HIV Care Continuum after Rapid Start with Patient Navigation in a Large Urban HIV Clinic in Dallas,TX

Abstract

Rapid start of antiretroviral therapy (ART) has been recommended by multiple international and US-based groups because of its ability to decrease time to initiation of ART, virologic suppression, and improve retention in care. However, there has been heterogeneity in the components of rapid start programs. This prospective study examines HIV care continuum outcomes after patient navigation (PN) was added to our large rapid start program in Dallas, TX. Newly diagnosed patients were enrolled in rapid start + PN (n = 199) for 12 months, and HIV care continuum outcomes were compared with historical pre-rapid start (N = 295) and rapid start (N = 244) cohorts. Cumulative incidence curves were used to summarize time to virologic suppression, incorporating death as a competing risk, and stepwise competing-risk regression and logistic regression models were used to build multivariate models to evaluate the associations between rapid start, PN, and time to virologic suppression, and between retention in care and sustained virologic suppression, respectively. Rapid start + PN was associated with increased odds of being retained in care [adjusted odds ratio (aOR) 2.00; 95% confidence interval (CI) 1.32, 3.02; p < 0.01] compared with pre-rapid start. The time to virologic suppression was not significantly different between the three cohorts [pre-rapid (2.50 months), rapid start (2.56 months), and rapid start + PN (2.78 months), p = 0.41], while the odds of sustained virologic suppression in the rapid start + PN group was significantly lower (aOR 0.65, 95% CI 0.44, 0.97; p = 0.04). Though PN had mixed HIV care continuum outcomes, it can be adapted to augment rapid start programs and improve retention in care.

Keywords

patient navigation rapid start rapid initiation of ART HIV care continuum

Introduction

Morbidity and mortality due to HIV persist in the United States, particularly in the US South, where almost half of new HIV diagnoses occur.¹ Interventions to increase early diagnosis and linkage to care and expedite treatment initiation are key components of ending the HIV epidemic and reducing health disparities. Rapid start of antiretroviral therapy (ART) has been shown to be a successful strategy in the United States, decreasing time to virologic suppression^2–4 and improving retention in care.⁵ Multiple international and US-based groups have now recommended rapid start in their guidelines, including the World Health Organization,⁶ International Antiviral Society—USA,⁷ and US Department of Health and Human Services.⁸

While rapid start is now widely recommended, implementation strategies and components of rapid start programs have been heterogeneous. Rapid start is generally implemented as a package of interventions, making it difficult to assess the contribution of each component to HIV care continuum outcomes. Aside from the common provision of same-day access to an HIV provider and medications, successful US programs have reported numerous components including taxi-vouchers to bring patients to clinic on the date of diagnosis,³ case management (CM) and a 24-h linkage navigator,⁴ in-clinic administration of the first dose of ART^3,4 provision of free ART samples,⁹ nursing telephone follow-up³ and education sessions,² and a peer navigator to assist with obtaining documents for medical coverage.²

In our own program, which focuses on the rapid start visit day, patients have sequential visits with CM, financial counselor, HIV provider, medication access specialist, and pharmacy pickup of ART on the day the patient is referred to our clinic. We have demonstrated that rapid start was associated with improved medication adherence but did not improve the time to virologic suppression. Other outcomes after the rapid start visit such as retention in care and sustained virologic suppression had mixed outcomes depending on the definition of the metric used and were not necessarily improved by our rapid start intervention alone.¹⁰ Thus, patient navigation (PN) was added to our rapid start program to extend the intervention beyond the first clinic visit day. In this second iteration of our rapid start program, we examine the impact of adding a PN component on HIV care continuum outcomes. Specifically, we aim to (1) compare linkage to care, ART initiation, virologic suppression, retention in care, and sustained virologic suppression for newly diagnosed individuals receiving rapid start with PN, to historical cohorts (pre-rapid start, rapid start alone); and (2) assess for independent associations between demographic, socioeconomic, and behavioral factors and HIV clinical outcomes in these cohorts.

Methods

Study design, population, and context

This study evaluates a prospective, nonrandomized patient navigator intervention for patients entering care after a new diagnosis of HIV. Since October 2018, Parkland HIV Services (PHS) has implemented a rapid start program for patients newly diagnosed with HIV to be seen and initiated on ART within 7 days (most within 1 day) of the day of referral. PHS is a large HIV program embedded in the county safety-net hospital system in Dallas, TX, comprised of four HIV clinics that care for ∼6700 patients per year. PHS receives funding from the Ryan White HIV/AIDS Program (RWHAP) parts A–D, and more than half of patients are funded primarily through this program, with the remainder on Medicare and Medicaid, and a small number on commercial insurance programs. The data from the prospective patient navigator intervention are compared with retrospective data (pre-rapid start and rapid start only) previously collected.¹⁰

