Health insurance coverage, clinical outcomes, and health-related quality of life among youth born to women living with HIV

Abstract

Background:

While sustained access to health care is essential, little is known about the relationship between insurance coverage and health among people born to women living with HIV (WLHIV).

Setting:

Prospective cohort studies of youth and young adults born to WLHIV from 2007-2019.

Methods:

We used adjusted generalized estimating equation models to estimate mean differences in, and relative risks of, health-related quality of life (HR-QoL) and HIV disease measures over time by insurance status. HR-QoL scales with limited variability were dichotomized. Modified Poisson models were used to estimate relative risks (RRs).

Results:

Six hundred sixty-nine Adolescent Master Protocol (AMP) youth (66% living with perinatally-acquired HIV [PHIV], 72% Black) and 939 AMP Up/AMP Up Lite young adults (89% PHIV, 68% Black) reported insurance. Most were publicly insured (87% youth, 67% young adults). Privately insured young adults living with PHIV (YAPHIV) had lower risk of antiretroviral therapy (ART) nonadherence (adjusted RR [aRR]: 0.82, 95%CI: 0.70, 0.97) than those with public insurance. There was a lower risk of suboptimal role functioning for young adults with private insurance (aRR: 0.58, 95% CI: 0.35, 0.97) and those unaware of their coverage (aRR: 0.41, 95%CI: 0.21, 0.78). Young adults with private insurance had higher health perception scores than those with public insurance (adjusted mean difference: 3.87, 95%CI: 0.37, 7.38). For youth, we observed no differences in HR-QOL and HIV disease measures by insurance.

Conclusion:

These findings suggest meaningful differences in ART adherence and some HR-QoL outcomes by health insurance coverage among young adults born to WLHIV.

INTRODUCTION

Youth and young adults living with perinatally acquired HIV (YPHIV and YAPHIV) have greater risk of morbidity and mortality than those in the general population.¹ Advancements in HIV-related care and antiretroviral treatment have dramatically improved the survival of YPHIV into adulthood and reduced the incidence of persistent viremia, opportunistic infections, and other AIDS-related complications.² As the lifespans of YPHIV and YAPHIV have lengthened, with HIV increasingly managed as a chronic disease, experiences of metabolic complications and emotional and behavioral health challenges associated with prolonged HIV infection, antiretroviral treatment (ART), and social determinants of health (e.g., poor education access, poverty, and limited social mobility) are increasingly common.³ Ensuring viral suppression, immune health, and prevention of comorbidities requires consistent, sustained access to care. However, YAPHIV may face challenges in clinical care retention, especially in the transition from pediatric to adult care when the loss of support systems built into pediatric care can be disorienting.^4–7

In the US, health insurance coverage is a critical determinant of health and health care access.^8–11 Among people living with HIV, those with insurance are more likely to adhere to ART¹², experience sustained viral suppression^13,14, and access mental health care¹⁵ for emotional and behavioral needs—all factors associated with improved HIV-related outcomes.^16,17 Not only is having insurance important, but the type of insurance one has could determine access to comprehensive, quality care. Studies have found mixed results comparing private and public coverage. Kreider et al. observed a higher percentage of health care needs met and less out-of-pocket costs among children with public coverage compared to private¹⁸; however, Zickafoose et al. found that privately insured youth had higher prevalence of having a medical home, a source of primary care focused on delivery of preventive care and managing chronic disease.¹⁹ Differences among insurance types include cost and service provision. An overview of insurance coverage financing and services by type among people living with HIV in the US has been previously described.²⁰

Insurance coverage is intergenerational and socioeconomically patterned. For many youth with low incomes, public Children’s Health Insurance Program (CHIP) eligibility begins at birth and ends at age 19; young adults with low incomes have limited eligibility for public coverage, especially in states that have not expanded Medicaid. Private insurance may be available to youth and young adults through age 26, if covered as a dependent by guardian private plans. Due to income ineligibility for public coverage and access to employer-based plans, those with private insurance typically have higher socioeconomic status. Moreover, youth and young adults with disabilities may be eligible for public coverage through Medicare. As HIV disproportionately affects racial, ethnic and sexual minority communities and people living in poverty, barriers to insurance coverage may impact not only YPHIV and YAPHIV but also youth and young adults perinatally exposed to HIV but uninfected (YPHEU and YAPHEU), also born to women living with HIV (WLHIV).^21,22

