Abstract

Introduction

Despite technical, scientific, and political strides in HIV/AIDS prevention, diagnosis, and treatment, an estimated 35 million infected individuals are unaware of their serological status, and approximately half of the persons diagnosed are not enrolled in any kind of health care service ¹1. Fonjungo PN, Osmanov S, Kuritsky J, Ndihokubwayo JB, Bachanas P, Peeling RW, et al. Ensuring quality: a key consideration in scaling-up HIV-related point-of-care testing programs. AIDS 2016; 30:1317-23.^,22. Bemelmans M, Baert S, Negussie E, Bygrave H, Biot M, Jamet C, et al. Sustaining the future of HIV counselling to reach 90-90-90: a regional country analysis. J Int AIDS Soc 2016; 19:20751..

In 2014, this scenario prompted the Joint United Nations Programme on HIV/AIDS (UNAIDS) to urge all countries to set targets aimed at maximizing the individual and collective benefits of HIV/AIDS diagnosis and treatment ³3. Joint United Nations Programme on HIV/AIDS. 90-90-90 An ambitious treatment target to help end the AIDS epidemic. Geneva: Joint United Nations Programme on HIV/AIDS; 2014.^,44. Nosyk B, Montaner JS, Colley G, Lima VD, Chan K, Heath K, et al. The cascade of HIV care in British Columbia, Canada, 1996-2011: a population-based retrospective cohort study. Lancet Infect Dis 2014; 14:40-9..

The goal, known as 90-90-90, proposes to expand diagnosis and treatment so that 90% of PLWA (persons living with HIV/AIDS) are aware of their serological status, of these, 90% are using ART (antiretroviral therapy), and 90% of persons using specific drug therapy achieve viral suppression. To reach these goals, UNAIDS recommended expanding the start of treatment immediately after diagnosis, that is, rapid and sustained administration of ART with a view towards improved immune status and reduction of the risks of HIV transmission, progression to AIDS, and premature death ²2. Bemelmans M, Baert S, Negussie E, Bygrave H, Biot M, Jamet C, et al. Sustaining the future of HIV counselling to reach 90-90-90: a regional country analysis. J Int AIDS Soc 2016; 19:20751.^,33. Joint United Nations Programme on HIV/AIDS. 90-90-90 An ambitious treatment target to help end the AIDS epidemic. Geneva: Joint United Nations Programme on HIV/AIDS; 2014.^,44. Nosyk B, Montaner JS, Colley G, Lima VD, Chan K, Heath K, et al. The cascade of HIV care in British Columbia, Canada, 1996-2011: a population-based retrospective cohort study. Lancet Infect Dis 2014; 14:40-9.^,55. Departamento de DST, Aids e Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Boletim Epidemiológico Aids e DST 2015; VI(3)..

In order to achieve the goal and control the epidemic by 2030, pillars were defined to be monitored worldwide in order to identify the interventions’ efficacy in each setting. These pillars represent the series of stages experienced by PLWA, from diagnosis of the infection to viral suppression (considered the ultimate objective of the continuum of care and treatment cascade). In Brazil, the cascade has three components, totaling the number of persons included in the provision of specific services in all the stages of: PLWA, persons diagnosed, enrolled in care, retained in health care, in use of ART, and with viral suppression. The latter is obtained with specific drug therapy ⁴4. Nosyk B, Montaner JS, Colley G, Lima VD, Chan K, Heath K, et al. The cascade of HIV care in British Columbia, Canada, 1996-2011: a population-based retrospective cohort study. Lancet Infect Dis 2014; 14:40-9.^,55. Departamento de DST, Aids e Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Boletim Epidemiológico Aids e DST 2015; VI(3).^,66. Departamento de Vigilância, Prevenção e Controle das Infecções Sexualmente Transmissíveis, do HIV/Aids e das Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Manual técnico de elaboração da cascata de cuidado contínuo do HIV. Brasília: Ministério da Saúde; 2017..

A PLWA is defined as retained in care when enrolled in an HIV care service, with continuous access to adequate care according to their health care needs and permanent follow-up by the service ⁶6. Departamento de Vigilância, Prevenção e Controle das Infecções Sexualmente Transmissíveis, do HIV/Aids e das Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Manual técnico de elaboração da cascata de cuidado contínuo do HIV. Brasília: Ministério da Saúde; 2017.. Evidence indicates that this is a behavior that should be encouraged in PLWA in order to establish timely ART, reduce drug resistance and viral load, increase the CD4+ T lymphocyte count, improve clinical conditions, and increase survival ⁷7. Lourenço L, Nohpal A, Shopin D, Colley G, Nosyk B, Montaner JSG, et al. Non-HIV-related healthcare utilization, demographic, clinical and laboratory factors associated with time-to-initial retention in HIV care among HIV-positive individuals linked to HIV care. HIV Med 2016; 17:269-79..

