Clinical Characteristics and In-Hospital Outcomes in Patients with Iliopsoas Abscess: A Multicenter Study

(1) Background: Iliopsoas abscess (IPA) is usually overlooked due to its nonspecific symptoms and signs. The resulting delayed diagnosis and treatment can increase morbidity and mortality. The purpose of the present study was to identify the risk factors for the unfavorable outcomes associated with IPA. (2) Methods: We included patients who presented to the emergency department and were diagnosed with IPA. The primary outcome was in-hospital mortality. Variables were compared, and the associated factors were examined with Cox proportional hazards model. (3) Results: Of the 176 patients enrolled, IPA was of primary origin in 50 patients (28.4%) and of secondary origin in 126 (71.6%). Skeletal origin was the most common source of secondary IPA (n = 92, 52.3%). The most common pathogens were Gram-positive cocci. Eighty-eight (50%) patients underwent percutaneous drainage, 32 (18.2%) patients underwent surgical debridement, and 56 (31.8%) patients received antibiotics. Multivariate analyses indicated that age > 65 (year) (HR = 5.12; CI 1.03–25.53; p = 0.046), congestive heart failure (HR = 5.13; CI 1.29–20.45; p = 0.021), and platelet < 150 (103/μL) (HR = 9.26; CI 2.59–33.09; p = 0.001) were significant independent predictors of in-hospital mortality in Model A, while the predictors in Model B included age > 65 (year) (HR = 5.12; CI 1.03–25.53; p = 0.046) and septic shock (HR = 61.90; CI 7.37–519.46; p < 0.001). (4) Conclusions: IPA is a medical emergency. Our study reported that patients with advanced age, congestive heart failure, thrombocytopenia, or septic shock had a significantly higher risk of mortality, and the recognition of the associated factors may aid in risk stratification and the determination of the optimal treatment plan for IPA patients.


Introduction
The iliopsoas abscess (IPA), a suppurative collection within the compartment of the psoas and iliacus muscles, was first reported by Mynter et al. in 1881 [1]. IPA is an easily overlooked but potentially serious infectious disease due to its insidious onset and rarity [2,3]. The classic triad of fever, flank pain, and hip movement limitation-referred to as the psoas-muscle sign-is only present in 30% of its patients, resulting in difficult diagnosis [4]. Delays in the appropriate treatments may lead to increased morbidity and mortality [5,6].
Given that the iliopsoas muscles are adjacent to certain organs, they are susceptible to infections from these contiguous structures via the blood supply and overlying lymphatic system [5]. Primary IPA occurs when the causative organism spreads through the bloodstream or lymphatic system from a distant site of infection, and it accounts for approximately 30% of all IPA cases. In contrast, secondary IPA develops as a result of direct extension of an adjacent infectious process into the iliopsoas muscle, and it accounts for the remaining 70% of cases [7]. The associated risk factors included diabetes, intravenous drug use, renal failure, human immunodeficiency virus (HIV) infection, and other forms of immunosuppression. Trauma and hematoma formation can lead to the development of psoas abscesses, especially for secondary IPA [8]. The mortality rates were significantly different among the two categories of IPA, with approximately 2.4% for primary IPA, and up to 18.9% for secondary IPA [9].
In recent years, because imaging modalities, e.g., gallium-67 inflammation scan, computed tomography, and magnetic resonance imaging, have undergone major advances in their speed, resolution, and multiplanar capacity, a growing body of literature reports cases of IPA [2,10,11]. Besides the improved diagnostic techniques, the aging population and the increase in immunocompromised hosts contribute to the rise of the incidence rate of IPA as well [2,12].
The first line treatment of IPA typically involves administering broad-spectrum antibiotics that provide coverage against potential microorganisms, with a focus on Staphylococcus aureus (S. aureus) [7]. Surgical drainage, by acclamation, was the treatment of choice, and previous studies have reported quicker recovery following the open procedure [3]. More recently, image-guided percutaneous drainage (PCD) has been established as another effective and safe alternative procedure [3,7].
Even though there are established treatment procedures, the mortality rate of IPA is still high due to the difficulty of early recognition. Although several studies have emphasized the importance of the early recognition and timely treatment of IPA, the prognostic factors for poor outcomes have yet to be documented in detail due to the small number of patients in the published studies. The aim of this study was to evaluate the clinical characteristics and identify the predictors of poor outcomes in patients with iliopsoas abscess.