Rapid start program and PN protocol

The PHS rapid start program process and structure have been previously described.¹⁰ Briefly, after being referred to PHS, newly diagnosed patients are scheduled within 1–7 days to undergo a rapid start visit which includes appointments with a case manager for intake and evaluation of psychosocial needs, a financial counselor for assessment of insurance coverage or enrollment in RWHAP, a medical provider for evaluation and initiation of ART, a medication access specialist to enroll the patient in appropriate programs for ART coverage or copay assistance, the laboratory for initial blood tests, and the pharmacy for pickup of ART. The rapid start visit takes an average of 4.5 h to complete. Follow-up lab appointments are typically made for 4–6 weeks, with a provider visit shortly thereafter.

A patient navigator (PN) was added in February 2021 to help systematize the follow-up process for patients and add an extra layer of communication from someone who could help them navigate the complex clinic processes and environment. The patient navigator was a bilingual Latino male in his twenties who had an interest in HIV care. He received individualized training in clinic processes and rotated with each of the major clinic teams (e.g., CM, medical providers, medication access specialists) before beginning his role. His role included screening patients for eligibility, meeting patients during their rapid start visit after seeing the provider, and obtaining verbal consent for study participation. Those who could not be met in person were contacted by phone and enrolled if agreeable. The PN performed telephone outreach at 1 week after the rapid start visit, and every 3 months for the first 12 months after the first clinic visit. The calls included a checklist of topics including medication refills, medication issues, medical coverage, changes to contact information, and remaining questions or needs. Participants filled out a survey as a part of their first contact with the PN and were given a $10 grocery store gift certificate for their time and were later offered this again if they filled out a follow-up survey months later. These survey results have been reported separately.¹¹

Eligibility criteria

Patients were eligible if they were: (a) establishing care for the first time at PHS; (b) at least 18 years old; (c) ART-naive; (d) diagnosed with HIV within the last 12 months; and (e) spoke English or Spanish. Participants were excluded from peer navigation if they were unable to provide informed consent, were pregnant (since they transfer care to maternal–fetal medicine clinic), or had an AIDS-defining condition at the time of enrollment that would contraindicate rapid initiation of ART. Individuals diagnosed within the preceding 12 months were included to account for delays in linkage to care and to better reflect the population served by PHS, a regional HIV referral center that serves eight counties.

Data collection

Data were collected from the electronic medical record (EMR), Epic (Verona, WI). All endpoints including demographics and baseline characteristics were abstracted from routinely collected data in the EMR.

Baseline characteristics

Age was determined on the date of the first clinic visit. Gender was defined as male, female, or transgender. Race and ethnicity were combined into non-Hispanic Black, non-Hispanic White, and Hispanic/other. Primary language was defined as English or non-English. HIV risk factor was categorized into men who have sex with men (MSM), heterosexual, injection drug use (IDU), or other. IDU was considered the primary risk factor for those who reported multiple risk factors. Marital status was considered as single/separated for those who reported being single, widowed, legally separated, divorced, other, or unknown. Highest education level was classified as less than high school, completed high school (including a General Education Development credential), or greater than high school. Insurance was classified as charity (including RWHAP)/self-pay (referring to those whose coverage has lapsed), Medicare/Medicaid, and commercial insurance. History of mental illness (MI) and substance use [SA; using the Substance Abuse and Mental Illness Symptoms Screener SAMISS],¹² as well as monthly family income, were determined by a case manager as part of the intake process. SAMISS scores were subsequently categorized as positive or negative for substance use (SAMISS-SU) and mental illness (SAMISS-MI) according to standard criteria.^13,14 PHQ-9 score of ≥ 10 was used to define major depression.¹⁵ Unstable housing includes those who were unhoused, lived in shelters, substance use treatment facilities, or transitional housing. Initial CD4 count was used to classify those with baseline < 200 cells/µL.

Outcomes

Linkage to care is reported categorically and as the time to completing a visit with an HIV provider within 12 months of CM intake. Initiated on ART is defined as having any prescription for ART sent to a pharmacy within 12 months of CM intake. Time to virologic suppression is defined as the number of days between intake visit with CM and the first VL < 200 copies/mL. Retained in care (“annual retention in care”) is defined by the US Health Resources and Services Administration (HRSA) as completing at least two HIV medical care encounters [medical visit with a prescribing provider or HIV viral load (VL) test] at least 90 days apart within a 12-month measurement year, at least one of which needs to be a medical visit with a provider.¹⁹ Continuous retention in care was also calculated using the HRSA HIV medical visit frequency performance measure, which requires at least one medical visit in each 6-month period with a minimum of 60 days between medical visits.¹⁹ Sustained virologic suppression is defined as a VL < 200 copies/mL between 6 and 12 months after CM intake date. Those without a VL measurement in this period are considered to not be suppressed.

