Get OUT: Factors Associated With a Longer Length of Stay in Patients Admitted With an Acute Gout Flare, and the Effects of Anakinra on Length of Stay

ABSTRACT

Objective

Acute gout flares (AGF) are a significant burden to healthcare systems globally, due to the prolonged length of stay (LOS). The study's aim was to audit patients admitted to our facility with AGF, then re‐audit after the introduction of interleukin‐1 receptor antagonist anakinra in select patients to assess whether it would result in a reduced LOS.

Methods

A single‐center retrospective audit was conducted on 62 hospitalizations for AGF in 2019 that met inclusion criteria. Data on patient demographics, comorbidities, and clinical parameters (number of joints involved, C‐reactive protein (CRP), urate levels, and tophi) were analyzed using multivariable Poisson regression models to determine predictors of LOS. We then compared patients that received anakinra in 2020 (n = 7) with a historical control group (n = 13).

Results

The average LOS in 2019 for AGF was 3.36 days (1.22–4.28). CRP, prior use of allopurinol, and impaired mobility levels were significant predictors of prolonged LOS, with respective coefficients of 0.0036 (95% CI: 0.0023–0.0050, p < 0.0001), 0.321 (95% CI: 0.043–0.599, p = 0.023), and 0.708 (95% CI: 0.416–1.000, p < 0.0001). When CRP was treated as a controlled variable, patients treated in 2020 with anakinra demonstrated a 0.493‐day reduction in LOS compared to the control group, which was statistically significant (p = 0.043).

Conclusion

CRP levels, allopurinol use, and impaired mobility are key predictors of prolonged LOS in AGF. Although the use of anakinra shows promise in reducing LOS and costs, larger studies are needed to confirm its efficacy.

Summary.

• The main predictors of extended LOS for gout patients in this audit were found to be significantly increased CRP levels, previous use of urate‐lowering therapy, and mobility impairment.

• Interleukin‐1 inhibition using anakinra may be a promising adjuvant therapy for patients admitted to the hospital for severe or refractory gout flares.

• High CRP levels may point to the appropriateness of starting advanced anti‐inflammatory treatment, such as anakinra, at an earlier stage to prevent complications of impaired mobility, prolonged hospitalization, and prolonged high inflammatory state.

1. Introduction

Gout, the most common form of inflammatory arthritis, is characterized by recurrent episodes of acute arthritis resulting from monosodium urate (MSU) crystal deposition in joints and periarticular tissues. It has emerged as a public health issue, with acute flares often leading to hospital admissions. In Australia, the incidence of hospitalizations associated with an acute gout flare is increasing, with 1 gout episode per 1000 total hospital admissions [1]. Although there have been advances in treatment options, the length of stay (LOS) for patients admitted with acute gout remains a critical concern. The average LOS in Australian hospitals in 2021–2022 for patients with a primary diagnosis of gout was 5.1 days, in keeping with international studies that have noted LOS ranges of 3–5 days [1, 2, 3].

Despite this, few studies have specifically analyzed factors associated with LOS in hospitalized gout patients, and those that have primarily focus on patient‐related factors rather than the severity of the gout flare itself [2, 3, 4, 5, 6]. A 2012 abstract presented at the American College of Rheumatology and the Association for Rheumatology Health Professionals Annual Meeting found that a delayed diagnosis of gouty arthritis and female sex were associated with an increased LOS at their institution [4]. Similarly, Singh and Yu, who used United States (US) National Emergency Departments (ED) Sample data to search for gout admissions from EDs across the US, identified that older age, renal failure, heart failure, diabetes, or osteoarthritis were associated with a longer LOS [2].

Understanding the factors associated with increased LOS in patients hospitalized with acute gout flares can help to tailor individual treatment plans based on the presence or absence of both patient‐related and gout flare‐specific factors associated with an increased LOS. Current treatment options for an acute gout flare include nonsteroidal anti‐inflammatory drugs (NSAIDs), colchicine, corticosteroids (either systemic or through local injection), and interleukin‐1 (IL‐1) inhibitors.