Study Design and Setting
This study was a retrospective, multicenter, observational study that utilized regularly collected electronic medical records (EMRs) from the emergency departments (EDs) of six hospitals in Taiwan that sharing the same EMR system. These hospitals included two tertiary medical centers and four regional hospitals. The combined capacity of these study sites exceeded 10,000 beds, and the EDs recorded an annual patient visit count of more than 500,000. The data collection period extended over six years, from 1 January 2016 to 31 December 2021. This study was conducted in accordance with the Declaration of Helsinki and received approval from the institutional review board of the Chang Gung Medical Foundation (IRB no. 202201959B0) with qualification for a waiver of informed consent.

Patient Selection and Data Collection
Through searching the electronic medical records (EMRs) during the study period, all the patients diagnosed with iliopsoas abscess in the EDs were first identified. Patients with an age under 18, incomplete medical records, or duplicated data were excluded. The patients selected by the EMRs were further reviewed by three physicians (L.-J.J., L.-Y.C., Y.-C.C.) for their inclusion eligibility.
A diagnosis of iliopsoas abscess was made if the patient showed the clinical symptoms of iliopsoas abscess and met one of the following criteria. (1) One or more positive culture or gram stain on iliopsoas compartment aspiration or surgical specimens was detected; or (2) radiological findings-including evidence from computed tomography (CT) or magnetic resonance imaging (MRI)-revealed the presence of an iliopsoas abscess, which was consistent with the patient's clinical presentation. The characteristics of an iliopsoas abscess in CT or MRI typically include: (1) enlargement of the involved muscle; (2) rim enhancement around the abscess with decreased density in the center; and (3) the presence of gas within the abscess, visible as an air-fluid interface or small pockets of air [1]. The images were reviewed by experienced radiologists to ensure accurate diagnosis of iliopsoas abscess ( Figure 1). which was consistent with the patient's clinical presentation. The characteristics of an iliopsoas abscess in CT or MRI typically include: (1) enlargement of the involved muscle; (2) rim enhancement around the abscess with decreased density in the center; and (3) the presence of gas within the abscess, visible as an air-fluid interface or small pockets of air [1]. The images were reviewed by experienced radiologists to ensure accurate diagnosis of iliopsoas abscess (Figure 1). For each identified patient, the demographic information (i.e., age and sex), initial vital signs upon admission, and comorbidities (i.e., hypertension, diabetes mellitus, chronic kidney disease, malignancy, intravenous drug abuse, coronary artery disease, liver cirrhosis, prior stroke, congestive heart failure, HIV infection, and prior tuberculosis (TB), were retrieved. Information regarding the initial presentations, site of involved iliopsoas muscle, pattern, and size of abscess formation, laboratory findings, the source of origin, organisms identified from blood or abscess cultures, treatment modalities, and length of hospital stay were collected. Laboratory findings, which included the results of blood tests that covered white blood count (WBC), hemoglobin, platelet count, international normalized ratio, C-reactive protein (CRP), and culture were obtained. We also classify IPA into primary or secondary types based on the presence of other sources of infection. Primary IPA usually originates within the muscle itself without any other infection focus, while secondary IPA occurs as a result of an infection that spreads from a nearby organ or tissue.
The primary outcome was in-hospital mortality. The secondary outcomes were the length of hospital stay (LOS) longer than 30 days and IPA recurrence. Patients were followed up with for one year to identify IPA recurrence, which was determined by reviewing imaging findings and surgical records in the EMRs in both outpatient and inpatient settings. For each identified patient, the demographic information (i.e., age and sex), initial vital signs upon admission, and comorbidities (i.e., hypertension, diabetes mellitus, chronic kidney disease, malignancy, intravenous drug abuse, coronary artery disease, liver cirrhosis, prior stroke, congestive heart failure, HIV infection, and prior tuberculosis (TB), were retrieved. Information regarding the initial presentations, site of involved iliopsoas muscle, pattern, and size of abscess formation, laboratory findings, the source of origin, organisms identified from blood or abscess cultures, treatment modalities, and length of hospital stay were collected. Laboratory findings, which included the results of blood tests that covered white blood count (WBC), hemoglobin, platelet count, international normalized ratio, Creactive protein (CRP), and culture were obtained. We also classify IPA into primary or secondary types based on the presence of other sources of infection. Primary IPA usually originates within the muscle itself without any other infection focus, while secondary IPA occurs as a result of an infection that spreads from a nearby organ or tissue.
The primary outcome was in-hospital mortality. The secondary outcomes were the length of hospital stay (LOS) longer than 30 days and IPA recurrence. Patients were followed up with for one year to identify IPA recurrence, which was determined by reviewing imaging findings and surgical records in the EMRs in both outpatient and inpatient settings.