Although retention is a priority in care for PLWA and an important pillar in the cascade of care, it is a challenge in Brazil and in the world. To improve retention requires a better understanding of barriers and facilitators, dealing with the origins in the individual patient, in the health services, and in the external environment ⁷7. Lourenço L, Nohpal A, Shopin D, Colley G, Nosyk B, Montaner JSG, et al. Non-HIV-related healthcare utilization, demographic, clinical and laboratory factors associated with time-to-initial retention in HIV care among HIV-positive individuals linked to HIV care. HIV Med 2016; 17:269-79.^,88. Fiuza MLT, Lopes EM, Alexandre HO, Dantas PB, Galvão MTG, Pinheiro AKB. Adherence to antiretroviral treatment: comprehensive care based on the care model for chronic conditions. Esc Anna Nery Rev Enferm 2013; 17:740-8.^,99. Holtzman CW, Brady KA, Yehia BR. Retention in care and medication adherence: current challenges to antiretroviral therapy success. Drugs 2015; 75:445-54..

Retention of PLWA in health care is known to be essential for the epidemic to be controlled in the next two decades. The necessary consideration of local factors that impact retention in the cascade of care and the importance of equipping health professionals to encourage retention were the motivations for the current study, aimed at creating a predictive model for retention of PLWA in health care and intending to contribute to the identification of possible strategies to optimize retention of PLWA in care and to help achieve the goals.

To predict retention in care, the study analyzed the inherent conditions in the enrollment and permanence of persons with HIV/AIDS in follow-up, with the specialized outpatient service as the reference. The choice was based on the relevance of this service in the supply of activities targeted to follow-up of health conditions in PLWA and the service’s impact on how these persons related to other services in the network of care.

Method

The study was designed to produce a decision tree, a statistical model using a data mining technique to classify the data, capable of leading the research to predict a target outcome with a view towards supporting the decision on this outcome ¹⁰10. Medeiros ARC, Araujo YB, Vianna RPT, Ronei MM. Modelo de suporte à decisão aplicado à identificação de indivíduos não aderentes ao tratamento anti-hipertensivo. Saúde Debate 2014; 38:104-18..

The model enables the identification of independent variables, available in a database, consisting of decision variables that provide the basis for producing classification rules (text representations obtained from the model’s structure), the graphic display of which produces a structure resembling an inverted tree ¹¹11. Medeiros LB, Trigueiro DRSG, Silva DM, Nascimento JA, Monroe AA, Nogueira JA, et al. Integração entre serviços de saúde no cuidado às pessoas vivendo com aids: uma abordagem utilizando árvore de decisão. Ciênc Saúde Coletiva 2016; 21:543-552..

The tree thus consists of decision variables displayed hierarchically, starting from the root node (first variable), whose classification rules lead to internal nodes and their branches, towards an end node with the decision. This is a low-cost resource that facilitates the data interpretation via graphic visualization with the potential to support health-related decisions ¹⁰10. Medeiros ARC, Araujo YB, Vianna RPT, Ronei MM. Modelo de suporte à decisão aplicado à identificação de indivíduos não aderentes ao tratamento anti-hipertensivo. Saúde Debate 2014; 38:104-18.^,1111. Medeiros LB, Trigueiro DRSG, Silva DM, Nascimento JA, Monroe AA, Nogueira JA, et al. Integração entre serviços de saúde no cuidado às pessoas vivendo com aids: uma abordagem utilizando árvore de decisão. Ciênc Saúde Coletiva 2016; 21:543-552..

To construct the model, variables were used from the study’s database, Dropout from Specialized Outpatient Clinical Follow-up of Persons Living with HIV/AIDS¹²12. Medeiros LB. Fatores influentes no abandono do acompanhamento clínico ambulatorial por pessoas vivendo com HIV/Aids, João Pessoa - PB [Dissertação de Mestrado]. João Pessoa: Universidade Federal da Paraíba; 2016., obtained from a secondary source (patient charts). The sample consisted of 260 AIDS cases, selected from a population of 1,941 adults recorded in the Brazilian Information System for Notificable Diseases/AIDS (SINAN/AIDS) in the state of Paraíba, Brazil, from January 2007 to December 2013, all over 18 years of age and enrolled in the state’s specialized outpatient referral clinic for clinical follow-up.

