Nutrition Services, Viral Suppression, CD4, and Retention in Ryan White Program Participants.

1

Chapter 1: Introduction to the Study

Background

In this study, I examined the potential association between nutrition services of

medical nutrition therapy and food assistance available through the Ryan White

HIV/AIDS Program (RWHAP) and HIV-related health outcomes of clients receiving

those services, including viral suppression, retention in care, and CD4. Maintaining

adequate nutrition is an important part of HIV care. Poor nutrition can cause the disease

by another qualified individual as nutrition counseling, is one way a patient could

improve his or her nutritional status. Almeida, Segurado, Duran, and Jaime (2011) found

that a 6-session nutrition intervention held over the course of 1 year had a statistically

significant improved the consumption of fiber by over 10 grams. Lazzaretti, Kuhmmer,

Sprinz, Polanczyk, and Ribeiro (2012) discovered that a 6-month dietary intervention

achieved statistically significant changes in total blood cholesterol, triglycerides, and

LDL cholesterol (bad cholesterol) between the intervention and control groups, as well as

increases in HDL (good cholesterol) in both groups. The increases in HDL levels were

due to viremic control associated with antiretroviral therapy. In addition, dietary

2

counseling can be used to address concerns regarding the role of medications on nutrition

status (including drug-nutrient interactions), body composition changes associated with

HIV, and other issues related to nutrition, malnutrition, and food that present with HIV

infection (Fields-Gardner, 2010).

Food access also influences nutritional status. Lack of healthy, nutritious food

(also known as food insecurity) can have a significant impact on a person’s health. For

HIV patients, Mendoza et al. (2013) concluded that food insecurity in pediatric patients

a food insecure HIV infected population compared to HIV positive patients who were

food secure.

At the same time, being HIV positive increases the likelihood that a family will be

food insecure. Anema et al., (2011) found very high prevalence of food insecurity in the

HIV positive population receiving antiretroviral treatment in a resource rich environment,

with 71% of the study participants being ranked as food insecure. Food insecurity can

also lead to negative uses of the health care system, such as increased usage of

emergency rooms and recent hospitalizations (Weiser et al., 2013). Food assistance, such

3

as food bags or home delivered meals, may potentially help address food insecurity

issues.

The RWHAP provides a safety net for low-income HIV infected patients who

lack adequate health insurance. The program plays an important role in ensuring high

quality HIV care to clients who might not have access to health care and support services

(Sood et al., 2014). The grant-based program funds states and metropolitan areas with

high incidence of HIV. Funds are used to provide a variety of health care services

assistance.

Despite the documented importance of medical nutrition therapy and food

assistance, there is a gap in the research as to whether patients who receive these services

through the RWHAP have different levels of viral suppression, CD4 counts, or retention

in care (being engaged in regular HIV-related health care; Mugavero, Davila, Nevin, &

Giordano, 2010) when compared with RWHAP clients who do not receive these services.

This study was conducted to fill that gap and to understand whether these services are

associated with improved HIV-related patient outcomes. I used a quantitative cross-

4

sectional design to assess whether there is a relationship between food assistance and

medical nutrition therapy and the health outcomes of viral suppression, CD4 count, and

retention in care.

Research Question1: Is there an association between receiving food assistance or

medical nutrition therapy through the RWHAP and health outcomes such as viral

suppression and CD4 counts?

Research Question 2: Is there an association between receiving food assistance or

medical nutrition therapy through the RWHAP and retention in care?

Null Hypothesis (H

2): There is no statistically significant association between

receiving food assistance or medical nutrition therapy through the RWHAP and retention

in care.

Alternative Hypothesis (H

between receiving food assistance or medical nutrition therapy through the RWHAP and

retention in care.

5

Framework

Andersen’s behavioral model of health services use was developed in the 1960s to

predict and explain health care usage by families (Andersen, 1995). Over time the model

was revised to focus more on individual usage. Additional revisions were made to

accommodate the acknowledgement of the role environment plays in health care

utilization and to include outcomes of health services.

