Long-term outcomes of patients affected by Guillain-Barré syndrome in Colombia after the Zika virus epidemic

Abstract

Background:

Guillain-Barré Syndrome (GBS) can lead to significant functional impairments, yet little is understood about the recovery phase and long-term consequences for patients in low- and medium-income countries.

Objective:

To evaluate the functional status and identify factors influencing outcomes among patients with GBS in Colombia.

Methods:

Between 2016 and 2020, telephone interviews were conducted with GBS patients enrolled in the Neuroviruses Emerging in the Americas Study. The investigation encompassed access to health services and functional status assessments, utilizing the modified Rankin Scale (mRS), GBS Disability Score (GDS), Barthel Index (BI), and International Classification of Functioning (ICF). Univariate analysis, principal component analysis, linear discriminant analysis, and linear regression were employed to explore factors influencing functional status.

Results:

Forty-five patients (mean age=50[±22] years) with a median time from diagnosis of 28 months (IQR=9–34) were included. Notably, 22% and 16% of patients did not receive rehabilitation services during the acute episode and post-discharge, respectively. Most patients demonstrated independence in basic daily activities (median BI=100, IQR=77.5–100), improved disability as the median mRS at follow-up was lower than at onset ( 1 [IQR=0–3] vs. 4.5 [IQR=4–5], p<0.001), and had better ability to walk without assistance (median GDS=2, IQR=0–2). A shorter period from disease onset to interview was associated with worse mRS (p=0.015) and ICF (p=0.019). Negative outcomes on GDS and ICF were linked to low socioeconomic status, ICF to the severity of weakness at onset, and BI to an older age.

Conclusions:

This study underscores that the functional recovery of GBS patients in Colombia is influenced not only by the natural course of the disease but also by socioeconomic factors, emphasizing the crucial role of social determinants of health.

INTRODUCTION

Guillain-Barré Syndrome (GBS) is an immune-mediated polyneuropathy characterized by acute flaccid paralysis affecting the extremities, facial muscles, and respiratory muscles, with diverse clinical variants (1). Its course follows a monophasic trajectory, encompassing a progressing phase reaching nadir within four weeks from symptom onset. Afterward, a plateau phase, lasting days to months, is observed, leading to subsequent recovery and potential residual disability (1). Approximately 80% of patients regain unaided walking ability six months post-onset (2,3). However, individuals with severe disease and prolonged mechanical ventilation may endure lasting disability (4,5). Predictors of poor outcomes in GBS may include advanced age, preceding diarrhea, dysautonomia, bulbar nerve involvement, severe disability at onset, and mechanical ventilation requirement (1,2). While predictive tools like the modified Erasmus GBS Outcome Score exist (3), their applicability varies across different populations (7–9). The International GBS Outcome Study revealed global variations in acute complications and long-term motor impairment, emphasizing worse outcomes in resource-limited countries (10). Disparities in genetic, microbiological, and demographic factors contribute to this variability. Despite these challenges, a substantial portion of GBS patients in low-income countries receive only supportive care, potentially influencing disease progression negatively (10,11). The unique biological and social factors in each region necessitate an understanding of the local behavior of GBS and its sequelae.

In Colombia, the Zika virus epidemic coincided with a surge in GBS cases (12–14). Subsequent efforts aimed at delineating the local epidemiology, clinical features (15), and outcomes (16) of GBS patients in Colombia, including a study that found GBS patients likely associated with Zika virus infection exhibited higher long-term disability and depression compared to healthy controls (16). Despite the transition of the epidemic of Zika to an endemic phase, GBS incidence persisted at a baseline rate. This study investigates the long-term outcomes of GBS patients in Colombia following the end of the Zika virus epidemic. It sheds light on specific factors impeding the recovery and rehabilitation of GBS patients in resource-limited countries such as Colombia.

