Prescribing patterns of echinocandins in adult patients in a private hospital in Gauteng, South Africa

Background Worldwide, the leading cause of invasive candidiasis and the fourth leading cause of hospital-acquired infections are the Candida species (spp.) group. One of the most important tools in fighting such drug-resistant fungi is the appropriate use of antifungal agents. Objectives The study aimed to determine echinocandins’ general prescribing patterns and how they are associated with the treatment period. Method A quantitative, observational, and descriptive was used, and included patients receiving antifungal treatment in a private hospital in Gauteng, South Africa between 01 January 2015 to 31 December 2015. Results Of the 146 patient files included, 102 patients (69.9%) received caspofungin and 44 patients (30.1%) were treated with anidulafungin. For the former, 99 (97.1%) patients received a loading dose (LD) of 70 mg, while 200 mg anidulafungin was only prescribed to 30 patients (68.2%). In line with maintenance dose guidelines, the majority (98.1%) of caspofungin-treated patients received 50 mg IV daily, whereas 4 (3.9%) patients were treated at higher doses (70 mg daily). Anidulafungin was administered at various maintenance doses, including 400 mg (2.3% of patients), 200 mg (52.3%), 100 mg (43.2%) and 50 mg (2.3%) IV daily. Conclusion Our results can be utilised to produce a hospital-specific algorithm in terms of Candida-infected patients. Contribution These findings contribute to our understanding of prescribing patterns of antifungal agents and the impact thereof on treating Candida spp. Infections.


Introduction
Candida species are the main source of invasive candidiasis globally; they are the fourth leading cause of infections acquired within hospital settings. In addition, they are associated with dramatic mortality and morbidity rates. 1 According to recent reports, while the leading fungal threat to public health 2 remains Candida albicans resistance, the number of patients infected with nonalbicans Candida spp. is steadily growing. These species include Candida glabrata, Candida parapsilosis, Candida krusei and Candida tropicalis, 1,3,4,5,6 which potentially contribute to the increased drug resistance to commonly prescribed pharmacological antifungal drugs. As such, it is widely recognised that the epidemiology-related changes of the Candida spp. contribute significantly to antifungal drug resistance.
While research acknowledges that the correct use of antifungal agents is one of the most important factors contributing to drug resistance, inappropriate antifungal use results in fungal species' resistance developing over time. Negative usage factors include sub-therapeutic dosing and inappropriate or ineffective treatment periods aimed at reaching expected clinical outcomes. 7,8 http://www.sajid.co.za Open Access Despite fluconazole being one of the most commonly prescribed antifungals, an increase in resistance has been observed in Candida spp. This rise in numbers may be because of increased prescription rates, previous exposure and resistant mechanisms, 1,10 such as genetic alterations.
Consequently, this has necessitated the use of alternative treatments such as the use of echinocandins, such as caspoand anidula-fungin, which inhibit fungal cell wall synthesis. They do this by inhibiting the β-D -(1-3)-glucan synthase enzyme to prevent the formation of β-D -(1-3)-glucan polymers in the cell wall, and, ultimately, render the cells vulnerable to external stressors. 11 In addition to proving effective against Candida spp., they further possess an attractive safety profile and are associated with minimal drug-drug interactions and adverse effects at therapeutic doses, 12 particularly in relation to other commonly prescribed antifungals, such as fluconazole.
Despite the aforementioned, a recent South African (SA) report observed multidrug resistance, especially for C. albicans, C. glabrata and C. tropicalis. This finding indicates a formidable threat to the public-health sector, which could be indicative of the inappropriate prescription and use of antifungal treatments. 1 While the narrow selection of available treatments complicates Candida infections, the situation is exacerbated by these organisms having biofilm activity within indwelling devices and catheters. 13 This unique characteristic of Candida spp. has led to many clinical guidelines recommending the removal or replacement of indwelling devices after diagnoses of invasive candidiasis or candidemia. 5,14,15 In 2012, the European Society for Clinical Microbiology and Infectious Diseases (ESCMID), recommended that treatment be simplified to an oral agent after 10 days of intravenous (IV) treatment if an individual patient is stable and able to tolerate an oral form and when the species is susceptible to the treatment. 15 Furthermore, in its 2016 guidelines, the Infectious Diseases Society of America (IDSA) recommend that clinically stable patients' treatments be changed from an IV form to an oral agent (i.e. fluconazole) within 5-7 days. In this regard, however, it is imperative that both suspected or confirmed Candida spp. be susceptible to at least one of the available oral agents. In addition, individual patients should have negative blood cultures repeatedly. 14 In general, this approach, namely the changing of IV to oral therapy, is aimed at minimising treatment and hospitalisation periods, among others. In SA, a local Quality Alert was issued by a private hospital group that based its recommendations on the ESCMID guidelines, which pharmacists working for this particular chain were to implement to ensure that echinocandins are appropriately used.
The duration of therapy for patients with the previously mentioned Candida infections remains a challenge because of various reasons. These factors include the following 10,14 : firstly, candidemia is associated with a high mortality rate; secondly, there are several limitations related to an early diagnosis; thirdly, the variety of different Candida spp. are geographically distributed, each having its own unique challenges; fourthly, some species possess biofilm activity; and, lastly, the Candida can be deep-seated within different organs. While these treatment-related challenges have hampered guidelines and the literature in terms of reaching an accurate consensus regarding the recommended length of treatment period, research has established that a 14-day course of IV treatment should be implemented after the first negative blood culture. Consequently, this study sought to determine whether doctors comply with the available guidelines.