All endpoints were measured from the date of CM intake since this represents the first time the patient interfaces with the clinic.

Data analysis

Sample size justification: Halperin et al. (2018) showed that the median time to viral suppression was 68 days in the historical cohort and 30 days in the CrescentCare Start Initiative (CCSI), a published rapid start program in the US South.⁸ We estimated the required sample size based on the primary end-point to determine whether there is a significant difference in time to virologic suppression before and after implementation of the rapid start program. A sample size of 200 patients in each of the retrospective and prospective cohorts would provide more than 90% power, with a two-sided significance level of 0.017 (0.05 ÷ 3) to adjust for multiple comparisons using the Bonferroni correction, based on a log-rank test accounting for competing risks. Power was estimated using the PASS sample size software.

Patients were stratified based on whether they initiated care in PHS clinics before rapid start (October 1, 2016, to September 31, 2018), after implementation of the rapid start protocol (October 1, 2018, to September 31, 2019), or after implementation of the PN intervention (February 1, 2021, to October 13, 2022). Outcome data were collected for the first 12 months after each patient’s CM intake. Baseline characteristics are reported as proportions based on available data for each variable. Since nearly 100% achieved linkage to care and were initiated on ART within the 12 months of the PN study, pre-rapid start and rapid start groups were compared with chi-squared tests only for these two outcomes, without multivariate modeling. Stepwise competing-risks regression, with death treated as a competing risk, was used to develop a multivariate model evaluating the association between rapid start and time to virologic suppression. Stepwise logistic regression was used to model retention in care, continuous retention in care, and sustained virologic suppression. Stepwise models included baseline variables with a univariate association with the outcome at p ≤ 0.20, with pre- or post-rapid start status included a priori to assess the strength of the association between the rapid start program and the outcomes while adjusting for significant covariates. Cumulative incidence curves were used to summarize the distribution of time to first virologic suppression. A sensitivity analysis was conducted by calculating the HIV care continuum outcomes among patients with any successful phone contact with the patient navigator after study enrollment, to better understand the effect of the patient navigator intervention. All analyses were performed in SAS, version 9.4.

Ethical approval and informed consent

The institutional review board (IRB) at UT Southwestern waived the need for full IRB approval and the need to obtain written informed consent for the study.

Results

Baseline characteristics and demographics

The study population overall included N = 738 individuals: rapid start + PN (N = 199) and two historical control groups, pre-rapid start (N = 295) and rapid start (N = 244) (Table 1). Mean age was 35.5, comprised primarily of cisgender men (75.8%), non-Hispanic Black (44.2%), and Hispanic (41.6%) individuals, with the majority speaking English (76.3%). The most reported HIV risk factor was heterosexual sex (54.6%), followed by MSM (40.4%) and IDU/other (5.0%). There was a broad distribution of education level among participants, with no significant differences between the three cohorts. Greater than 80% reported single/separated marital status in the three cohorts. The proportion with CD4 < 200 cells/mm³ was ∼28% across all three cohorts.

Table 1.

Baseline Characteristics for Individuals Newly Diagnosed with HIV Before Rapid Start, During Rapid Start, and Rapid Start with Patient Navigation