METHODS

Study design and population

A retrospective cohort study was conducted within the Neuroinfections Emerging in the Americas Study (NEAS) framework, a multicenter initiative established during the Zika epidemic in Colombia starting in January 2016 and continuing to the present day (17). The NEAS encompassed patients newly diagnosed with GBS across 11 hospitals in seven cities throughout Colombia. Comprehensive clinical, epidemiological, and biological data were systematically collected at the time of enrollment and subsequently every week for up to a month or until the point of hospital discharge. Patients who had completed the NEAS protocols (14) and neurological examination by NEAS clinicians and had consented to clinical follow-up were included in the survey. Trained researchers at each participating hospital entered the data into REDCap (Research Electronic Data Capture, Vanderbilt University, Nashville, Tennessee, USA), a secure platform hosted at Johns Hopkins University School of Medicine. Eligibility criteria for inclusion in the study were based on the Brighton criteria for GBS, with a diagnostic certainty level ranging from 1 to 3, as outlined in the NEAS database (18).

Data collection

Baseline variables were extracted from the NEAS database, including demographics, clinical characteristics at diagnosis, and modified Rankin Scale (mRS) scores. Socioeconomic status was assessed using the social strata system defined by the Colombian National Statistics Agency, classifying housing conditions and the surrounding environment on an ordinal scale from 1 to 6. On this scale, 1 represents the poorest material conditions, while 6 signifies higher economic resources. Each stratum reflects the degree of access to public services such as health, education, and recreation (19). To evaluate clinical outcomes during follow-up, eligible participants were contacted via telephone and interviewed by one of three researchers (DOAG, DZL, and DOM) using a pre-designed structured questionnaire (Supplementary material). The questionnaire explored patients’ experiences with the rehabilitation process during hospitalization and post-discharge. Additionally, the patient’s functional status was assessed using mRS (20), GBS Disability Score (GDS) (5), Barthel Index (BI) (21), and International Classification of Functioning core set (ICF) (22). For the latter, we evaluated seven features of body function: one of body structure, nine of activities and participation, and three of environmental factors. Consequently, 20 features were evaluated in adults and 19 in children, excluding “keeping and maintaining a job” as it was deemed inappropriate for children. The percentage functioning score was determined by dividing the number of unimpaired functional features by the total number of features evaluated and multiplying by 100. A higher percentage indicates better functioning. Positive outcomes were defined as a score of 0/1 in mRS, 0/1 in GDS, 100 in BI, and equal to or greater than 70% in ICF, while the remaining were considered negative.

Statistical Analysis

For the descriptive analysis, categorical data were presented using absolute and relative frequencies, while normally distributed variables were summarized using means and standard deviations. Skewed variables were described using medians and ranges. In multivariate descriptive statistics, Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) were employed, with the latter specifically applied to each functional scale. To compare categorical variables, the chi-squared test was used, and for quantitative variables, we employed the t-test or the non-parametric rank sum test when the data did not follow a normal distribution. Statistical significance was set at p < 0.05. The Gardner-Altman estimation plot was employed to visually illustrate the difference in mRS and GDS scores between the onset and the time of the interview (23). A linear regression model was used to explore the association between International Classification of Functioning (ICF) scores at follow-up and baseline variables,. All data analyses were conducted using Epidat version 4.2 (24) and R version 4.0.3 (25).

Ethical considerations

The study was approved by the Ethical Review Board at the Universidad del Valle (ethics committee approval #034–016). All participants of the NEAS provided written informed consent to participate in the study at the time of diagnosis. Participation in this study was voluntary, and consent was obtained verbally before the interviews.