The sample size was set by sampling calculation for finite populations and the sample was collected by simple random sampling. Exclusion criteria were pregnant women and persons in detention or incarceration. Retention was defined from records on the patient chart pertaining to the individual’s appearance at the clinic for consultation and/or tests, at a maximum interval of seven months. This period was defined according to the reasoning that clinically stable individuals would return for their clinical follow-up appointment within a maximum of six months, with one additional month added for cases that required rescheduling an appointment, according to the clinic’s routine.

The decision tree was built using the Waikato Environment for Knowledge Analysis (Weka) package, version 3.7.8 (https://www.cs.waikato.ac.nz/ml/weka/downloading.html), and selection of variables from the database for inclusion in the model was based on variable’s IG (information gain) in relation to the outcome variable (retention in health care). The IG was obtained by calculating the probability of decisions in the database set and in the subsets of independent variables related to the decision. Based on the numerical distribution of each variable’s IG, the variables selected were those with the largest information gain in relation to the outcome. These variables were distributed hierarchically on the tree, in decreasing order of IG.

Based on the above-mentioned criteria, the following sociodemographic variables were included in the model: age (< 20 years, 20-39, 40-59, ≥ 60), marital status (single, married or in a stable union, separated/divorced/widow or widower), and residence in the same municipality as the outpatient service conducting the follow-up (yes, no). The lifestyle variables and associated health problems included in the model were: sexual orientation (homosexual, heterosexual, bisexual, or information not recorded on the patient chart) and alcohol use (yes, no). The clinical variables were: in use of ART (yes, no), result of last viral load test (undetectable, detectable, not performed), hospitalizations at the service (yes, no), and number of pills (ART) taken per day (< 4; ≥ 4).

The variables from the database that did not obtain a significant IG and were thus not included in the model, although the literature suggests their relationship to retention in care, were: sex, race, schooling, illegal drug use, time since diagnosis, and presence of chronic diseases.

The study complied with the ethical guidelines for research in human subjects, as specified in Resolution n. 466/2012, including approval by the Institutional Review Board, of the Federal University of Paraíba (UFPB), under case review CAAE 41019115.7.0000.5188/2015.

Results

The study showed an overall retention rate of 68.5%. The sample was predominantly male (64.6%), single (45%), heterosexual (69.3%), with age ranging from 18 to 71 years and a mean of 42.3 years, predominantly in the 20-39-year bracket (60.4%). Only 37.7% lived in the same municipality as the health service responsible for the follow-up; 92.3% of users were in ART, 91.2% presented a detectable viral load (VL) at the start of outpatient clinical follow-up; 65% had an undetectable VL at the last test; and 53.1% were taking more than four pills a day.

The decision tree (Figure 1) showed 80.4% correct answers and the capacity to correctly classify 209 individuals according to the decision matrix (Table 1), presenting correct answers along the main diagonal and the errors identified outside of it.

Validation measures were also used to check the model’s quality (Table 2), the results of which showed good sensitivity (approximately 80%) for identifying individuals with probability of non-retention in care and positive likelihood ratio, indicating that the odds of identifying likely non-retention in care is four times higher when the model is applied. These results were the basis for the model’s validity in predicting the target outcome.

The tree allowed establishing 24 decision rules (Box 1) for the study sample. This highlights the variables’ importance as decisive for predicting retention of PLWA in health care.

The variables closest to the tree’s root or main node were those with the highest IG. This means that their results lead to greater or lesser likelihood of the outcome occurring. Thus, “use of ART” was the variable with the highest information gain, constituting the tree’s main node, branching to the internal nodes (formed by the other variables) and the leaves that contain the decision variable - likely versus unlikely retention in health care.

To illustrate the model’s applicability, its rules were applied to a clinical case (Box 2).

Discussion

The study’s sociodemographic profile corroborates the epidemiological pattern of HIV infection in Brazil and the world and indicates the predominance of males and heterosexuals ⁵5. Departamento de DST, Aids e Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Boletim Epidemiológico Aids e DST 2015; VI(3).^,88. Fiuza MLT, Lopes EM, Alexandre HO, Dantas PB, Galvão MTG, Pinheiro AKB. Adherence to antiretroviral treatment: comprehensive care based on the care model for chronic conditions. Esc Anna Nery Rev Enferm 2013; 17:740-8.^,1313. Stricker SM, Fox KA, Baggaley R, Negussie E, de Pee S, Grede N, et al. Retention in care and adherence to ART are critical elements of HIV care interventions. AIDS Behav 2016; 18 Suppl 5:S465-75..

