Enhanced care coordination in New York City that leveraged surveillance data with a health plan’s Medicaid managed care roster improved its HIV viral load suppression rate.
ABSTRACTObjectives: Optimizing HIV treatment benefits the health of the individual and the community at large. Health department HIV surveillance data matched with Medicaid managed care rosters can be used to target people with HIV infection who have an unsuppressed viral load or are unengaged in care. MetroPlus Health Plan, a Medicaid managed care organization, implemented a 2-pronged approach: street outreach and peer care connection interventions.
Study Design: A cohort study that included demographics, program contact type and frequency, antiretroviral therapy refill pattern, and CD4 count and HIV viral load values/ranges and dates.
Methods: Members without a viral load test result during the prior 9 months (not engaged) received outreach, and those with unsuppressed viral loads received intensified care coordination and peer support. A retrospective statistical analysis was conducted on cohort members with sufficient viral load data. A subanalysis excluded members who had suppressed viral loads at baseline.
Results: A total of 1429 (82%) members in the state cross-referenced list were still enrolled in the plan at study initiation. Successful contact with targeted members by outreach was 60% compared with 40% by care coordination and peer support combined. Members who were successfully contacted by the program had a 44% suppression rate (<200 copies/mL) and a greater likelihood of achieving viral load suppression (odds ratio, 1.55; 95% CI, 1.23-1.95; P <.01) than those who were not.
Conclusions: Surveillance data were successfully used to target HIV-positive Medicaid members who had an unsuppressed viral load or were unengaged in care. Individuals with an unsuppressed viral load can achieve suppression through intensified outreach, care coordination, and peer support by a Medicaid managed care plan.
Am J Manag Care. 2019;25(6):e167-e172Takeaway Points
Untreated HIV infection results in loss of immune function, which ultimately leads to opportunistic infection or neoplasms and death.1 Antiretroviral therapy (ART) is highly efficacious in both research studies and real-world populations in restoring or preserving immune function, extending life span, and improving the quality of life for HIV-positive individuals.2,3 For individuals to fully benefit from ART, they need to know they are infected, engage in regular HIV care, and receive and adhere to ART. These are the elements of the HIV care continuum.4-6
At the end of 2015, the latest year for which data are reported, approximately 1.1 million individuals were living with HIV in the United States.7 New York City has been an epicenter of the AIDS epidemic in the United States.8-10 In 2016, approximately 110,000 individuals were living with HIV in New York City and almost 50% of those with a diagnosis were not engaged in HIV care programs.11
Upon initiation of ART, plasma virus concentration (viral load) declines rapidly to undetectable levels with high medication adherence.12 With poor or no medication adherence, viral load becomes detectable again, and over time, HIV infection progresses to AIDS with poor health outcomes such as AIDS-related morbidity and hospitalizations.13,14 Suboptimal medication adherence can lead to loss of immunologic benefit and viral resistance, limiting future treatment options.15 Moreover, despite good adherence to medication, some patients with HIV will have detectable viral loads even when being treated with ART.
Low medication adherence is detrimental not only to the individual but also to the community, as the increase in viral load poses an increase in the risk of transmission,16 which is associated with the level of viremia.17 Thus, the goal of HIV care is to achieve and maintain viral load suppression through a high level of medication adherence. However, a portion of the treated population continues to have difficulty achieving or maintaining viral load suppression. Social determinants of health play an important role in viral suppression.18,19 Factors that affect medication adherence include depression, adverse effect severity, self-efficacy, and social support.20 Low levels of engagement in care, especially in the early stages of treatment (eg, missed visits), are correlated with poor medication adherence.13
New York State (NYS) created a task force to develop specific recommendations to improve the diagnosis, treatment, and prevention of HIV in its citizens. One recommendation in the resulting blueprint to end the epidemic was to “use client-level data to identify and assist patients either lost to care or not virally suppressed.”21 This was an example of the CDC’s Data to Care initiative, a novel public health strategy that aimed to use HIV surveillance data to identify HIV-diagnosed individuals not in care, link them to care, and support the HIV care continuum.22
The NYS Department of Health (NYSDOH) maintains 2 large databases that are operationally separate: the HIV Surveillance Registry, which contains individual identifiers, viral load, and other HIV-related laboratory results; and an active roster of all Medicaid managed care recipients, which contains individual identifiers, contact information, and Medicaid plan assignments. The surveillance registry had been previously used only for epidemiologic monitoring on a population health and not an individual health basis. In April 2014, the NYS Public Health Law was amended to allow for the information within the registry, which was created with strict confidentiality protections, to be cross-referenced with its Medicaid roster in an individually identified manner. This identified some people who were not engaged in HIV care or who were known to have an unsuppressed viral load at last observation. The comparison showed that a few Medicaid managed care plans insured a large number of the HIV-positive individuals in New York City. In August 2015, the NYSDOH AIDS Institute shared the resultant comparison with 5 plans and funded a pilot program to allow the plans to target the identified population with specific enhanced care coordination, which began in January 2016.