RESULTS

Patient characteristics at baseline and follow-up

One hundred and five GBS patients enrolled in the NEAS registry between 2016 and June 2020, who fulfilled the criteria established for the study, were selected for the phone follow-up interview survey. Of these, 51 were reachable by telephone, and 45 patients (43% of all GBS patients) were included in the analysis (Figure 1). The sociodemographic and clinical characteristics of the study participants are detailed in Table 1. No significant differences were observed between the included and excluded patients regarding gender, age, socioeconomic status, or mRS scores at onset (Supplementary Material, sTable1). However, those included in the analysis had a shorter time from disease onset (28 vs. 35 months) and were more frequently recruited in Barranquilla than those not included. The cohort consisted predominantly of males (male-to-female ratio of 1.8), with ages at follow-up ranging from 4 to 96 years. Notably, children aged <18 years and adults aged >65 years represented 9% (n= 4) and 22% (n= 10) of the included patients, respectively. The majority of patients (87%) belonged to the lowest socioeconomic strata (1–3), were married (39% of adults), and were cared for by their partner/spouse (42%). Acute Inflammatory Demyelinating Polyradiculoneuropathy (AIDP) was the most common electrophysiological pattern (57%), and most patients (62%) reported preceding infectious symptoms. Most patients (32/45, 71%) underwent a microbiological assessment to assess infectious risk factors at the time of diagnosis. Microbiological studies at the time of diagnosis revealed serological findings consistent with recent C. jejuni infection in seven out of 32 (22%) patients, M. pneumoniae in four out of 29 (14%), recent flavivirus in two out of 32 (6%) cases, and recent Chikungunya infection in one of 29 cases (3%). All patients tested negative for Zika (n=32) and Chikungunya using PCR (n=32), and for Hepatitis E (n=26), Epstein-Barr Virus (n=28), Cytomegalovirus (n=29), and Varicella Zoster Virus (n=29) using serology. Most patients received treatment, although five (11%) did not undergo Intravenous Immunoglobulin (IVIG) or plasma exchange. During the acute phase of GBS, the majority experienced severe disability (82% mRS 4 and 5, 76% GDS 4 and 5), with 9% requiring invasive mechanical ventilation. However, a significant proportion (78%) underwent physical rehabilitation during hospitalization (Table 1).

Figure 1.

Participant inclusion flowchart.

Table 1.

Sociodemographic information, rehabilitation experience, and clinical characteristics of GBS patients

Sociodemographic Information and Clinical Characteristics
Patients	n= 45
Male, n (%)	29 (64)
Age at onset (years), median (IQR)	53 (34–65)
Socioeconomic stratum, n (%)
1 (low-low)	14 (31)
2 (low)	14 (31)
3 (medium-low)	11 (24)
4 (medium)	2 (4)
5 (medium-high)	3 (7)
6 (high)	1 (2)
Caregiver, n (%)
Partner/Spouse	19 (42)
Son/Daughter	8 (18)
Mother	7 (16)
Other	11 (24)
Received Inpatient Rehabilitation, n (%)	35 (79)
Received Outpatient Rehabilitation, n (%)	38 (84)
Clinical Diagnosis, n (%)
Guillain-Barré Syndrome	39 (87)
Miller Fisher Syndrome	5 (11)
Pharyngeal-cervical-brachial variant	1 (2)
Clinical features at presentation, n (%)
Lower limb paralysis	34 (76)
Paresthesia	28 (62)
Upper limb paralysis	18 (40)
Cranial nerve compromise	12 (27)
Autonomic dysfunction	9 (20)
Urinary retention	5 (11)
Ataxia	4 (9)
Comorbiditiesa, n (%)	15 (33)
Preceding symptoms of infectionb, n (%)	28 (62)
ICU admission, n (%)	29 (64)
Invasive mechanical ventilation, n (%)	4 (9)
Treatment, n (%)
IV Immunoglobulins	20 (44)
Plasma exchange	20 (44)
None	5 (11)
Second cycle of treatment	5 (11)
Electrophysiologic subtype, n (%) n= 28
Acute Inflammatory Demyelinating Polyneuropathy	16 (57)
Acute Motor Axonal Neuropathy	4 (14)
Other	8 (30)
Others, median (IQR)
Medical Research Council- SUM score	36 (20.5 – 44)
Modified Rankin score at inclusion	4.5 (4 – 5)
GBS Disability Scale at inclusion	4 (4 – 4)
Time from initial diagnosis, months	28 (9 – 34)

^aHypertension (n=12), diabetes mellitus (n=5), cancer (n=1), stroke (n=1), previous Guillain-Barre syndrome (n=1), facial palsy (n=1)

^bUpper respiratory tract infection/flu-like illness (n=16), diarrhea (n=10), urinary tract infection (n=1), fever (n=1), and pneumonia (n=1).