Although gender was not considered in the model as a significant variable for predicting retention, an analysis of the proportion, by gender, in the study scenario (ratio of 1.8:1, i.e., 18 men for every ten women) corroborates the results of a study in another state of Northeast Brazil, reaffirming the increase in the number of cases in women, which the literature attributes to the increase in heterosexual transmission, among other factors ¹⁴14. Silva RAR, Silva RTS, Nascimento EGC, Gonçalves OP, Reis MM, Silva BCO. Perfil clínico-epidemiológico de adultos hiv-positivo atendidos em um hospital de Natal/RN. Rev Pesqui Cuid Fundam (Online) 2016; 8:4689-96..

As for age bracket, studies have suggested higher likelihood of retention with increasing age, since for every additional year in age, the risk of non-retention in care decreases by 4 to 5%. According to the model, younger individuals (under 20 years) run higher risk of non-retention, which the literature suggests may be due to the fact that young people experience greater difficulty in adapting their lifestyle to the new reality required by treatment, with changes in eating habits, the routine of taking medications, and other psychological issues related to living with the infection ¹²12. Medeiros LB. Fatores influentes no abandono do acompanhamento clínico ambulatorial por pessoas vivendo com HIV/Aids, João Pessoa - PB [Dissertação de Mestrado]. João Pessoa: Universidade Federal da Paraíba; 2016.^,1515. Schilkowsky LB, Portela MC, Sá MC. Factors associated with HIV/AIDS treatment dropouts in a special care unit in the city of Rio de Janeiro, RJ, Brazil. Rev Bras Epidemiol 2011; 14:187-97.^,1616. Yehia BR, Fleishman JA, Metlay JP, Korthuis PT, Agwu AL, Berry SA, et al. Comparing different measures of retention in outpatient HIV care. AIDS 2012; 26:1131-9.^,1717. Naguchi NEO. Adesão ao tratamento dos portadores de HIV/AIDS: compartilhando desafios [Monografia de Especialização]. Florianópolis: Universidade Federal de Santa Catarina; 2016..

As for marital status, a study in Santa Maria, Rio Grande do Sul State, Brazil, suggests that single individuals are more prone to ART dropout, considering the associations with social isolation, lack of social support, and treatment dropout in general ¹⁸18. Padoin SMM, Zuge SS, Santos EEP, Primeira MR, Aldrighi JD, Paula CC. Adesão à terapia antirretroviral para HIV/Aids. Cogitare Enferm 2013; 18:446-51.. Considering the intersection between ART adherence and retention in care, this result would contradict what the model proposes.

However, a study in Rio de Janeiro that specifically involved a male population (the predominant gender in the current study) mentions that in cases where serological status is undisclosed, having more persons residing in the same household can hinder regular taking of medications ¹⁹19. Moraes DCA, Oliveira RC, Costa SFG. Adherence of men living with HIV/AIDS to antiretroviral treatment. Esc Anna Nery Rev Enferm 2014; 18:676-81.. This finding could explain the higher retention of single and separated males and widowers in the model, considering the likelihood of these individuals living with fewer persons in the household and the close relationship between ART adherence and retention in care. However, other studies are needed to elucidate this issue and the factors that influence higher retention by these groups, regardless of disclosure of their serological status.

As for place of residence, the prevalence of individuals residing in other municipalities not only expresses the phenomenon of interiorization of HIV/AIDS in the state of Paraíba, but also calls attention to the negative influence of geographic barriers on the follow-up of health status in these individuals. It is known that the increase in infection in smaller, lower-income municipalities includes financial, housing, and transportation problems experienced by the inhabitants. These difficulties are further aggravated when the individual needs to travel away from home to obtain access to ART, follow-up appointments, and specific tests ¹⁶16. Yehia BR, Fleishman JA, Metlay JP, Korthuis PT, Agwu AL, Berry SA, et al. Comparing different measures of retention in outpatient HIV care. AIDS 2012; 26:1131-9.^,2020. Silva JVF, Nascimento JFJM, Rodrigues APRA. Fatores de não adesão ao tratamento antirretroviral: desafio de saúde pública. Cadernos de Graduação: Ciências Biológicas e da Saúde 2014; 2:165-75.^,2121. Souza CC, Mata LRF, Azevedo C, Gomes CRG, Cruz GECP, Toffano SEM. Interiorização do HIV/Aids no Brasil: um estudo epidemiológico. Rev Bras Ciênc Saúde 2013; 11:25-30., which can compromise retention in outpatient care, as indicated by the model.

