Analysis of the clinical complexity of people living with HIV based on the GeSIDA stratification system

Abstract

Introduction

Thanks to improvements in antiretroviral therapy (ART), HIV infection has become a chronic disease in developed countries. The stratification of the complexity of chronic patients is a procedure used in other diseases that allows for identifying the risk level of patients and organizing care based on their characteristics.

The objective of this study is to analyze the complexity of a large cohort of people living with HIV (PLWH) on ART and in regular follow-up consultations.

Methods

Descriptive study of complexity according to the GeSIDA stratification scale (GSS), which includes social aspects, risk behaviors, comorbidities, psychological and cognitive status, and viral control in a cohort of PLWH on ART. Patients were classified as extreme, high, medium, or low complexity based on their GSS score.

Results

A total of 1127 patients were included, 83.7% male, with a mean age of 48 years (SD: 13.10). 65.3% had low complexity, 18.5% medium, 8.3% high, and 7.9% extreme complexity. The most determining factors for higher complexity on the GSS were lower social support, drug use and risky sexual behaviors. Female sex, lower educational level, history of AIDS, a lower CD4 count and a lower CD4/CD8 ratio were associated with higher complexity.

Conclusions

The GSS identifies most PLWH on ART as low complexity. There are clinical factors associated with greater complexity, suggesting that there are subgroups of PLWH who should receive differentiated clinical care to prevent complications and improve their quality of life.

Resumen

Introducción

Gracias a las mejoras del tratamiento antirretroviral (TAR), la infección por VIH se ha convertido en nuestro medio en una enfermedad crónica. La estratificación de la complejidad de los pacientes crónicos es un procedimiento usado en otras enfermedades, que permite identificar el grado de riesgo de los pacientes y organizar la asistencia en función de sus características.

El objetivo de este estudio es analizar la complejidad de una serie de personas que viven con VIH (PVVIH) en TAR y en seguimiento regular en consultas.

Métodos

Estudio descriptivo de la complejidad según la escala de estratificación la complejidad de GeSIDA (EECG), que incluye aspectos sociales, conductas de riesgo, comorbilidades, estado psicológico y cognitivo y control virológico, en una serie de PVVIH en TAR. Los pacientes fueron clasificados como de complejidad extrema, alta, media o baja, en función de la puntuación en la EECG.

Resultados

Se incluyeron 1127 pacientes, 83,7% varones, con edad media 48 años (DE:13,10). El 65,3% presentaba una complejidad baja, el 18,5% media, el 8,3% alta y 7,9% extrema. Los factores más determinantes de una mayor complejidad en la EECG son el menor apoyo social, el consumo de drogas y las conductas sexuales de riesgo. Otros factores identificados con mayor grado de complejidad fueron el sexo femenino, el menor nivel educativo, tener historia de sida, una cifra menor de linfocitos CD4 y un menor cociente CD4/CD8.

Conclusiones

La EECG identifica a la mayoría de las PVVIH en TAR como de baja complejidad; existen factores clínicos asociados a una mayor complejidad, que sugiere que existen subgrupos de PVVIH en los que se debería proporcionar una atención clínica diferenciada para prevenir complicaciones y mejorar su calidad de vida.

Introduction

HIV infection has become a chronic disease in developed countries thanks to the improvements in ART.^1,2 Recent studies suggest that the life expectancy of people living with HIV (PLWH), especially those diagnosed in recent years with a CD4 count ≥350 cells/ml at diagnosis, is similar to that of the general population.³ There are approximately 150,000 PLWH in Spain, or about 7.71 cases per 100,000 inhabitants.⁴ In addition to the follow-up required for the HIV infection itself, there is a higher prevalence of comorbidities compared to the general population. This poses a challenge for healthcare systems.⁵

Stratification is a recommended strategy for managing chronic patients as it allows classifying patients based on the complexity of their condition and, accordingly, planning resources for their care.^6,7 A panel of experts linked to GeSIDA has recently developed a stratification system to optimize the medical care of PLWH based on their complexity level to provide homogeneous and efficient care.^8,9 The utility of this tool has not been widely verified in clinical practice, and some authors have found limitations with this stratification system.¹⁰ If this model could adequately discriminate the complexity of PLWH, practical actions in the management of these patients could be deduced, and care could be provided with varying intensity based on the estimated complexity.

