This study was conducted at the Shandong Mental Health Center (Jinan, China) from January 2023 to September 2024. This research included the participation of Chinese individuals, both male and female, aged from 18 to 65 years, who met the criteria for schizophrenia as outlined by the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10), under the category of F20.X. The inclusion was specific to those in the acute phase of schizophrenia, with the stipulation that they could be administered IM medication for a minimum of 3 days based on the clinical judgement of the conducting researchers. Criteria for entry into the study included a Positive and Negative Syndrome Scale (PANSS) overall score of 70 or higher, a Positive and Positive and Negative Syndrome Scale Excited Component (PANSS-EC) score of 15 or higher, along with an Agitation-Calmness Evaluation Scale (ACES) score not exceeding 3.

Exclusion criteria: (1) alcohol and/or substance abuse; (2) a previous diagnosis of another mental disease; (3) thyroid illness, gout, kidney disease, major organic brain disease, immune disease, diabetes, etc.; (4) an infection in the previous 4 weeks or use of anti-inflammatory drugs, antibiotics, or glucocorticoids; (5) breastfeeding or pregnant; (6) regular use of psychiatric drugs in the past 2 weeks; (7) validated clinical significant abnormal laboratory values; (8) QTc prolongation or a prodrug QTc over or equal to 450 milliseconds; (9) known allergy to ziprasidone or haloperidol; and (10) those who were treated with electroconvulsive therapy during the previous 4 weeks. All eligible patients underwent a psychiatric assessment to determine if they satisfied the inclusion or exclusion criteria.

The subjects were randomly divided into equally sized groups to receive either ziprasidone (20 to 40 mg/d) (Jiangsu Nhwa Pharmaceutical Co., Ltd., lot No.: XN05AEQ010B001010101435, Xuzhou, Jiangsu, China) or haloperidol (5 to 10 mg/d) (Shandong Lukang Pharmaceutical Group Set limited liability company, lot No.: XN05ADF085B002020104166, Taian, Shandong, China) for 72 hours (3 days). This research used an unfixed-dose design; the dose range for ziprasidone reached 20 to 40 mg/d, and for haloperidol, 5 to 10 mg/d. These doses were regulated in accordance with the investigators’ clinical decisions. Research visits included a baseline visit, and at 72 hours after the first-dose time points.

Psychiatric evaluation utilized an array of scales: the PANSS, the PANSS-EC, the ACES and Barnes Akathisia Rating Scale (BARS) [11]. Scale evaluation is based on the performance of the patients on the day of testing. The PANSS-EC includes five items, with each item rating between 1 and 7, culminating in a total score range of 5 to 35 points. This scale is specifically designed to assess the intensity of hostile and restive symptoms. The PANSS features 30 items, rated similarly from 1 to 7 points, resulting in an overall score range from 30 to 210, and is employed to appraise the severity of psychotic manifestations.

Additionally, multiple scales were employed to assess the side effects linked to pharmacological interventions. These include the Rating Scale for Extrapyramidal Side Effects (RSESE), consisting of 10 items with scores ranging from 1 to 5 on each, the Barnes Akathisia Scale (BAS) comprising four items, each rated 0 to 3 for the initial three items and 0 to 5 for the last, resulting in a total possible score of 0 to 14. The Treatment Emergent Symptoms Scale (TESS) has 36 items, with each item graded on severity (0–4 points), the relationship to medication (0–4 points), and actions taken as a response to symptoms (0–6 points).

Laboratory examinations were also performed, such as complete blood cell counts, and electrocardiography at baseline and study endpoint, with the intention of evaluating the patients’ physical condition.

The primary measure of effectiveness was to compare the mean alterations in the PANSS total and PANSS-EC scores from the study’s onset to its completion between the groups receiving IM ziprasidone and haloperidol. The principal measure for analysis was the rate of reduction in the PANSS-EC and PANSS scores, calculated as [(baseline score – endpoint score)/baseline score] $\times$ 100%.

Secondary measures of efficacy involved comparisons of the differences in the changes observed in ACES and BARS scores between the two treatment cohorts. Safety outcomes were primarily focused on contrasting the frequency of anticholinergic medication use and the occurrence of adverse effects between the ziprasidone and haloperidol groups.

Blinding was maintained for subjects and all members of the research team regarding treatment allocation, except for the onsite rater. The evaluators responsible for assessing scale scores at each site visit were blinded to the treatment conditions; they also underwent standardized training before the commencement of the study to ensure consistency in evaluation. The same evaluator at each site conducted all assessments for each participant to maintain uniformity in ratings.

The usage of the other antipsychotics and parenteral benzodiazepines, or the prophylactic treatment of extrapyramidal reactions (EPS) with anticholinergics or propranolol was not permitted during the study period. Benzodiazepines were allowed for patients with insomnia. Anticholinergics or propranolol were allowed to be used in patients with EPS.