IL1, an inflammatory cytokine, plays a key role in the inflammatory response associated with an acute gout flare [7]. Anakinra, an IL‐1 receptor antagonist, has shown success in managing acute gout flares, serving both as an effective treatment option and as an alternative for patients with comorbidities, such as patients with diabetes or chronic kidney disease [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18].

Pharmacologically, anakinra reaches its peak concentration within 3–7 h, and it demonstrates faster efficacy than other treatments, with significant reductions in pain typically observed by 48 h. In an analysis of 72 published cases with data on the timeframe to improvement, 41 (57% reported) reached significant improvement after 24 h, and 62 (86% reported) after 48 h [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]. Janssen et al. [15], in a randomized‐controlled trial, showed that anakinra was noninferior to usual treatment with regard to reductions in mean pain scores on Days 2–4 of treatment during an acute gout flare, but by Day 3 there was a significant difference in favor of anakinra, with more patients achieving a ≥ 50% decrease in mean pain scores (odds ratio [OR] 2.56; 95% confidence interval [CI], 1.03–6.37; p = 0.04).

The most common dosage of anakinra is 100 mg/day administered subcutaneously for 3–5 days, and at this dose, it appears to be as safe as other treatments used to manage acute gout flares. Janssen et al. [15] recorded 15 adverse events among 43 patients treated with anakinra, all of which were nonsevere, compared to 21 adverse events in 45 patients in patients who received conventional treatment (NSAIDs, corticosteroids, or colchicine). Furthermore, Thueringer et al. reported safely using anakinra to treat acute gout flares in critically ill patients with severe infections and included patients requiring renal replacement therapy [10].

Campbelltown Hospital is a major metropolitan hospital located in the Macarthur region of New South Wales, Australia, with a catchment of 324,000 people. As of February 2020, anakinra cost Campbelltown Hospital $53 per 100 mg syringe, compared to $0.22 per colchicine 500 μg tablet, $2.22 per vial of methylprednisolone acetate 40 mg for injection, $0.10 per prednisolone 25 mg tablet, and $0.05 per indometacin 25 mg capsule. In contrast, the approximate cost per night for a gout‐related hospital admission was $1904. Therefore, while costs relative to other treatments may limit its use in the community, the evidence that anakinra may act faster at inducing remission would make its use in hospitalized patients favorable and cost‐effective. Despite this, there is currently no published data on the effect it has on hospital LOS in patients with acute gout flares.

The purpose of this study was to audit the inpatient LOS and management of patients admitted to Campbelltown Hospital in 2019 with an acute gout flare as the primary reason for hospital admission. The study aimed to identify factors associated with an increased LOS at our institution to help guide clinical decisions regarding which patients may benefit from more intense therapy. The study also aimed to identify whether anakinra was an appropriate management option alongside conventional therapy for treatment of an acute gout flare in select hospitalized patients.

2. Patients and Methods

The retrospective aspect of this audit was conducted at Campbelltown Hospital over a 12‐month period, from January 2019 to December 2019. All patients admitted to the hospital for an acute gout attack were included in the study.

Eligible participants were identified through an International Classification of Diseases (ICD)‐10 code search of discharge summaries using gout‐specific codes where gout was the primary diagnosis for the admission. A retrospective review of the electronic and laboratory results was performed. Patients were excluded if they developed an acute gout flare while in hospital for another primary diagnosis. Patients were also excluded if they were admitted with complications of gout unrelated to an acute flare, including patients admitted with an infected tophus or joint and patients admitted for an elective gout‐related surgical procedure such as tophi removal. If a patient had multiple admissions for acute gout flares, each admission was analyzed as a separate case. Patients who were classified as having two separations of admission but had not left hospital were counted as only one admission.

The length of hospital stay was calculated from date and time of admission to date and time of discharge, as recorded in the electronic medical record. The number of joints affected was calculated based on the clinical description of joints involved and treated as a continuous variable. It was also categorized into monoarticular (one joint), oligoarticular (two to four joints), or polyarticular (five or more joints) flares. The presence of tophi, as described in the clinical notes, was also noted. Use of NSAIDs, colchicine, and systemic steroids, as well as intra‐articular steroids, was recorded, as was the use of urate lowering therapy (ULT) prior to admission. Nursing documented pain scores, which are a subjective numerical rating from 0 to 10 (with 10 indicating the most severe pain), were recorded. Serum C‐reactive Protein (CRP) and urate levels, and results of any arthrocentesis, were documented. Co‐morbidities, including diabetes, ischaemic heart disease, heart failure, and chronic kidney disease, were also recorded.