The main objective of our study, therefore, is to analyze the clinical complexity of PLWH according to this stratification model and understand the clinical characteristics of patients based on their estimated complexity level.

Methods

This was an observational, cross-sectional study that included PLWH aged 18 and over who were followed in outpatient clinics between September 1, 2023, and February 1, 2024. Patients were consecutively included after attending their consultation, either in person or by phone. Using the information collected from their electronic medical records (EMR) from their last visit, patients were classified based on their complexity level using the GSS. This scale consists of seven items, which include clinical and psychosocial aspects with various response options, each with an assigned score. Based on the results, patients were categorized into four complexity levels: low (score ≤ 5), medium (>5 and ≤15), high (<15 and <23), and extreme (≥23), with 99 being the maximum possible score. The full questionnaire is shown in Annex 1.

All clinical information was obtained from interviews with patients and their electronic medical records (EMRs). Behavioral questions (substance use and usage patterns) were explicitly asked for the period since the last consultation. Dependency was eventually confirmed with information provided by relatives or close individuals who could offer relevant details. Risky sexual behaviors were defined as unprotected sex with individuals outside their stable partner. Other variables collected from the EMR included age, gender, educational level, country of origin, previous AIDS diagnosis, time since HIV diagnosis, current CD4 count, current ART regimen and type of care received (in-person, phone consultation or unscheduled without an appointment). All information obtained from the EMR was coded and recorded in an anonymized database. The study protocol was approved by the Clinical Research Ethics Committee (CEIC/HCSC ref. 24/342-E).

A univariate analysis of the sample was conducted. Qualitative variables were described using frequency distributions, and quantitative variables with measures of central tendency (mean or median) and dispersion (standard deviation or interquartile range), depending on the distribution of the variables. The Chi-square test or Fisher's exact test was used to compare categorical variables when appropriate, and the Wilcoxon rank-sum test for quantitative variables. Measures of association were expressed as odds ratios (OR) with a 95% confidence interval (CI). A two-tailed p-value < 0.05 was considered statistically significant.

Results

A total of 1127 PLWH were included in the study, of which 83.7% (N = 943) were men and 14.9% (N = 168) were women (of which 15 were transgender, 1.3%). The average age was 48 years (SD: 13.10). The youngest patient included in the study was 20 years old, and the oldest was 93 years old. The country of origin was recorded for 91.57% of cases (N = 1032). Of the patients analyzed, 53.1% (N = 598) were Spanish and 38.5% (N = 434) were foreign, with the largest group (347 patients, 79.95%) coming from Latin America. The average CD4 count was 813.13 cells/μL (SD: 377.51). Information on previous AIDS diagnosis was available for 779 cases (69.12%), of which 29.26% (N = 228) had experienced an AIDS-defining illness at some point in their progression. Regarding antiretroviral therapy (ART), the most commonly used combination was Dolutegravir/Lamivudine in 425 patients (37.7%), followed by Bictegravir/FTC/TAF in 358 patients (31.76%) and Dolutegravir/Rilpivirine in 93 patients (8.3%). Sixty-two patients (5.6%) were receiving intramuscular Cabotegravir and Rilpivirine injections. Table 1 summarizes the most relevant baseline characteristics of the patients included in the study.

Table 1.

Baseline characteristics.

Global sample size N (%)	1127 (100)
Gender N (%)
Man	943 (83.67)
Transgender man	1 (0.08)
Woman	168 (14.90)
Transgender woman	15 (1.33)
Age (mean, SD)	48 (13.10)
Region of origin N (%)
Spain	598 (35.1)
Other	434 (38.5)
LT-CD4 (mean, SD)	813.13 cel/μL (377.51)
Previous AIDS diagnosis N (%)	228 (29.26)
Current ART N (%)
Dolutegravir/lamivudine	425 (37.71)
Bictegravir/tenofovir/emtricitabine	358 (31.76)
Dolutegravir/rilpivirine	93 (8.25)
Cabotegravir/rilpivirine	58 (5.14)
Darunavir/cobicistat/tenofovir/emtricitabine	45 (3.99)
Rilpivirine/tenofovir/emtricitabine	26 (2.30)
Elvitegravir/cobicistat/tenofovir/emtricitabine	6 (0.53)
Not recorded	1 (0.1)

Of the 1127 PLWH studied, 65.3% (N = 737) had low complexity; 18.5% (N = 208) had medium complexity; 8.3% (N = 93) had high complexity; and 7.9% (N = 89) had extreme complexity. A comparison was made by grouping patients with extreme and high complexity (>15 points) and comparing them with patients with low complexity (≤5 points). The results of the GSS are reflected in Table 2.