The behavioral model for vulnerable populations is a modification of Andersen’s

Predisposing → Enabling → Need → Health Behavior → Outcomes

Figure 1. Behavioral model for vulnerable populations.

The predisposing component includes characteristics that are not easily

changeable: age, gender, marital status, and social structure. For vulnerable populations,

the predisposing component also includes characteristics such as immigration status,

living conditions, criminal behavior, mental illness, and substance abuse. The enabling

component includes characteristics that help support a person, such as personal resources,

insurance status, income, and health services resources. For vulnerable populations, this

6

might include receiving public benefits, community crime rates, and the availability of

social services within the person’s community. The need component focuses on both

perceived need and objective need for health conditions of the general population. For

vulnerable populations, it includes the perceived and evaluated needs, such as sexually

transmitted diseases and HIV. Personal health practices include activities such as diet,

exercise, and adherence. They also include sources of food and food safety, as well as

unsafe sexual behaviors. Outcomes include perceived and evaluated health status and

impacts personal health behaviors by providing skills to help patients improve their

eating habits, as well as food to address health issues. The program also may help keep

patients in care (retention) by ensuring that they go to their medical appointments so that

they can get referrals or access to these other services. These two services then

theoretically could lead to better health outcomes in the form of higher CD4 counts and

viral suppression.

7

I conducted a quantitative study with a cross-sectional design because the data

used for the analysis were gathered on an annual basis and a direct statistical relationship

between the independent and dependent variables could not be established.

The independent variables in this study were receipt of medical nutrition therapy,

food assistance, or both. The dependent variables were viral load, CD4 count, and

data. I examined the variables of receipt of nutrition services (medical nutrition therapy

and food assistance), CD4 count, retention in care, and viral load.

CD4 count: A laboratory test measuring the number of T lymphocytes (CD4 cells)

in a person’s blood, which are a type of white blood cell that plays an important role in

fighting infection. In HIV infection, the virus destroys CD4 cells, lowering the body’s

defenses to infection. The CD4 count is important because it allows clinicians to monitor

how well a patient’s body is responding to antiretroviral treatment (ART) and provides an

overall status of the patient’s immune system. The CD4 count is used clinically in

8

determining how quickly ART needs to be initiated and when to start and discontinue

prophylaxis for opportunistic infections and is a predictor in disease progression and

survival (DHHS Panel on Antiretroviral Guidelines for Adults and Adolescents, 2015).

For this study, the Centers for Disease Control and Prevention’s case definition for HIV

was used to categorize program participants into levels by CD4 count. Stage 1 is a CD4

count of greater than or equal to 500 cells/µL or a CD4 percentage of total lymphocytes

of greater than 26, Stage 2 is a CD4 count of 200 to 499 or CD4 percentage of total

category for data collection (Health Resources and Services Administration, 2015a). This

category also includes nutrition supplements not provided under the care of a registered

dietitian and the provision of vouchers for food. Because these services cannot be broken

apart, they were referred to as food assistance throughout this document.

Medical nutrition therapy (MNT): The treatment and prevention of disease by a

registered dietitian or nutrition professional through assessment, setting nutrition-related

goals and making a plan that includes counseling, therapy, education, and modification of

diet (Fields-Gardner, 2010). In the RWHAP, MNT also includes the provision of

9

nutrition supplements and food pursuant to a medical professional’s recommendation and

a nutrition care plan (Health Resources and Services Administration, 2015a).

Outpatient/ambulatory medical service: Defined by the Health Resources and

Services Administration (2015a, p. 7) as “the provision of professional diagnostic and

therapeutic services directly to a client by a physician, physician assistant, clinical nurse

specialist, nurse practitioner, or other health care professional certified in his or her

jurisdiction to prescribe antiretroviral (ARV) therapy in an outpatient setting.

annual basis by the Health Resources and Services Administration to monitor impacts

and outcomes in the RWHAP (Health Resources and Services Administration, 2015b).