The goal of this cohort study was to assess the effectiveness of the first 2 years of a Medicaid managed care plan’s program. Using surveillance program viral load data, care coordinators and peer counselors reached out to viremic members to address barriers to medication adherence and to engagement in care.
MetroPlus, a participant plan in the pilot, established 2 interventions: street outreach, designed to target members who were not engaged in care; and peer care connection, designed to target members who were engaged in care but had an unsuppressed viral load.
Inclusion criteria for the street outreach intervention were (1) actively enrolled, (2) no viral load test or primary care provider visit in the prior 9 months, and/or (3) no ART refill in the prior 6 months. Exclusion criteria were (1) the discovery of a negative HIV antibody test or (2) disenrollment from the plan after only 1 month of enrollment. MetroPlus partnered with the Alliance for Positive Change, an AIDS service organization, to conduct street-based outreach using trained peers to seek out these lost-to-care members either by telephone or through face-to-face interaction. When contact was made, the peers discussed returning to care with the member and, with member consent, helped to make an appointment and escort the member to an HIV-related primary care appointment. Some of these visits occurred on the same day the contact was made.
Inclusion criteria for the peer care connection intervention were (1) included in the target population, (2) actively enrolled and engaged in care, and (3) with an unsuppressed viral load, defined as 200 copies/mL or greater, at last available result. Exclusion criteria were (1) the discovery of a negative HIV antibody test after program initiation or (2) disenrollment from the plan after only 1 month of enrollment. Once a member in the street outreach intervention group became engaged in care, they were also included in the peer care connection intervention. Care coordinators, working together with trained peer educators and peer counselors, sought to contact these members through telephone and/or face-to-face interactions at the clinics or hospitals that the members attended. Comprehensive psychosocial assessments were conducted whenever possible. Activities within this intervention included educational workshops, creative arts workshops, individual adherence counseling, referrals to community programs and other supportive services, and individual navigation to appointments.
MetroPlus received a cross-referenced list in August 2015. NYS purposely sent names of individuals who were enrolled with MetroPlus at any time in the prior 4 years because members may return to the plan (beneficiaries have the right to switch Medicaid managed care plans once every 12 months). Monthly, MetroPlus reconciled the list with its active enrollment roster.
Viral Load Data Handling and Collection
For listed members with surveillance viral load results, baseline values were as recent as July 2015. To maintain some measure of confidentiality, NYS chose not to share exact numeric values and dates for viral load results but instead reported them with month/year only and in predefined logarithmic ranges categorized as suppressed (<200 copies/mL) and unsuppressed (200-999, 1000-9999, 10,000-99,999, and ≥100,000 copies/mL). MetroPlus reconciled the list with its internal care coordination database, which included available viral load results collected from provider medical records. If the internal database contained a quantified viral load value that matched the range, month, and year of the surveillance viral load value, the quantified value was kept. When there was a range value for which MetroPlus was unable to obtain a corresponding quantified value, the range value was assigned a quantified value for statistical analysis as follows: 999.99, 9999.99, 99,999.99, or 100,000.99. Using the “.99” within the value allowed for clear recognition by staff working with the members that the result was an approximation and originated from the state list. Throughout the 2-year study interval, quantified viral load values were collected and recorded from available medical records.
Collected data included demographics, program contact type and frequency, ART usage (refill pattern), CD4 cell counts and dates, and HIV viral load values and dates over 2 years. A successful program contact was defined as direct contact with a member who agreed to speak with the person attempting the contact either by telephone or face-to-face.
Not all members in the program had viral load data. Hence, the sample selection methodology required that eligible members for analysis had at least 2 viral load data points to measure the change in viral load from baseline. The closest viral load value to the program initiation date (+/— 90 days) was labeled the baseline viral load value. The current viral load value was selected based on the viral load available at last observation. Finally, the viral load values were categorized into suppressed or unsuppressed logarithmic ranges.
This study analyzed 2 groups: one including the derived sample of members with comparable viral loads and a subset who had an unsuppressed viral load at baseline. Because of the participation overlap of the outreach and peer care connection interventions, as well as the small number of members referred to the street outreach intervention, members from both interventions were combined for analysis.