Studies have shown increased burden of mental health challenges among both children and adolescents living with perinatally-acquired HIV and those who were perinatally exposed to HIV but uninfected compared to the general U.S. youth population, with some evidence for greater burden among YPHEU.^23,24 Studies have also highlighted that adverse health outcomes may be attributed to sociocontextual factors like education, access to care, and social support among those perinatally exposed to HIV but uninfected compared to those unexposed.^25,26 Health-related quality of life (HR-QoL), has been increasingly considered in UNAIDS efforts to end the HIV epidemic ^27,28 through the promotion of holistic well-being.²⁹ Among YPHEU/YAPHEU, optimizing HR-QoL could contribute to attenuating disparities in health driven by poor socioeconomic status. Health insurance is an important facilitator of health and health access; however, very little is known about insurance coverage and its effects on wellness among youth and young adults born to WLHIV in the U.S. While studies of insurance coverage among the general US population have reported mixed results^18,30–32,no studies have assessed the association between insurance coverage and HR-QoL among YPHIV/YAPHIV or YPHEU/YAPHEU. Our primary aim was to estimate the associations between insurance coverage and HIV-related outcomes and HR-QoL in three multi-site cohorts of YPHIV/YAPHIV and YPHEU/YAPHEU.

METHODS

Study Population

We analyzed data from the Adolescent Master Protocol (AMP) Up Series of the multi-site Pediatric HIV/AIDS Cohort Study (PHACS) network, which includes three prospective cohorts: AMP, AMP Up, and AMP Up Lite (see Table S1, Supplemental Digital Content). The PHACS AMP Up Series has followed youth and young adults living with PHIV or PHEU for over a decade. These studies assess the long-term effects of perinatally-acquired HIV and perinatal HIV exposure and the impact of HIV infection and ART on youth and young adults as they transition into adulthood.

AMP enrolled youth aged 7 to 16 years born to WLWH (n=451 YPHIV, n=227 YPHEU) from clinical sites in the US between March 2007 and October 2009. AMP sites were in the Northeast (NY, MA), Mid-Atlantic (PA, NJ, MD), Midwest (IL, CO), South (FL, LA, TN), West (CA) and Puerto Rico. Follow-up visits initially occurred every six months (2007-2010) and then every year (2010-2020), and evaluations included clinical, demographic, and developmental outcomes. YPHIV were eligible for enrollment if currently engaged in medical care with documentation of key medical data related to their HIV infection since birth. AMP also enrolled caregivers of participants and assessed their sociodemographic characteristics, along with mental and physical health. AMP follow-up ended in early 2021.

AMP Up is a cohort of YAPHIV and YAPHEU aged 18 years and older that began enrolling in 2014 at the AMP sites. Most AMP youth transitioned to AMP Up once they turned 18 (n=422); other eligible YAPHIV are also being enrolled. AMP Up participants complete an annual online survey and attend in-person visits at entry and every three years to complete physical exams and psychosocial, mental, and neurocognitive assessments.

AMP Up Lite opened in 2017 as a streamlined version of AMP Up, enrolling only YAPHIV age 18 years and older in MA, CA, IL, NY, MD, and WA. Sites were later added in PA, AL, FL, TX, and Puerto Rico. Participants complete an in-person visit at entry and annual online surveys thereafter.

For participants in all cohorts, clinical outcomes such as viral load are assessed via chart abstraction; other characteristics are assessed by survey. Across all three cohorts, our analyses included youth and young adults with information available on at least one outcome of interest.

Outcomes

The primary outcomes of interest were HR-QoL and HIV disease status [PHIV only: viremia, CD4+ T cell count (cells/mL), and ART nonadherence]. Outcomes were lagged by one visit period (1 year) to ensure temporal ordering of the exposures and outcomes. Viremia was defined as HIV-1 RNA >400 copies/mL for AMP participants and >200 copies/mL for AMP Up and AMP Up Lite participants. Differences in detection thresholds reflect technological capacity at start of follow up for each cohort (see Table S1, Supplemental Digital Content, for cohort-specific enrollment dates). For HIV-1 RNA and CD4+ T cell count measures in AMP Up/AMP Up Lite, eligible laboratory records were defined as those taken on the date closest to and within 9 months before or after a visit. Across all cohorts, we annualized the data by taking the closest HIV-1 RNA and CD4+ T cell count to each annual visit, maintaining temporality across cohorts. Among YPHIV and YAPHIV, ART nonadherence was defined as self-report of missing one or more doses to any component of an ART regimen in the week prior to a visit.

Among AMP participants, HR-QoL was self-reported using the General Health Assessment for Children (GHAC) with the following scales: general health ratings, physical functioning, psychological well-being, social and role functioning (ability to participate in usual social and work/school-related activities), health care utilization, and symptoms.³³ For AMP Up/AMP Up Lite participants, HR-QoL was assessed via online survey using the 20-item Short Form Health Survey (SF-20) and included scales for physical functioning, social functioning, role functioning, mental health, health perception, and pain (see Table S2, Supplemental Digital Content for scale items).³⁴ The GHAC and the SF-20 both demonstrate good validity and reliability for measuring multiple domains of health and psychosocial wellbeing.^33,34 Responses to items in individual domains were reverse coded if necessary and standardized to range from 0 to 100, with higher scores denoting better HR-QoL for both measures. If most questions were completed within a domain but some were missing (<5%), missing items were assigned the mean value of completed items for that participant, as described in prior work.³⁵ Some scales were dichotomized at the 25th percentile due to limited variability and ceiling effects (score distributions skewed to higher values), resulting in binary outcomes denoting suboptimal HR-QoL (i.e., suboptimal physical functioning).