Viral load: A laboratory test measuring Plasma HIV-1 RNA in a person’s blood.

It is useful for measuring sustained response to ART (DHHS Panel on Antiretroviral

Guidelines for Adults and Adolescents, 2015).

Viral suppression: Having a Plasma HIV-1 viral load of less than 200 copies on

the most recent viral load test. A recent meta-analysis confirmed that in heterosexual

serodiscordant couples (couples where one partner is HIV positive and the other is HIV

Predisposing factors are exogenous factors that play into the use of health care

(Andersen, 1995). The predisposing traditional factors are primarily demographic

characteristics and include things like gender, age, and social structure factors such as

education, employment, and family size (Gelberg et al., 2000). The predisposing

vulnerable factors are social structures such as living conditions, immigration status, and

substance abuse. The predisposing domain items generally impact a person’s social and

community status and ability to deal with health problems, and to find resources to solve

available if a patient is to utilize a health care service.

Finally, the need traditional domain includes perceived need and the evaluated

need of general health conditions of the population (Gelberg et al., 2000). The need

vulnerable domain focuses on perceived and evaluated need of those disease conditions

that are of importance to vulnerable populations, such as HIV infection or tuberculous.

Andersen’s models suggest that these three sets of characteristics then lead to

health behaviors, which include personal health practices such as diet and exercise, as

well as choices around food sources, safe/unsafe sexual practices, and which providers

the person sees (Andersen, 1995; Gelberg et al., 2000). A person’s health behaviors in

turn affect health outcomes, which include perceived and evaluated health status and

satisfaction with care. Perceived status could be the person’s perception of his or her

heath, while the evaluated status would include items that concretely measure health

status.

Andersen’s various models have been used as a framework to support studies on

health access and health outcomes in many vulnerable populations including PLHW

(Andersen et al., 2000; Anthony et al., 2007; Brennan-Ing et al., 2014; Datti & Conyers,

health care provider in individuals recently diagnosed (Anthony et al., 2007). Factors that

were associated with being seen by a provider include having more symptoms, being

infected with viral hepatitis, having public health insurance, or case management. The

behavioral model for vulnerable populations has been used to identify variables that

impact viral suppression in HIV positive individuals (King et al., 2009). These variables

include the predisposing variables of homelessness and drug use, both of which had a

negative impact on the health outcome of viral suppression. Andersen’s original model

has been used to identify variables that would predict use of vocational rehabilitation

services in HIV positive Latino men (Datti & Conyers, 2010).

This theory was chosen because it has demonstrated use with HIV populations.

Researchers have used this model as a framework to explore health care utilization in

various segments of this population, including the newly diagnosed PLWH; older PLWH;

gay, lesbian, bisexual, and transgender PLWH; and injection drug users (Anthony et al.,

2007; Brennan-Ing et al., 2014; Mizuno et al., 2006). The model has also been used to

describe how the various factors (enabling, predisposing, and need) impact HIV

treatment (Andersen et al., 2000). In addition, the model has been used to examine

factor of being HIV positive to examine the evaluated health outcomes of viral

suppression, CD4 count, and retention in care.

Food security and good nutrition are important elements in HIV care. In the

following sections, I review the current literature related to the impact of a lack of food

security (food insecurity) and the role of nutrition in HIV-related health outcomes. I also

discuss the literature on food assistance (the receipt of food or meals to improve nutrition

or reduce food insecurity) and nutrition counseling (including medical nutrition therapy)

and the health outcomes of interest in this study (CD4 counts, retention in care, and viral

suppression).