Additional factors thought to have the potential to increase length of stay for reasons unrelated to the severity of the flare were also collected. Mobility assessments by physiotherapists or other allied health professionals were recorded, and patients were classified as having impaired mobility if it was documented that they were below their baseline level of mobility and required more time in hospital owing to this fact. If the patient was discharged prior to any allied health assessments, they were classified as having no mobility impairment. The presence of support people at home was also recorded, and patients were classified as either home alone or home with support (including those residing in aged care facilities).

Upon completion of the initial audit, a change to practice was made to include the use of anakinra as an add‐on therapy in addition to conventional therapies in selected patients admitted to our facility with an acute gout flare in 2020 onward. It is important to note that not all patients admitted with a gout flare were given anakinra. The decision to use anakinra was made at the attending physicians' discretion. Reasons for using anakinra were based on the perceived severity of the flare and were influenced by the number of joints involved, inflammatory markers, and mobility impairment. Patients' comorbidities were also considered, as was failure to respond to usual treatments either during hospitalization or prior to admission. For the purposes of this audit, data collection included patients treated with anakinra through to April 2021. Anakinra was given subcutaneously at a daily dose of 100 mg. No routine screening for latent infections was performed prior to the use of anakinra. Similar data to the original audit was collected from these patients prospectively, as well as dates and number of doses of anakinra. To measure whether anakinra resulted in any difference in LOS, the senior rheumatologist in the department, who was responsible for inpatient care in both 2019 and 2020, was presented with deidentified data from the 2019 group. The rheumatologist was blinded to the actual LOS and asked which patients he would have chosen to treat with anakinra had it been available at the time, based on the patient characteristics including CRP levels, mobility impairments, and joint count.

All statistical analyses were performed using Minitab for Windows version 21.1.0. To assess factors associated with an increased LOS, Spearman's rank correlation coefficient (r _s) was used to assess for significant correlation using an initial univariable Poisson regression model (LOS data did fit the Poisson model (p = 0.001)). The level of significance (alpha) was set to 0.05. Subsequently, a multivariable Poisson regression model was created with all predictors that were significant or tended towards significance (p < 0.15), and backwards elimination was used to assess for significant factors associated with LOS. For the anakinra subgroup, further analysis was completed by combining both the pre‐anakinra and anakinra cohorts and performing multivariable analysis using a Poisson regression model with anakinra included as a covariate. Missing data were replaced using mean substitution, which included 11 missing serum urate levels and 9 missing nursing documented pain scores. Student t‐tests were used for continuous variables, and the chi‐squared test was used for categorical variables.

This audit was undertaken as a quality improvement activity by employees of the organization, and approval from the Quality Manager of the facility was granted. A formal ethics review was not required.

3. Results

Over the initial 12‐month period of the 2019 audit, 75 individual acute hospital admissions for gout were identified. Nine episodes were excluded as they were not classified as gout flares, and four additional episodes were excluded as they were different separations of the same admission. This resulted in a final cohort of 62 separate hospital admissions for an acute gout flare as the primary diagnosis in 2019, with a mean LOS of 3.36 days (1.22–4.28) and a median of 2.40 days (Figure 1). Four patients (6%) were readmitted during the study period with recurrent gout flares.

FIGURE 1.

Patient flowchart.

The average age of patients was 67.45 years (95% CI, 63–72), and 76% of patients were male. 29% of patients were on ULT prior to admission (allopurinol in all cases), and 29% had documented presence of tophi, as represented in Table 1. 34% had diabetes, 48% had chronic kidney disease, 39% had a history of ischemic heart disease, and 23% had a history of heart failure. Twenty‐nine percent were classified as having an acute mobility impairment during their admission, and the majority (77%) had support people at home.

TABLE 1.

Analysis of correlation between variables and LOS for admissions for acute gout flares in 2019.