Exposure

The primary exposure of interest was self-reported health insurance coverage. At each study visit, AMP caregivers responded to the question, “Which of the following are you using to pay all or part of your child’s medical bills, including medications?” with the following options: “Private insurance,” “Medicaid,” “Medicaid managed care plan,” “AIDS drug assistance program (ADAP),” “Medicare,” “Social Security,” “Self-pay,” or “Other.” AMP Up/AMP Up Lite participants were asked a similar question, with the additional response option, “Don’t know.” For all participants, insurance coverage was categorized as “public” (Medicaid, Medicare, Social Security, or other public program), “private” (private insurance coverage), or “uninsured” (self-pay). Among AMP Up/AMP Up Lite participants, the additional category of “unaware” was used to classify those answering “Don’t know” to any insurance coverage option, in the absence of a “Yes” response to any coverage option. Ryan White program assistance, including AIDS Drug Assistance Program (ADAP) support, was assessed independently of insurance coverage for YPHIV/YAPHIV. Notably, Ryan White funds are allocated to institutions and individuals by state and may contribute to purchasing insurance plans. ADAP is often utilized as a supplement for those who are uninsured or underinsured. Individual eligibility for these programs varies based on state-defined criteria.

Covariates

Time-fixed covariates included race (White, Black, or other), ethnicity (Hispanic or non-Hispanic), biological sex at birth, region of residence, and HIV status. Region of residence was based on US Census-defined regions and specified in the analysis as the South, Puerto Rico, and other (Northeast, Midwest, and West). Time-varying covariates included age, income (AMP household income: <$20,000, $20,001 to $40,000, $40,001 to $70,000, >$70,000; AMP Up/AMP Up Lite personal income: <$20,000, $20,001 to $40,000, >$40,000, don’t know), education (<high school, high school diploma or equivalent, and associate degree or higher), marital status, and employment. Analyses of YAPHIV enrolled in AMP Up/AMP Up Lite also included covariates for transition to adult care and ability to self-manage healthcare. Additionally, for both YPHIV and YAPHIV, HIV disease severity was assessed using US Centers for Disease Control and Prevention (CDC) class indicators and classified as Class C (prior AIDS-defining diagnosis) versus not Class C. Notably for AMP participants, sociodemographic characteristics like education and income reflect that of the caregiver.

Statistical Analysis

We described the distribution of demographic and other health-related characteristics of participants by insurance coverage at baseline using percentages, medians, and interquartile ranges.

For continuous outcomes (CD4+ T cell count and HR-QoL scales), we used repeated measures generalized estimating equation (GEE) models for linear regression, with an exchangeable correlation structure to account for repeated measures and robust variance. We used modified Poisson GEEs to estimate associations between insurance coverage and each binary outcome.³⁶ Models for continuous outcomes were linear GEE models. Adjusted models included the covariates listed above. To address missingness, we conducted all analyses using multiple imputation including all covariates from adjusted models based on a fully conditional specification (FCS) method for arbitrary missingness patterns.³⁷ We pooled estimates and chi-squared statistics across 40 imputations.³⁸

For all participants, follow up began at enrollment and continued to death, loss to follow up, or administrative censoring (AMP: 18 years of age; AMP Up/AMP Up Lite: July 1, 2021). We assessed modification of the association between insurance coverage and HR-QoL by HIV status by interaction terms. We also conducted sensitivity analyses to assess the robustness of results, including lab eligibility window (6 months and 3 months), viral load sensitivity (≤50 copies/mL for YAPHIV), and attention checks to further assess the authenticity of self-report. All analyses were conducted using SAS 9.4 (SAS Institute, Cary, NC).

RESULTS

Study Population

AMP

We included 674 (YPHIV: n=448, YPHEU: n=226) AMP youth aged 7-16 at enrollment. The median number of visits per participant was 5 (range: 1 – 11), and median length of follow up was 4.2 years. Of 669 youth who reported insurance coverage status at baseline, most youth had public insurance (87%) or private insurance coverage (10%), while only 19 (2.8%) had no insurance coverage. The distribution of insurance coverage remained similar across follow up; 70% of participants had the same coverage throughout follow-up. Compared to youth with public coverage, youth with private coverage more often reported White race and had caregivers with an associate degree or higher, higher income, employment, and a spouse at study entry (Table 1). A greater proportion of uninsured youth reported Black race, lived in the South, and had a caregiver born outside the US, compared to youth with public insurance. We observed only minor differences by age category and HIV status.

Table 1 -.