The null hypothesis was that the program had no impact on lowering viral load values from program initiation to termination. We conducted a retrospective statistical analysis on viral load values to evaluate the change of member viral loads in each logarithmic range at baseline compared with current viral load. We visualized the change from baseline to current viral load using a kernel density estimation (KDE) plot. A χ2 analysis was performed with a P value α of .05 as a cutoff for significance for the above comparisons. Additionally, an odds ratio (OR) analysis was conducted on variables of program contact and viral load suppression, with a P value α of .05 as the cutoff for significance.
Study Population Derivation
The cross-referenced state list contained 1741 members (Figure 1). After eliminating members who were disenrolled from the plan as of January 1, 2016, the study population consisted of 1429 actively enrolled members during the 2-year study interval. Of those, 1410 members had at least 1 viral load value, 1216 had more than 1 viral load value, and 901 had more than 2 viral load values. Of those, 500 members had an initial viral load value within 90 days before or after the program initiation date of January 1, 2016 (comparable group). Of those, 316 members had an initial unsuppressed viral load value (unsuppressed-at-baseline group).
The targeted population represented 24% (1429/5919) of the total identified HIV-positive Medicaid population actively enrolled with the plan at the time of the program initiation. The baseline characteristics of all targeted members, those with comparable (baseline and current) viral loads, and those with unsuppressed viral loads at baseline are summarized in Table 1. Because of the nearly 6-month gap between receipt of the list from the state and the program initiation, 184 listed members had already achieved viral load suppression. The comparable and unsuppressed groups were representative of the total population with respect to age, gender, and baseline CD4 count. A relative proportion (13%-15%) of those referred to the street outreach intervention contributed to the composition of all 3 groups.
Contact with either the outreach or peer care connection interventions is summarized in Table 2. The total population, comparable, and unsuppressed groups had 56%, 61%, and 61% successful contacts with the outreach team, respectively, compared with 40%, 44%, and 44% successful contacts by the care coordinators and the peer educators/counselors combined, respectively. Thus, a notable portion of members from all groups were engaged in care but did not have any successful contact with the program staff. Despite our best efforts, the program could not contact every targeted member.
Viral Load Suppression
The viral load suppression rates at last observation were 40.9%, 47.4%, and 38.6% for the total population, comparable, and unsuppressed-at-baseline groups, respectively, suggesting that members in the program experienced significantly improved viral load suppression (P <.01 for both groups) (Figure 2). Members with viral loads in the suppressed range increased by 10.6% in the comparable group. However, members with viral loads of 100,000 copies/mL or greater increased by 3%. The observed increase in viral load suppression was even greater in the unsuppressed-at-baseline group, as seen in Figure 2B. Members with viral loads in the suppressed range increased by 38.61% at current viral load measurement, whereas members with viral loads of 100,000 copies/mL or greater increased by 3.16%. The fact that some members’ viral loads increased to more than 100,000 copies/mL indicates poor, if any, medication adherence in this small percentage of the population.
Viral load movement patterns in the comparable group are visualized in a KDE plot (Figure 3) (P <.01). This plot confirms that many members experienced viral load suppression, even those in the high viral load ranges (≥10,000 copies/mL). Nonetheless, the plot also demonstrates that members who had higher viral loads at program initiation were somewhat less likely to lower their viral loads at current viral load measurement.
Approximately one-third of members with unsuppressed viral loads at baseline in the higher viral load ranges (>10,000) achieved viral load suppression at current viral load measurement, compared with 40% to 51% of members in the lower viral load ranges (eAppendix Figure [available at ajmc.com]) (P <.01).
However, of the 1410 members with at least 1 viral load value, 44% (417/945) who were successfully contacted achieved viral suppression compared with 34% (157/465) who were not (OR, 1.55; 95% CI, 1.23-1.95; P <.01). Therefore, successful contact was associated with an improved health outcome.
Within our plan’s total HIV-positive Medicaid managed care population, approximately 76% were engaged in care and had viral suppression at the start of 2016. This is consistent with the citywide viral load suppression rate of 74% as reported by the New York City Department of Health and Mental Hygiene at the end of 2016.23 However, a quarter of the population has a viral load that remains or becomes unsuppressed at any given time. Of that quarter, 44% achieved viral load suppression over the program’s first 2 years.