Characteristic	Totaln = 738	Pre-rapid startn = 295	Rapid startn = 244	Rapid start + PNn = 199	P value^a
Mean age, years (SD)	35.5 (11.3)	35.7 (11.3)	34.7 (11.1)	36.0 (11.6)	0.45
Gender					0.61
Male	559 (75.8%)	225 (76.3%)	186 (76.2%)	148 (74.4%)
Female	148 (20.1%)	61 (20.7%)	48 (19.7%)	39 (19.6%)
Transgender	31 (4.2%)	9 (3.1%)	10 (4.1%)	12 (6.0%)
Race/Ethnicity					0.1
Non-Hispanic Black	326 (44.2%)	124 (42.0%)	123 (50.4%)	79 (39.7%)
Hispanic/Other	307 (41.6%)	127 (43.1%)	86 (35.3%)	94 (47.2%)
Non-Hispanic White	105 (14.2%)	44 (14.9%)	35 (14.3%)	26 (13.1%)
Primary language					0.19
English	563 (76.3%)	227 (77.0%)	193 (79.1%)	143 (71.9%)
Non-English	175 (23.7%)	68 (23.1%)	51 (20.9%)	56 (28.1%)
HIV risk factor					0.1
Heterosexual	403 (54.6%)	167 (56.6%)	137 (56.2%)	99 (49.8%)
MSM	298 (40.4%)	118 (40.0%)	89 (36.5%)	91 (45.7%)
IDU/Other	37 (5.0%)	10 (3.4%)	18 (7.4%)	9 (4.5%)
Marital status: single/separated	633 (85.8%)	249 (84.4%)	221 (90.6%)	163 (81.9%)	0.02
Highest education level					0.98
Less than high school	155 (21.4%)	60 (20.4%)	52 (21.9%)	43 (22.4%)
Completed high school	269 (37.2%)	110 (37.4%)	87 (36.7%)	72 (37.5%)
Greater than high school	299 (41.4%)	124 (42.2%)	98 (41.4%)	77 (40.1%)
Insurance					<0.01
Charity/Self-pay	590 (80.0%)	213 (72.2%)	198 (81.2%)	179 (90.0%)
Medicare/Medicaid	88 (11.9%)	51 (17.3%)	28 (11.5%)	9 (4.5%)
Commercial	60 (8.1%)	31 (10.5%)	18 (7.4%)	11 (5.5%)
Unstable housing	66 (8.9%)	16 (5.4%)	28 (11.5%)	22 (11.1%)	0.02
Employed	291 (39.4%)	127 (43.1%)	100 (41.0%)	64 (32.2%)	0.04
Baseline CD4 < 200 cells/mm³	204 (28.0%)	83 (28.2%)	68 (28.2%)	53 (27.3%)	0.97
Baseline PHQ9 score ≥ 10	53 (7.3%)	14 (4.8%)	14 (5.9%)	25 (12.8%)	<0.01
Baseline SAMISS—MI positive	329 (44.7%)	110 (37.3%)	109 (44.7%)	110 (55.8%)	<0.01
Baseline SAMISS—SU positive	269 (36.6%)	106 (35.9%)	91 (37.3%)	72 (36.6%)	0.95

Note: Percentages are based on the denominator of persons with available data for each characteristic.

Student’s t-tests and chi-square tests were performed for continuous and categorical data, respectively.

IDU, injection drug use; MSM, men who have sex with men; PHQ9, Patient Health Questionnaire-9; PN, patient navigation; SAMISS-MI, Substance Abuse and Mental Illness Symptoms Screener–Mental Illness; SAMISS-SU, Substance Abuse and Mental Illness Symptoms Screener–Substance Use; SD, standard deviation.

The major differences between the three cohorts were in other social determinants of health. During the time that the clinic progressed from pre-rapid start to rapid start and then to rapid start + PN, the proportion of patients without health insurance (charity/self-pay) increased from 72% to 90%, and consequently the proportion with public or commercial insurance decreased significantly. The rapid start cohorts had a significantly higher proportion with unstable housing (∼11%) compared with pre-rapid start (5.4%), and significantly less employment in the rapid start + PN cohort. Regarding mental health, there were significantly higher PHQ9 scores and rates of SAMISS-MI positivity at baseline across the three cohorts over time, though the rates of SAMISS-SU positivity were stable across time.

HIV care continuum

The overall HIV care continuum is depicted in Figure 1 for newly diagnosed patients initiating care pre-rapid start, during rapid start, and rapid start + PN.

Linkage to care

Although nearly 100% of patients across all three eras were linked to care within the 12 months after CM intake, the speed with which this is occurring has accelerated greatly. While initially only about 1% of patients were seen within 7 days of HIV diagnosis, 37.7% are now able to be seen within this time frame, and close to 80% within 30 days of diagnosis (the current federal metric) in the rapid start + PN era [median 13 days, interquartile range (IQR) 5, 25]. When starting from the point of CM intake, now 95.5% of patients were seen within 7 days (Table 2).

Table 2.

Selected Linkage and ART Initiation Outcomes Before Rapid Start, During Rapid Start, and Rapid Start with Patient Navigation

Outcome	Totaln = 738	Pre-rapid start n = 295	Rapid startn = 244	Rapid start + PN n = 199	Overallp value	Pre-rapid start vs. rapid start	Pre-rapid start vs. RapidStart + PN	Rapid start vsRapidStart + PN
Seen by provider within 7d of diagnosis	123 (16.8%)	3 (1.0%)	45 (18.6%)	75 (37.7%)	<0.01	<0.01	<0.01	<0.01
Seen by provider within 30d of diagnosis	342 (46.8%)	60 (20.7%)	123 (50.8%)	159 (79.9%)	<0.01	<0.01	<0.01	<0.01
Seen by a provider within 7d of CM intake	451 (61.7%)	55 (19.0%)	206 (85.1%)	190 (95.5%)	<0.01	<0.01	<0.01	<0.01
Median (IQR) days from diagnosis to first provider visit	35 (13, 118)	71 (37, 225)	30 (11, 93)	13 (5, 25)	<0.01^#	<0.01^$	<0.01^$	<0.01^$
Median (IQR) days from CM intake to first provider visit	1 (0, 14)	14 (9, 24)	0 (0, 1)	0 (0, 0)	<0.01^#	<0.01^$	<0.01^$	<0.01^$
Median (IQR) days from CM intake to ART initiation	1 (0, 14)	14 (8, 23)	0 (0, 5)	0 (0, 0)	<0.01^#	<0.01^$	<0.01^$	<0.01^$

Note: Percentages are based on the denominator of persons with available data for each outcome ^#by Kruskal–Wallis test; ^$by Wilcoxon rank-sum test; otherwise by chi-square test.

