript publicly available in PubMed Central upon the Official Date of Publication, as defined by NIH. Data availability. Data from the NA-ACCORD can be accessed by submitting a Concept Sheet Proposal to NA-ACCORD, available at https://naaccord.org. NA-ACCORD Collaborating Cohorts and Representatives: AIDS Clinical Trials Group Longitudinal Linked Randomized Trials: Constance A. Benson and Ronald J. Bosch. AIDS Link to the IntraVenous Experience: Gregory D. Kirk. Emory-Grady HIV Clinical Cohort: Vincent C. Marconi and Jonathan Colasanti. Fenway Health HIV Cohort: Kenneth H. Mayer. HIV Outpatient Study: Kate Buchacz and Jun Li. HIV Research Network: Kelly A. Gebo and Richard D. Moore. Johns Hopkins HIV Clinical Cohort: Richard D. Moore. John T. Carey Special Immunology Unit Patient Care and Research Database, Case Western Reserve University: George Yendewa. Kaiser Permanente Mid-Atlantic States: Michael A. Horberg. Kaiser Permanente Northern California: Michael J. Silverberg. Longitudinal Study of Ocular Complications of AIDS: Jennifer E. Thorne. MACS/WIHS Combined Cohort Study: Todd Brown, Phyllis Tien, and Gypsyamber D'Souza. Maple Leaf Medical Clinic: Graham Smith, Mona Loutfy, and Meenakshi Gupta. The McGill University Health Centre, Chronic Viral Illness Service Cohort: Marina B. Klein Multicenter Hemophilia Cohort Study–II: Charles Rabkin. Ontario HIV Treatment Network Cohort Study: Abigail Kroch, Ann Burchell, Adrian Betts, and Joanne Lindsay. Parkland/UT Southwestern Cohort: Ank Nijhawan. Retrovirus Research Center, Universidad Central del Caribe, Bayamon Puerto Rico: Angel M. Mayor. Southern Alberta Clinic Cohort: M. John Gill. Study of the Consequences of the Protease Inhibitor Era: Jeffrey N. Martin and Steven G. Deeks. Study to Understand the Natural History of HIV/AIDS in the Era of Effective Therapy: Jun Li and John T. Brooks. University of Alabama at Birmingham 1917 Clinic Cohort: Michael S. Saag, Michael J. Mugavero, and Greer Burkholder. University of California at San Diego: Laura Bamford and Maile Karris. University of North Carolina at Chapel Hill HIV Clinic Cohort: Joseph J. Eron and Sonia Napravnik. University of Washington HIV Cohort: Mari M. Kitahata and Heidi M. Crane. Vanderbilt Comprehensive Care Clinic HIV Cohort: Timothy R. Sterling, David Haas, Peter Rebeiro, and Megan Turner. Veterans Aging Cohort Study: Lesley Park, Kathleen McGinnis, and Amy Justice. NA-ACCORD Study Administration: Executive Committee: Keri N. Althoff, Richard D. Moore, Mari M. Kitahata, Catherine R. Lesko, Michael A. Horberg, Marina B. Klein, Rosemary G. McKaig, and Aimee M. Freeman. Administrative Core: Keri N. Althoff, Richard D. Moore, and Aimee M. Freeman. Data Management Core: Mari M. Kitahata, Heidi M. Crane, Joseph (Chris) Delaney, Liz Morton, Justin McReynolds, and William B. Lober. Epidemiology and Biostatistics Core: Catherine R. Lesko, Sally B. Coburn, Brenna Hogan, Elizabeth Humes, Wendy (Chunyan) Zheng, Jason Haw, Yawei Cheng, and Lucas Gerace. Contributor Information Yue Zhang, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. Brenna C Hogan, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. Kelly A Gebo, Department of Epidemiology, Milken Institute School of Public Health, The George Washington University, Washington, D.C., USA. Peter F Rebeiro, Departments of Medicine, Biostatistics, and Biomedical Informatics, Vanderbilt University Medical Center, Nashville, Tennessee, USA. Milton L Wainberg, Department of Psychiatry, Columbia University Medical Center, New York State Psychiatric Institute, New York, New York, USA. Catherine R Lesko, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. Vincent C Marconi, Division of Infectious Diseases, School of Medicine, Emory University, Atlanta, Georgia, USA; Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA. N Sarita Shah, Division of Infectious Diseases, School of Medicine, Emory University, Atlanta, Georgia, USA; Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, Georgia, USA. Raynell Lang, Department of Medicine, University of Calgary Cumming School of Medicine, Calgary, Alberta, Canada. Nel Jason L Haw, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. Yawei Cheng, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. Michael A Horberg, Mid-Atlantic Permanente Research Institute, Kaiser Permanente Mid-Atlantic States, Washington, DC, USA. Jennifer O Lam, Division of Research, Kaiser Permanente Northern California, Pleasanton, California, USA. Michael J Silverberg, Division of Research, Kaiser Permanente Northern California, Pleasanton, California, USA. George Yendewa, Division of Infectious Diseases and HIV Medicine, Case Western Reserve University, Cleveland, Ohio, USA. Joseph J Eron, Division of Infectious Diseases, University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, North Carolina, USA. Richard D Moore, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA; Department of Medicine, School of Medicine, The Johns Hopkins University, Baltimore, Maryland, USA. Angela M Parcesepe, Department of Maternal and Child Health, Gillings School of Global Public Health, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA. Keri N Althoff, Department of Epidemiology, Bloomberg School of Public Health, The Johns Hopkins University, Baltimore, Maryland, USA. for the North American AIDS Cohort Collaboration on Research and Design (NA-ACCORD): Constance A Benson, Ronald J Bosch, Gregory D Kirk, Vincent C Marconi, Jonathan Colasanti, Kenneth H Mayer, Kate Buchacz, Jun Li, Kelly A Gebo, Richard D Moore, Richard D Moore, George Yendewa, Michael A Horberg, Michael J Silverberg, Jennifer E Thorne, Todd Brown, Phyllis Tien, Gypsyamber D’Souza, Graham Smith, Mona Loutfy, Meenakshi Gupta, Marina B Klein, Charles Rabkin, Abigail Kroch, Ann Burchell, Adrian Betts, Joanne Lindsay, Ank Nijhawan, Angel M Mayor, M John Gill, Jeffrey N Martin, Steven G Deeks, Jun Li, John T Brooks, Michael S Saag, Michael J Mugavero, Greer Burkholder, Laura Bamford, Maile Karris, Joseph J Eron, Sonia Napravnik, Mari M Kitahata, Heidi M Crane, Timothy R Sterling, David Haas, Peter Rebeiro, Megan Turner, Lesley Park, Kathleen McGinnis, Amy Justice, Keri N Althoff, Richard D Moore, Mari M Kitahata, Catherine R Lesko, Michael A Horberg, Marina B Klein, Rosemary G McKaig, Aimee M Freeman, Keri N Althoff, Richard D Moore, Aimee M Freeman, Mari M Kitahata, Heidi M Crane, Joseph (Chris) Delaney, Liz Morton, Justin McReynolds, William B Lober, Catherine R Lesko, Sally B Coburn, Brenna Hogan, Elizabeth Humes, Wendy (Chunyan) Zheng, Jason Haw, Yawei Cheng, and Lucas Gerace Supplementary Data Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author. References Associated Data Supplementary Materials