Comparative Analysis of Monotherapy and Bi-Therapy in Antipsychotic Treatment ()
1. Introduction
Antipsychotic medications, including typical (first-generation) and atypical (second-generation) antipsychotics, are essential in managing psychiatric disorders such as schizophrenia [1]. Typical antipsychotics like Haloperidol effectively reduce psychotic symptoms but are often associated with extrapyramidal side effects (EPS) [2]. Atypical antipsychotics, such as Risperidone and Olanzapine, are preferred for their lower risk of EPS and their ability to treat both positive and negative symptoms of schizophrenia, although they can cause metabolic issues [3]. The choice between monotherapy (a single antipsychotic) and bi-therapy (a combination of two antipsychotics) is critical. Monotherapy is simpler and has fewer drug interactions, while bi-therapy can be more effective for treatment-resistant cases but carries a higher risk of side effects [4] [5]. This study compares the efficacy and outcomes of these two approaches in treating schizophrenia. By analyzing a cohort of 100 patients over 12 months using tools like the Positive and Negative Syndrome Scale (PANSS) and the Global Assessment of Functioning (GAF), we aim to identify which treatment method provides better symptom control and fewer side effects. Machine learning techniques will help highlight key predictors of treatment success, providing insights to optimize therapeutic strategies. Understanding the comparative benefits of monotherapy versus bi-therapy will help clinicians make informed decisions, improving treatment outcomes for patients with schizophrenia.
2. Methods
This study involved 100 participants diagnosed with schizophrenia, who were randomly assigned to two treatment groups: Group A and Group B, with 50 participants in each group. Group A received monotherapy, consisting solely of an atypical antipsychotic, such as Risperidone, administered at a standard therapeutic dose. Group B was treated with bi-therapy, a combination of a typical antipsychotic (e.g., Haloperidol) and an atypical antipsychotic (e.g., Risperidone), both administered at standard therapeutic doses [6]. The duration of the study was 12 months. Throughout this period, participants were assessed using several clinical tools to evaluate the efficacy and side effects of the treatments. The Positive and Negative Syndrome Scale (PANSS) was used to measure symptom severity, the Global Assessment of Functioning (GAF) was employed to assess overall functional ability, and a Side Effect Rating Scale was utilized to monitor and quantify adverse effects. These assessments were conducted at baseline and after 12 months of treatment to determine changes in clinical status and treatment outcomes.
2.1. Sample Size, Dosage, and Participant Characteristics
2.1.1. Sample Size Justification
A total of 100 patients diagnosed with schizophrenia were included in this study, with 50 patients assigned to the monotherapy group (Group A) and 50 to the bi-therapy group (Group B). The sample size was selected based on previous studies exploring antipsychotic treatments using similar methodologies. Additionally, a power analysis was conducted to ensure that the study had sufficient statistical power to detect moderate effect sizes. Specifically, the analysis showed that a sample size of 100 participants would provide an 80% chance of detecting a medium effect size (Cohen’s d = 0.5) with a significance level of p < 0.05.
2.1.2. Dosage Information
Participants in Group A received monotherapy consisting of the atypical antipsychotic Risperidone, administered at a standard daily dose of 4 mg. This dosage was selected in accordance with established treatment guidelines for schizophrenia and is known for its balance between efficacy and a manageable side effect profile. Participants in Group B were treated with a combination therapy (bi-therapy), which involved a typical antipsychotic (Haloperidol, 5 mg/day) and an atypical antipsychotic (Risperidone, 2 mg/day). The dosage was chosen based on clinical best practices for treatment-resistant cases and supported by previous research on combination antipsychotic therapy.
2.1.3. Participant Demographics and Baseline Characteristics
The baseline demographic and clinical characteristics of the participants are presented in Table 1. The patients had a mean age of 35.7 years (SD ± 10.1) and an average illness duration of 5.5 years (SD ± 2.1). The gender distribution was approximately equal, with 58 male and 42 female participants. The baseline clinical assessments included the Positive and Negative Syndrome Scale (PANSS) and the Global Assessment of Functioning (GAF) scores. The mean baseline PANSS score was 90.5 (SD ± 15.0), and the mean baseline GAF score was 51.0 (SD ± 9.4), indicating moderate to severe functional impairment.
