Prescription Patterns of Drugs Given to Hospitalized COVID-19 Patients: A Cross-Sectional Study in Colombia

The impact of COVID-19 prompted a race to find a treatment that would reduce its mortality. Most studies have not shown favorable results for many of these drugs, but they are still used. The aim as to determine the differences and similarities in the hospital pharmacological management of patients with COVID-19 according to sex, age group, and geographical region of Colombia, 2020–2021. Descriptive cross-sectional study was conducted on the prescription patterns of the medications given to patients diagnosed with COVID-19 treated in eight clinics in Colombia between 6 March 2020 and 31 May 2021. We performed a descriptive analysis of the sociodemographic, clinical, and pharmacological variables of the patients. A total of 8596 patients from 170 cities were identified, with a median age of 53.0 years and 53.3% of them men. A total of 24.3% required care in the intensive care unit, and 18.7% required invasive mechanical ventilation. The most commonly used drugs for the treatment of COVID-19 were systemic corticosteroids (63.6%), followed by colchicine (12.8%), azithromycin (8.9%), and ivermectin (6.4%). Corticosteroids, anticoagulants, colchicine, azithromycin, ivermectin, and hydroxychloroquine were prescribed more frequently in men, and their overall use increased with age. There were differences in prescriptions between geographic regions. The majority of patients were managed with medications included in the management guidelines. There were differences between sexes, age groups, and geographical regions.


Introduction
The type 2 coronavirus that causes severe acute respiratory syndrome (SARS-CoV-2) is the cause of the 2019 coronavirus disease (COVID- 19) [1], which has had a great global impact as a severe public health problem, such that on 12 March 2020, the World Health Organization declared it a pandemic [2]. More than 238 million cases of COVID-19 have been reported, with an approximate mortality of 2.0% worldwide [3]. In Colombia, according to the National Institute of Health, almost 5 million people have contracted and more than 126 thousand people have died from COVID-19 [4]. This pandemic has generated an unprecedented burden on health systems around the world in the form of a greater number of hospital admissions and high demands for intensive care unit (ICU) beds, advanced respiratory support, renal replacement therapy, and other interventions for life support and medical care [5]. The impact of the COVID-19 pandemic on the health system of each country has been different, depending on the balance between the supply and demand of services, which has been associated with the ability to expand the number of hospital beds, in particular in the ICU, and public health policies to contain the pandemic [5,6].
In the midst of the need to find an effective treatment for COVID-19, antiviral drugs, antibiotics, antimalarials, immunosuppressants, immunomodulators, and corticosteroids, among others, have been chosen as therapeutic measures that help reduce the mortality of mechanical ventilation, severe acute respiratory syndrome (SARS), and other complications caused by the virus [7,8]. Many of the proposed drugs have a possible mechanism of action on SARS-CoV-2 or on the pathophysiology of the disease [9,10]. However, most drugs have not demonstrated efficacy through clinical studies, and others still do not have enough evidence to be recommended [7,8]. Even so, they are used in the general population, bringing the risk of drug interactions and adverse drug reactions [11,12].
The objective of this study was to determine the differences and similarities in the hospital pharmacological management of patients with COVID-19 according to sex, age group, and geographical region of Colombia.

Comparisons of Drug Use
Systemic corticosteroids, anticoagulants, colchicine, azithromycin, ivermectin, and hydroxychloroquine were prescribed significantly more often to men than women (Table 1). In general, the proportion of medications used for the management of COVID-19 increased with age, especially after 40 years of age. The antimalarials lopinavir/ritonavir and tocilizumab were not prescribed in children under 18 years of age (Table 2). Corticosteroids were prescribed mainly in the Pacific and Central regions. Colchicine was prescribed in the Pacific, Central, and Caribbean regions. Azithromycin, ivermectin, and chloroquine were prescribed with greater frequency in the Pacific region (Table 3). Tables 1-3 show the comparisons of other sociodemographic, clinical, and pharmacological variables according to sex, age group, and geographic region.