ART, antiretroviral therapy; CM, case management; IQR, interquartile range; PN, patient navigation.

Initiated on ART

The median time from CM intake to ART initiation has decreased significantly from 14 days in the pre-rapid start era to the point where almost all patients are now started on ART on the same day as the CM intake (median 0 days, IQR 0, 0) (Table 2).

Time to virologic suppression

There was no significant difference in time to virologic suppression during the pre-rapid start (median, 2.50 months), rapid start (2.56 months), and rapid start + PN (2.79 months) periods (p = 0.41). The corresponding probabilities of achieved virologic suppression within the first 12 months were 86%, 84%, and 78%, respectively (Fig. 2). In the Fine and Gray competing-risks regression model, rapid start and rapid start + PN were not significantly associated with time to virologic suppression. However, non-English Language [adjusted subdistribution hazard ratio (aSHR) 1.30; 95% confidence interval (CI) 1.08, 1.56; p < 0.01] and Medicaid/Medicare coverage (aSHR 1.32; 95% CI 1.01, 1.71; p = 0.04) were significantly associated with increased likelihood of achieving virologic suppression, while unstable housing was associated with a lower likelihood of virologic suppression (aSHR 0.69; 95% CI 0.50, 0.96; p = 0.03) (Table 3A).

Table 3.

Multivariate Analyses for Various HIV Care Outcomes

Characteristic	aSHR	95% CI	P value
(A) Time to virologic suppression, stepwise competing-risks regression, death as competing risk, n = 735
Group
Pre-rapid start	ref
Rapid start	0.99	(0.83, 1.19)	0.93
Rapid start + PN	0.91	(0.74, 1.13)	0.39
Employment	1.20	(1.01, 1.42)	0.04
Unstable housing	0.69	(0.50, 0.96)	0.03
Non-English Language	1.30	(1.08, 1.56)	0.01
Coverage
Charity/self-pay	ref
Commercial	1.28	(0.97, 1.69)	0.08
Medicaid/Medicare	1.32	(1.01, 1.71)	0.04
Characteristic	aOR	95% CI	P value
(B) Retention in care at 12 months, stepwise logistic regression, n = 728
Group
Pre-rapid start	ref
Rapid start	1.70	(1.16, 2.48)	<0.01
Rapid start + PN	2.00	(1.32, 3.02)	<0.01
HIV Risk Factor
Heterosexual	ref
MSM	0.79	(0.56, 1.11)	0.17
IDU/other	0.34	(0.16, 0.69)	<0.01
Unstable housing	0.56	(0.33, 0.97)	0.04
(C) Continuous retention in care at 12 months, stepwise logistic regression, n = 731
Group
Pre-rapid start	ref
Rapid start	0.72	(0.49, 1.07)	0.10
Rapid start + PN	0.60	(0.40, 0.89)	0.01
Race/ethnicity
Hispanic/other	ref
Non-Hispanic Black	0.52	(0.36, 0.76)	<0.01
Non-Hispanic White	0.44	(0.27, 0.73)	<0.01
Unstable housing	0.49	(0.29, 0.84)	<0.01
(D) Sustained virologic suppression at 12 months, stepwise logistic regression, n = 729
Group
Pre-rapid start	ref
Rapid start	0.83	(0.57, 1.20)	0.32
Rapid start + PN	0.65	(0.44, 0.97)	0.04
Race/ethnicity
Hispanic/other	ref
Non-Hispanic Black	0.42	(0.30, 0.60)	<0.01
Non-Hispanic White	0.43	(0.27, 0.70)	<0.01
Coverage
Charity/self-pay	ref
Commercial	1.69	(0.92, 3.11)	0.09
Medicaid/Medicare	1.78	(1.06, 2.97)	0.03
Unstable housing	0.54	(0.31, 0.93)	0.03
Baseline SAMISS-MI positive	0.63	(0.46, 0.87)	<0.01

aOR, adjusted odds ratio; aSHR, adjusted subdistribution hazard ratio; CI, confidence interval; IDU, injection drug use; MSM, men who have sex with men; PN, patient navigation; SAMISS-MI, Substance Abuse and Mental Illness Symptoms Screener–Mental Illness.