Baseline characteristics of 669 youth enrolled in the PHACS AMP study by type of health insurance coverage

		Health insurance coverage
	Total (N=669)	Public (N=582)	Private (N=68)	Uninsured (N=19)
Participant characteristics	N(%)a	N(%)	N(%)	N(%)
Age at entry
9 and under	182 (27.2%)	159 (27.3%)	19 (27.9%)	4 (21.1%)
10-11	151 (22.6%)	129 (22.2%)	15 (22.1%)	7 (36.8%)
12-13	147 (22.0%)	128 (22.0%)	14 (20.6%)	5 (26.3%)
14 and over	189 (28.3%)	166 (28.5%)	20 (29.4%)	3 (15.8%)
Female sex at birth	344 (51.4%)	299 (51.4%)	33 (48.5%)	12 (63.2%)
Race
Black	459 (71.6%)	409 (72.9%)	35 (54.7%)	15 (93.8%)
White	173 (27.0%)	146 (26.0%)	26 (40.6%)	1 (6.3%)
Other	9 (1.4%)	6 (1.1%)	3 (4.7%)	0 (0.0%)
Hispanic ethnicity	188 (28.3%)	168 (29.0%)	16 (23.9%)	4 (21.1%)
Living with HIV	443 (66.2%)	382 (65.6%)	48 (70.6%)	13 (68.4%)
Caregiver: Highest level of education
High school diploma	241 (36.0%)	212 (36.4%)	25 (36.8%)	4 (21.1%)
Less than HS diploma	239 (35.7%)	224 (38.5%)	7 (10.3%)	8 (42.1%)
Associates degree or higher	189 (28.3%)	146 (25.1%)	36 (52.9%)	7 (36.8%)
Caregiver: Household income
Less than $20K	333 (52.4%)	320 (57.7%)	6 (9.5%)	7 (41.2%)
$20,001 to $40K	172 (27.1%)	153 (27.6%)	12 (19.0%)	7 (41.2%)
$40,001 to $70K	84 (13.2%)	56 (10.1%)	26 (41.3%)	2 (11.8%)
Greater than $70K	46 (7.2%)	26 (4.7%)	19 (30.2%)	1 (5.9%)
Caregiver: Married	357 (53.4%)	291 (50.1%)	53 (77.9%)	13 (68.4%)
Caregiver: Employed	255 (38.2%)	198 (34.1%)	48 (70.6%)	9 (47.4%)
Region of residence
Non-South	420 (62.8%)	362 (62.2%)	49 (72.1%)	9 (47.4%)
South	198 (29.6%)	174 (29.9%)	14 (20.6%)	10 (52.6%)
Puerto Rico	51 (7.6%)	46 (7.9%)	5 (7.4%)	0 (0.0%)
Caregiver: Birthplace
US	543 (81.3%)	481 (82.8%)	51 (75.0%)	11 (57.9%)
Other country	125 (18.7%)	100 (17.2%)	17 (25.0%)	8 (42.1%)
Caregiver living with HIV	326 (53.4%)	281 (53.1%)	32 (50.8%)	13 (72.2%)
Health-related quality of life scaleb indicators
Suboptimal physical functioning	109 (19.2%)	97 (19.9%)	10 (16.4%)	2 (11.1%)
Suboptimal social and role functioning	84 (14.8%)	71 (14.5%)	11 (18.0%)	2 (11.1%)
Higher health care utilization	61 (10.8%)	55 (11.3%)	5 (8.2%)	1 (5.6%)
Health-related quality of life scales [median (IQR)]
General health ratings	86.1 (72.2, 94.4)	86.1 (72.2, 94.4)	80.6 (72.2, 88.9)	86.1 (77.8, 97.2)
Psychological well-being	83.9 (71.4, 92.9)	83.9 (71.4, 92.9)	85.7 (73.2, 92.9)	89.3 (82.1, 96.4)
Symptoms	97.0 (92.0, 100)	97.0 (92.0, 100)	96.0 (89.5, 98.0)	97.5 (89.0, 100.0)

^aMissing values included in percent total

^bHealth-related quality of life was first assessed at week 48

AMP Up/AMP Up Lite

We included 981 young adults (PHIV: n=870, PHEU: n=111) aged 18-40 years at enrollment. The median number of visits per participant was 3 (range: 1 – 4), and median follow up was 2 years. At baseline, 939 young adults reported insurance coverage status. Person-visits with unidentifiable insurance coverage were excluded (n=24). Most participants reported public insurance (67%), while fewer reported private insurance (15%), being uninsured (7.2%), or being unaware of their coverage (11%). Twenty-two percent of participants reported receiving ADAP or other Ryan White program assistance, 59% reported no support, and 18% reported being unaware of any support received. The distribution of insurance coverage remained similar across follow-up; 64% of AMP Up/AMP Up Lite participants had the same coverage throughout follow-up. Compared to young adults with public insurance, those with private coverage more often had at least an associate degree, higher income, and were employed at entry (Table 2). Among young adults unaware of their insurance coverage, a greater proportion were <21 years and living with a parent than those reporting public insurance. YAPHIV unaware of their insurance coverage more often reported being in pediatric care instead of adult care and lower ability to self-manage healthcare than YAPHIV with public insurance coverage (Table 3). A greater proportion of YAPHEU reported being unaware of their insurance coverage (Table S3, Supplementary Digital Content).