Characteristic	Descriptive statistics (count (percentage) or mean (95% confidence interval))	r s	r s 95% CI	p
Male	47 (76%)	−0.14	−0.45 to 0.17	0.37
Age	67.45 (63.39–71.52)	0.007	−0.002 to 0.015	0.138
Tophi	18 (29%)	0.624	0.35 to 0.90	< 0.0001
CRP (mg/L)	109.5 (86.79–132.3)	0.005	0.004 to 0.006	< 0.0001
Urate (mmol/L)	0.4810 (0.34–0.60)	0.831	−0.27 to 1.93	0.14
Allopurinol use prior to admission	18 (29%)	0.523	0.25 to 0.80	< 0.0001
Mobility impaired	18 (29%)	0.981	0.71 to 1.25	< 0.0001
Diabetes	21 (34%)	0.056	−0.23 to 0.34	0.70
CKD	29 (48%)	0.506	0.23 to 0.78	< 0.0001
IHD	24 (39%)	0.32	0.05 to 0.59	0.02
Heart failure	14 (23%)	0.33	0.03 to 0.63	0.032
Presence of support at home	48 (77%)	0.057	−0.26 to 0.38	0.73
Pain score on day of admission	6.1 (5.3–6.9)	−0.013	−0.06 to 0.03	0.552
Joint count	2.5 (1.9–3.2)	0.0762	0.04 to 0.11	< 0.0001

Abbreviations: CKD, chronic kidney disease; CRP, C‐reactive protein; IHD, ischemic heart disease.

The mean CRP level was 109.5 mg/L (95% CI, 86.8–132.3), and mean serum urate levels at admission with an acute flare of gout were 0.48 mmol/L (95% CI, 0.34–0.60), with 42% of patients having a urate level within the normal range. The mean number of joints involved was 2.6 (95% CI, 1.9–3.2), and the majority of flares were monoarthritis (52%), followed by oligoarthritis (27%) and polyarthritis (21%). The mean nursing‐assessed pain score on the day of admission was 6.1 (95% CI, 5.3–6.9).

Twenty‐nine patients underwent arthrocentesis during their admission, with 28 aspirates confirming the presence of MSU crystals. Despite this, only two patients received intra‐articular steroid injections. A vast majority (79%) of patients received oral corticosteroids during their admission, whereas 37% received colchicine and 18% received NSAIDs.

Among the 34 episodes without an aspirate confirming the presence of MSU crystal during their admission, 10 had previous aspirate results available on our electronic pathology records that confirmed MSU crystal deposition, and another 10 had the presence of tophi documented on clinical examination. This left 14 patients without laboratory or clinical confirmation of MSU deposition. Of these 14 patients, nine had elevated serum urate, and two were treated with allopurinol prior to admission. In total, 10 of these patients met the American College of Rheumatology (ACR)/European League Against Rheumatism (EULAR) Gout Classification Criteria [19] based on the available clinical information.

Univariable Poisson regression models were created to identify which variable could be significant predictors when compared with LOS. Higher joint counts (treated as a continuous variable, r _s = 0.0762 (95% CI: 0.04–0.11, p < 0.0001)), elevated CRP (r _s = 0.005 (95% CI: 0.004–0.006, p < 0.0001)) levels, impaired mobility (r _s = 0.981 (95% CI: 0.71–1.25, p < 0.0001)), presence of tophi (r _s = 0.624 (95% CI: 0.35–0.90, p < 0.0001)), allopurinol use prior to admission (r _s = 0.523 (95% CI: 0.25–0.80, p < 0.0001)), chronic kidney disease as a comorbidity (r _s = 0.506 (95% CI: 0.23–0.78, p < 0.0001)), known ischemic heart disease (r _s = 0.32 (95% CI: 0.05–0.59, p = 0.02)) and heart failure as a comorbidity (r _s = 0.33 (95% CI: 0.03–0.63, p = 0.032)) correlated with a longer LOS (Table 1). Subsequently, multivariable analysis was conducted including the above variables and other variables that approached significance (age and urate levels), and on backward elimination, three variables were found to be significant predictors including CRP (r _s = 0.0036 (95% CI: 0.0023–0.0050, p < 0.0001)), allopurinol use prior to admission (r _s = 0.321 (95% CI: 0.043–0.599, p = 0.023)), and impaired mobility (r _s = 0.708 (95% CI: 0.416–1.000, p < 0.0001)).