Retained in care

The odds of being retained in care significantly increased during each time period (Table 3B). In the logistic regression model, rapid start had an adjusted odds ratio (aOR) of 1.70 (95% CI 1.16, 2.48; p < 0.01), while rapid start + PN had an even stronger association (aOR 2.00; 95% CI 1.32, 3.02; p < 0.01) with retention in care compared with pre-rapid start. IDU (aOR 0.34, 95% CI 0.16, 0.69; p < 0.01) and unstable housing (aOR 0.56, 95% CI 0.33, 0.97; p = 0.04) were significantly associated with lower odds of retention in care.

Continuous retention in care

In contrast, the odds of being continuously retained in care were decreased in the rapid start + PN period (aOR 0.60, 95% CI 0.40, 0.89; p = 0.01) compared with pre-rapid start (Table 3C). Non-Hispanic Black (aOR 0.52, 95% CI 0.36, 0.76; p < 0.01) and non-Hispanic White (aOR 0.44, 95% CI 0.36, 0.76; p < 0.01) race/ethnicity and unstable housing (aOR 0.49, 95% CI 0.29, 0.84; p < 0.01) were associated with decreased odds of being continuously retained in care.

Sustained virologic suppression

There was a trend toward decreased odds of sustained virologic suppression during the rapid start periods, with rapid start + PN having a significant association (aOR 0.65, 95% CI 0.44, 0.97; p = 0.04) (Table 3D). Non-Hispanic Black (aOR 0.42, 95% CI 0.30, 0.60; p < 0.01) and non-Hispanic White (0.43, 95% CI 0.27, 0.70; p < 0.01) race/ethnicity, unstable housing (aOR 0.54, 95% CI 0.31, 0.93; p = 0.03), and baseline SAMISS-MI positivity (aOR 0.63, 95% CI 0.46, 0.87; p < 0.01) were all associated with decreased odds of sustained virologic suppression, while Medicaid/Medicare coverage (aOR 1.78, 95% CI 1.06, 2.97; p = 0.03) was associated with increased odds of sustained virologic suppression.

Sensitivity analysis

Among those who had at least one successfully completed phone call with the patient navigator after study enrollment (n = 152), the time to virologic suppression was 2.63 months and was not statistically different from the pre-rapid start or rapid start groups (p = 0.84; data not shown). All patients were linked to care and initiated on ART, 84.9% attained virologic suppression, 86.2% were retained in care, and 63.2% had sustained virologic suppression.

Discussion

In this study, adding PN to a rapid start program in a large urban safety-net clinic in Texas was associated with mixed outcomes across the HIV care continuum. We observed significantly faster linkage to care and ART initiation, as well as improved retention in care. However, PN was not associated with improved rates or speed of virologic suppression or continuous retention in care and was associated with a decreased likelihood of sustained virologic suppression at 12 months.

Rapid start programs primarily aim to reduce time to ART initiation. While many programs report downstream improvements such as faster virologic suppression,^2–4 the immediate goal is compressing steps between diagnosis and ART initiation. Our study demonstrates marked improvement in linkage and ART initiation compared with pre-rapid start, consistent with this objective. However, as seen previously,¹⁰ downstream outcomes after the initial visit did not improve to the same extent. PN was not associated with faster virologic suppression, likely due to variability in timing of follow-up VL testing and missed blood draws. Because suppression can only be documented after blood draws, inconsistent follow-up limits observed improvements. Since completing this study, we have standardized follow-up with labs at 3 weeks and clinic visits at 4 weeks to limit variability due to provider practice.

Several trends in our clinic population may also explain these discordant findings. Over time, our rapid start cohorts included patients with higher-risk social determinants of health, including significantly higher rates of unstable housing, MI, unemployment, and lack of insurance. Rapid start prioritizes same-day or near–same-day provider visits, whereas pre-rapid start processes required multiple steps over several weeks before linkage. This earlier model likely excluded more vulnerable patients who fell off the care continuum before they were even officially linked to care, thus excluding them from the denominator and leaving a more engaged pre-rapid start comparison group. Additionally, expanded routine opt-out HIV testing at Parkland Health in emergency and outpatient settings in 2021–2022 coincided with the rapid start periods, increasing the number of patients diagnosed when not actively seeking testing. Our prior qualitative study found high levels of stigma and shock among newly diagnosed patients, which may reflect this shift in testing practices.¹¹

Our cohort also reflects unique population characteristics that may influence outcomes. Heterosexual contact was the most commonly reported HIV risk factor (55%), followed by MSM (40%), contrasting with national data in which 70% of new diagnoses in 2022 occurred among MSM and 22% among those with heterosexual contact.¹⁶ This likely reflects expanded opt-out testing, which captures individuals less likely to request testing. Nationally, individuals with heterosexually acquired HIV have worse care engagement and viral suppression outcomes.¹⁷ Additionally, 28% of our cohort presented with CD4 < 200, exceeding the national estimate of 21.6% in 2023,¹⁸ and 24% preferred a language other than English, reflecting high levels of immigration among Hispanics—a group with higher prevalence of stage 3 HIV.¹⁹