HIV’s primary mode is to destroy the immune system. It does this by infecting

and destroying specific white blood cells in the body, called T cells, which are the body’s

primary defense against infection (Moir, Chun, & Fauci, 2011). The virus enters these

white blood cells via CD4 coreceptors. The virus then uses mechanisms in the white

mechanism for interrupting the virus’s reproduction cycle. For example, nucleoside

reverse transcriptase inhibitors (NRTI) inhibit viral reverse transcriptase, which is an

enzyme that is needed for the virus to replicate. Nonnucleoside reverse transcriptase

inhibitors work by binding to the viral reverse transcriptase in a different spot than

NRTIs. Protease inhibitors block HIV protease and prevent the virus from infecting other

cells. Entry inhibitors or fusion inhibitors block HIV from entering the T cell at all. Other

drugs work to block integrase, which is the enzyme used to integrate HIV RNA into the

host cell.

Regular, consistent ART use can lead to viral suppression and a reduced risk of

viral transmission. Retention in care is an important component of viral suppression

(Yehia et al., 2014) because access to care leads to access to ART. Once PLWH are

linked to and then retained in care, they may be able to achieve viral suppression and

decrease the likelihood of poor health outcomes (Reveles et al., 2015).

Food Insecurity

Food insecurity is an important variable in HIV-related health outcomes.

food, some patients suggested that the side effects of the antiretroviral are exacerbated.

Further, Weiser et al. noted that patients often worried about starting an ART regime

because they were concerned they would not be able to complete it because of a lack of

food.

For food insecure patients prescribed ART that is taken with food, there was

statistically significant disparity between outcomes compared to patients who were

prescribed ART that does not need to be taken with food (Kalichman et al., 2015). Of the

63% of study subjects who were food insecure, 57% had been prescribed ART to be

taken with food (Kalichman et al., 2015). These patients had lower CD4 counts, higher

viral loads, and more HIV-related symptoms than the food secure study participants.

Food insecurity has also been linked to not being virally suppressed (Wang et al.,

2011). Wang et al. (2011) found that in a population of HIV positive veterans on ART,

24% were food insecure. Of those, their viral load levels were higher, with a median viral

load of 400 (IQR 75, 15,572) versus 359 (IQR 75, 3384) for their food secure

counterparts (p < 0.0001) and were more likely to not be virally suppressed (adjusted OR

Agriculture Food Security Survey Module. Food security has also been associated with

increased acute care utilization in HIV positive homeless and marginally housed patients

(Weiser et al., 2013). Weiser et al. (2013) found that food insecurity was statistically

significantly associated with the likelihood of having been hospitalized or to have visited

an emergency room in the most recent 3 months.

A study of PLWH living in Atlanta, Georgia, concluded that those that reported

food insecurity over the last twelve months also had reported more HIV symptoms, as

identified in a scale of 14 related symptoms (Kalichman et al., 2010). While this study

observed it less likely that these individuals were virally suppressed (OR 1.7; CI 95%, 1.1

– 3.0) and were more likely to have a lower CD4 count (t=2.2, p<0.05), according to

Kalichman et al. (2015) the type of ART a food insecure patient was prescribed (ART

needing to be taken with food versus ART taken without food) impacts adherence, which

would impact viral suppression and CD4 counts. Those patients prescribed ART that had

to be taken with food fared worse than those that were prescribed ART that did not

require food. In addition, according to Kalichman et al., (2010), those patients reporting

food insecurity were much more likely to report barriers to staying in HIV treatment,

changes in CD4 counts or other HIV-related health outcome measures.

One meeting abstract examined food security and RWHAP. The study found that

77% of the RWHAP food and nutrition service clients in the New York City jurisdiction

were food insecure and this was associated with an unsuppressed viral load (Alexy,

Feldman, Thomas, & Irvine, 2013).

Nutrition in HIV

Nutrition plays a key role in HIV care. Deficiencies of certain micronutrients,

malabsorption, wasting disease, and weight loss each impact health outcomes in HIV

infection and are important nutrition considerations in care of a PLWH (de Pee & Semba,

2010). In addition, there are significant drug-nutrient interactions that impact

bioavailability of ART in HIV treatment that must receive consideration (Raiten, 2011).