Between January 2020 and April 2021, seven patients were treated with anakinra whilst admitted to the study site with an acute gout flare as the primary reason for hospital admission. Three patients received 3 days and four received 4 days of anakinra 100 mg daily by subcutaneous injection. The mean LOS for these patients was 5.0 days, with a mean LOS of 3.2 days calculated from the first dose of anakinra. Thirteen patients from the 2019 cohort were retrospectively selected by the senior rheumatologist as individuals they would have treated with anakinra for their acute gout flare had it been available at the time—functioning as our historical control group (Table 2). Baseline characteristics between the two groups were similar. However, the joint count, when treated as a continuous variable, was significantly higher in the anakinra group, with a mean joint count of 20.14 compared to 6.46 in the control group. When joint counts were categorized as mono‐, oligo‐, or polyarthritis, the results were similar (Table 2), with the majority of patients admitted having five or more joints involved. Treatment with conventional medications was also similar between the two groups (Table 3).

TABLE 2.

Baseline characteristics between anakinra group and control group.

Characteristic	Anakinra group (n = 7, count (percentage); mean (95% confidence interval))	Control group (n = 13, count (percentage); mean (95% confidence interval))	p
Male	7 (100%)	12 (92%)	0.25
Age	53.14 (39.08–67.2)	66.54 (57.03–76.05)	0.08
Tophi	4 (57%)	9 (69%)	0.59
CRP (mg/L)	159.8 (72.32–247.2)	205.3 (148.0–262.5)	0.32
Urate (mmol/L)	0.48 (0.37–0.58)	0.46 (0.39–0.53)	0.77
Allopurinol use prior to admission	4 (57%)	7 (54%)	0.89
Mobility impaired	6 (86%)	7 (54%)	0.15
Diabetes	1 (14%)	4 (31%)	0.41
CKD	2 (29%)	8 (62%)	0.16
IHD	1 (14%)	6 (46%)	0.15
Heart failure	2 (29%)	2 (15%)	0.48
Presence of support at home	5 (71%)	9 (69%)	0.92
Pain score on day of admission	6.3 (2.1–10)	5.4 (3.5–7.5)	0.62
Joint count (continuous variable)	20.1 (3.6–36.7)	6.5 (4.7–8.2)	0.01
Joint count (categorical variable)	Monoarthritis: 0 (0%) Oligoarthritis: 1 (14%) Polyarthritis: 6 (86%)	Monoarthritis: 0 (0%) Oligoarthritis: 2 (15%) Polyarthritis: 11 (85%)	0.95

Abbreviations: CKD, chronic kidney disease; CRP, C‐reactive protein; IHD, ischemic heart disease.

TABLE 3.

Conventional treatments per group.

Treatment during admission	Anakinra group (n = 7)	Control group (n = 13)	p
NSAIDs	2 (29%)	1 (8%)	0.21
Colchicine	3 (43%)	2 (15%)	0.18
Systemic corticosteroids	5 (71%)	13 (100%)	0.11
Intra‐articular corticosteroids	1 (14%)	1 (8%)	0.64

Abbreviation: NSAIDs, nonsteroidal anti‐inflammatories.

The mean LOS for patients receiving anakinra was 5 days, compared to 6.3 in the control group, which did not reach significance (p = 0.27) (Figure 2). When the mean LOS was analyzed from the time of the first dose of anakinra to the time of discharge, a mean LOS of 3.2 days was observed compared to 6.3 days in the control group (p = 0.076) (Figure 3). A further multivariable regression model was created using the combined populations (anakinra and historical control groups) with the anakinra group used as a covariate, which found that there was a 0.662‐day difference between the cohorts (p = 0.005), in addition to a 0.3543 day increase in length of stay for both cohorts for every 100‐point increase in CRP levels (p = 0.001). If a patient had the same CRP level in both cohorts (CRP treated as a controlled variable), there was noted to be a reduction in LOS of 0.493 days (95% CI 0.016–0.969, p = 0.043) with R ² = 0.32, meaning that this regression model would be significant for 32% of the population described.

FIGURE 2.

