Introduction

Medications are directly used to prevent and treat diseases. However, all medications can also cause harmful effects¹. According to the World Health Organization, an adverse drug reaction (ADR) is "a harmful and unwanted reaction that occurs after the administration of a drug at doses commonly used in humans, to prevent, diagnose or treat a disease, or to modify any biological function"². Although some ADRs are detected during clinical trials; others, in the post-marketing stage³. ADRs are a significant cause of morbidity and mortality, responsible for up to 6% of hospital admissions with an associated mortality of 2%, and represent a substantial financial burden for patients and health systems. Additionally, they affect the patient's quality of life, confidence in the healthcare system, and length of hospital stay⁴.

While some ADRs are unpredictable, many can be prevented with proper foresight and control⁵. Continuous and constant surveillance, through pharmacovigilance programs, has allowed the reporting of suspected ADRs to generate alerts and prevent or avoid greater harm caused by medications⁶. Unfortunately, underreporting and under-notification remain a key challenge, as it has been estimated that less than 5% of all ADRs are reported in routine practice. This limits the ability of systems to provide accurate incidence data⁵.

One group of medications that may be associated with a significant incidence of ADRs is antipsychotics⁷, due to their pharmacodynamics and direct effect on the delicate balance of neurotransmitters that control behavior and brain function⁸. Psychiatric disorders are chronic in nature and often require prolonged and continuous medication treatments, increasing the likelihood of an ADR occurring during their use. Monitoring and prevention are crucial to improving clinical practice, enhancing medication safety, and supporting public health programs⁹.

In Mexico, there is the Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz (INPRFM), which is a specialized health center of national and international reference that provides care to patients suffering from mental disorders¹⁰. It is a public sector institute that belongs to the Mexican Ministry of Health and provides outpatient medical consultations and hospitalization services to psychiatric patients over the age of 13¹¹; it is one of the most important and representative health centers in the country. Given this, this study aimed to determine the pattern of ADRs due to antipsychotic drugs, detected at the INPRFM during the period from December 2021 to May 2022.

Materials and methods

2.3. Data Evaluation

A description of the manifestation and type of problem caused and classified as ADR was made. The accumulated incidence of ADR occurrence during the study period was calculated using the following equation:

The severity of ADRs, defined according to NOM-220-SSA1-2016 "installation and operation of pharmacovigilance"¹², was classified as "serious" and "not serious". According to the same standards, the severity of ADRs was classified as mild, moderate, and severe. On the other hand, the quality of the information from the ADR notification was also evaluated according to the same standards as grade 0 when the notification only includes the identified patient, at least one suspected adverse reaction, the suspected drug, and the notifier's data. Grade 1 when, in addition, it includes the dates of the start of the suspected adverse reaction, as well as the start and end of the treatment: day, month, and year. Grade 2 when it also includes the generic and distinctive name of the medication used, its posology, the route of administration, the reason for its prescription, the consequence of the event, and the data from the medical history. And grade 3 when, in addition, it includes the reappearance of the clinical manifestation consequent to a new administration of the drug in question.

Finally, the causality of ADRs was determined using the Naranjo algorithm and were also classified according to NOM-220-SSA1-2016¹²as: 1) Certain when the clinical event manifested with a plausible temporal sequence in relation to drug administration, and could not be explained by concurrent disease, nor by other drugs or substances. The response to drug withdrawal (discontinuation) must have been clinically plausible. The event must have been definitive from a pharmacological or phenomenological point of view, using, if necessary, a conclusive re-exposure procedure. 2) Probable when the event manifested with a reasonable temporal sequence in relation to drug administration; it was unlikely to be attributed to concurrent disease, nor to other drugs or substances, and withdrawing the drug, a clinically reasonable response occurred. Information about drug re-exposure was not required. 3) Possible when the event manifested with a reasonable temporal sequence in relation to drug administration, but could also be explained by concurrent disease, or by other drugs or substances. Information regarding drug withdrawal may have been missing or unclear. 4) Improbable when the event manifested with an improbable temporal sequence in relation to drug administration, and could be explained more plausibly by concurrent disease, or by other drugs or substances. 5) Conditional to a clinical event, reported as an adverse reaction, for which it was essential to obtain more data for a proper assessment, or additional data were under examination. And 6) Not assessable to a notification that suggested an adverse reaction but could not be judged, as the information was insufficient or contradictory, and could not be verified or completed in its data.