Similarly mixed effects were observed with PN addition. Although fewer patients achieved or sustained virologic suppression and continuous retention, a higher proportion remained retained when laboratory visits were included. Sensitivity analyses showed that patients who had post–first visit contact with the PN experienced improved rates of virologic suppression, retention in care, and sustained suppression, suggesting benefit among those reached. In contrast, patients without PN contact had worse outcomes, lowering overall cohort performance. These findings are consistent with prior studies showing improved retention but no effect on virologic suppression.²⁰

Several implementation lessons emerged. A key strength was having a linguistically and culturally concordant PN. Our bilingual English- and Spanish-speaking PN served as a critical link to care, particularly for Spanish-speaking patients, who initiated more frequent contact. Prior qualitative work underscores the importance of personal connection and regular check-ins.¹¹ Our PN was not living with HIV, and it is unclear whether a peer navigator might have further improved outcomes. While both peer and PN “provide linkages and patient-centered support to overcome health system barriers, improving health care engagement and clinical care outcomes,”²¹ a peer navigator typically comes from the same communities as patients and is living with HIV as well. Peer navigation has shown mixed effects on linkage, retention, and viral suppression,²² likely reflecting intervention heterogeneity. In our study, specific challenges faced by our PN included communication, boundary setting, and patient volume, which we addressed through texting protocols, crisis triage training, and workflow optimization.

This study has limitations. PN is not a standardized intervention, and effectiveness may depend on individual navigator characteristics and responsibilities, limiting generalizability. Our setting is also unique, with a predominantly heterosexual population and a high proportion of Black and Hispanic patients. As one of the largest urban HIV clinics in the United States embedded within a large county hospital, system complexity may heighten the need for navigation services. Finally, the retrospective comparison introduces potential secular trends, including clinic relocation from a small clinic building to a large multi-specialty ambulatory care building in 2021 and broader policy changes affecting insurance coverage and immigration, which may have influenced patient characteristics over time. Nevertheless, as a rapid start program in the US South, our findings are relevant to similarly affected regions.

In conclusion, adding PN to a rapid start program improved linkage to care, ART initiation, and overall retention but did not improve continuous retention or virologic suppression among newly diagnosed patients. These findings support incorporating PN into rapid start models, though patients with unstable housing, MI, or lack of insurance may require more intensive, longitudinal support. As programs adopt rapid start, PN can play a critical role but should be adapted to local patient populations to optimize both short- and long-term outcomes across the HIV care continuum.

Author Contributions

J.Y.C. and A.E.N. share responsibility for the design and conceptualization of this study. G.P. contributed to the acquisition of data and project administration. A.G. and C.A. are responsible for the statistical design, data curation, and formal analyses. All authors have contributed to the interpretation of results, have critically reviewed and approved the article, and have participated sufficiently in the work to take public responsibility for its content.

Data Availability Statement

Data from this study are not publicly available. Please contact the authors with specific inquiries.

Footnotes

Acknowledgments

The authors thank the study participants for their participation and the clinic staff for their contributions in the implementation of the rapid start program.

Author Disclosure Statement

A.E.N. receives additional research funding from Gilead Sciences.

Funding Information

This work was supported by an Investigator Sponsored Research grant from Gilead Sciences (grant number IN-US-380-5711 to J.Y.C.).

References

1.Anonymous. HIV Surveillance Supplemental Report: Estimated HIV Incidence and Prevalence in the United States, 2018–2022. 2024. Available from: https://stacks.cdc.gov [Last accessed: November 6, 2025].

2. Colasanti

, Sumitani

, Mehta

, et al. Implementation of a rapid entry program decreases time to viral suppression among vulnerable persons living with HIV in the Southern United States. Open Forum Infect Dis 2018;5(6):ofy104; doi: 10.1093/ofid/ofy104

3. Pilcher

, Ospina-Norvell

, Dasgupta

, et al. The effect of same-day observed initiation of antiretroviral therapy on HIV viral load and treatment outcomes in a U.S. public health setting. J Acquir Immune Defic Syndr 2017;74(1):44–51; doi: 10.1097/QAI.0000000000001134

4. Halperin

, Butler

, Conner

, et al. Linkage and antiretroviral therapy within 72 hours at a federally qualified health center in new orleans. AIDS Patient Care STDS 2018;32(2):39–41; doi: 10.1089/apc.2017.0309

5. Halperin

, Conner

, Butler

, et al. A care continuum of immediate ART for newly diagnosed patients and patients presenting later to care at a federally qualified health center in new orleans. Open Forum Infect Dis 2019;6(4):ofz161; doi: 10.1093/ofid/ofz161

6.Anonymous. Guidelines for Managing Advanced HIV Disease and Rapid Initiation of Antiretroviral Therapy. n.d. Available from: https://www.who.int/publications/i/item/9789241550062 [Last accessed: November 6, 2025].