Length of stay (LOS) for anakinra vs. control group (column graph refers to group mean). ●, Data point for patients who received anakinra. ○, Data point for historical control patients who did not receive anakinra. Height of the column graph refers to mean LOS of that particular group.

FIGURE 3.

Length of stay (LOS) from time of first anakinra dose vs. control group (column graph refers to group mean). ●, Data point for patients who received anakinra. ○, Data point for historical control patients who did not receive anakinra. Height of the column graph refers to the mean LOS of that particular group.

No adverse events were noted in patients receiving anakinra, with no injection site reactions or infections documented. One patient (14%) was readmitted with a relapse of his gout flare 3 weeks after discharge in the anakinra group, compared to one patient out of 13 in the control group.

4. Discussion

This audit establishes that acute gout flares are an important cause of hospital admission in the Macarthur region, New South Wales, Australia, with 62 inpatient admissions for an acute gout flare as the primary reason for admission in 2019, comparable to published rates Australia‐wide and in other regions [1, 20]. These findings highlight that gout imposes a substantial burden on both our hospital and the wider community, particularly as there is a proportionately larger population of high‐risk gout ethnicities living in the multiculturally diverse region, such as New Zealand Māori and Pacific Islander ethnic groups.

The patient demographics in our cohort were similar to that of other published studies, with males being disproportionately affected, patients being of an older age, and frequently having comorbidities such as diabetes, heart failure, renal impairment, and ischaemic heart disease [1, 2, 20].

In analyzing factors associated with a longer LOS for an acute gout flare requiring hospital admission, we found higher CRP levels, allopurinol use prior to admission, and impaired mobility to be associated with longer LOS. This is in contrast to the findings of Sharim et al. [4], who observed that in a hospital inpatient population, a later diagnosis of gouty arthritis and female sex were associated with an increased LOS in their study conducted at a single center. Similarly, our results differed from those reported by Singh et al. [2], whose study focused on reasons for emergency department (ED) visits and included other critical patient and hospital characteristics such as age, sex, insurance status, location, and comorbidities. Singh et al. found that older age and the presence of renal failure, heart failure, diabetes, or osteoarthritis were associated with longer LOS. Our finding of higher CRP levels correlating with a longer LOS suggests that CRP serves as an important objective marker of disease severity during acute gout flares. This aligns with the findings from Janssen et al. [6], who reported that CRP levels exceeding 30 mg/L were associated with increased risk of a recurrence of gout flare in the first 3 months of starting ULT during an acute flare. Impaired mobility, unsurprisingly, was associated with a longer LOS, as it limits people's ability to function independently at home and in the community. This is particularly relevant for older patients, who are becoming more frequently admitted with acute gout flares. In this context, the effects of a gout flare on mobility appear to reflect the broader impact that the flare has on a patient's functioning, rather than a direct marker of disease severity. In addition to this, patients on ULT prior to admission tended to have a longer length of stay. This contrasts with data from Jatuworapruk's cohort study [21] (n = 36 047), where they observed a 10% reduction in LOS with patients on ULT prior to admission.

In the analysis of LOS among patients treated with anakinra, there were no significant differences observed between patients who received anakinra compared to the control group who would have been considered for anakinra therapy had it been available, apart from the joint count when taken as a continuous variable. When joint count was taken as a categorical variable, there was no observed difference between groups with most patients with polyarticular disease, demonstrating the loss of data clarity when clustered. Hence, this makes it an insignificant variable, noting that in the initial retrospective cohort analysis, joint count, when taken as a continuous variable, did not have a significant correlation with LOS regardless. The control group exhibited higher CRP levels, whilst the anakinra group had higher rates of impaired mobility, neither of which reached statistical significance. Patients in the anakinra group who were on ULT prior to admission tended to have a lower mean LOS compared to those in the control group who were not (5.25 (3.78–6.72) and 7.41 (4.18–10.64) days, respectively), but this difference is not statistically significant (p = 0.449).