During the hospitalization phase for managing the acute stage of GBS, a significant majority of patients (84%) received some form of rehabilitation. However, nearly half of them (43%) encountered barriers hindering access to comprehensive and sustained rehabilitation activities. For instance, a subset (n= 3) only received physical therapy without access to other modalities of therapy. Notably, a considerable proportion of patients (40%) reported a lack of education on home therapy or rehabilitation services during their hospital stay. The period from hospital discharge to the initiation of outpatient therapy displayed notable variability, with 11% (4/38) reporting waiting periods exceeding a month. Transitioning from hospital to outpatient therapy presented challenges for 42% of patients (n=16/38), including administrative hurdles imposed by their healthcare insurance companies (n=6). Once engaged in therapy, half of the patients (50%, n=19/38) faced continuity issues, primarily stemming from administrative barriers imposed by their healthcare insurance providers. Importantly, none of the patients required new hospital admission related to GBS or its complications.

Functional outcomes

Details of the outcome measures at the time of the survey are outlined in Table 2. Most patients (58%) exhibited improvement to no or minimal residual disability (mRS score 0–1). Notably, follow-up mRS scores were significantly lower than baseline scores, with a mean difference of 2.8 (p < 0.001) (Figure 2A). Almost all patients (98%) showed improvement at follow-up, with a median change of three points (IQR= 2–4). Only one patient had no change in the mRS score (score 5) at the 10-month follow-up. Similarly, the GBS at follow-up was significantly lower than at baseline, with a median difference of 2 (IQR 1–3, p<0.001) (Figure 2B). 91% of patients showed improvement of at least one point in the GDS. No or minimal disability in the GDS (scores 0–1) were reported in for 47% of patients at follow-up. Correspondingly, BI scores were skewed towards the highest values. However, the activities of daily living, such as “going up and downstairs” and “transferring their own body from chair to bed,” were reported as “dependent” in 14 (31%) and 10 (22%) subjects, respectively.

Table 2.

Functional Outcomes of Patients with GBS at follow-up survey.

Outcome scales	Total 45 (100%)	Male 29 (64%)	Female 16 (36%)	p-value
Modified Rankin Scale/scorea
0-No symptoms	14 (31)	8 (28)	6 (38)
1-No significant disability despite symptoms	12 (27)	7 (24)	5 (31)
2-Slight disability	9 (20)	6 (21)	3 (19)
3-Moderate disability	5 (11)	4 (14)	1 (6)
4-Moderate to severe disability	4 (9)	3 (10)	1 (6)
5-Severe disabiilty	1 (2)	1 (3)	0 (0)
Median (IQR)	1 (0 – 2)	1 (0 – 3)	1 (0 – 2)	0.3**
GBS disability scoreb, n (%)
0- A healthy state.	14 (31)	8 (28)	6 (37)
1- Minor symptoms and capable of running	7 (15)	3 (10)	4 (25)
2- Able to walk >10m but unable to run	12 (27)	9 (31)	3 (19)
3- Walk 10m in an open space with help	8 (18)	6 (21)	2 (13)
4- Bedridden or chairbound	4 (9)	3 (10)	1 (6)
Median (IQR)	2 (0 – 3)	2 (0 – 3)	1 (0 – 2)	0.2**
Barthel index, n (%)
Independent (80–100)	36 (80)	22 (76)	14 (88)
Minimally dependent (60–79)	4 (9)	3 (10)	1 (6)
Partially dependent (40–59)	3 (7)	3 (10)	0 (0)
Very dependent (20–39)	1 (2)	1 (3)	0 (0)
Totally dependent (<20)	1 (2)	0 (0)	1 (6)
Median (IQR)	100 (85 – 100)	100 (78 – 100)	100 (90 – 100)	0.3**
International Classification of Functioning - Core set
Percentage functioning, median (IQR)	60 (43–75)	60 (40–78)	55 (45–74)	0.9**
Dysfunction in Body Function, n (%)
Energy and drive functions	10 (22)	7 (24)	3 (19)	0.9*
Sensation of pain	36 (80)	22 (76)	14 (88)	0.5*
Lower limbs	29/36 (81)	16/22 (73)	13/14 (93)
Lumbar	11/36 (31)	6/22 (27)	4/14 (29)
Upper limbs	10/36 (28)	6/22 (27)	5/14 (36)
Headache/facial	4/36 (11)	2/22 (9)	2/14 (14)
Exercise tolerance function	10 (22)	6 (21)	4 (25)	0.7*
Muscle power functions	33 (73)	22 (76)	11 (69)	0.7*
Lower limbs	26/33 (79)	17/22 (77)	9/11 (82)	1.0*
Upper limbs	17/33 (52)	12/22 (55)	5/11 (45)	0.2*
Orofacial	3/33 (9)	1/22 (5)	2/11 (18)	0.5*
Muscle tone functions	22 (49)	13 (45)	9 (56)	0.5*
Muscle endurance function	21 (47)	13 (45)	8 (50)	0.7*
Gait pattern functions	26 (58)	18 (62)	8 (50)	0.4*
Dysfunction in Body Structure, n (%)
Structure of lower extremity	10 (22)	4 (14)	6 (38)	0.1*
Dysfunction in Activities and Participation, n (%)
Transferring oneself	10 (22)	8 (28)	2 (13)	0.3*
Fine hand use	5 (11)	4 (14)	1 (6)	0.6*
Walking	4 (9)	3 (10)	1 (6)	0.9*
Moving around	18 (40)	13 (45)	5 (31)	0.4*
Driving	25 (56)	14 (48)	11 (69)	0.2*
Complex interpersonal interactions	5 (11)	4 (14)	1 (6)	0.6*
Acquiring keeping and terminating a job	19/41 (46)	13/26 (50)	6/15 (40)	0.6*
Community life	34 (76)	18 (62)	16 (100)	0.004 *
Recreation and leisure	17 (38)	11 (38)	6 (38)	0.9*
Dysfunction in Environmental factors, n (%)
Immediate family (supports & relationships)	1 (2)	1 (3)	0 (0)
Individual attitudes of immediate family members	27 (60)	16 (55)	11 (69)	0.4*
Health services, systems and policies	29 (64)	21 (72)	8 (50)	0.1*