Statistical Analysis
Patients were divided into survivors and non-survivors. Patient characteristics, comorbidities, initial presentations, laboratory findings, organisms, treatment modalities, and outcomes were reported as numbers (percentages) for the categorical variables and mean ± standard deviation (SD) for the continuous variables. For the categorical variables, the characteristics of the survivors and non-survivor were compared using the Chi-square test or Fisher's exact test as appropriate. For the continuous variables, independent Student's t-tests were used for the normally distributed variables, and Mann-Whitney U-tests were used for the skewed variables. The optimal cutoff values of IPA diameter for treatment modality were determined using Youden's index from ROC curve analysis. To identify the predictors for in-hospital mortality, the length of hospital stays (LOS) longer than 30 days, and IPA recurrence, a univariate logistic regression was first performed, and the statistically significant risk factors (p < 0.05) were then selected and used in a multivariate logistic regression model. All analyses were performed using SPSS software v26.0 (SPSS Inc., Chicago, IL, USA). A two-sided p value of <0.05 was considered statistically significant.

Patient Characteristics between Survivors and Non-Survivors
A total of 176 patients met the inclusion criteria and were included in the study. The patient characteristics are presented in Table 1. The number of patients who survived hospitalization was 163 (92.6%), and the number who did not survive hospitalization was 13 (7.4%). The distributions of age and sex did not differ between survivors and non-survivors. Twelve (6.8%) patients were bedridden, with significantly more patients in the non-survivor group (23.1 vs. 5.5%, p = 0.047). The non-survivor group showed a significantly higher rate of congestive heart failure (30.8% vs. 3.1%, p = 0.002) and renal replacement therapy (46.2% vs. 9.2%, p = 0.001) than the survivor group. The most frequent presenting symptom was flank or back pain (n = 130, 73.9%) and the traditionally described psoas abscess triad of fever, flank or back pain, and limp were present only in 45 (25.6%) patients. Flank or back pain was significantly less frequent in the non-survivor group (38.5 vs. 76.7%, p = 0.006). Multiple lobulated IPA was observed in 89 (50.6%) patients, while gas-forming IPA were found in 71 (40.3%) patients. The average maximal IPA diameter was 7.4 ± 4.1 cm. Regarding the laboratory findings, a significantly higher proportion of patients in the non-survivor group had thrombocytopenia (53.8% vs. 17.8%; p = 0.006).

Treatment and Outcomes
Each patient received intravenous antibiotics once IPA was diagnosed. Eighty-eight (50%) patients underwent PCD, 32 (18.2%) patients underwent surgical debridement, and 56 (31.8%) patients received antibiotics alone. Overall, 81 (46.0%) patients received primary antibiotic treatment. Of these patients, 50 survived hospitalizations without an interventional procedure, six passed away, and 25 failed the initial antibiotic treatment with additional intervention. The failure rate of antibiotics-only treatment was 38.3% (31/81). Ninety (51.1%) patients underwent primary PCD, of whom 65 survived, five passed away, and 20 failed PCD with additional salvage surgery. The failure rate of primary PCD was 27.8% (25/90). In contrast, five patients underwent primary surgical debridement, and all of them survived with a failure rate of 0% (0/6) (Figure 2). When the abscess diameter was larger than 8cm, primary PCD or surgery was significantly more frequently performed than primary antibiotics in patients with IPA (41.7% vs. 19

Treatment and Outcomes
Each patient received intravenous antibiotics once IPA was diagnosed. Eighty-eight (50%) patients underwent PCD, 32 (18.2%) patients underwent surgical debridement, and 56 (31.8%) patients received antibiotics alone. Overall, 81 (46.0%) patients received primary antibiotic treatment. Of these patients, 50 survived hospitalizations without an interventional procedure, six passed away, and 25 failed the initial antibiotic treatment with additional intervention. The failure rate of antibiotics-only treatment was 38.3% (31/81). Ninety (51.1%) patients underwent primary PCD, of whom 65 survived, five passed away, and 20 failed PCD with additional salvage surgery. The failure rate of primary PCD was 27.8% (25/90). In contrast, five patients underwent primary surgical debridement, and all of them survived with a failure rate of 0% (0/6) (Figure 2). When the abscess diameter was larger than 8cm, primary PCD or surgery was significantly more frequently performed than primary antibiotics in patients with IPA (41.7% vs. 19.6%, p = 0.004). Among all the patients with IPA, 86 (48.9%) and 30 (17%) patients presented with sepsis and septic shock, with the percentage being significantly higher in the non-survivors than in the survivors (100% vs. 44.8%, p < 0.001; 92.3 vs. 11%, p < 0.001). The mean LOS was 32.4 ± 26.9 days. There were 38 (21.6%) patients admitted to the intensive care units, and 13 (7.4%) passed away during hospitalization (Table 4)

Univariate and Multivariate Analyses of Predictors of In-Hospital Mortality, Longer LOS, and Recurrence
To identify the predictors of in-hospital mortality, we employed univariate Cox regression followed by multivariate Cox regression to create two separate models. Model A consisted of all relevant variables but did not include septic shock. In contrast, Model B included septic shock and the same variables as Model A, except for systolic blood pressure, respiratory rate (RR), and platelet count, which are components used to diagnose sepsis [13].