As for lifestyle variables and related health problems, individuals with a history of alcohol use (abusive or otherwise) are known to have higher odds of non-adherence to ART and non-retention in care ¹⁶16. Yehia BR, Fleishman JA, Metlay JP, Korthuis PT, Agwu AL, Berry SA, et al. Comparing different measures of retention in outpatient HIV care. AIDS 2012; 26:1131-9.^,1818. Padoin SMM, Zuge SS, Santos EEP, Primeira MR, Aldrighi JD, Paula CC. Adesão à terapia antirretroviral para HIV/Aids. Cogitare Enferm 2013; 18:446-51..

As for clinical variables, the model highlighted ART use as having the highest information gain for retention. According to a previous study ¹¹11. Medeiros LB, Trigueiro DRSG, Silva DM, Nascimento JA, Monroe AA, Nogueira JA, et al. Integração entre serviços de saúde no cuidado às pessoas vivendo com aids: uma abordagem utilizando árvore de decisão. Ciênc Saúde Coletiva 2016; 21:543-552., the risk of likely non-retention decreases by approximately 61% when individuals are in ART. This reaffirms the close relationship between these two pillars in the cascade of care ¹³13. Stricker SM, Fox KA, Baggaley R, Negussie E, de Pee S, Grede N, et al. Retention in care and adherence to ART are critical elements of HIV care interventions. AIDS Behav 2016; 18 Suppl 5:S465-75.. Considering that the Brazilian Ministry of Health currently recommends initiating ART immediately after diagnosis, regardless of CD4+ T lymphocyte count, knowing the interrelationship between ART adherence and retention in care ⁴4. Nosyk B, Montaner JS, Colley G, Lima VD, Chan K, Heath K, et al. The cascade of HIV care in British Columbia, Canada, 1996-2011: a population-based retrospective cohort study. Lancet Infect Dis 2014; 14:40-9.^,55. Departamento de DST, Aids e Hepatites Virais, Secretaria de Vigilância em Saúde, Ministério da Saúde. Boletim Epidemiológico Aids e DST 2015; VI(3). and implications for inadequate retention and adherence for individual and population-based results ¹³13. Stricker SM, Fox KA, Baggaley R, Negussie E, de Pee S, Grede N, et al. Retention in care and adherence to ART are critical elements of HIV care interventions. AIDS Behav 2016; 18 Suppl 5:S465-75., this study’s identification of 92.3% adherence to ART and only 68.5% retention in care emphasizes the need for interventions capable of improving retention.

The model showed that individuals whose last viral load was undetectable tended to be retained in care. For this variable, there was an increase in the percentage of individuals with undetectable viral load, from the start of treatment (8.8%) to the most recent viral load test (65%), which can be attributed to adequate use of ART. Although positive, this percentage is still insufficient to reach the 90-90-90 targets.

According to the study, the use of four or more pills a day contributed to retention in care. Reporting on the interrelationship between ART and retention in care, as mentioned previously, this finding contradicts some studies that suggest that the more medications in the treatment regimen, the lower the adherence and thus the lower the retention in care ⁸8. Fiuza MLT, Lopes EM, Alexandre HO, Dantas PB, Galvão MTG, Pinheiro AKB. Adherence to antiretroviral treatment: comprehensive care based on the care model for chronic conditions. Esc Anna Nery Rev Enferm 2013; 17:740-8.^,1919. Moraes DCA, Oliveira RC, Costa SFG. Adherence of men living with HIV/AIDS to antiretroviral treatment. Esc Anna Nery Rev Enferm 2014; 18:676-81.^,2222. Oliveira EF, Paes MSL. Adesão ao tratamento antirretroviral de pessoas com HIV/Aids. Revista Enfermagem Integrada 2013; 6:1154-66..