Table 2.

Complexity of PLHIV on the GeSIDA complexity stratification scale: univariate analysis.

Variable	Complexity	Value	N (%)	p
Social support	Low	Deficient	0 (0)	<0.001
		Sufficient	737 (100)
	High-extreme	Deficient	78 (43.1)
		Sufficient	103 (56.9)
Drug use	Low	Yes	0 (0)	<0.001
		No	887 (100)
	High-extreme	Yes	65 (35.9)
		No	116 (64.1)
Risky sexual behaviors	Low	Yes	201 (27.3)	<0.001
		No	536 (72.7)
		Frequent chemsex	0 (0)
	High-extreme	Yes	49 (27.1)
		No	106 (58.6)
		Frequent chemsex	26 (14.4)
Inmuno-virological status	Low	ART failure/toxicity/intolerance	0 (0)	<0.001
		ART initiation < 6 months	0 (0)
		Undetectable VL and good tolerance to ART	737 (100)
	High-extreme	ART failure/toxicity/intolerance	17 (9.4)
		ART initiation < 6 months	20 (11.0)
		Undetectable VL and good tolerance to ART	144 (79.6)
Comorbidities	Low	Active cancer, neurological sequelae, transplant, severe COPD and/or advanced chronic heart failure	0 (0)	<0.001
		Active disease or limiting sequelae	0 (0)
		Active disease or non-limiting sequelae	126 (17.1)
		No	611 (82.9)
	High-extreme	Active cancer, neurological sequelae, transplant, severe COPD and/or advanced chronic heart failure	6 (3.3)
		Active disease or limiting sequelae	53 (29.3)
		Active disease or non-limiting sequelae	31 (17.1)
		No	91 (50.3)
Psychological and cognitive status	Low	Limiting psychiatric diagnosis and/or cognitive impairment	0 (0)	<0.001
		Non-limiting psychiatric diagnosis and/or cognitive impairment	32 (4.3)
		Good mental health	705 (95.7)
	High-extreme	Limiting psychiatric diagnosis and/or cognitive impairment	16 (8.8)
		Non-limiting psychiatric diagnosis and/or cognitive impairment	69 (38.1)
		Good mental health	96 (53.0)

Regarding the other variables analyzed in the study, the results are summarized in Table 3 and Fig. 1.

Table 3.

Other differences between low and high-extreme complexity patients.

Variable	Complexity	N (%)	p
Non-native Spaniard	Low	266 (39.5)	0.102
	High-extreme	79 (46.7)
Previous AIDS diagnosis	Low	120 (23.6)	<0.001
	High-extreme	57 (44.5)
Primary-level education	Low	22 (19.6)	0.024
	High-extreme	44 (33)
Rilpivirine ART	Low	111 (15.1)	0.023
	High-extreme	8 (4.4)
Women	Low	93 (12.6)	<0.001
	High-extreme	36 (19.9)
Unscheduled visit	Low	83 (11.7)	<0.001
	High-extreme	27 (15.7)

Fig. 1.

Immunovirological status vs. low and high-extreme complexity. The first chart represents the differences in the total CD4+ T lymphocyte count. The second chart represents the differences in the CD4/CD8 ratio.

The main differences found based on complexity include the following:

• 1.

  Women presented higher complexity due to less social support (present in 118 [84.9%] women vs. 722 [92.7%] men, p < 0.004), more comorbidities (43 [30.9%] vs. 173 [22.2%], p = 0.034), and neuropsychiatric conditions (33 [23.7%] vs. 84 [10.8%], p < 0.001).

• 2.