Discussion
This analysis allowed us to characterize the hospital pharmacological treatment of a group of patients with a confirmed diagnosis of COVID-19 treated in eight cities of Colombia. In general, most of the findings of this report regarding sociodemographic data, comorbidities and pharmacological treatment were consistent with what has been described in other studies carried out in the world [13][14][15][16][17][18][19][20][21][22]. The median age was similar to that found in other studies (55.0-60.0 years) [13,14,23] and lower than that reported by others (50.4-68.1 years) [19][20][21][22]24], with a predominance of men, as identified in the majority of studies [13,14,20,[22][23][24][25][26], which may be due to biological factors such as genes, sex hormones and microbiome that may influence the host immune responses to infections [27]; though some reports have had a higher proportion of women (51.2-64.6%) [15,16,28]. The most frequently found comorbidities were arterial hypertension, diabetes mellitus, and chronic obstructive pulmonary disease, in line with other studies [13,14,[20][21][22][23][24]29]. These are pathologies that have been associated with a greater probability of presenting complications, severe forms of the disease and death [30,31]. For this reason, almost one-fifth of the patients required invasive mechanical ventilation, consistent with other reports [32,33], and approximately a quarter needed to be managed in the ICU, a proportion that was similar to that found in Greece (20.0%) [32] and lower than that described in other Colombian studies (32.4-47.8%) [14,33,34].
Most patients received some systemic corticosteroids, similar to earlier reports (56.5-69.7%) [15][16][17]23,34] but contrasting with others that prescribed them less often [19,24,25]. The most often prescribed one was dexamethasone, consistent with what was found in the United States and Colombia [16,24,33,34] but different from what was found in Spain and Peru, where the use of methylprednisolone predominated [17,25,28], and in Pakistan, where hydrocortisone predominated [26]. Its use is based on the findings of the RECOVERY study, which showed that dexamethasone reduced the risk of death by 36% in patients with invasive mechanical ventilation and by 18% among patients who required supplemental oxygen [35]. Dexamethasone is currently recommended by different international [36][37][38] and national clinical practice guidelines for patients [39], especially those with hypoxemia or with oxygen or mechanical ventilation requirements [36][37][38][39]. Although dexamethasone has been shown to be effective in these patients, some serious side effects are associated with its use, such as hyperglycemia, fluid retention, weight gain, bacterial superinfection, confusion, and behavioral changes, so its use must be weighed against the benefits and risks [36][37][38][39]. Another therapeutic group with strong evidence of their worth in patients hospitalized for COVID-19 has been anticoagulants [37,39], used in more than half of the patients in this report, a finding consistent with several studies (50.4-75.0%) [14,16].
Tocilizumab is another drug that has growing evidence in favor of its use [36,37]. In this report, it was prescribed in a small portion of the patients, contrasting with the 3.1-10.8% in other studies [13,15,18,24,28]. Some guidelines recommend it in combination with systemic corticosteroids for hospitalized adults who have rapid respiratory decompensation due to COVID-19 [36,37]. The last update of the Colombian consensus did not give a recommendation for or against the drug due to its cost, its limited availability, and its lack of current approval by the National Institute of Food and Drug Surveillance (INVIMA, for its name in Spanish) for said indication [39], which explains its low use. On the other hand, remdesivir is approved for use in certain patients [36,37], but it is not available in Colombia [39]. Likewise, tofacitinib may be recommended for some adults hospitalized for severe COVID-19 but without mechanical ventilation [36], while other guidelines do not recommend this medication [37,39], which helps explain its absence in this cohort of patients.
Colchicine, an anti-inflammatory agent that is approved for gout, recurrent pericarditis, Behçet's disease, and familial Mediterranean fever [40], is not recommended for use in patients hospitalized for COVID-19 [37,39], and in other clinical practice guidelines, it is not even considered a treatment [36,38]. However, it was found to be prescribed in more than a tenth of the present patients with COVID-19, which is consistent with what was found in Greece (8.2%) [32] and Peru (14.3%) [18] but is notably lower than its prescription rate in two other studies conducted in Colombia (48.2-54.1%) [14,33] because the drug was recommended in the management protocols of the institutions involved in the studies [14,33]. As previously mentioned, this drug is not recommended for the management of patients with COVID-19, but preliminary data from some meta-analyses have shown that colchicine can reduce mortality, although most of the studies included in the analyzes were observational. Therefore, clinical trials are required to corroborate these findings [41][42][43].
Mainly at the beginning of the pandemic, other drugs were also used without solid scientific knowledge but that sought to impact the morbidity and mortality of patients.
The evidence was increasingly robust, showing that medications such as azithromycin, chloroquine, hydroxychloroquine, lopinavir/ritonavir, and ivermectin did not offer benefit in the treatment of SARS-CoV-2 infection and could even be related to a higher risk of pharmacological interactions and adverse events such as QT interval prolongation, arrhythmias, elevated liver enzymes, blood dyscrasias, seizures, skin rash [36][37][38][39]. Thus, there are reports of studies from different countries that show a wide use of azithromycin (60.6-88.6%) [13,25,26,28], which contrasts with the low use found in this report. Likewise, ivermectin was prescribed in 6.4% of our patients, contrasting with that found in Pakistan (2.2%) [26] and Peru (37.0%) [18]. Some of these prescriptions could be founded in sound prophylaxis to avoid hyperinfection by Strongyloides stercoralis when given before the use of corticosteroids [39]. Finally, antimalarials and lopinavir/ritonavir were used in fewer than 2% of patients, which contrasts significantly with other studies [13,17,25,28,32,34,44].
There were evident differences in the prescription of these drugs between men and women, by age group, and by geographical region of the country. This dynamic is not unheard of [15,24]. For example, in two studies conducted in the US, Stroever et al. found that the prescription of corticosteroids predominated in women, and the use of medications such as hydroxychloroquine, tocilizumab, lopinavir/ritonavir, and corticosteroids increased with age up to 65 years. The patterns of use of these drugs varied according to geographical area [15]. Best et al. found such differences in terms of geographic region and age group, and differences were also found depending on race or ethnic group [24]. Several factors can influence the differences in drug use patterns, such as medical decisions, which are affected by demographic, cultural, and economic aspects, as well as the academic training of the doctor. In addition, the differences between countries may reflect the availability or absence of the drug, its costs, the approved indications, the characteristics of the patients, the health system itself, the local management guidelines, and the influence of the media [45,46]. Some limitations should be considered when interpreting our results. First, we did not have access to medical records to verify all pathologies and complications during hospitalization, so we could not verify the accuracy of the diagnoses assigned by the doctor, nor the severity of COVID-19, among other clinical and paraclinical variables. Similarly, any medications prescribed outside the health system or not delivered by the dispensing company are unknown. One strength is that the study included a significant number of patients distributed in most geographic regions of the national territory.