Results

A total of 74 ADRs were detected during the study period, presented in 52 patients out of a total of 88 who were being treated with antipsychotics. The accumulated incidence of ADRs in the analyzed population during the study period was 59%. The average number of ADRs per patient was 1.42 (range 1-5). The detected ADRs were mostly in women (54%) and in the adult population between 30 and 59 years old. Also, most of the ADRs were detected in patients diagnosed with schizophrenia (65%). Table 1 shows these results.

The 74 ADRs were caused by 24 different types: Olanzapine, risperidone, clozapine, aripiprazole, haloperidol, quetiapine, and paliperidone were the drugs that caused the detected ADRs (Table 2).

Figure 1-A shows the percentage distribution of antipsychotic-associated ADRs detected. It can be seen that the most frequent were hyperprolactinemia (34%), somnolence (12%), weight gain (7%), and sialorrhea (7%). On the other hand, olanzapine (25%), risperidone (23%), and clozapine (15%) were the drugs that caused the most ADRs (Figure 1-B ).

Figure 1. A: Percentage distribution by type of ADR detected in the study period. B: Percentage distribution of ADRs detected by suspected antipsychotic medication during the study period.

Table 3 shows that of the 74 ADRs found, none were severe in intensity and the majority were mild in severity (55%). Only 3 ADRs: hypotension, bradycardia, and sedation were classified as serious, which occurred in the same patient, and olanzapine was the suspected drug. In all cases, the quality of the information was at least grade 2. When analyzing the causality of ADRs using the Naranjo algorithm, 55% were of possible causality and 45% of probable causality.

The factors associated with ADRs were unknown in 91% of cases; only in seven cases was it possible to know this information, six being of etiology due to dose increase and one due to a change in the route of administration.

Discussion

This study provides current information on ADRs associated with antipsychotics, a group of drugs related to various adverse reactions, detected in the Mexican population attended at one of Mexico's most important and reference health centers, where people from various parts of the country come. We found that olanzapine was the drug responsible for most of the detected ADRs, and hyperprolactinemia was the most incident.

The incidence of ADRs found during the analysis period was 59%, which is higher than what was observed in a study conducted at the CAISAME Long Stay Department, the largest hospital in the western region of Mexico, where 29.2% of the patients presented at least one ADR, 17.8% presented extrapyramidal effects, 15% non-extrapyramidal effects, and 3.57% both types of side effects. Although in said study a larger number of patients were analyzed (n = 140), the analysis period was shorter than the one used in our study¹³, which may explain why the accumulated incidence of ADRs was higher in the present work. In the same trend, the incidence of ADRs estimated in our study was also higher than that reported in other parts of the world; Lucca et al. reported, in 2014, an incidence of ~42% (n= 517 patients) over a two-year period⁹, while Chawla et al., in 2017, reported an incidence of ~17% (n= 224 patients) over a three-month period¹⁴. Both studies were conducted in India, which may explain the difference found, given that it is another geographical context.

Previously, it has been reported that ADRs in psychiatric patients are more frequent in women than in men¹⁵, and the data derived from our study do not differ from this observation. The group of people most affected by ADRs was adults between 30 and 59 years old, with an average age of 38 years; according to other reports, the higher incidence in this age group may be due to the onset of psychiatric disorders such as schizophrenia and psychosis, which were the most prevalent diagnoses in our study; typically occur in early adulthood⁹, so it is expected that the prevalence of these disorders is high in adulthood.

Hyperprolactinemia was the most frequently detected ADR in the analyzed patients. In the literature, it has been estimated that it is induced in up to 70% of patients with schizophrenia who consume antipsychotics¹⁶. In our study, the incidence was 48%. Hyperprolactinemia caused by antipsychotics is due to blocking the dopaminergic D2 receptors, which in turn are responsible for inhibiting the hormone prolactin, which causes hyperprolactinemia¹⁷, which has short- and long-term consequences that can seriously affect the patient's quality of life, commonly causing menstrual disorders, sexual dysfunction, galactorrhea, amenorrhea, among others¹⁸. In addition, hyperprolactinemia can lead to other pathologies such as osteoporosis¹⁹. Therefore, pharmacovigilance programs are important within public institutions to propose risk management plans for antipsychotic-induced hyperprolactinemia and its possible clinical implications.