The novel Data to Care public health strategy, which leveraged surveillance data to target HIV-positive individuals who are not engaged in care or who have unsuppressed viral loads despite being in care, was originally conceptualized as a state and local health department exercise. However, NYS took an innovative approach by involving Medicaid managed care plans that already had proven care coordination programs, and this approach proved to be successful in this study.
This pilot program resulted in other unexpected benefits. Our plan developed a good working collaboration with a community agency that continues to the present day. The placement of our care coordinators and peers on-site at provider clinics has engendered an improved collaboration with our providers and brought us closer to our members. Our care coordinators are no longer just a voice on the telephone. Finally, the process of hiring and working alongside HIV-positive peer educators and peer counselors has humanized the disease for our care coordinators and resulted in improved functioning as a team to support our HIV-positive membership.
Although the analysis focused on those participants with available data, it is likely that all members who received the interventions benefited. This analysis focuses on the first cohort we received, as it has the longest observation time. We have received additional lists annually from NYS in 2017, 2018, and 2019. We plan to report on these cohorts in the future. The program is ongoing and still being funded.
The results of this study suggest that HIV-positive members not engaged in care and those with high viral loads at baseline can achieve viral load suppression with outreach and enhanced care coordination from a Medicaid managed care organization. Such efforts positively contribute to overall improved engagement in HIV primary care and ART medication adherence.
This study has some important drawbacks that chiefly involve data collection. Although significant effort was devoted to the collection of viral load values, collection was incomplete. Surveillance viral loads were reported in ranges and not in absolute values. For 8.6% of members, the range value was the only available viral load value. This caused the values in the high unsuppressed viral load ranges to be tightly centered around the 100,000 level, resulting in a distinct concentration within the KDE plot, especially in members whose values did not change. In addition, not all members of the cohort were able to be engaged, and some members achieved viral load suppression without program contact. Additionally, 13.6% of members were already engaged and had viral load suppression at program initiation because of the time lag of reporting viral loads to the HIV Surveillance Registry and the inability of MetroPlus to collect more real-time viral load data.
Although a large majority of identified people living with HIV can achieve and maintain viral load suppression with routine HIV care and support, there remains an important minority who do not. Viral load suppression is transient if adherence to ART is not maintained. Cross-referencing Medicaid plan rosters with health department surveillance data can help identify higher-risk populations to target. Medicaid managed care organization programs that provide enhanced care coordination and support can successfully contribute to improving viral load suppression within an urban HIV population. Such programs will continue to be needed as long as social determinants of health exist.Author Affiliations: MetroPlus Health Plan (RGH, DW, RA), New York, NY; New York State Department of Health AIDS Institute (IF), Albany, NY; Annex Clinical Corporation (MA), New York, NY.
Source of Funding: New York State Department of Health C30813GG.
Author Disclosures: Dr Hewitt is employed with MetroPlus Health Plan, which contracts with New York State (NYS) to administer its Medicaid managed care plans; has received a grant for peer certification recruitment from the NYS Health Department; and attended the International Conference on AIDS in July 2018. Mr Adule is employed with MetroPlus Health Plan. Mr Alsumidaie was paid by MetroPlus to conduct statistical analysis. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (RGH, DW, IF); acquisition of data (RGH, DW, RA, IF); analysis and interpretation of data (RGH, RA, IF, MA); drafting of the manuscript (RGH, IF, MA); critical revision of the manuscript for important intellectual content (RGH); statistical analysis (MA); provision of patients or study materials (RGH, DW); obtaining funding (RGH, DW); administrative, technical, or logistic support (DW, RA); and supervision (RGH, DW).
Address Correspondence to: Ross G. Hewitt, MD, MetroPlus Health Plan, 160 Water St, Ste 18W-009, New York, NY 10038. Email: email@example.com.REFERENCES
1. Phillips AN, Lundgren JD. The CD4 lymphocyte count and risk of clinical progression. Curr Opin HIV AIDS. 2006;1(1):43-49. doi: 10.1097/01.COH.0000194106.12816.b1.
2. Gandhi M, Gandhi RT. Single-pill combination regimens for treatment of HIV-1 infection. N Engl J Med. 2014;371(3):248-259. doi: 10.1056/NEJMct1215532.
3. May MT, Gompels M, Delpech V, et al; UK Collaborative HIV Cohort (UK CHIC) Study. Impact on life expectancy of HIV-1 positive individuals of CD4+ cell count and viral load response to antiretroviral therapy. AIDS. 2014;28(8):1193-1202. doi: 10.1097/QAD.0000000000000243.
4. Bradley H, Hall HI, Wolitski RJ, et al. Vital Signs: HIV diagnosis, care, and treatment among persons living with HIV—United States, 2011. MMWR Morb Mortal Wkly Rep. 2014;63(47):1113-1117.