^*Fisher’s exact test

^**Wilcoxon / U Mann Whitney

Figure 2. Comparison of mRS and GDS at onset and follow-up interview.

A. The mean difference between the mRS at onset (median 4.5 [IQR 4–5]) and at follow-up interview (median 1 [IQR 0–2]) and B. the mean difference between the GDS at onset (median 4 [IQR 4–4]) and at follow-up interview (median 2 [IQR 0–3]) are shown in the Gardner-Altman estimation plot(23). The mRS and GDS scores at both times are plotted on the left axis, and the mean difference is plotted on a floating axis on the right as a bootstrap sampling distribution. The mean difference is depicted as a dot, and the ends of the vertical error bar indicate the 95% confidence interval.

The functional scale based on the ICF exhibited wide fluctuations from 41% (indicating a low level of functioning) to 96% (functional in almost all evaluated aspects of life). Half of the patients reported dysfunction in four or more features of body dysfunction, with pain (80%) and weakness (73%) being the most commonly reported symptoms, especially in the lower limbs. Structural involvement, as assessed by deformity in the lower limbs, was present in 22% of patients. Three or more activities were reported as dysfunctional by half of the patients, with participation in community life being the most affected (76%), including all interviewed women. Additionally, half of the patients reported at least one dysfunctional environmental factor, with the most common barriers to accessing health services being administrative hurdles imposed by their healthcare insurer (24%) and living in distant/rural areas with inadequate infrastructure (15%). We found no differences in the outcome measures in univariate analysis when comparing cases according to age or access to rehabilitation; however, a shorter time from disease onset was associated with a better outcome in the mRS, as stated above, and in the ICF (Table 3).

Table 3.