On the other hand, the three drugs most frequently associated with the ADRs detected in the study were olanzapine, risperidone, and clozapine. This could be because olanzapine and risperidone were the most frequently used drugs in the clinical practice of schizophrenia at the INPRFM, a place that treated the most patients and where most ADRs were detected. This finding coincides, both in order and frequency, with the results of the study conducted by Piparva et al. regarding the suspected drugs related to antipsychotic ADRs²⁰and with the publication of Prajapati et al. in 2013, who found clozapine and risperidone among the three main drugs that caused the most appearance of ADRs²¹.

On the other hand, regarding the characteristics of the ADRs found, all were mild or moderate in intensity, and it was not necessary to withdraw the suspected antipsychotic drug or change the treatment. However, the cases of hypotension, bradycardia, and sedation detected were considered serious, as they prolonged hospital stay and endangered the patient's life. Continuous monitoring and timely detection of all ADRs are important, as rare or infrequent ADRs can be identified²², and for those that are already known, the manifestation from patient to patient can be variable²³. Chawla et al. reported, in 2017, the analysis of ADRs associated with antipsychotic drugs and observed that the causality of all ADRs analyzed using the Naranjo algorithm was classified as possible and probable¹⁴; we obtained similar results, as all the ADRs detected were classified in the same causality categories and no definite causality was identified.

It is important to note that all the cases of ADRs found had an information quality classification above grade 1 and have sufficient information about the patient, the drug, the start date of the suspicion and the treatment used and, for the cases classified with grade 3, data on re-exposure to the suspected drug, complying with international and national recommendations for ADR notifications.

Conclusion

This study provides additional information to that currently existing on the incidence and frequencies of ADRs of antipsychotic drugs in Mexico.

In general, a high incidence of ADRs was found in patients treated at the INPRFM, over 50%, most of them found in schizophrenic patients. Most were mild in severity. ADRs of the gastrointestinal and endocrine systems were the most incident, due to the use of atypical antipsychotic drugs. Olanzapine and clozapine were the drugs that caused the most ADRs. The most frequent gastrointestinal system ADRs were sialorrhea and weight gain, while in the endocrine system it was hyperprolactinemia. It is necessary to give importance to the monitoring of hyperprolactinemia, since it was an ADR caused by all the antipsychotics analyzed in this study. A protocol should be implemented that clearly establishes the prolactin concentration, which should begin to be gradually suspended and, in a timely manner, the drug that is causing this ADR or switch to antipsychotics that do not cause an increase in prolactin in the blood: the so-called prolactin-sparing antipsychotics or consider the use of dopamine agonists. It is recommended to promote the culture of ADR reporting at the INPRFM, both expected and unexpected, and to strengthen the follow-up of ADRs caused by antipsychotic drugs.

GSV: Great Saphenous Vein, SSV: Small Saphenous Vein

The most frequent CEAP clinical class was C2, representing 44.9%, a group that mostly exhibited insufficiency in both the superficial and deep venous systems. See table 2.

C: Clinical Class (p-value ≤0.001, using the Monte Carlo test)

39.7% of all evaluated lower limbs were C1 (telangiectasias); 35.1% of them had an insufficient venous system. (See Table 3)

In the saphenous veins, it was found that 44.1% of cases had insufficiency of the GSV; 3.7% of the SSV and 9.6% of both saphenous veins. In lower limbs with CEAP C2, half had GSV insufficiency. ( See Table 4)

GSV: Great Saphenous Vein; SSV: Small Saphenous Vein; p-value = 0.227, using the Monte Carlo test

As shown in Table 5, there is a significant association between the CEAP clinical classification and the insufficiency of the SFJ, superficial and deep venous systems.

SFJ: Saphenofemoral Junction; p-value ≤ 0.05 was considered statistically significant, using the Monte Carlo test.