This emphasizes the need for more frequent follow-up of drug-drug interactions and side effects when the regimen contains more pills, a possible contributing factor to closer contact between users and the service responsible for their clinical management ¹⁵15. Schilkowsky LB, Portela MC, Sá MC. Factors associated with HIV/AIDS treatment dropouts in a special care unit in the city of Rio de Janeiro, RJ, Brazil. Rev Bras Epidemiol 2011; 14:187-97. and retention, in case the service is active from this perspective. The team should thus value care with the expectation of simplifying the regimens and supporting the adaptations to the routine, which requires closer linkage between the health service and PLWA to strengthen retention ²³23. Programa Nacional de DST e Aids, Secretaria de Vigilância em Saúde, Ministério da Saúde. Diretrizes para o fortalecimento das ações de adesão ao tratamento para pessoas que vivem com HIV e Aids. Brasília: Ministério da Saúde; 2007.^,2424. Tietzmann DC, Béria JU, Santos GM, Mallmann DA, Trombini ES, Schermann LB. Prevalências de adesão à terapia antirretroviral e fatores associados em pacientes adultos de três centros urbanos do Sul do Brasil. Aletheia 2013; 41:154-63.^,2525. Moraes DCA, Oliveira RC, Motta MCS, Ferreira OLC, Andrade MS. Terapia antirretroviral: a associação entre o conhecimento e a adesão. Rev Pesqui Cuid Fundam (Online) 2015; 7:3563-73..

The retention rate of 68.5% was higher than the Brazilian national average, which is 66% (among 83% of HIV-positive diagnosed individuals), according to the country’s latest HIV/AIDS epidemiological bulletin ⁴4. Nosyk B, Montaner JS, Colley G, Lima VD, Chan K, Heath K, et al. The cascade of HIV care in British Columbia, Canada, 1996-2011: a population-based retrospective cohort study. Lancet Infect Dis 2014; 14:40-9.^,1313. Stricker SM, Fox KA, Baggaley R, Negussie E, de Pee S, Grede N, et al. Retention in care and adherence to ART are critical elements of HIV care interventions. AIDS Behav 2016; 18 Suppl 5:S465-75.. This scenario underscores the relevance of planning activities and local strategies capable of improving all the pillars in the cascade of care, especially those related to ART adherence and retention in care to achieve the targets ²⁶26. Montaner JS. Treatment as prevention: toward an AIDS-free generation. Top Antivir Med 2013; 21:110-14.^,2727. Silva JAG, Dourado I, Brito AM, Silva CAL. Fatores associados à não adesão aos antirretrovirais em adultos com AIDS nos seis primeiros meses da terapia em Salvador, Bahia, Brasil. Cad Saúde Pública 2015; 31:1188-98.. Early identification of triggering factors for retention and non-retention is essential for health professionals in the specialized service to act by maximizing facilitators of retention and reducing barriers ²⁸28. Figueiredo LA, Lopes LM, Magnabosco GT, Andrade RLP, Farias MF, Goulart VC, et al. Provision of health care actions and services for the management of HIV/AIDS from the users' perspective. Rev Esc Enferm USP 2014; 48:1026-34..

It is important to extend the care provided by the health care team to include the patients’ families, considering their role in this context as a motivator for PLWA to proceed with treatment and improve their quality of life ²⁹29. Pereira PE, Espírito SCC, Tosoli GAM, Santos EI, Oliveira, Pontes APM. Adherence to antiretroviral therapy and its representations for people living with HIV/AIDS. Esc Anna Nery Rev Enferm 2014; 18:32-40..

Final remarks

Retention in health care for PLWA has gained attention thanks to the cascade of continuing care, so it is important to know the factors that help promote retention in order to achieve the goals of controlling the epidemic.

Despite the study’s difficulties in approaching the topic of “retention” in the Brazilian literature for discussion of the results in the Brazilian reality, the close relationship between this theme, and other inherent phenomena in treatment and care of PLWA (such as the health team-patient bond and adherence to treatment and care), the findings indicate that retention in care is influenced by sociodemographic factors, lifestyle, and clinical variables, which merit attention in order to promote effective intervention by health services in promoting retention.

In this sense, the decision tree created here is a model that can support health professionals at the specialized service to identify aspects by which it is possible to intervene in the bond and retention (follow-up) of PLWA, supporting the identification and planning of interventions to prevent non-retention. For example, the model can be used in nursing consultations to improved nursing care as an important component of this specialized care as a whole.

A limitation to the study, for purposes of generalization, was the fact that the sample came from a single service, although this service is the reference for specialized HIV/AIDS care in the state of Paraíba. Other limitations include the use of a secondary data source, where the richness of the results depends on the records’ quality, and the fact that the model did not consider inherent aspects of the service’s organization.

Thus, considering the importance of retention in care in the coming years, especially for reaching the UNAIDS targets for controlling the epidemic, other studies are needed on individual, social, and programmatic variables to produce a wider picture of factors involved in retention in care for persons living with HIV/AIDS.

References

Publication Dates

History