Table 1. Baseline demographic and clinical characteristics of study participants.
Characteristic |
Monotherapy group (n = 50) |
Bi-therapy group (n = 50) |
Mean age (years) |
35.2 ± 10.5 |
36.1 ± 9.8 |
Gender (male/female) |
28/22 |
30/20 |
Duration of illness (years) |
5.4 ± 2.1 |
5.6 ± 2.0 |
PANSS baseline score |
91.2 ± 15.4 |
89.8 ± 14.7 |
GAF baseline score |
51.3 ± 9.2 |
50.7 ± 9.5 |
This Table 1 demographic and clinical data provides insight into the study population and ensures that the sample is representative of the broader population of schizophrenia patients, allowing for generalization of the study results.
3. Results
Table 2 presents descriptive statistics for key clinical outcomes, including PANSS and GAF scores, side effects, and treatment success. PANSS scores, which assess symptom severity, show an increase from baseline (mean = 40.88) to 12 months (mean = 89.70), reflecting symptom improvement. GAF scores, which measure overall functioning, slightly decreased from baseline (mean = 62.11) to 12 months (mean = 50.26), suggesting a decline in functional ability. The side effects score (mean = 70.81) varied considerably, with scores ranging from 50 to 89. Treatment success had a low mean of 0.20, indicating that only a few participants achieved high treatment success. These statistics summarize the clinical outcomes of monotherapy and bi-therapy over 12 months.
Table 2. Descriptive statistics: The descriptive statistics of the study samples.
|
|
PANSS_baseline |
PANSS_12 months |
GAF_baseline |
GAF_12 months |
Side_effects |
Treatment_success |
Count |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
100.0 |
Mean |
40.88 |
89.70 |
62.11 |
50.26 |
70.81 |
4.87 |
0.20 |
Std |
13.99 |
17.12 |
16.05 |
11.78 |
11.07 |
2.61 |
0.40 |
Min |
18.00 |
60.00 |
30.00 |
30.00 |
50.00 |
1.00 |
0.00 |
25% |
30.50 |
77.00 |
52.00 |
41.00 |
61.00 |
3.00 |
0.00 |
50% |
41.00 |
91.00 |
62.00 |
52.00 |
71.00 |
5.00 |
0.00 |
75% |
53.25 |
105.50 |
75.00 |
61.00 |
80.25 |
7.00 |
0.00 |
Max |
64.00 |
119.00 |
89.00 |
69.00 |
89.00 |
9.00 |
1.00 |
Correlation Analysis: The correlation matrix (Figure 1) reveals significant negative correlations between PANSS_12 months and treatment_success (r = −0.51), indicating that lower PANSS scores after 12 months are associated with higher treatment success. Similarly, side effects negatively correlate with treatment success (r = −0.46).
PANSS Score Comparison: Box plots comparing PANSS baseline and 12-month scores between the two treatment groups (Figure 2 and Figure 3) show that both groups experienced reductions in PANSS scores, with bi-therapy showing a slightly greater reduction.
GAF Score Comparison: GAF scores improved in both groups, with bi-therapy demonstrating slightly higher improvements in baseline and 12-month GAF scores (Figure 4 and Figure 5).
Figure 1. Correlation matrix of clinical and functional variables.
Figure 2. PANSS baseline score comparison between monotherapy and Bi-therapy treatments.
Figure 3. PANSS 12-month score comparison between monotherapy and Bi-therapy treatments.
Figure 4. GAF baseline score comparison between monotherapy and bi-therapy treatments.
Figure 5. GAF 12-month score comparison between monotherapy and Bi-therapy treatments.
Side Effects Analysis: Side effects were more pronounced in the bi-therapy group, as shown in Figure 6. The median side effects score was higher for bi-therapy compared to monotherapy.
Figure 6. Side effects comparison between monotherapy and Bi-therapy treatments.
Treatment Success Distribution: The distribution of treatment success (Figure 7) indicates that while both groups had similar numbers of successful treatments, monotherapy had a slightly higher success rate.