Despite the use of anakinra in selected patients, no differences in the use of conventional treatments were observed between groups. While there was a reduction in mean LOS in patients who received anakinra in comparison to the control group (5.0 and 6.3 days respectively), this did not reach statistical significance. However, when the control group was compared to LOS analyzed from the time of the first dose of anakinra to the time of discharge (6.3 and 3.2, respectively), this approached statistical significance (p = 0.076). This 1.8‐day difference may be attributable to the delay in commencing anakinra due to permission being required on each occasion for usage, and hence there would be a potentially clearer advantage if started earlier rather than waiting for an inadequate response to oral corticosteroids. However, on further multivariable analysis with the creation of a Poisson regression model, it was shown that there is a reduction in LOS of 0.493 days in the anakinra cohort (p = 0.043) if the only significant covariant, CRP, was taken as a controlled variable. In addition to this, as CRP was a significant variable, it was shown that for every 100‐point increase in CRP levels, LOS increased by 0.3543 days, making it an important variable to note when patients are admitted to the hospital. If CRP was not taken as a controlled variable, there would be a 0.662‐day difference in LOS. This would represent a cost saving of AUD $1048 per patient in 2019, and annual savings of AUD $64 976, based on the admission rate of 62 patients from our 2019 data, which considers the added cost of anakinra to the hospital. If this finding could be attributed to Australia as a whole, there would be a potential cost saving of AUD $7 440 800 based on the 2022 Australian Institute of Health and Welfare's reported 7100 gout hospitalizations in a year [22].

Anakinra appeared safe to use in our cohort, with no adverse events noted. Overall, readmissions were relatively common, with four readmissions (6.5%) from 62 admissions in 2019, which is in keeping with the 7.8% readmission rate that Kennedy et al. [20] found in their audit. There was one readmission 3 weeks following discharge in a patient treated with anakinra; however, it is important to note that this patient had a history of difficult‐to‐control gout. There was no significant difference in the rates of readmission between groups.

The strengths of our project are that it is the first Australian study to examine factors associated with an increased LOS, and the first to analyze the impact of anakinra on LOS. However, several limitations should be noted. The retrospective nature of the study introduces potential bias, particularly in regard to the retrospective clinical decisions for inclusion into the historical control group. Additionally, a major limitation of this study was its small sample size, which limits the applicability of this study to larger populations. If the applicable population proportion was set to 32% given R ² = 0.32 of the final Poisson regression model, the minimum sample size would have to be n = 53 to apply to the Campbelltown Hospital area, and n = 335 for global application of the analyzed data [23]. Several patients were already on treatment prior to commencing anakinra, and the timing of anakinra varied during admissions, which may have affected the LOS data.

Gout represents a significant burden on the community and hospitals, and finding ways to improve the management of acute flares is justified. We found higher CRP levels, prior use of ULT, and impaired mobility to be associated with an increased LOS during hospitalization for an acute gout flare. Although the reduction in LOS observed in this study did not achieve statistical significance, the results suggest its potential economic and therapeutic utility. If validated, these findings could support the integration of anakinra as an effective adjunctive treatment, particularly in those with the highest cost savings potential, for example, patients with tophi, polyarticular gout, and mobility impairment. Clinical markers such as CRP and impaired mobility could be used to identify candidates for anakinra use. Future research should focus on larger, multicenter randomized‐controlled trials to confirm these findings and establish standardized treatment protocols. Additionally, such studies could also explore the optimal timing and duration of anakinra administration to maximize its clinical benefits. Future studies could investigate the long‐term outcomes of patients treated with anakinra through flare recurrence rates, as well as long‐term safety in diverse patient groups, including those with significant comorbidities.

Author Contributions

Hansaja Weerawardena: further data and cost analysis, writing of the manuscript, background research, editing and liasing with authors. Cameron Louis Adams: study design, collecting of data, initial data analysis, drafting of manuscript. Narainraj Kamalaraj: study design, editing, providing feedback. Kevin Pile: study design, editing, providing feedback.

Conflicts of Interest

The authors declare no conflicts of interest.

Weerawardena H., Adams C. L., Kamalaraj N., and Pile K., “Get OUT: Factors Associated With a Longer Length of Stay in Patients Admitted With an Acute Gout Flare, and the Effects of Anakinra on Length of Stay,” International Journal of Rheumatic Diseases 28, no. 9 (2025): e70418, 10.1111/1756-185x.70418.

Data Availability Statement

Deidentified data that support the findings of this study are available from the corresponding author, [H.W.], upon reasonable request.