Further, a Kenyan study determined there was an association between higher CD4 counts

in male patients and adequate protein intake, which may indicate a role for medical

nutrition therapy to educate on diet and macronutrient consumption to support

improvements in HIV-related outcomes (Vaughan, Cardenas, & Keiser, 2013). Further

exploration of this topic is warranted to provide information on the role of diet, especially

treat HIV and related health concerns either require specific dietary interventions (i.e.,

require that medications be taken with food or certain amount of fat) or there are dietary

implications due to these medications, such as dyslipidemia. In particular, for people who

are food insecure, Kalichman et al., (2015) found that these patients were more often

prescribed ART that was required to be taken with food and had worse adherence, lower

CD4 counts, and were less likely to be virally suppressed than their food secure

counterparts.

There are many benefits to nutrition education and medical nutrition therapy for

PLWH. For example, HIV positive patients in South Africa receiving nutrition education

are less likely to be food insecure (p=0.014), as well as less likely to have reported being

sick in the last thirty days (p=0.037) or, if they were sick, to have milder symptoms

(p=0.038) (Oketch, Paterson, Maunder, & Rollins, 2011). Nutrition education can also

help improve eating habits, such as increasing consumption of healthy fats and fiber,

even if associated health outcomes are not seen in the short term (Almeida et al., 2011).

Nutrition education can also improve treatment adherence. In a study in Honduras

that provided nutrition education, it was found that treatment adherence, which included

infection and ART usage. A study by Lazzaretti, Kuhmmer, Sprinz, Polanczyk, and

Ribeiro, (2012) found that patients receiving targeted nutrition education had

significantly less hypercholesterolemia and Hypertriglyceridemia (p=0.001 for both).

This study also concluded that significantly fewer patients developed lipid profiles

capable with the diagnosis of dyslipidemia.

Food Assistance

Food assistance is another potential way to address food insecurity, not only in

PLWH, but also in the general food insecure population. It may also help with retention

in HIV care by decreasing the hunger and helping patients to meet their food needs

(Aberman, Rawat, Drimie, Claros, & Kadiyala, 2014). Depending on the setting, it could

include receiving a bag of groceries to prepare at home (often called a food pantry, food

shelf, or food bank), a hot meal eaten onsite (e.g., a soup kitchen), or a prepared meal

delivered to the patient’s house (e.g., a Meals on Wheels type program). In the RWHAP,

there is not a clear sense as to what format food assistance takes, but anecdotally, food

pantry services are the most likely form for it to take. In the RWHAP, food assistance,

like the other support services provided by the program, is intended to support the patient

subscale score for general health perceptions in people receiving assistance (Maluccio,

Palermo, et al, 2015, Table 3), as well as a statistically significant decrease in the number

of self-reported symptoms (p≤0.01).

In a study in rural Rwanda, community accompaniments, which included a

monthly food package for family of four, have been shown to increase retention in care

and viral suppression than those patients who did not receive the community

accompaniments (Franke et al., 2013). In this study those patients receiving the added

food and support benefit had a 15% greater chance of being retained in care and virally

factors, demographics, health status of the infected person, and access to care and

treatment. The three outcomes of interest in this study, CD4 counts, viral load, and

retention in care, are all impacted by these factors.

Retention in care and viral suppression are two important points that measure

progress along the HIV care continuum, whereas CD4 counts are emphasized less these

days, but provide a basis measure of the immune function of a PLWH. Retention in care

provides a measure for how well PLWH stay in care and have access to life-saving

been clearly studied.

Chapter 3: Research Method

I examined HIV-related health outcomes of viral suppression, CD4 counts, and

retention in care of PLWH who receive nutrition-related services in RWHAP. The

primary analysis was conducted using the RSR, which is a data set of all PLWH who

receive services through the RWHAP, as reported by program recipients (e.g., states,

cities, universities, and HIV clinics).