Figure 7. Treatment success distribution by monotherapy and Bi-therapy.
Machine Learning Feature Importance: The Random Forest model identified the most important features influencing treatment success (Figure 8). PANSS_12 months and side effects were the most significant predictors, followed by age and PANSS_baseline scores.
Figure 8. Feature importances in predicting treatment success (Machine learning analysis).
3.1. Side Effects Analysis
The side effects of both treatment groups were measured using the Side Effect Rating Scale. To provide a more comprehensive view, side effects were categorized into three levels: mild, moderate, and severe. The Bi-therapy group experienced significantly more side effects compared to the monotherapy group, with a higher proportion of moderate to severe cases. A chi-square test was performed to assess the statistical significance of the differences in side effect severity between the two groups.
A significant difference was observed in the distribution of moderate and severe side effects in Table 3 between the two groups (p < 0.01). Patients in the bi-therapy group experienced higher rates of severe side effects, particularly extrapyramidal symptoms and weight gain, which are known complications of typical antipsychotic use. These findings suggest that while bi-therapy may offer better symptom control, the increased burden of side effects must be managed carefully to optimize treatment outcomes.
Table 3. Comparison of side effect severity between monotherapy and Bi-therapy groups.
Side effect severity |
Monotherapy group (n = 50) |
Bi-therapy group (n = 50) |
Mild (Score 1 - 3) |
35% (17/50) |
24% (12/50) |
Moderate (Score 4 - 6) |
18% (9/50) |
46% (23/50) |
Severe (Score 7 - 9) |
2% (1/50) |
30% (15/50) |
3.2. Machine Learning Model for Predicting Treatment Success
To identify the key predictors of treatment success, a Random Forest classifier was implemented. The model was built using 500 decision trees, with a maximum tree depth of 10. The Random Forest model was chosen for its robustness against overfitting and ability to handle interactions between variables. The dataset was split into training (80%) and testing (20%) sets. Five-fold cross-validation was used during training to ensure the generalizability of the results.
The model’s accuracy on the test set was 84%, and its performance was evaluated using the area under the receiver operating characteristic curve (AUC = 0.87), indicating strong predictive power. Feature importance was calculated using the Gini impurity index, which revealed that the most significant predictors of treatment success were PANSS_12 months and side effects. Other notable features included age and baseline PANSS scores. The machine learning analysis not only confirmed the importance of clinical factors like symptom improvement (PANSS) and side effects but also provided insight into how these features interact to predict overall treatment success.
4. Discussion
The development of obsessive-compulsive symptoms in patients treated with neuroleptics, as observed in these cases, highlights a significant yet often overlooked side effect. Both groups (A and B) showed improvements in PANSS and GAF scores, with bi-therapy demonstrating slightly higher efficacy at the cost of increased side effects. The machine learning analysis further emphasized the importance of PANSS_12 months and side effects as key predictors of treatment success [7]-[10].
4.1. Case Reports
4.1.1. Case Report for Group A (Monotherapy)
Mr. T, a 35-year-old male diagnosed with schizophrenia, was selected for monotherapy with Risperidone (4 mg/day). His case illustrates a typical response to monotherapy, with a significant reduction in PANSS scores (from 92 to 63) and improvement in GAF scores (from 51 to 71) over 12 months. His side effects were minimal, with only mild sedation (side effect rating of 3/10), which was manageable and did not affect his overall quality of life. This case highlights the efficacy of monotherapy in managing symptoms with a favorable side effect profile, making it an ideal option for patients who are sensitive to adverse reactions.
4.1.2. Case Report for Group B (Bi-Therapy)
Ms. R, a 40-year-old female treated with a combination of Haloperidol (5 mg/day) and Risperidone (2 mg/day), demonstrated the typical outcome associated with bi-therapy. Her PANSS score improved markedly from 88 to 54, indicating superior symptom control. However, she experienced notable side effects, including extrapyramidal symptoms and significant weight gain (side effect rating of 6/10), requiring adjunctive treatment to manage these issues. Despite the greater efficacy of bi-therapy in her case, the side effects reduced her overall quality of life. This case exemplifies the trade-off between efficacy and side effects in combination therapy, underscoring the need for careful side effect management in clinical practice.