Table 2 -.

Baseline characteristics of 939 young adults enrolled in PHACS AMP Up/AMP Up Lite by type of health insurance coverage

		Health insurance coverage
	Total (N=939)	Public (N=629)	Private (N=137)	Unaware (N=105)	Uninsured (N=68)
Participant characteristics	N(%) a	N(%)	N(%)	N(%)	N(%)
Age group
21 and under	430 (45.8%)	258 (41.0%)	56 (40.9%)	83 (79.0%)	33 (48.5%)
22 to 26	215 (22.9%)	156 (24.8%)	27 (19.7%)	15 (14.3%)	17 (25.0%)
27+	294 (31.3%)	215 (34.2%)	54 (39.4%)	7 (6.7%)	18 (26.5%)
Female sex at birth	578 (61.6%)	404 (64.2%)	80 (58.4%)	55 (52.4%)	39 (57.4%)
Race
Black	610 (65.0%)	418 (66.5%)	82 (59.9%)	67 (63.8%)	43 (63.2%)
White	246 (26.2%)	161 (25.6%)	39 (28.5%)	29 (27.6%)	17 (25.0%)
Other	36 (3.8%)	19 (3.0%)	8 (5.8%)	7 (6.7%)	2 (2.9%)
Hispanic ethnicity	259 (27.6%)	182 (28.9%)	30 (21.9%)	28 (26.7%)	19 (27.9%)
Living with HIV	832 (88.6%)	566 (90.0%)	127 (92.7%)	81 (77.1%)	58 (85.3%)
Highest level of education
High school diploma	400 (42.6%)	273 (43.4%)	49 (35.8%)	51 (48.6%)	27 (39.7%)
Less than HS diploma	252 (26.8%)	175 (27.8%)	27 (19.7%)	36 (34.3%)	14 (20.6%)
Associate degree or higher	278 (29.6%)	174 (27.7%)	60 (43.8%)	18 (17.1%)	26 (38.2%)
Personal income
Less than $20K	516 (55.0%)	367 (58.3%)	60 (43.8%)	50 (47.6%)	39 (57.4%)
$20,001 to $40K	70 (7.5%)	40 (6.4%)	25 (18.2%)	1 (1.0%)	4 (5.9%)
Greater than $40K	34 (3.6%)	9 (1.4%)	17 (12.4%)	2 (1.9%)	6 (8.8%)
Don’t know	272 (29.0%)	180 (28.6%)	28 (20.4%)	50 (47.6%)	14 (20.6%)
Married	78 (8.3%)	51 (8.1%)	16 (11.7%)	2 (1.9%)	9 (13.2%)
Employed	447 (47.6%)	272 (43.2%)	91 (66.4%)	47 (44.8%)	37 (54.4%)
Living with a parent	518 (55.2%)	318 (50.6%)	77 (56.2%)	88 (83.8%)	35 (51.5%)
Region of residence
Non-South	627 (66.8%)	441 (70.1%)	86 (62.8%)	67 (63.8%)	33 (48.5%)
South	256 (27.3%)	150 (23.8%)	47 (34.3%)	30 (28.6%)	29 (42.6%)
Puerto Rico	56 (6.0%)	38 (6.0%)	4 (2.9%)	8 (7.6%)	6 (8.8%)
Health-related quality of life scale indicators
Suboptimal physical functioning	235 (25.0%)	168 (26.7%)	33 (24.1%)	19 (18.1%)	15 (22.1%)
Suboptimal social functioning	180 (19.2%)	130 (20.7%)	26 (19.0%)	13 (12.4%)	11 (16.2%)
Suboptimal role functioning	163 (17.4%)	126 (20.0%)	19 (13.9%)	6 (5.7%)	12 (17.6%)
Health-related quality of life scales [median (IQR)]
Mental health	72.0 (56.0, 88.0)	72.0 (56.0, 88.0)	76.0 (56.0, 88.0)	68.0 (60.0, 84.0)	72.0 (52.0, 88.0)
Health perception	75.0 (55.0, 90.0)	75.0 (55.0, 90.0)	80.0 (65.0, 95.0)	70.0 (55.0, 85.0)	75.0 (60.0, 90.0)
Pain	80.0 (60.0, 100)	80.0 (60.0, 100)	80.0 (60.0, 100)	80.0 (60.0, 100)	80.0 (50.0, 100)

^aMissing values included in percent total

Table 3 -.