5. What is the HIV care continuum? HIV.gov website. hiv.gov/federal-response/policies-issues/hiv-aids-care-continuum. Updated December 30, 2016. Accessed December 13, 2018.
6. National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention. Understanding the HIV care continuum. CDC website. cdc.gov/hiv/pdf/library/factsheets/cdc-hiv-care-continuum.pdf. Published June 2018. Accessed December 18, 2018.
7. HIV in the United States and dependent areas. CDC website. cdc.gov/hiv/statistics/overview/ataglance.html. Updated January 29, 2019. Accessed May 8, 2019.
8. The HIV/AIDS epidemic in the United States: the basics. Kaiser Family Foundation website. kff.org/hivaids/fact-sheet/the-hivaids-epidemic-in-the-united-states-the-basics. Published March 25, 2019. Accessed May 8, 2019.
9. Auerbach C, Beckerman NL. HIV/AIDS prevention in New York City: identifying sociocultural needs of the community. Soc Work Health Care. 2010;49(2):109-133. doi: 10.1080/00981380903158011.
10. HIV in the United States by region. CDC website. cdc.gov/hiv/statistics/overview/geographicdistribution.html. Updated November 27, 2018. Accessed December 13, 2018.
11. Local data: New York City. AIDSVu website. aidsvu.org/state/new-york/new-york-city. Accessed September 23, 2018.
12. Mugavero MJ, Amico KR, Westfall AO, et al. Early retention in HIV care and viral load suppression: implications for a test and treat approach to HIV prevention. J Acquir Immune Defic Syndr. 2012;59(1):86-93. doi: 10.1097/QAI.0b013e318236f7d2.
13. Gardner EM, McLees MP, Steiner JF, Del Rio C, Burman WJ. The spectrum of engagement in HIV care and its relevance to test-and-treat strategies for prevention of HIV infection. Clin Inf Dis. 2011;52(6):793-800. doi: 10.1093/cid/ciq243.
14. Bangsberg DR, Perry S, Charlebois ED, et al. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS. 2001;15(9):1181-1183.
15. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection [erratum in Ann Intern Med. 2002;136(3):253. doi: 10.7326/0003-4819-136-3-200202050-00020]. Ann Intern Med. 2000;133(1):21-30. doi: 10.7326/0003-4819-133-1-200007040-00004.
16. Flaks RC, Burman WJ, Gourley PJ, Rietmeijer CA, Cohn DL. HIV transmission risk behavior and its relation to antiretroviral treatment adherence. Sex Transm Dis. 2003;30(5):399-404.
17. Skarbinski J, Rosenberg E, Paz-Bailey G, et al. Human immunodeficiency virus transmission at each step of the care continuum in the United States. JAMA Intern Med. 2015;175(4):588-596. doi: 10.1001/jamainternmed.2014.8180.
18. Dean HD, Fenton KA. Addressing social determinants of health in the prevention and control of HIV/AIDS, viral hepatitis, sexually transmitted infections, and tuberculosis. Public Health Rep. 2010;125(suppl 4):1-5. doi: 10.1177/00333549101250S401.
19. Saracino A, Zaccarelli M, Lorenzini P, et al; Icona Foundation Study Group. Impact of social determinants on antiretroviral therapy access and outcomes entering the era of universal treatment for people living with HIV in Italy. BMC Public Health. 2018;18(1):870. doi: 10.1186/s12889-018-5804-z.
20. Catz SL, Kelly JA, Bogart LM, Benotsch EG, McAuliffe TL. Patterns, correlates, and barriers to medication adherence among persons prescribed new treatments for HIV disease. Health Psychol. 2000;19(2):124-133. doi: 10.1037/0278-6184.108.40.206.
21. End AIDS: 2015 blueprint. New York State Department of Health website. health.ny.gov/diseases/aids/ending_the_epidemic/docs/blueprint.pdf. Published March 30, 2015. Accessed December 18, 2018.
22. Data to care: using HIV surveillance data to support the HIV care continuum. CDC website. effectiveinterventions.cdc.gov/docs/default-source/data-to-care-d2c/pdf-of-important-considerations.pdf. Accessed December 18, 2018.
23. New York City Department of Health and Mental Hygiene. HIV surveillance annual report, 2017. NYC website. www1.nyc.gov/assets/doh/downloads/pdf/dires/hiv-surveillance-annualreport-2017.pdf. Published December 1, 2018. Accessed May 8, 2019.