7. Gandhi

, Landovitz

, Sax

, et al. Antiretroviral drugs for treatment and prevention of HIV in adults: 2024 recommendations of the international antiviral society–USA panel. JAMA 2025;333(7):609–628; doi: 10.1001/jama.2024.24543

8.Anonymous. Initiation of Antiretroviral Therapy | NIH. 2025. Available from: https://clinicalinfo.hiv.gov/en/guidelines/hiv-clinical-guidelines-adult-and-adolescent-arv/initiation-antiretroviral-therapy [Last accessed: November 25, 2025].

9. Mercado

, Quiroga

, Cantu

, et al. The AHORA Study: A Real-World Mixed-Methods Study Investigating the Effectiveness of, and Experiences with, Rapid Antiretroviral Therapy Initiation in People with Advanced HIV in the Rio Grande Valley, Texas. AIDS Patient Care STDS 2025;39(11):427–440; doi: 10.1177/10872914251379972

10.

10. Chow

, Gao

, Ahn

, et al. Rapid Start of Antiretroviral Therapy in a Large Urban Clinic in the US South: Impact on HIV Care Continuum Outcomes and Medication Adherence. J Int Assoc Provid AIDS Care 2024;23:23259582241228164; doi: 10.1177/23259582241228164

11.

11. Lee

, Nijhawan

, Tiruneh

, et al. Patient Perspectives on Rapid Antiretroviral Treatment Initiation for HIV: Implications for Best Practices. AIDS Patient Care STDS 2025;39(10):405–417; doi: 10.1177/10872914251364667

12.

12. Pence

, Gaynes

, Whetten

, et al. Validation of a brief screening instrument for substance abuse and mental illness in HIV-positive patients. J Acquir Immune Defic Syndr 2005;40(4):434–444; doi: 10.1097/01.qai.0000177512.30576.9c

13.

13. Whetten

, Reif

, Swartz

, et al. A Brief Mental Health and Substance Abuse Screener for Persons with HIV. AIDS Patient Care STDS 2005;19(2):89–99; doi: 10.1089/apc.2005.19.89

14.

14. Ahmed

, Nijhawan

, Gao

, et al. Association Between Substance Abuse and Mental Illness Symptoms Screener (SAMISS) Scores and HIV Care Continuum Outcomes in People Newly Diagnosed with HIV in the US South. AIDS Behav 2024;28(5):1731–1740; doi: 10.1007/s10461-024-04287-1

15.

15. Kroenke

, Spitzer

, Williams

JBW

. The PHQ-9. J Gen Intern Med 2001;16(9):606–613; doi: 10.1046/j.1525-1497.2001.016009606.x

16.

16.CDC. Fast Facts: HIV in the United States. 2024. Available from: https://www.cdc.gov/hiv/data-research/facts-stats/index.html [Last accessed: November 25, 2025].

17.

17.Centers for Disease Control and Prevention. HIV Surveillance Supplemental Report: Monitoring Selected National HIV Prevention and Care Objectives by Using HIV Surveillance Data United States and 6 Territories and Freely Associated States, 2022. 2024. Available from: https://stacks.cdc.gov/view/cdc/156511 [Last accessed: November 25, 2025].

18.

18.CDC. National HIV Prevention and Care Objectives: 2025 Update. 2025. Available from: https://www.cdc.gov/hiv-data/nhss/national-hiv-prevention-and-care-objectives-2025.html [Last accessed: November 25, 2025].

19.

19.Anonymous. HIV/AIDS and Hispanic/Latino Americans | Office of Minority Health. n.d. Available from: https://minorityhealth.hhs.gov/hivaids-and-hispaniclatino-americans [Last accessed: January 29, 2026].

20.

20. Cabral

, Davis-Plourde

, Sarango

, et al. Peer support and the HIV continuum of care: Results from a multi-site randomized clinical trial in three urban clinics in the united states. AIDS Behav 2018;22(8):2627–2639; doi: 10.1007/s10461-017-1999-8

21.

21. Krulic

, Brown

, Bourne

. A scoping review of peer navigation programs for people living with HIV: Form, function and effects. AIDS Behav 2022;26(12):4034–4054; doi: 10.1007/s10461-022-03729-y

22.

22. Kennedy

, Yeh

, Verster

, et al. Do peer navigators improve initiation and retention in HIV/VH/STI treatment programs for people from key populations? a systematic review of effectiveness, values and preferences, and cost. J Acquir Immune Defic Syndr 2024;95(4):305–312; doi: 10.1097/QAI.0000000000003364