Baseline HIV-related outcomes of youth and young adults with HIV enrolled in the AMP Up series by health insurance coverage

AMP
				Health insurance coverage
			Total (N=443)	Public (N=382)	Private (N=48)	Uninsured (N=13)
Participant characteristics			N(%) a	N(%)	N(%)	N(%)
Ryan White or ADAP			8 (1.8%)	2 (0.5%)	2 (4.2%)	4 (30.8%)
CDC Class C			108 (24.4%)	95 (24.9%)	9 (18.8%)	4 (30.8%)
Viremia (>400 copies/mL)			139 (31.4%)	128 (33.5%)	8 (16.7%)	3 (23.1%)
ART nonadherence
No missed doses in past week			235 (53.0%)	196 (51.3%)	30 (62.5%)	9 (69.2%)
One or more missed doses in past week			112 (25.3%)	100 (26.2%)	10 (20.8%)	2 (15.4%)
Not taking ART			18 (4.1%)	16 (4.2%)	2 (4.2%)	0 (0.0%)
CD4+ T cell countb	N Mean (SD)		397 758 (360)	339 746 (360)	46 840 (379)	12 781 (257)
AMP Up/AMP Up Lite
			Health insurance coverage
		Total (N=832)	Public (N=566)	Private (N=127)	Unaware (N=81)	Uninsured (N=58)
Participant characteristics		N(%)	N(%)	N(%)	N(%)	N(%)
Ryan White or ADAP
No		488 (58.7%)	370 (65.4%)	78 (61.4%)	19 (23.5%)	21 (36.2%)
Yes		186 (22.4%)	114 (20.1%)	33 (26.0%)	6 (7.4%)	33 (56.9%)
Unaware		153 (18.4%)	80 (14.1%)	15 (11.8%)	55 (67.9%)	3 (5.2%)
Limited health care management		339 (40.7%)	234 (41.3%)	45 (35.4%)	42 (51.9%)	18 (31.0%)
Transitioned to adult care		329 (39.5%)	231 (40.8%)	57 (44.9%)	17 (21.0%)	24 (41.4%)
CDC Class C		239 (28.7%)	172 (30.4%)	42 (33.1%)	13 (16.0%)	12 (20.7%)
Viremia (> 200 copies/mL)		263 (31.6%)	186 (32.9%)	31 (24.4%)	24 (29.6%)	22 (37.9%)
ART nonadherence
No missed doses in past week		231 (27.8%)	149 (26.3%)	41 (32.3%)	27 (33.3%)	14 (24.1%)
One or more missed doses in past week		461 (55.4%)	326 (57.6%)	66 (52.0%)	37 (45.7%)	32 (55.2%)
Not taking ART		48 (5.8%)	32 (5.7%)	5 (3.9%)	4 (4.9%)	7 (12.1%)
CD4+ T cell count	N Mean (SD)	783 608 (543)	535 595 (587)	117 654 (507)	79 640 (365)	52 593 (332)

^aMissing values included in percent total

^bcells/mL

HIV-related Health Outcomes

AMP

Among 443 AMP YPHIV, 31% had viremia at baseline (Table 3). The mean CD4+ T cell count was 757.7 (SD: 360.3) cells/mL. About one-fourth (25%) of participants had missed ≥1 doses in the past week (Table 3). While youth with private insurance and youth with no insurance had a lower risk of viremia compared to those with public coverage, the confidence intervals for these estimates were wide (Table 4).

Table 4.

Measures of association (MoA) and 95% confidence intervals for fully adjusted analyses of insurance coverage and HIV-related outcomes

AMP	Health Insurance Coverage
Outcomes	Private vs Public			Uninsured vs Public
	aRR	95% CI	p	aRR	95% CI	p				Overall pa
Viremia (>400 copies/mL)	0.72	(0.41, 1.26)	0.24	0.88	(0.50, 1.57)	0.67				0.43
ART nonadherence	0.97	(0.88, 1.06)	0.49	0.96	(0.81, 1.14)	0.65				0.69
	aMD	95% CI	p	aMD	95% CI	p
CD4+ T cell countb	−14.1	(−97.8, 69.5)	0.74	−49.3	(−154.2, 55.7)	0.36				0.64
AMP Up/AMP Up Lite	Health Insurance Coverage
Outcomes	Private vs Public			Uninsured vs Public			Unaware vs Public
	aRR	95% CI	p	aRR	95% CI	p	aRR	95% CI	p	Overall p
Viremia (> 200 copies/mL)	0.71	(0.49,1.03)	0.069	1.26	(0.96,1.65)	0.091	0.83	(0.59,1.18)	0.30	0.054
ART nonadherence	0.82	(0.70,0.97)	0.019	0.95	(0.80,1.13)	0.55	0.92	(0.76,1.10)	0.36	0.11
	aMD	95% CI	p	aMD	95% CI	p	aMD	95% CI	p
CD4+ T cell count	50.85	(−10.32,112.02)	0.10	−7.08	(−70.56,56.39)	0.83	30.62	(−25.07,86.32)	0.28	0.32