  Lower educational level was associated with higher complexity. These patients had greater complexity due to less social support (present in 107 [81.1%] patients with a lower education level vs. 103 [92%], p = 0.023).

• 3.

  Regarding ART, patients with lower complexity were more frequently treated with regimens based on Rilpivirine (Dolutegravir/Rilpivirine, 65 patients with low complexity [8.8%] vs 3 with high complexity [1.7%], p = 0.071; and Cabotegravir/Rilpivirine, 46 with low complexity [6.2%] vs 5 with high complexity [2.31%], p = 0.023). Similarly, high genetic barrier ART regimens, such as Darunavir/Cobicistat/TAF/FTC, were used more in patients with high complexity (14 cases [7.7%] vs. low complexity, 25 [3.39%], p = 0.071).

• 4.

  The type of consultation varied depending on the complexity level. Patients with lower complexity were more frequently seen through phone consultations (125/709, 17.6% vs 12/171, 7.0%), while a higher percentage of more complex patients required an unscheduled consultation.

• 5.

  Patients with higher complexity had worse immunological status, defined by a lower absolute CD4 count (median 816, [IQR 618–1047], vs 678 [393–928], p < 0.001) and a lower CD4/CD8 ratio (low complexity, CD4/CD8 ratio 0.93 [IQR 0.65–1.25] vs high complexity 0.72 [IQR 0.44–1.01], p < 0.001). The viral load (VL) was detectable in 10.8% of patients (N = 98), which was significantly more frequent in patients with high and extreme complexity compared to those with low complexity (28.7% [N = 51] vs. 6.44% [N = 47], p < 0.001). Among those with detectable VL, the median VL was 399 (copies/mL) (IQR 100–2000), which was higher in complex patients (610 [IQR 140–13429] vs. 180 [IQR 88–880], p = 0.050). Additionally, patients with low complexity were less likely to have a previous AIDS diagnosis.

Discussion

Our study describes an extensive series of PLWH on ART, composed predominantly of men between 45 and 50 years of age. The majority of the PLWH studied present a low level of complexity, and only 16.2% exhibited a high or extreme degree of complexity, according to the GSS. We believe that this scale is highly useful and may contribute to improving care for these individuals.

We have identified that women exhibit a greater degree of complexity, which suggests that the women in our HIV-positive cohort face additional challenges in terms of health and treatment, as well as receiving less social support. Barriers within the European healthcare system for women have been described, which could explain a higher proportion of late diagnoses,¹¹ reduced access to PrEP¹² or less access to harm reduction strategies among female drug users.¹³ In our series, women also present a higher number of comorbidities—as has been described in other series—especially in those over 60 years old.¹⁴ Other factors associated with increased complexity in women in our series are lower social support and, although this study did not yield statistically significant results, other series have also described higher rates of treatment failure related to poorer adherence, which is associated with increased rates of depression and alcohol consumption.¹⁵ Therefore, women with HIV could be a particularly vulnerable population to higher clinical complexity for multiple reasons. From this information, it could be deduced that closer monitoring of comorbidities—especially psychiatric ones—and treatment adherence would be advisable for women.

In our series, we have found that PLWH with lower educational levels show greater complexity, especially those who do not attain higher education. These differences could reflect barriers in access to healthcare resources and, more generally, in obtaining information or understanding the treatment. It has been described that low educational attainment and greater stigma may be partly responsible for reduced participation in healthcare services and poorer HIV care.¹⁶ PLWH with lower educational levels also have less social support. Educational level is generally associated with higher poverty, which in people with HIV—especially in women—is linked to increased mortality.¹⁷ These findings highlight the importance of the social determinants of disease in this patient group. Therefore, PLWH with low educational levels could benefit from specialized management that emphasizes social needs, treatment adherence, and the monitoring of comorbidities.

The GSS identifies a population of PLWH with high clinical complexity that shows a notable association with immunological status. In our study, we have identified that those patients with a higher degree of complexity on the scale have a worse immunological status, defined by a lower CD4 lymphocyte count and a lower CD4/CD8 ratio. A relationship between this CD4/CD8 ratio and the prognosis of HIV infection has been described,¹⁸ especially due to its association with certain comorbidities related to HIV infection, such as cancer¹⁹ or cardiovascular disease,²⁰ which may reflect a state of persistent immune activation. Although the GSS is a scale based solely on clinical data, this association between complexity and immunological dysfunction could indicate a greater vulnerability to HIV complications. This poor response might be related to a reduced effect of ART, perhaps associated with poorer adherence to treatment; further studies are needed to clarify this relationship and to investigate the effect of improvements in complexity on the CD4/CD8 ratio.