Research Design and Rationale

assistance. Although a cross-sectional design cannot be used to determine a causal

relationship, it can be used to identify an association between the variables. The RSR data

set is the only national data set that has health outcomes for PLWH and includes PLWH

who have received the nutrition assistance services of interest in this study. A cross-

sectional design was appropriate due to the limitations imposed by the use of existing

data. It would not have been practical to collect additional data at the national level to

look at this issue across the entire RWHAP. There is limited literature on the impact of

nutrition services on HIV-related health outcomes, especially CD4 counts, viral

suppression, and retention in care. Given this gap in the literature, this was an appropriate

starting place for this research.

The target population was PLWH ages 13 and older who receive medical nutrition

therapy or food bank services and outpatient/ambulatory medical service through the

RWHAP. Under RSR reporting requirements, only clients who receive an

outpatient/ambulatory medical service visit (Health Resources and Services

Administration, 2015b).

Choosing an appropriate sampling design is important because it allows statistical

results to be more generalizable to the larger populations of interest in a particular study

(Frankfort-Nachmias & Nachmias, 2008). The total data set was divided into four groups:

RWHAP clients who received food assistance, RWHAP clients who received medical

nutrition therapy, RWHAP clients who received both food assistance and medical

nutrition therapy, and RWHAP clients who received neither service. Given the

potentially large size of the data set for this study, a systematic sample was applied to

each group to identify a sample for analysis from each group to ensure similar size for

each of the groups for analysis. A systematic sample means that every K

th

entry will be

selected (K = total population/sample size needed). It is often used with large populations

or when a larger sample is needed (Frankfort-Nachmias & Nachmias, 2008). No

additional stratification of demographic or variables occurred. In this type of sample,

each sampling unit in the data set has a 1/K probability of being selected. Although this

Sample Size and Power Analysis

Sample size is also important in a study design because it impacts the researcher’s

ability to correctly accept or reject the null hypothesis (Cohen, 1992). Cohen’s (1992)

paper on statistical power was used as a guide to determine the minimum sample size. I

determined that for the chi-square statistical test with a power of .80, α of .05, medium

effect size, and 2 degrees of freedom, the appropriate sample size was 107 for each

group, equaling a total of 428 to ensure statistical power.

Variables

The independent variables were receipt of medical nutrition therapy and food

assistance. Both of these variables were dichotomous (yes/no). The dependent variables

were viral suppression, CD4 count, and retention in care. Viral suppression was

dichotomous and was defined as having a Plasma HIV-1 viral load of less than 200

copies on the most recent viral load test. Those clients with the most recent viral load of

less than 200 copies/mL were coded as 1 for viral suppression and 0 if the latest viral

load was greater than or equal to 200 copies/mL and the client was not virally suppressed.

The variable of CD4 count was an interval variable, but was coded as categorical

using the CDC surveillance case definition: Stage 1 is a CD4 count of 500 cells/mL or

retained in care if they had at least two outpatient/ambulatory medical service visits at

least 90 days apart during the course of a year, with the first occurring by September 1 of

the reporting year and the second at least 90 days after. Retention in care was coded as 1

for retained and 0 for not retained. If the patient had only one outpatient/ambulatory

medical service in a year, he or she was coded as not retained unless the visit occurred

after September 1, and then it appeared as if the variable was missing for that case. If the

client received two or more visits at least 90 days apart, the client was coded as retained.

Data Analysis Plan

The following research questions were addressed in this study:

Research Question 1: Is there an association between receiving food assistance or

medical nutrition therapy counseling through the RWHAP and health outcomes such as

viral suppression and CD4 counts?

Null Hypothesis (H

1): There is no statistically significant association between

receiving food assistance or medical nutrition therapy through the RWHAP and health

outcomes such as viral suppression and CD4 counts.

Alternative Hypothesis (H

in care.

Alternative Hypothesis (H

between receiving food assistance or medical nutrition therapy through the RWHAP and

retention in care.