^aOverall p-value based on pooled Chi-square statistics

^bcells/mL

aRR: adjusted relative risk, aMD: adjusted mean difference, CI: confidence interval

Upper confidence limit and lower confidence limit based on 95% confidence intervals

All models adjusted for sex, age, race, ethnicity, income*, employment*, marital status*, education*, and region of residence, and CDC Class C Models for AMP/AMP Up Lite additionally adjusted for ability to manage healthcare and transition from pediatric to adult care

^*Caregiver factors for AMP models

AMP Up/AMP Up Lite

At baseline, 32% of YAPHIV had viremia (Table 3). The mean CD4+ T cell count was 607.5 (SD: 543.2) cells/mL, and 55% of participants reported missing ≥1 doses in the past week. The risk of viremia varied by insurance coverage in both unadjusted (Table S5, Supplemental Digital Content) and adjusted models (Table 4). In unadjusted models, young adults with private insurance were at lower risk of viremia [Relative risk (RR): 0.62, 95% CI: 0.44, 0.87] compared to those with public insurance. In adjusted models, this association was attenuated (adjusted RR [aRR]: 0.71, 95% CI: 0.49, 1.03). In adjusted models, uninsured YAPHIV had higher risk of viremia than those with public insurance (aRR: 1.26, 95% CI: 0.96, 1.65). YAPHIV with private coverage had lower risk of ART nonadherence compared to those publicly insured (aRR: 0.82, 95% CI: 0.70, 0.97).

Health Related Quality of Life (HR-QoL) Outcomes

AMP

The range of reported scores was limited for physical functioning, social and role functioning, and health care utilization (Table 1). Insurance coverage was not associated with any HR-QoL metrics in unadjusted or adjusted models (Table S4, Supplemental Digital Content). In adjusted pairwise comparisons, privately insured youth had greater risk of suboptimal physical functioning than publicly insured youth (aRR: 1.42, 95% CI: 0.97, 2.07).

AMP Up/AMP Up Lite

Variability was limited for physical functioning, role functioning, and social functioning (Table 2). In adjusted models, privately insured young adults (aRR: 0.58, 95% CI: 0.35, 0.97) and those unaware of their coverage (aRR: 0.41, 95% CI: 0.21, 0.78) had lower risk of suboptimal role functioning than publicly insured young adults. Compared to young adults with public coverage, we also observed lower risk of suboptimal physical functioning among uninsured young adults and lower risk of suboptimal social functioning for other insurance coverage types (Figure 1). Privately insured young adults also had higher adjusted mean scores for health perception than those with public insurance coverage (adjusted mean difference: 3.87, 95% CI: 0.37, 7.38). We also observed a higher mean score for pain among young adults who were unaware of their coverage compared to those who were publicly insured. Notably, confidence intervals for several of these estimates were wide (see Figure 1).

Figure.

Associations between health insurance coverage and health-related quality of life scales for AMP Up/AMP Up Lite

The reference group for all comparisons is public insurance coverage

aRR: adjusted relative risk, aMD: adjusted mean difference, LCL: lower confidence limit, UCL: upper confidence limit

Upper confidence limit and lower confidence limit based on 95% confidence intervals

All models adjusted for sex, age, race, ethnicity, income, employment, marital status, education, region of residence, and HIV status

In analyses assessing modification by HIV status, there were only minor differences in adjusted measures of association (Tables S6, S7; Supplemental Digital Content) including lower risk of suboptimal physical functioning among YAPHEU with private compared to public coverage; however, these analyses are likely underpowered to detect meaningful differences between these groups.

Sensitivity analyses

Restricting the eligible laboratory result window to 6 months and 3 months resulted in findings that were qualitatively similar to the main results, but slightly less precise (Tables S8, S9; Supplemental Digital Content). Reducing the viremia threshold for YAPHIV to >50 copies/ml strengthened findings of lower risk of viremia with private insurance compared with public insurance (aRR: 0.73, 95% CI: 0.53, 1.00) (Table S10, Supplemental Digital Content). Lastly, 378 person-visits with incorrect attention check answers were excluded from the analysis. The results of these sensitivity analyses were qualitatively similar to the main results (Table S11, Supplemental Digital Content).

DISCUSSION

In this longitudinal study of youth and young adults with PHIV, we found that few youth <18 years of age were uninsured. However, the proportion with private insurance was 10%, lower than the 54.9% of US youth with private insurance in the general population.³⁹ This difference highlights that HIV disproportionately impacts communities of socio-economic disadvantage. Among those uninsured, the higher proportion of participants reporting Black race, living in the South, and having caregivers born outside the US underscores how policies and structural barriers experienced based on marginalized identities (race, ethnicity, immigration status, linguistic barriers, etc.) may impact youth insurance coverage.