Discussion
This retrospective study examined the clinical characteristics in patients with iliopsoas abscess and analyzed the predictors of in-hospital outcomes-in-hospital mortality, LOSand recurrence, which have not been previously reported. Our results showed that patients with an advanced age, congestive heart failure, and thrombocytopenia had a higher risk of in-hospital mortality. The independent predictors of LOS longer than 30 days were tachypnea, gas-forming abscess, chronic kidney disease, skeletal origin, and the need of PCD or surgery, and the independent predictor of IPA recurrence was HIV infection.
Similar to the rate of 30% in the study by Chern et al. [4], the traditionally described psoas abscess triad-fever, flank/back pain, and limp [2]-presented in around only onefourth of the patients in our study (45 patients, 25.6%), and flank or back pain was especially less presented in the non-survivor group (38.5 vs. 76.7%, p = 0.006). This is compatible with the results demonstrated by Hamano et al., which showed that the classic presentation was significantly infrequent in patients with septic shock than those without (p = 0.012) [13]. This observation signifies that the timely recognition of IPA was difficult in the patients without the typical presentation, which can lead to poor prognosis.
Because of the non-specific symptoms and signs of IPA, it is often difficult to diagnose, and this can result in life-threatening situations [3]. Various studies have analyzed the clinical features and risk factors associated with IPA [3,[14][15][16]. However, owing to the rarity of the disease, the majority of research related to IPA presented a limited number of patients and only in the form of case reports and short case series [4,10,17]. To the best of our knowledge, our multicenter-based study had the largest patient number in the current literature. Moreover, this is the first study to report the predictors for poor outcomes, which consist of not only in-hospital mortality, but also length of stay (LOS) and recurrence, in patients with IPA. With a longer LOS, patients with IPA may undergo a longer period of antibiotic therapy and management of potential complications. IPA recurrence may cause further antibiotics resistance or dissemination of the infection to adjacent organs. Patients with recurrent IPA may go through more aggressive treatment modalities and a longer period of treatment. The aforementioned consequence implies a greater healthcare burden for the hospital and higher medical expenses for the patients. To ensure effective and efficient care, it is crucial to take these factors into account when assessing each patient's risk and determining the optimal treatment plan. By recognizing these associated consequences, we can improve patient outcomes while reducing healthcare costs.
IPA is classified into primary and secondary origins by whether there is an initial site other than the iliopsoas compartment.  [3,14,16,18,19]. Our results reflected an identical trend with the majority of IPA cases being of secondary origin (71.6%), and skeletal origin being the most common source of secondary IPA (n = 92, 52.3%), followed by intra-abdominal origin (n = 12, 6.8%). A study in Argentina illustrated that skeleton was the leading origin of IPA [14] and suggested that this could be explained by the low incidence of inflammatory bowel disease in their population, which is compatible with the pattern in our study.
The microbiological corroboration rate in IPA patients was 52.8% (93/176) with positive blood culture and 78.3% (94/120) with positive pus culture in our study. The causative organisms were predominantly Gram-positive bacteria, with S. aureus attributable to more than 60% of these cases, and mostly isolated from primary and skeletal origin. Of S. aureus cases, MRSA was accountable in 26.8% (15/56) of blood cultures, and 35% (14/40) of abscess aspirate cultures; both of these prevalence rates were comparable to that of the general population [3,16,20]. Whereas some of the literature suggested that empiric coverage of MRSA is not warranted for all IPA patients due to the low frequency [2,14], other recent articles indicated a significant increase in MRSA isolated from IPA [16,20]. This may be due to several reasons, such as drug-resistant organisms, repeat surgery, and increasing immunocompromised hosts [7,20,21]. Even though most cases of primary and skeletal IPA were caused by Gram-positive cocci, IPA from urinary or gastrointestinal origins usually involved the isolation of Gram-negative bacilli, and the abscess aspirate culture revealed that more than half of these cases involved polymicrobial infections [22]. Therefore, the timely initiation of empirical antimicrobial therapy with broad-spectrum coverage for iliopsoas abscess is crucial, and regional epidemiologic patterns of susceptibility should always be considered.