IBM SPSS Statistics (SPSS), version 21, was used to conduct the statistical

analysis. Data were checked for inconsistencies such as outliers and missing data. The

chi-square statistical test was used to answer the research questions. Chi-square tests

were chosen for this analysis because the outcomes for the research questions were all

categorical variables.

The RSR contains annual client level data reported by funded recipients on the

clients they serve. It is used to monitor outcomes of PLWH and families who get medical

care and support services through the RWHAP (Health Resources and Services

Administration, 2015b). It is also used to monitor progress toward achieving the goals of

the National HIV/AIDS Strategy, Updated to 2020, as well as to provide feedback to the

Department of Health and Human Services, Congress, and other key stakeholders on the

To gain access to the RSR data set, a formal data request was made to the Health

Resources and Services Administration, HIV/AIDS Bureau, Division of Policy and Data

via an internal process (V. Rao, personal communication, April 7, 2016). An analyst

worked directly with the requestor to clarify the request and ensure that the correct data

had been pulled. In addition, the request included the specific variables needed, the

reason for the data request, and who will have access to the data. Once the request was

approved, a data use agreement (DUA) was required (see Appendix A). Once an

approved DUA was on file, access was granted for the requested data and was allowed

for the duration of the project. Data were handled per the instructions in the DUA.

One of the major threats to the validity of this study was the concern that the

dependent variables and the independent variables could not be directly linked. The RSR

is reported on an annual basis; however, receipt of medical nutrition therapy and food

assistance are only reported as a yes/no variable received during the year, while dates are

reported for all outpatient/ambulatory medical services, viral load testing, and CD4

results (Health Resources and Services Administration, 2015a). Because the dates of

data or incorrectly reported data that might impact the quality of what is reported and

analyzed.

To access RSR data, the Health Services and Resources Administration gave

permission to use the data (A. Dempsey, personal communication, November 3, 2015).

Once permission was granted, a data use agreement was signed and approved. Once the

DUA was signed, the agency granted access to the requested data. Each record was

identified by an encrypted unique client identifier. Data were handled, stored, and

destroyed as described in the DUA.

One ethical issue of note is that I am employed by the Health Services and

Resources Administration, HIV/AIDS Bureau, which is the organization that administers

the RWHAP nationally. In this role, I have access to privileged, predecisional

information. I ensured that my role as a student researcher and an employee were

separated so that my professional position did not bias the study or interpretation of the

results. I abided by all IRB approved protocol and the legally binding DUA. Walden

University IRB approval #08-12-16-0197756 was obtained on August 12, 2016 to

The purpose of this study was to examine the association between the nutrition

services of food assistance and medical nutrition therapy provided by the RWHAP, and

the HIV-related health outcomes of viral suppression, CD4 count, and retention in care

using the 2014 RSR. Research has indicated that nutrition services may have a positive

impact on HIV-related health outcomes including helping to keep PLWH retained in

care. This chapter presents the results of the 2014 RSR cross-sectional data analysis and

Data analysis included descriptive statistics on demographic characteristics of the sample

and clinical variables in the study. In addition, this chapter provides results of statistical

analyses to answer each of the research questions.

IRB approval was obtained on August 12, 2016. Once approval was obtained, a

data request was submitted to the Health Resources and Services Administration. The

data were received within a week, along with a variable codebook. The data file was

uploaded into SPSS 21.0 for analysis. The only discrepancies from the original data plan