A higher proportion of young adults than youth were uninsured and had private coverage. Lack of insurance coverage is a well-documented barrier to health care transition for young adults living with HIV.^40–44 Over 15% of young adults in the US who were 19-34 years of age have no health insurance, exceeding any other age group in 2019.⁴⁵ The proportion uninsured was lower at 7.2% for young adults in our cohorts, but 11% reported being unaware of their insurance coverage. Most of these young adults were aged 18 to 21 years, lived with a parent, and remained in pediatric care. This lack of awareness may be attributable to having a primary caregiver coordinating care or uncertainty about coverage options after becoming ineligible for public programs like CHIP.

While insurance was not associated with HIV-related outcomes among youth, we observed that young adults with private insurance had better adherence than those with public insurance. We also observed higher risk of viremia among uninsured YPHIV, although the confidence interval was wide. While studies of insurance among young adults with HIV are rare⁴⁶, the adult literature suggests that having private insurance is associated with better retention in care⁴⁷, viral suppression^13,48, and reduced mortality⁴⁹, compared to public or no insurance coverage. Early studies of the association between insurance coverage and ART uptake and adherence have yielded mixed results.^50–52 Compared to older adults, and to children, young adults with HIV have worse outcomes along the HIV care continuum. These young adults may also face unique barriers to care and treatment in the transition from pediatric to adult care.^4,53 Therefore, more research is needed to explore age-specific impacts of insurance coverage.

We observed similar HR-QoL scores and risk of suboptimal HR-QoL in youth born to WLHIV across types of insurance coverage, perhaps due to the abundance of care resources at AMP sites attended by most youth participants. However, privately insured young adults and those unaware of their coverage had lower risk of suboptimal role functioning compared to those with public insurance. Young adults with private insurance also had higher mean scores for health perception. People with HIV, including youth, report lower HR-QoL compared to individuals without HIV.^54–56 Among YPHIV, studies have highlighted associations between HR-QoL and do-not-resuscitate orders and hospice enrollment⁵⁷, while suggesting no association with HIV status disclosure.³⁵ Lower risk of suboptimal role functioning for the privately insured could be attributed to the burden of health care coordination. For example, some clinics may not accept Medicaid coverage⁵⁸, resulting in more time spent coordinating care. Also, eligibility changes and renewal processes may further burden young adults in transition. For young adults unaware of their insurance coverage, this lesser risk of suboptimal role functioning could be attributed to caregivers or other support systems managing their care. Although insurance-associated burden of care and coverage interruptions may immediately impact role functioning and ART adherence, HIV-specific metrics like HIV-1 RNA and CD4+ T cell count may be affected over time (beyond one year of follow-up). We also observed better health perception scores among privately insured young adults, which could be attributed to quality of care and extensiveness of supports available. Young adults born to WLHIV with private insurance may have enhanced access to care resources and fewer barriers to care that are associated with social disadvantage, both of which could translate to better perceived health. Moreover, perceptions of relative privilege with may influence health outlook and perception^59,60, highlighting the potential for unmeasured confounding by sociocontextual factors such as wealth and social mobility.

Several limitations apply to this study. Yearly, self-reported assessments for insurance coverage introduce the possibility of misclassification and do not capture more frequent changes in coverage. For most participants, however, insurance coverage remained stable throughout follow-up. Second, given sparse data at higher incomes within certain insurance types, we relied on interpolation to address random positivity violations. Moreover, given the small number of patients with private insurance coverage in AMP, we take caution in interpreting estimates. Third, support from programs like Ryan White may be underreported, especially when given directly to PHACS clinics. Fourth, generalizability of these results is limited given that AMP Up Series participants are engaged in care at enrollment and are encouraged to continue engagement. Lastly, while we were able to adjust for income and some broad factors, other unmeasured common causes of insurance coverage and HR-QoL, such as caregiver health-seeking support and engagement in community-based health initiatives may limit the interpretation of our findings.

Despite these limitations, this study has several strengths. The AMP Up Series of studies have followed YPHIV/YAPHIV and YPHEU/YAPHEU for over 15 years. Study sites are strategically located across all regions of the US and Puerto Rico, yielding a diverse sample of youth and young adults. Moreover, the longitudinal data collection from both participants and their caregivers supports robust adjustment for a wide range of characteristics associated with health insurance status and outcomes of interest. Another key strength of this study is the focus on both clinical and HR-QoL outcomes, emphasizing the importance of holistic health assessments.

In summary, we provide the first description of insurance coverage among a cohort of youth and young adults born to WLHIV. Collectively, these findings provide more insight into how YPHIV/YAPHIV and YPHEU/YAPHEU in PHACS pay for healthcare costs and suggest that insurance coverage may have implications for health and well-being. Our findings also underscore the need for opportunities to supplement care provision during the transition period with tailored education and training on understanding and managing health insurance.