Comparison of outcomes in patients with GBS by access to rehabilitation, age, and time to follow-up survey

Functional scale	Outpatient rehabilitation (n= 38)	No outpatient rehabilitation (n= 7)	p (0.05)
ICF percentage, mean (SD)	61 (20)	55 (17)	0,470a
Hughes, median (Q1-Q3)	1,5 (0–2.8)	2 (1–2.5)	0,572b
mRS, median (Q1-Q3)	1 (0–2)	1 (0.5–2.5)	1,000b
Barthel, median (Q1-Q3)	100 (86–100)	100 (80–100)	1,000b
	Age <65 years (n= 35)	Age ≥65 years (n= 10)	p (0.05)
ICF percentage, mean (SD)	59 (20)	61 (19)	0,876a
Hughes, median (Q1-Q3)	1 (0–2)	2.5 (0.8–3)	0,147b
mRS, median (Q1-Q3)	1 (0–2)	2 (0.8–2.3)	0,249b
Barthel, median (Q1-Q3)	100 (85–100)	90 (78.8–100)	0,212b
	Time from symptoms onset <12 months (n= 30)	Time from symptoms onset ≥12 months (n= 15)	p (0.05)
ICF percentage, mean (SD)	50 (18)	64 (19)	0,019 a
Hughes, median (Q1-Q3)	2 (1–3)	1 (0–2)	0,059b
mRS, median (Q1-Q3)	2 (1–4)	1 (0–2)	0,015 b
Barthel, median (Q1-Q3)	100 (55–100)	100 (90–100)	0,146b

We selected sociodemographic and clinical variables that could explain differences between patients with positive and negative outcomes (Supplementary Material, sTable 2 and sTable 3). The individual principal components generated by the PCA explained a low proportion of the variance (PC1 22%, PC2 11%, PC3 10%)(Supplementary Material, sTable 4). No evident clustering was observed when plotting the first three principal components, and separation according to positive or negative outcomes in each scale was unsatisfactory (Supplementary Material, sFigure 1). Then, a LDA was conducted using the same variables. The accuracy of LDA in discriminating positive and negative functional outcomes was relatively high at all scales (89%, CI 95% 76–96% in mRS; 80%, 65–90% in LDA; 84%, 70–94% for BI; and 82%, 68–92% for ICF [Supplementary Material, sFigure 2]). The time from the initial diagnosis to follow-up evaluation exhibited a positive correlation with a positive outcome in the mRS, the socioeconomic stratum with a good outcome in the GDS, and socioeconomic strata and Medical Research Council (MRC) muscle strength sum score was associated with the ICF (Supplementary Material, sTable 5). Conversely, age showed a negative correlation with a positive outcome in BI.

In the univariate linear regression for ICF score, socioeconomic stratum emerged as the only significantly associated variable(p= 0.001). However, this association was not significant (p=0.076) in the multivariate model when adjusted for the variables included in the PCA (Supplementary Material, sTable 6 and sTable 7).

DISCUSSION

This study illustrates the lasting challenges faced by GBS patients in Colombia, revealing pervasive long-term disabilities that impact various facets of their daily lives. Despite a notable improvement in outcome measures such as mRS scores from baseline to long-term follow-up, patients experienced persistent complaints related to motor impairment, difficulties in performing activities of daily living, and diminished engagement in community life. The definition of “long-term” and the proportion of patients achieving favorable outcomes exhibit considerable variability across studies. The proportion of patients with no or minimal residual symptoms (GDS score of 1–0) fluctuated widely, ranging from 27 to 80% in different studies (26–28). GBS patients typically experience a rapid functional recovery within the first month after disease onset, reaching a functional nadir around the first year, beyond which only modest improvement is anticipated (26,29,30). The extent of improvement appears linked to individual factors such as younger age, lower disease severity, and timely medical interventions during both the acute and rehabilitation phases (3,26,27,31). Considerable efforts have been directed toward developing predictive tools for poor long-term outcomes, often assessed at six months using the GDS (3,5). Other studies conducted on the American continent have delved into the long-term outcomes of GBS patients. Notably, a study on Zika virus-related GBS reported a comparable low percentage (9%) of individuals with severe disability (mRS score 4–5) but a higher proportion (74%) with scores of 0–1 (3). The sole predictor of an adverse outcome in that study was the molecular identification of viral RNA in the patient’s fluids. In our study, the variability in microbiological testing and infectious etiologies limited the ability to analyze the role of infections on outcomes.