Methods
The study took place within the context of a private hospital in the SA province, of Gauteng. The target population comprised of all patients receiving antifungal treatments from 01 January 2015 to 31 December 2015. Furthermore, in order to take part, patients had to comply with the study's inclusion criteria. Patients, therefore, had to be adults admitted to a specific private hospital who were receiving antifungal treatment during their hospitalisation. A total of 146 met the criteria and, consequently, took part in the study.
This particular study followed a quantitative research design method and was observational and descriptive in nature. This approach was chosen because the data were retrospective and had no influence on the findings.
Using a data collection form, this study captured the following data fields, namely patients' demographic information, including age and ward admitted in hospital; type of the IV echinocandin that the patient was initiated on; loading dose (LD); prescribed daily dose (PDD); starting date and end date of echinocandin treatment; and cost of antifungal treatment, including the blood tests and blood cultures performed.
Further inclusion criteria included: (1) all patients 18 years of age and older and (2) the period during which all patients 18 years of age and older were started on IV echinocandin treatment from 01 January 2015 to 31 December 2015. The specific time frame was chosen because anidulafungin only became available at the end of 2014 in SA. Patients were excluded if (1) they were admitted while on any acute or chronic antifungal treatment, (2) pregnant patients and (3) patients whose echinocandin treatment was changed during their hospital stay to some other IV antifungal.
Important variables in this study included the type of antifungal, the dosage of antifungal treatment, the duration of treatment, the de-escalation of therapy, blood results and cultures and the average cost of antifungal treatment.
In consultation with the Statistical Consultation Services of the NWU, the Statistical Analysis System ® , SAS 9.3 ® (SAS Institute Inc., 2009) was used to evaluate the data. In this regard, descriptive statistics, such as frequencies (n), percentages (%), means, standard deviations (s.ds), and 95% confidence intervals (CIs), were used to express all variables. In addition, the study used Cohen's d-value to define the practical significance of the results (with d ≥ 0.8 seen as a large effect with practical significance). To determine the association between two groups, the two-sample t-test was used and the association between two categorical variables were determined by the Pearson's chi-square test. Lastly, Cramér's V was used to determine the practical significance of the results (with V ≥ 0.5 accepted as a large effect with practical significance).

General prescribing patterns of echinocandins
One hundred and forty-six patient records were retrieved. These were examined to determine echinocandin-related prescription patterns.

A comparison of the average duration of treatment of patients with blood cultures and those without
An independent t-test and Cohen's d value were used to compare the average treatment duration of patients with blood cultures to those without. The results are presented in Table 2.
Here, of the 146 patients in the study, the mean days on therapy were 9.7, regardless of whether blood cultures were performed.

