Healthcare costs and mortality associated with serious fluoroquinolone‐related adverse reactions

Abstract The aim of this study was to estimate healthcare costs and mortality associated with serious fluoroquinolone‐related adverse reactions in Finland from 2008 to 2019. Serious adverse reaction types were identified from the Finnish Pharmaceutical Insurance Pool’s pharmaceutical injury claims and the Finnish Medicines Agency’s Adverse Reaction Register. A decision tree model was built to predict costs and mortality associated with serious adverse drug reactions (ADR). Severe clostridioides difficile infections, severe cutaneous adverse reactions, tendon ruptures, aortic ruptures, and liver injuries were included as serious adverse drug reactions in the model. Direct healthcare costs of a serious ADR were based on the number of reimbursed fluoroquinolone prescriptions from the Social Insurance Institution of Finland’s database. Sensitivity analyses were conducted to address parameter uncertainty. A total of 1 831 537 fluoroquinolone prescriptions were filled between 2008 and 2019 in Finland, with prescription numbers declining 40% in recent years. Serious ADRs associated with fluoroquinolones lead to estimated direct healthcare costs of 501 938 402 €, including 11 405 ADRs and 3,884 deaths between 2008 and 2019. The average mortality risk associated with the use of fluoroquinolones was 0.21%. Severe clostridioides difficile infections were the most frequent, fatal, and costly serious ADRs associated with the use of fluoroquinolones. Although fluoroquinolones continue to be generally well‐tolerated antimicrobials, serious adverse reactions cause long‐term impairment to patients and high healthcare costs. Therefore, the risks and benefits should be weighed carefully in antibiotic prescription policies, as well as with individual patients.


| INTRODUC TI ON
Fluoroquinolones (FQs) are a group of antimicrobial drugs, which have activity against gram-positive and gram-negative bacteria and are used to treat genitourinary, respiratory, gastrointestinal, skin, and soft tissue infections. FQs have been in wide clinical use for over 30 years. In Finland, the number of FQ prescriptions has steadily remained at 5-6% of all antibiotic prescriptions. 1 Although FQs are still considered to be mostly well tolerated, there have been increasing concerns about their safety. Their most common adverse reactions are mild and reversible, such as diarrhea, nausea, or headaches, and lead to termination of treatment in less than two percent of patients. However, fluoroquinolones are also associated with serious adverse drug reactions, such as recurring clostridioides difficile infections (CDI), cardiovascular toxicity, musculoskeletal, renal, and liver disorders, and reactions involving the central nervous system. 2 Although the mechanisms behind FQ-related adverse effects remain unclear, for example, mitochondrial damage has been implicated in the process. 3,4 Serious adverse reactions are defined as events that either result in death, are life-threatening, require inpatient hospitalization or prolong current hospitalization, or result in persistent or significant disability or incapacity. 5 Due to reports of serious ADRs associated with the use of FQs, the European Medicines Agency has recommended restrictions on their use several times. 6 Although the total burden of FQ-related ADRs has previously not been calculated, ADRs are generally recognized as an important cause of morbidity and even death. For example, according to Finnish studies, about 8% of hospital emergency visits and 3%-5% of in-hospital deaths are explained by ADRs, corresponding to approximately 0.05% of all hospital admissions. [7][8][9] Apart from direct health consequences, estimates of the economic burden of unsafe care due to ADRs have varied between 0.2% and 6.0% of the total health expenditure. 10 The impact of the ongoing COVID-19 pandemic on FQ-related ADRs is still unknown. For example, according to a systematic review from 2021, 1% of COVID-19 patients were reported to have clostridioides difficile infections (CDI). 11 However, the authors conclude that there is little change in the overall incidence of CDI compared to the time before the pandemic.
While it is important to understand the economic burden of ADRs, research of the costs is challenging since most of them remain unreported. 12 The findings of a systematic review from 2016 propose that published literature does not provide sufficient information on ADRs and that the majority of information is found in 'grey literature', such as government reports, working papers, press releases, theses, and conference proceedings. 13 The aim of this study was to estimate healthcare costs and risk of mortality associated with serious FQ-related adverse reactions in Finland between 2008 and 2019 using registry-based and literature data and a decision tree modeling approach.  FQs prescribed in the study comprised ciprofloxacin, levofloxacin, moxifloxacin, ofloxacin, and norfloxacin.