The types of ART used by the cohort participants reflect treatment decisions tailored to each patient: patients with high complexity tend to receive regimens with a high genetic barrier, such as Darunavir/Cobicistat/TAF/FTC, whereas those with low complexity more often use Rilpivirine-based treatments. This can be interpreted as a medical strategy to adapt treatment to the specific needs and risks of each group; patients with greater complexity, who may have issues with treatment resistance or a delicate immunological status, are considered to require more robust treatments.

One of the objectives of this tool is to be useful in the care of PLWH, improving their care and eventually reducing morbidity and mortality while enhancing their quality of life. Patients with a lower degree of complexity do not require the same level of care as those with higher complexity, and this may influence the frequency or type of consultations. In our study, we found that patients with low complexity had more telephone visits than those with high complexity. This could suggest that patients with low complexity might be seen in specialized consultations at longer intervals (for example, once a year), with support and follow-up procedures on demand—such as via email or as-needed consultations—being established; eventually, primary care could assume the follow-up of comorbidities or other medical issues in patients with low complexity, as has been suggested.²¹

Our study has some limitations; among them, it is worth noting that although we believe the study population is representative of the reality of PLWH on ART in specialized units in Spain, a selection bias cannot be ruled out, since the participants were PLWH who were consecutively seen in the clinic, even though they represent 42% of the 2600 PLWH in our service. This may be considered a limitation of the study, as those not included might have poorer adherence or less frequent follow-up, which could eventually imply a different degree of complexity.

Furthermore, this study only provides information on the level of complexity of the sample at the time the data were collected, acknowledging that complexity is a dynamic dimension. The near absence of similar studies published to date makes it difficult to compare these results with those from other cohorts, although a recent study¹⁰ involving 94 PLWH at another hospital in Madrid reported similar sociodemographic characteristics (76.6% cisgender men with a median age of 41 years and a median CD4+ count of 785 cells/mm³, with a high proportion of foreign individuals) and a level of complexity (78.7% with low complexity) that is lower than what is observed in our sample.

Conclusions

In conclusion, the majority of the PLWH (65%) studied are considered to have low complexity according to the GSS. For the complex patients, it appears paramount to establish pathways to address situations of low social support and harmful substance use, as these are the two modifiable risk factors that increase most in frequency with rising complexity. PLWH with lower educational levels and women are generally more clinically complex groups. Regardless of the level of complexity, it is essential to conduct a thorough history regarding risky sexual behaviors at each consultation, given their high frequency across all groups, with the appropriate recommendations for screening for sexually transmitted infections as indicated. Further studies are needed to more precisely determine the determinants of complexity in PLWH in our country.

Authors’ contributions

JBC wrote the manuscript and conducted the literature search. AVC performed the statistical analysis. MJNO, NCC, JPSM, AMG, VVP, COM and VE were responsible for patient inclusion and their follow-up in the consultations. Additionally, VE supervised the development of the study and the writing of the manuscript.

Ethical considerations

The described work has been carried out in accordance with the World Medical Association's Declaration of Helsinki for experiments involving human beings; and in accordance with uniform requirements for manuscripts submitted to biomedical journals.

All information obtained from the electronic medical records was coded and recorded in an anonymized database. The study protocol was approved by the Clinical Research Ethics Committee (CEIC/HCSC ref. 24/342-E). Due to the observational and retrospective design of the study, we obtained a waiver of informed consent from the ethics committee.

Use of artificial intelligence

No artificial intelligence–based tools or technologies were used in the preparation, writing, analysis, or editing of this manuscript.

Conflicts of interest

No institution has provided financial funding for the completion of this work.

Acknowledgments

We would like to thank the staff at IdiSSC for their contribution to this work in collecting the data, and all the patients who agreed to participate in the study. Thank you for your generosity in helping improve healthcare for PLHIV.