BI increases as the recovery phase ensues, reflecting independence in basic activities of daily living and a reduction in disability. In a study conducted by Forsberg et al. 2004, most patients were completely independent six months after disease onset, as measured by the BI(3). Bathing was the most affected activity in this population, while in our population, the most commonly affected function was “going up and downstairs”(3). In a 10-year follow-up study, some patients with GBS-related disability one year after the onset continued to be, and some previously fully independent individuals had a decrease in BI scores, which was mainly attributed to age(4). In contrast with our findings, another study found that the increase in BI score was correlated with the duration of rehabilitation but not with age, although this study had a shorter follow-up than our study (3 vs. 28 months)(5).

We employed a comprehensive set of measurements to conduct a thorough assessment of overall functioning. While many studies predominantly rely on motor disability, as evaluated by the GDS, to gauge clinical recovery, such scales often lack sensitivity to subtle impairments that, though not directly impacting gait, significantly affect quality of life—examples being pain, fatigue, or social function (35–37). Our study revealed that 47% of patients achieved complete or near-complete recovery in GDS, surpassing the 26% reported in a previous study of Colombian patients (16) and falling short of the 62% reported in a Dutch study (38). Differences in the follow-up period likely contribute to these variations, with our cohort experiencing a more extended follow-up period since diagnosis (median 28 months) compared to the other Colombian study (17 months) but shorter than the Dutch study (3 to 6 years). Pain frequency was notably higher (80%), though similar in body distribution to other studies (36,39). Notably, our study did not assess pain intensity and the reported severity varied widely (36,40). Loss of muscle strength (73%), predominantly in the lower limbs, was a common symptom, contrasting with the lower frequency of motor impairment measured using the GDS. Although subjective sensation or weakness may not impede locomotion, it can still impact daily life activities and mental health (16,38). Reports in the literature have also highlighted loss of social participation and leisure activities due to poor physical condition (38,39), with lower extremity deformity being less common in our patients than in previous reports (41).

Certain factors associated with worse outcomes, such as age or severity of motor impairment at the initial assessment (MRC sum score), are nonmodifiable. However, lower socioeconomic status emerged as a modifiable factor linked to poorer outcomes. These socially modifiable factors likely mirror limited access to healthcare services in vulnerable, lower-income populations. In our study, socioeconomic status was a differentiator between patients with positive and negative outcomes in ICF and GDS. This underscores the barriers GBS patients face in resource-limited contexts in accessing healthcare during the acute phase and rehabilitation in the recovery phase (42). Evidence suggests worse outcomes, including mortality, in low-income compared to higher-income countries, partly attributed to structural barriers like delayed diagnosis, limited treatment access, and supportive care (42). Colombia’s status as the second most unequal country in Latin America exacerbates health inequalities tied to social determinants, highlighting the need to address these disparities for improved prognoses (43). A concerning 21% of patients did not receive inpatient rehabilitation, a proportion higher than the 10% reported in the United Kingdom (44). In our study, those who lacked inpatient rehabilitation tended to experience poorer outcomes, emphasizing the need for access to such services. Receiving rehabilitation during hospitalization has proven beneficial for GBS patients, including older and comorbid individuals (45,46). Despite most participants reporting only physical therapy, shown in a recent study of Colombian GBS patients (16), it remains crucial to adopt multidisciplinary rehabilitation approaches in GBS to enhance the improvement of disability beyond motor impairment (49). Such programs should persist throughout the chronic phase, as robust evidence supports their effectiveness in reducing disability (31). Speech therapy proves beneficial due to the high incidence of cranial nerve impairment in GBS (50). Occupational therapy is crucial in enhancing interactions with the environment and facilitating efficient functioning despite residual physical impairments (51). Neurorehabilitation units are scarce in many Low- and Middle-Income Countries (LMICs), including Colombia. Establishing or adapting rehabilitation centers in geographically strategic areas can extend coverage, enhancing access to rehabilitation resources. This model is particularly cost-effective, especially for highly impaired patients with GBS (46) and other neuromuscular disorders.