5. Conclusions
This case report underscores the nuanced benefits and limitations of monotherapy and bi-therapy in the antipsychotic treatment of schizophrenia. Monotherapy with atypical antipsychotics, particularly Risperidone, was shown to provide effective symptom management while minimizing side effects. This makes monotherapy a particularly suitable option for patients who are sensitive to adverse drug reactions or for those whose primary goal is to maintain a higher quality of life alongside symptom control [11]-[15]. The relatively low incidence of side effects such as sedation and metabolic disturbances highlights the safety profile of atypical antipsychotics, especially in long-term treatment scenarios where drug tolerability is crucial for adherence and overall success [16]. In contrast, bi-therapy, which combines typical and atypical antipsychotics, resulted in more significant symptom reduction, demonstrating its efficacy, particularly in treatment-resistant cases. However, this benefit comes at a cost. Patients receiving bi-therapy experienced more pronounced side effects, including extrapyramidal symptoms (EPS), such as muscle rigidity and tremors, as well as metabolic issues such as significant weight gain [17]. These side effects can substantially impact the patient’s quality of life, increasing the burden of care and the need for additional treatments to mitigate these adverse effects. The trade-off between enhanced symptom control and the management of side effects is a critical consideration, emphasizing that bi-therapy may be more appropriate for patients who prioritize rapid symptom reduction or those with more severe or refractory forms of schizophrenia [18]. The machine learning analysis offered valuable insights, identifying PANSS scores at 12 months and the presence of side effects as key predictors of treatment success. These findings suggest that while symptom improvement is a primary driver of treatment efficacy, the burden of side effects plays an equally important role in determining long-term outcomes. Patients who achieve lower PANSS scores but experience significant side effects may struggle with adherence, reducing the overall effectiveness of treatment. This highlights the importance of balancing efficacy with tolerability when choosing a treatment strategy. Machine learning approaches, such as Random Forest models, demonstrate potential in personalizing treatment by predicting which patients are more likely to benefit from monotherapy versus bi-therapy based on individual characteristics, allowing clinicians to optimize treatment plans early in the treatment process [19]. While this study provides essential insights into the comparative benefits of monotherapy and bi-therapy, it is limited by its relatively small sample size and homogeneous population. To strengthen the generalizability of these findings, future studies should include larger and more diverse populations that account for variations in age, gender, ethnicity, and co-occurring conditions. Additionally, longer follow-up periods are necessary to assess the sustainability of symptom control and the long-term impact of side effects on health outcomes [20].
The application of advanced machine learning techniques presents exciting possibilities for refining personalized treatment strategies. By incorporating more complex patient data, such as genetic markers, cognitive assessments, and real-time side effect monitoring, machine learning models could improve predictive accuracy, allowing clinicians to tailor treatment choices to the individual needs and risk profiles of patients [21]. Such personalized approaches could help optimize both efficacy and tolerability, potentially reducing the trial-and-error approach currently common in psychiatric care. Ultimately, this could lead to a paradigm shift in schizophrenia treatment, where decisions are guided by data-driven insights, improving both clinical outcomes and patient satisfaction. While bi-therapy may offer superior symptom control, careful management of side effects is essential for long-term success [22]. Monotherapy remains a viable option for many patients, particularly those who prioritize minimizing side effects [23]. As machine learning advances, it holds promise for further enhancing clinical decision-making, ensuring that each patient receives the most appropriate and effective treatment [24]. This approach has the potential to revolutionize psychiatric care, improving outcomes for patients with schizophrenia and beyond.
Acknowledgements
I would like to express my deepest gratitude to Dr. Rocco de Filippis for his invaluable guidance, expertise, and support throughout the course of this research. His insights have been instrumental in shaping the direction and success of this study. I also extend my appreciation to the Neuroscience Institute of Psychopathology, Rome, Italy, and the University of Genova, Italy, for their support.
Conflicts of Interest
The authors declare no conflicts of interest.