| Severe clostridioides difficile infections
Clostridioides difficile bacteria are members of normal gut microflora that are resistant to numerous antimicrobials and can thus colonize the human gut after antimicrobials have altered the proportions of the microbiota. The toxin expression of the bacteria results in gastrointestinal illness. 20 The overall incidence of CDI after the use of FQs is estimated to be between 3 and 10 percent. 16,20 Since the emergence of the epidemic Clostridioides difficile ribotype 027 clone, CDI has become more prevalent, severe, and more difficult to treat due to resistance to many antimicrobial agents. 21 The clinical course of the disease can range from asymptomatic colonization or mild diarrhea to severe infections characterized by pseudomembranous colitis, toxic megacolon, colonic perforation, or death. 20 Advancing age and a history of partial colostomy are among the risk factors associated with more severe disease. 22 The treatment options of CDI include vancomycin, metronidazole, fidaxomicin, fecal microbiota transplants, and, in life-threatening cases, the surgical management of CDI complications. 23 Severe CDI (SCDI) increases the risk of mortality to 57%. 24 Additionally, patients over 80 years and those with increased inflammatory parameters and diagnosed sepsis have an increased risk of mortality. 25 Previous research suggests that CDIs are associated with the highest costs among FQ-related ADRs, although published literature remains scarce. 26 According to the Finnish National Infectious Diseases Register, the overall number of CDI cases has reduced by 60% since 2008. 27

| Tendon ruptures
FQ s were first connected to tendon injuries in 1983 28 and these ADRs remain the most distinctive and reported FQ ADRs to this day. The incidence of fluoroquinolone-related tendon injuries has been estimated to average 15-20 per 100 000 patients. The risk of tendon injuries is higher with over 60-year-old people, in particular males, and patients on corticosteroid therapy. 29 Regardless of relieving pressure and avoiding activities that strain the tendon, fluoroquinolonerelated tendon injuries may result in tendon ruptures, which are treated with immobilization and/or corrective surgery. 30 Although tendon ruptures might be risk factors for the physical and psychological decline, there is no evidence of them being a direct cause of death.

| Severe cutaneous adverse reactions
Hypersensitivity to FQs can cause rare severe cutaneous adverse reactions (SCAR), which can be life-threatening and is often preceded by hypersensitivity to beta-lactams. 31 Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) represent a spectrum of severe reactions varying with respect to the proportion of skin affected. SJS affects less than ten percent of the skin, whereas TEN is more serious and involves more than 30 percent of the body surface area. 32 Overlapping of SJS/TEN is seen in 5% of cases. The mortality rates associated with SJS and TEN are between 10 and nearly 50 percent. 33 39 Yet, mortality associated with severe liver injuries is known to be high. 38 The use of corticosteroids may be beneficial, although long-term data for the treatment of DILI is still lacking. 40

| Aortic dissection and rupture
The array of FQ-associated cardiovascular toxicity has amplified in recent years. QT interval prolongation and torsade de pointes in connection with FQs were described in the 1990s in several publica-

| Modeling approach to adverse reactions
The decision tree is a commonly used tool for decision analysis in

| Sensitivity analyses
Due to varying estimates of FQ-related ADR and mortality incidence, the decision tree model contains uncertainty about the exact probabilities used as inputs. To deal with parameter uncertainty in the model, sensitivity analyses were conducted to assess their impact on direct healthcare costs and ADR-related mortality. 60 The lowest probability for an ADR and mortality generated the best-and highest probability yielded the worst-case scenario, respectively.     65 According to a recently published rapid review and meta-analysis, the global prevalence of antibiotic prescribing has been around 75% in COVID infections, while bacterial co-infections are only present in 6%-8% of patients, suggesting a fair amount of unnecessary use. 66 The increased use of broad-spectrum antimicrobials is expected to add to the burden of CDI, and rates were reported to have increased from 3.32/10,000 patient-days to 3.60/10,000 patient-days already during early 2020. 67 The risks and benefits of FQs should be weighed carefully also during the COVID pandemic since patients most vulnerable to FQ-related serious ADRs are often the ones most seriously affected by the virus.

| CON CLUS IONS
Although FQs continue to be generally well-tolerated antimicrobials, serious ADRs can cause long-term disabilities to patients and high healthcare costs. In fact, on average, FQ-related ADRs appear to be approximately ten times more expensive than the FQ medications themselves. The most impactful of the FQ-related ADRs are clostridioides difficile infections, which alone cause more than 95% of the total direct healthcare costs of the ADRs.
Accordingly, from the economical perspective, efforts should be focused on limiting the risk of FQ-associated CDI. Nevertheless, the different risks and benefits of these drugs should always be weighed carefully.

E TH I C S S TATEM ENT
This is a retrospective observational study with registered data, and neither an ethics approval nor patient consent was required.

ACK N OWLED G EM ENTS
This research received no funding from any funding agency in the public, commercial, or not-for-profit sectors.

D I SCLOS U R E
The authors of this manuscript have no conflicts of interest to declare.

AUTH O R CO NTR I B UTI O N
All persons who meet authorship criteria are listed as authors. All authors took part in conceptualizing the idea, designing the study, analyzing the data, and writing the manuscript. All authors have seen and approved the final version of the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.