In our study, the time elapsed since onset significantly correlates with better outcomes, especially when measured using the modified Rankin Scale (mRS). Given GBS’s monophasic course, patients with recent onset may still be in a recovery phase after acute disease (10). Those with less than one year of disease onset displayed significantly worse scores on all measured scales, aligning with prior research suggesting recovery mainly occurs in the first year after onset (30). However, a 10-year follow-up study revealed that 14% of patients exhibited higher GDS, implying that recovery extends beyond the first year after disease onset (33). The potential for prolonged recovery emphasizes the importance of early and sustained rehabilitation. Older age was associated with negative outcomes in our study, a finding consistent with previous research examining both motor function and quality of life (3,5,52–54). We observed that 11% of patients experienced impairments in interpersonal relationships. Psychosocial health status and depression significantly impact the initial disease stages but tend to improve over time. Social dysfunction is more prevalent in patients with less GBS recovery, indicating a reciprocal relationship between physical complaints and psychosocial distress (55). Interestingly, even patients deemed “fully recovered” or those with minor motor symptoms exhibited higher rates of depression, anxiety, fatigue, and compromised physical functioning (28). A study involving Colombian GBS patients reported a higher frequency of poor mental health than poor physical health, with a greater likelihood of moderate to severe depression compared to healthy controls (Odds Ratio = 3.9, 95% CI: 1.2 – 11.2) (16). Given the priority of mental health in ensuring an adequate rehabilitation process, every patient should undergo a prompt (and preventive) evaluation by a psychologist or psychiatrist early in the disease course, with follow-up as needed.

MRC-sum score, mRS, and upper limb weakness are recognized predictors of both short and long-term outcomes, often indicating more extensive spinal root involvement leading to respiratory failure (3–5,56). In our analysis, the MRC sum score discriminated patients with poor outcomes in the International Classification of Functioning (ICF), while the other two variables did not show discriminative power for any scale. Although facial or bulbar weakness is acknowledged as a predictor of short-term complications, including mechanical ventilation and long-term impairment in speech, oral hygiene, and facial expression, we did not find an association between cranial nerve dysfunction and functional outcomes (57,58). Furthermore, a higher utilization of plasma exchange was observed in patients with adverse mRS outcomes despite its proven effectiveness in the acute phase (59). Notably, most patients who underwent plasma exchange were admitted to the Intensive Care Unit (ICU), potentially reflecting the need for ICU admission to safely administer plasma exchange in severe cases. The elevated use of plasma exchange as the primary therapy in our study, compared to high-income countries (60), may be attributed to the constraints of our resource-limited context (10,42).

Our study has some limitations. Many patients could not be reached for follow-up, potentially impacting external validity, particularly in cases with extended follow-up periods where a more favorable functional status might be expected. The absence of a control group limits our ability to assess the potential influence of comorbidities and contextual factors on functional outcomes. Nevertheless, most of our cohort did not have comorbidities, and city of residence did not emerge as a significant factor influencing functional outcomes. Mental health and quality of life were not evaluated, representing an understudied aspect of our population. In addition, an analysis of the influence of infectious etiological factors on functional outcomes was limited by the small cohort, the variability of infectious etiologies identified, and the lack of a case-control design. Despite these limitations, telephone interviews have demonstrated high reliability in assessing outcomes in neurological diseases, including GBS, compared to in-person scale measurements in previous studies (22,54,61–63). The multicentric design allowed us to include patients from diverse backgrounds, providing a comprehensive national perspective. The NEAS platform facilitated access to extensive clinical and laboratory data, enabling an exploration of their influence on outcomes.

CONCLUSION

The functional recovery of Colombian GBS patients is shaped not only by factors intrinsic to the natural course of the disease but also by social and healthcare-related determinants. Lower socioeconomic status emerges as a significant factor associated with poorer functional outcomes, underscoring the impact of social determinants of health, including access to hospital-based healthcare and rehabilitation services. Addressing the influence of economic inequities on the long-term outcomes of GBS patients requires heightened efforts, especially considering the ongoing research exploring health outcome disparities between lower- and higher-income countries (42).

Highlights.

• The functional recovery of Colombian GBS patients is influenced not only by the natural progression of the disease but also by social and healthcare-related factors.

• Lower socioeconomic status is significantly associated with poorer functional outcomes, highlighting the impact of social determinants of health, such as access to hospital-based healthcare and rehabilitation services.

• Our study highlights the difficulties GBS patients in resource-limited settings encounter in accessing healthcare during the acute phase and obtaining rehabilitation during the recovery phase.