Population

N = 304,366

Sample

N = 428

f %

Retained in Care

Yes

No

Missing

217,764

53,320

33,282

80.3

19.7

323

72

33

81.8

18.2

Table 2

Each Nutrition Assistance Category

Food

Assistance and

Receiving

Food

Not Receiving

Any Nutrition

Assistance

SD=11.2

SD=10.9

26

55

1

0

17

69

1

0

21.0

54.2

0.6

0

24.8

41.9

1.0

0

Table 3

Representative Comparison of Full Study Sample to Population of RWHAP Clients

Variable

RWHAP

Population

N=304,366

Islander

American

Indian

/Alaska

Native

Multiracial

Hispanic

.004

.019

.246

Gender**

Male

.708

2.548

.280

None

Medicare

Only

Private

Employer

Private

Individual

Medicaid

Only

VA/TriCare

More than one

type

.304

.088

.062

.054

.311

.002

.075

39.654

<.001*

Poverty Level

>500%

401% - 500%

251% - 400%

139% - 250%

101% -138%

<100%

2.4

1.8

5.2

14.8

11.6

63.6

14.283

.027*

CD4 Case

Definition

Stage 1

Stage 2

Stage 3

.560

.333

.107

6.741

.034*

Retained in Care

Yes

No

.803

.197

.541

.462

clients receiving outpatient ambulatory medical services through the RWHAP in 2014.

For categorical variables, the chi square goodness of fit was used. For other variables, a

one sample t-test was used.

Table 4

Representative Comparison of Sample of RWHAP Clients Receiving Food Assistance to

RWHAP

N=304,366

t-test df p-value

Mean Last viral

Mean Last CD4

count 610.2 523.8 -2.939 100 .004*

Mean Age 44.6 45.8 1.110 106 .269

White

Black

Asian

Native

Hawaiian/

Pacific

Islander

American

Indian

/Alaska

.251

.466

.012

.001

.004

Female

Transgender

.281

.010

None

Medicare

Only

Private

Employer

Private

Individual

Medicaid

Only

VA/TriCare

More than one

type

.304

.088

.062

.054

.311

.002

.075

Poverty Level**

>500%

401% - 500%

251% - 400%

139% - 250%

101% -138%

2.4

1.8

5.2

14.8

11.6

10.990

CD4 Case

Definition

Stage 1

Stage 2

Stage 3

.560

.333

.107

2.946

.229

Retained in Care

Yes

No

.803

.197

.769

**One or more cells have expected counts less than 5

Table 5

Representative Comparison of Sample of RWHAP Clients Receiving MNT to Population

Variable

RWHAP

Population

N=304,366

Islander

American

Indian

/Alaska

Native

Multiracial

Hispanic

.004

.019

.246

Gender

Male

.708

.318 1 .573

Coverage**

None

Medicare

Only

Private

Employer

Private

Individual

Medicaid

Only

Medicaid

More than one

type

.304

.088

.062

.054

.311

.075

Poverty Level**

>500%

401% - 500%

251% - 400%

139% - 250%

101% -138%

2.4

1.8

5.2

14.8

11.6

1.599

CD4 Case

Definition

Stage 1

Stage 2

Stage 3

.560

.333

.107

.534

.766

Retained in Care

Yes

No

.803

.197

4.219

**One or more cells have expected counts less than 5

Table 6

Representative Comparison of Sample of RWHAP Clients Receiving Food Assistance and

Variable

RWHAP

Population

N=304,366

Islander

American

Indian

/Alaska

Native

Multiracial

Hispanic

.004

.019

.246

Gender

1.457 1 .227

Coverage**

None

Medicare

Only

Private

Employer

Private

Individual

Medicaid

Only

Medicaid

More than one

type

.304

.088

.062

.054

.311

.075

Poverty Level**

>500%

401% - 500%

251% - 400%

139% - 250%

101% -138%

2.4

1.8

5.2

14.8

11.6

7.687

CD4 Case

Definition

Stage 1

Stage 2

Stage

.560

.333

.107

3.804

.149

Retained in Care

Yes

No

.803

.197

1.718 1 .190

**One or more cells have expected counts less than 5

stage (χ

= .442, p = .802), or for receipt of MNT and food assistance together and stage

(χ

= .714, p = .700).

Table 10

Chi Square Analysis for CDC Case Definition (Based on CD4 count)

Yes 54 13.2 37 9.0 12 2.9