Results

66.7% of lower limbs with mild-moderate CVDLL had great saphenous vein (GSV) insufficiency and 9.0% had insufficiency in both saphenous veins. 7.7% of lower limbs with severe CVDLL had insufficiency in both saphenous veins, with a p-value of 0.011 and assessed by the Monte Carlo test. 50.7% of lower limbs with mild-moderate CVDLL had superficial venous system insufficiency, with a p-value of 0.005 and assessed by the Chi-square test. 29.4% of lower limbs with mild-moderate CVDLL had saphenofemoral junction (SFJ) insufficiency, with a p-value of 0.073 and assessed by the Chi-square test. 27.2% of lower limbs with mild-moderate CVDLL had deep venous system insufficiency, with a p-value of 0.001 and assessed by the Chi-square test. 16.2% of lower limbs with mild-moderate CVDLL had perforating venous system insufficiency, with a p-value of 0.020 and assessed by the Monte Carlo test.

As shown in Table 6, ultrasound findings showed a significant association between severe CVDLL and deep venous system insufficiency.

CI: Confidence Interval; p < 0.05 was considered statistically significant; the regression was bivariate logistic.

Discussion

This study demonstrated the predominance of the female gender in CVDLL, consistent with other authors^5,17,23. The superficial venous system was the most frequently insufficient system; GSV was the most affected, similar to Taengsakul⁵; GSV reflux was the most common in their study population. Andaç N et al.¹⁸observed that the most common segment of GSV with reflux was above the knee. Kanchanabat et al.¹⁹noted that although GSV reflux was present in most patients with lower limb CVI, SSV reflux could occur in a third of patients, especially those with lateral ulceration.

In this study, all dilated GSV and nearly half of the dilated SSV were insufficient, consistent with Choi et al.⁽²⁴⁾, who found that GSV and SSV diameters were significantly larger in patients with reflux, concluding that although vein diameter cannot be used as an absolute reference for venous reflux, it may have predictive value in patients with varicose veins. Kim et al.¹²reported that this relationship was only evident in the lower part of the thigh; Yang et al.⁹found that mean GSV diameters correlated with CEAP progression, but with SSV, the disease progression was less clear. In this study, the most common clinical category was C2: 44.8%, which aligns with Taengsakul⁵at 39%, unlike Porciunculla et al.⁷, who found C3 as the most frequent category at 60%.

It was found that a third of the CEAP clinical class C1 had venous system insufficiency, of which 12.8% was of the saphenous veins, similar to Hong¹⁷, who found a 19.2% prevalence of saphenous vein incompetence in CEAP C1 limbs; additionally, a considerable number of limbs without varices had incompetent saphenous veins.

In this study, 44.1% of lower limbs had GSV insufficiency, 3.6% SSV, and 9.5% both, similar to Hong¹⁷, who reported 71% GSV reflux; 11.9% SSV reflux, and 17.1% both GSV and SSV; however, Kanchanabat et al.¹⁹reported 47.2% GSV reflux; 8.1% SSV reflux, and 25.6% both. Yilmaz et al.²³reported that the most common reflux pattern in patients with GSV insufficiency involved the SFJ with competent malleolar region: 48.9%.

The study showed a relationship between SFJ incompetence and CEAP clinical class, unlike Porciunculla et al.⁷, who found no relationship, but Hong¹⁷did show the correlation between incompetent SFJ and the distribution of incompetent segments in the GSV.

This work found deep venous system insufficiency in 75.5% of mild-moderate grades, much higher than Taengsakul⁵: 57.8%. Hong¹⁷reported that among limbs with deep venous system insufficiency, 98% had popliteal vein insufficiency and 2% femoral vein insufficiency.

This study did not find an association between perforating venous system insufficiency and the CEAP clinical category. Tolu et al.⁶found that varicose veins of lower limbs were related to perforating vein insufficiency in 44.7% of cases and observed a significant relationship between increased diameter of the perforating vein and the presence of perforating vein insufficiency. Huang et al.²⁰found that incompetent perforating veins are a significant risk factor for dermal pigmentation.

One of the limitations of the study was the lack of uniformity in the Doppler reports, which prevented the analysis of other data such as reflux velocity, etc. The strength was that each venous system and its relationship with the clinical category were studied. It is suggested to conduct research on lower limb venous insufficiency in the Peruvian population using other classification systems such as HASTI and the Venous Clinical Severity Score, which are used to assess severity, quantify progression, and treatment outcomes of patients with CVI^2,9.

Conclusions

There is an association between the insufficiency of both superficial and deep venous systems and the CEAP clinical classification. One third of the lower limbs with CEAP clinical class C1 (telangiectasias) showed insufficiency of a venous system.