Materials and Methods
A cross-sectional study was conducted on the prescription patterns of drugs given to hospitalized patients with a diagnosis of COVID-19. These were identified from the report of confirmed positive cases by polymerase chain reaction or antigen reactivity performed by the Ospedale Group network, which consists of eight clinics located in the cities of Armenia, Barranquilla, Bogotá, Cartagena, Cali, Manizales, Pereira, and Popayán.
From this population, patients with a first confirmed diagnosis of COVID-19 were selected, of any age, sex, and city of residence, between 6 March 2020 and 31 May 2021, who were treated in-hospital in emergency services, in the general ward, or in the ICU (Figure 1). With this identified patient list, information on the use of medications was obtained from the dispensing company (Audifarma S.A.). A database was set up to gather the following patient variables:  The protocol was endorsed by the Bioethics Committee of the Technological University of Pereira in the category of "research without risk" (approval code: 30-070421). The principles of confidentiality of information established by the Declaration of Helsinki were respected.

Statistical Analysis
The data were analyzed with the statistical package SPSS Statistics, version 26.0, for Windows (IBM, USA). A descriptive analysis was performed by calculating frequencies and proportions for the qualitative variables and measures of central tendency and dispersion for the quantitative variables, depending on their parametric behavior as shown 16  The protocol was endorsed by the Bioethics Committee of the Technological University of Pereira in the category of "research without risk" (approval code: 30-070421). The principles of confidentiality of information established by the Declaration of Helsinki were respected.

Statistical Analysis
The data were analyzed with the statistical package SPSS Statistics, version 26.0, for Windows (IBM, USA). A descriptive analysis was performed by calculating frequencies and proportions for the qualitative variables and measures of central tendency and dispersion for the quantitative variables, depending on their parametric behavior as shown by the Kolmogorov-Smirnov test. Quantitative variables were compared by the Mann-Whitney U test and the χ 2 or Fisher's exact test for categorical variables. The level of statistical significance was p < 0.05.

Conclusions
From the above results, we can conclude that the majority of patients with a confirmed diagnosis of COVID-19 were managed with medications included in the management guidelines. The prescriptions differed by sex, age, and the geographic regions where they were treated. These findings can be useful for clinicians who treat these patients and for decision-makers to strengthen continuing education programs for physicians to optimize the quality of their prescriptions and thereby improve the quality of care and reduce the risks to patients. Funding: This study did not receive funding.
Institutional Review Board Statement: The protocol was endorsed by the Bioethics Committee of the Technological University of Pereira in the category of "research without risk" (approval code: 30-070421).

Informed Consent Statement: Not applicable.
Data Availability Statement: The database is available at protocols.io. Data access: dx.doi.org/10.1 7504/protocols.io.b2dnqa5e.