International guidelines and general patterns of prescription
As seen in Table 3, three international guidelines and a local Quality Alert by a private SA hospital group were studied to explore the correct LD and PDD for echinocandins (Messina A, 2017, personal communication, July 01). 5,14,15,16 These guidelines reached a consensus in terms of the dosing of this particular drug. Subsequently, they were used to measure the degree of compliance in the prescribing patterns of the drug; particularly, in adult patients who were privately hospitalised. Within the context of this study setting, we noted that these drugs were, on the whole, correctly prescribed, according to the LD and PDD. As mentioned, the caspofungin-related results were recorded as 98 patients who received daily 50 mg IV and four who received 70 mg IV per day. The literature recommends 1,3,4,5,7,8,14,17 that the daily dose of this drug should be 50 mg IV daily, while the anidulafungin-related dosage is 100 mg IV per day. The study, therefore, concluded that although most patients were receiving the correct daily dose of caspofungin, the daily dose of anidulafungin was not adhered to.

A comparison of the average duration of treatment in patients with blood cultures and those without
Out of the total number of patients (146), 129 were recorded as having positive blood cultures, while only 17 did not have a blood culture at all. Moreover, the mean duration of echinocandin therapy for the former was lower than those without, namely 9.7% and 10.0%, respectively. This mean was considered while measuring the blood cultures' performances because the literature states that therapy should be ceased after 14 days of IV treatment, after patients' first negative blood cultures. 5,14,15 This indicates that those patients with positive blood cultures were administered antifungal therapy for a shorter period than their counterparts without blood cultures.
The s.d. for positive blood culture patients was also determined to be lower than those without. This implies that their values have less variability than the values of those  without blood cultures. In this regard, both medians are smaller than the corresponding means, which suggests that there is a positive skew in their distribution. Consequently, the probability density function has a longer tail on the right.
The study further established that the p-value of 0.888 was greater than the p-value of significance (0.05) meaning that there was no statistically significant difference between the average duration of treatment of patients with blood cultures and those without. Moreover, it can be seen as a large effect when Cohen's d value is ≥ 0.8. In this study, Cohen's d value = 0.031, which indicates that there is a small effect of no practical significance (Messina A, 2017, personal communication, July 01). 5,14,15,16 In light of the given results, this study has proven important in terms of the relevance of antifungal stewardship and the practice of appropriate dosage.

Conclusion
From the given results and discussion, it is concluded that prescribing doctors working in this particular setting comply with international guidelines regarding the dosing of echinocandins. Moreover, it was determined that the presence or lack thereof of a blood culture does not greatly impact the duration of the treatment therapy. Consequently, we recommend that more research be conducted to accurately compare guidelines with the actual compliance with the duration period.
Various principles regarding echinocandins have been set according to the international guidelines pertaining to the prescription thereof (Messina A, 2017, personal communication, July 01). 5,14,15,16 In order to ensure pharmacists use echinocandins appropriately, the local hospital group set up a Quality Alert. This alert was already being implemented in 2012, at least 3 years before the onset of echinocandinresistant organisms, such as C. auris. 1,3,9,17,18 Subsequent reports have established that this type of resistance is increasingly common, and, therefore, the appropriate use of echinocandins should be promoted with particular attention paid to therapy durations, the correct dosages, and the uses of blood cultures to assist practitioners in de-escalating Candida species and the total cost of treatments relating to them. 9,17,18 This study aimed to investigate the available literature pertaining to the utilisation of echinocandins in comparison with the current patterns of prescription of these drugs within a private-practice setting. The results can be used to develop hospital-specific algorithms for Candida-infected patients. In addition, this study highlighted the value of guidelines related to available antifungals and how this implementation may lead to more positive outcomes.
This study had various limitations, namely its outcomes depending on how thoroughly patients' prescription charts were completed by doctors; the researcher's lack of clinical data in terms of changing catheters and lastly, the study's reliance on a random sample because of its inclusion and exclusion criteria. In the light of this, future studies may render more accurate results owing to more correct prescription charts.