Tolerability and safety assessments were carried out initially and again at the 72-hour mark, encompassing electrocardiograms, complete blood counts, and evaluations for any side effects. QT intervals were corrected for heart rate using QTc. Evaluations for EPS were conducted using the BAS and the RSESE both at the starting point and after 72 hours.

We reported side effect which took place during this research and validated the precision of the clinical report on the grounds of the associated scale evaluation and physical test indexes. When assessing the adverse effects of EPS upon the RSESE scale and Barnes scale, doctors frequently recommend medications like anticholinergics. Through the clinical reporting of adverse effects, insomnia, excessive sedation, and dizziness were detected. While the BARS index was less than 3, those subjects were identified as excessive sedation patients.

To rigorously analyze the collected data, we first conducted a preliminary examination to assess the distribution of the dataset. Recognizing that not all variables followed a normal distribution, we deployed descriptive statistics for a comprehensive summary. For variables that were normally distributed, we used the mean $\pm$ standard deviation (SD) to outline their central tendency and variability. Conversely, for those variables not aligning with normal distribution, we provided median and interquartile range (Q1, Q3) as representations, thus offering insight into the data’s distribution and spread. To delineate differences between groups where the data followed normal distribution, the independent samples t-test was applied, an addition aimed at addressing the oversight in initial specifications of our analytic approach. For comparative analysis between groups in instances of non-normal data distribution, we utilized the Mann Whitney U test. Categorical variables reported as n (%) underwent analysis through either the Pearson Chi-squared test or Fisher’s exact test, depending on suitability to discern any notable associations or divergences. Since the proportion of cells with theoretical frequency less than 5 was greater than 20%, and some cells even had 0, we used Fisher-Freeman-Halton test to compare the marital status of the two groups, and the same was used for drug side effects. An integral facet of our safety assessment—the frequency of side effects—was meticulously documented and subjected to statistical evaluation to establish any significant variations across the treatment cohorts. All statistical tests prescribed to a significance threshold established at $\alpha$ = 0.05, denoting statistical significance. This decision aligns with standard practices in clinical research, facilitating consistency in our interpretation of the findings. The comprehensive data analysis was carried out using SPSS version 25.0 (SPSS Inc., Armonk, NY, USA).

Seventy patients were screened, and 1 patient was excluded before grouping who had unstable medical conditions(=1). All 69 patients were treated with the study drugs, 35 patients with ziprasidone and 34 patients with haloperidol, and they underwent baseline assessment. None of them left the study before study completion. Over the course of the 72-hour timeframe, the most commonly administered dosage was 40 mg for ziprasidone, achieved by 30 (85.7%) of the subjects, and 5 mg for haloperidol, achieved by 30 (88.23%) of the participants.

The study cohort comprised individuals diagnosed with schizophrenia, presenting moderate to severe psychotic symptoms, with an average age of 31.5 years. The mean duration of schizophrenia among the participants was approximately 4 years. At the initiation of the study, there were no statistically significant differences in clinical and demographic characteristics, nor in electrocardiogram and hematological test results (except for white blood cell count, p = 0.018) between the two treatment groups (Table 1).

Main assessment measures: PANSS-EC score and PANSS total score demonstrated a decreasing trend after 72 hours of treatment and was statistically significant (haloperidol, p = 0.001; ziprasidone, p = 0.001) (Fig. 1). There were no significant between-group distinctions in the subtraction rate of PANSS total score (p = 0.159) and PANSS-EC score (p = 0.312) (Table 2).

Fig. 1.

Changes in PANSS Total Scores (A) and PANSS-EC Scores (B) from Baseline to Study Endpoint (using last observation carried forward methodology). Throughout the 72-hour period, a significant difference (p = 0.001) was observed between the two groups at all points. The error bars represent the standard deviation. **: p 0.01.

Secondary efficacy outcome: BARS score decreased significantly (p = 0.001) and ACES score increased significantly (p = 0.001) in the haloperidol and ziprasidone groups from baseline to the study endpoint (Fig. 2).

Fig. 2.

Changes in ACES (A) and BARS (B) from Baseline to Study Endpoint (using last observation carried forward methodology). Throughout the 72-hour period, a significant difference (p = 0.001) was observed between the two groups at all points. The error bars represent the standard deviation. **: p 0.01.

Side effects and the prescribed dosage were evaluated at baseline visit and at 72 hours. The presence of extrapyramidal syndrome in the two groups was detected. RSESE, BAS, and TESS were used for evaluating side effects that occurred during the study period.

Fewer patients suffered side effects in the ziprasidone group (n = 3, 8.57%) than in the haloperidol group (n = 10, 29.41%). The discrepancy between the groups was prominent ($\chi$² = 4.889, p = 0.027). No serious side effects or deaths occurred in either group. In the ziprasidone group, there was one case of extrapyramidal reactions and two cases of excessive sedation. In the haloperidol group, extrapyramidal reactions occurred in six cases, excessive sedation in three cases, and dizziness in one case. See Table 3. The use of anticholinergic drugs in the ziprasidone group was no more than that in the haloperidol group; the difference was not statistically significant (p = 0.055). See Table 3.