Antibiotic treatment combined with abscess drainage is the treatment of choice for IPA. In our cohort, 45.5% patients received primary antibiotic treatment in our study, of whom only 21.8% (50/80) survived hospitalization without an interventional procedure. The failure rate of antibiotics-only treatment was 37.5% (30/80) with six patients who have passed away. Some authors have advocated that targeted antibiotics may be adequate to treat abscesses less than 6 cm [23]. Our study showed that PCD or surgery was significantly more frequently performed than antibiotics alone in patients with IPA when the abscess was larger than 8 cm. Accordingly, we recommend careful evaluation for timely abscess drainage-including surgical intervention or PCD-in the presence of co-existent bowel lesions or retroperitoneal abnormalities, phlegmonous involvement of muscle without liquefaction, multiloculated, gas-forming, or large abscess [2,3,7,9,24,25].
Few studies have analyzed the risk factors for mortality, but some previous research revealed that advanced age, thrombocytopenia, elevated serum creatinine level, bacteremia, and cardiovascular disease were associated with in-hospital mortality [2,3,16,[26][27][28]. Our study indicated that advanced age, congestive heart failure, thrombocytopenia, and septic shock were significant independent risk factors for in-hospital mortality. Thrombocytopenia is not only an indicator of severe sepsis, but also suspends the administration of drainage [27]. The severity of the existing physical condition and clinical status (e.g., advanced age, congestive heart failure) may affect the prognosis in patients with IPA. The development of septic shock significantly increases the risk of mortality, which is attributed to several factors, including multiple organ dysfunction, immune system response, and treatment challenges [13,29]. For patients at a higher risk of mortality, a more aggressive therapeutic approach and more frequent monitoring should be considered in addition to empiric antibiotic treatment.
Our study revealed that tachypnea, gas-forming abscess, chronic kidney disease, skeletal origin, and the need of PCD or surgery are associated with LOS longer than 30 days. No studies have assessed the risk factors for longer LOS in patients with IPA. Although the relation between tachypnea and prognosis has not been reported, some studies have reported that at triage certain vital signs, such as low blood pressure, were associated with poor prognosis [27]. Tachypnea may be related to acidosis or respiratory failure and revealed the severity of sepsis, which would complicate the hospital course [27]. One study conducted by Hsieh et al. revealed that gas-forming IPA frequently requires aggressive drainage and, consequently, contributes to longer LOS [3]. Several studies have demonstrated that an elevated serum creatinine level or dialysis status was associated with poor prognosis [2,3,16,26,27,29,30]. Chronic kidney disease results in impairment of the immune system and aggravates the disease [31][32][33]. For patients with IPA from secondary origin, the infectious origin needs to be treated in addition to IPA. The standard treatment duration for skeletal infections, such as osteomyelitis or spondylitis, is typically longer than that of other types of infections, often lasting six to eight weeks. This is likely the reason behind the correlation between infections of skeletal origin and a longer LOS.
Our study was the first to examine the predictors of IPA recurrence. The results of our analysis indicated that HIV infection was the predictor for IPA recurrence. The study by Lopez et al. highlighted the specific clinical features of IPA in patients with HIV infection [21]. The immunosuppression in this population leads to a high susceptibility to opportunistic infections. The difficulty of completely eliminating pathogens, combined with their weakened immune status, means that patients with HIV infection are at a heightened risk for IPA recurrence.

Limitations
This study has several limitations. First, its retrospective design resulted in missing data and an uneven collection of clinical variables, such as a lack of abscess aspirate culture for all patients. Second, the rarity of IPA resulted in a small sample size, which might have limited the statistical power and generalizability of our findings. Additionally, the limited number of mortality events could disproportionately influence the model's estimates of the impact of specific variables. Finally, our study was conducted in Taiwan, a monoethnic country where most cases were Asian, so our results may not be representative of the broader population. Further prospective multi-center studies with large sample sizes would be needed to further verify our findings.

Conclusions
Iliopsoas abscess is a medical emergency requiring early recognition for the initiation of timely treatment and the reduction of the associated morbidity and mortality. Patients suffering from iliopsoas abscess with advanced age, congestive heart failure, thrombocytopenia, and septic shock had a significantly higher risk of mortality. Tachypnea, gas-forming abscess, chronic kidney disease, IPA from the skeletal origin, and the need of PCD or surgery were associated with a longer LOS. Patients with HIV infection were vulnerable to the recurrence of IPA. The recognition of these associated factors allows for risk stratification and the determination of the optimal treatment plan in patients with iliopsoas abscess.