Retrospective Analysis of Septic Arthritis Caused by Intra-Articular Viscosupplementation and Steroid Injections in a Single Outpatient Center.

BACKGROUND: Septic arthritis is defined by the presence of pathogen(s) in a joint by direct inoculation or hematogenous spread. Most common organisms include Staphylococcus aureus and Escherichia coli. Clinical presentation is fever, warmth and night pain, with most common joints involved being the knee and hip. Iatrogenic septic arthritis is an uncommon complication of intra-articular injection for osteoarthritis yet its complications can be devastating. We aim to highlight ten cases of iatrogenic septic arthritis in retrospective study reviewing symptoms, signs, laboratory data, causing organisms and reasons leading to those infections.
METHODS: Retrospective analysis of charts of ten patients who were admitted to Jersey Shore University Medical Center with diagnosis of iatrogenic septic arthritis.
RESULTS: Average age of patients is 69.9 years. Most common comorbidities seen in our patient were hypertension and diabetes mellitus. The most common intra-articular agents that were injected were cortisone and Synvisc. The mean incubation period was 11.9 days. Most common presenting symptoms were joint pain and swelling. The most common organism isolated in cultures was Streptococcus mitis. A total of 100% of patients underwent surgical intervention for septic arthritis. One case was complicated by sepsis.
CONCLUSIONS: Iatrogenic septic arthritis is not common; however its complications can be catastrophic to patients. Improper sterile techniques and untrained physicians are the main risks factors for this complication. Physicians should take proper sterile measures to avoid complications of intra-articular injections.

Introduction

Septic arthritis is defined by the presence of pathogen(s) in a joint by direct inoculation or hematogenous spread. It is a relatively uncommon disease with an annual incidence of 2 - 10 cases per 100,000 [1-2]. Delayed diagnosis or suboptimal treatment is associated with irreversible joint damage and permanent disability. It has about 10% mortality and significant morbidity [3]. Staphylococcus aureus including methicillin-resistant Staphylococcus aureus (MRSA) is the most common pathogen (51.4%) followed by gram negative enteric bacillus such as E.coli (5.7%), which is more prevalent in elderly patients, immunocompromised patients, and patients with intravascular devices, urinary catheters and Streptococcus pyogenes (4.7%) [1-2]. Risk factors for septic arthritis are immunosuppression (30%), previous knee surgery (14%), intravenous drug use, joint prostheses, alcoholism, diabetes, previous steroid injections, and cutaneous ulcers [2, 4-7]. Common clinical presentations are fever (30%), local warmth (100%), night pain (85%), and common laboratory abnormalities are elevated erythrocyte sedimentation rate (ESR) (85%) and C-reactive protein (CRP) (100%) [8]. A total of 22% of septic arthritis are poly-articular which often reflect bacteremia seeding multiple joints. The most common joints involved are the knee (48%) and hip (21%) [7]. Key for diagnosis of septic arthritis is microscopic analysis and cultures of the synovial fluid [5]. Gram stain and microscopy of synovial fluid has positive results in 50% [5]. Synovial fluid analysis of white blood cells (WBCs) > 50,000 has 50% sensitivity to diagnose septic arthritis.

Intra-articular injections have become popular as a treatment option for painful osteoarthritis where patients cannot tolerate non-steroidal anti-inflammatory drugs (NSAIDs) or inadequate pain control by NSAIDs or other analgesics [8]. Results from meta-analysis showed that intra-articular steroids and hyaluronic acid are associated with improvement of pain and functionality although it is not risk-free [9-10]. Risk factors associated with increasing chance of knee infection are corticosteroid injections, history of rheumatoid arthritis or other joint abnormalities, and injections performed by general practitioners [8]. In March 2017, an outbreak of multiple cases of septic arthritis occurred in NJ, USA. A total of 41 cases were identified by New Jersey Department of Health (NJDOH) [11] and the outbreak was mainly attributed to intra-articular injections administered at a single outpatient practice in NJ. Ten of these cases were treated in our hospital [11]. We retrospectively analyzed these cluster cases that were managed in our hospital to identify the major factors associated with this outbreak and to prevent future incidence.

Materials and Methods

Study was conducted at Jersey Shore University Medical Center (JSUMC), Neptune, NJ, USA. The medical records of patients with acute septic arthritis who had knee injection at local osteoarthritis clinic and was subsequently seen by Meridian Infectious Diseases as inpatient at JSUMC or outpatient from October 2016 to May 2017 were reviewed. Data was analyzed retrospectively from medical records at JSUMC where these patients were seen.

Demographic characteristics (age, gender), past medical history, clinical variables on admission, time of onset of symptoms, incubation period, bacterial pathogen, length of stay at the hospital, treatment duration, need for surgery, need for peripherally inserted central catheter (PICC) line, complications of hospitalization, and outcome (complete resolution, discharge to home, and rehabilitation) were reviewed.

Meridian Institutional review board approval was obtained for this study. All study procedures were carried out in accordance with the Declaration of Helsinki regarding research involving human subjects. All data analysis was conducted with Microsoft excel software.

Results

Demographics of patients who presented at our medical center are shown in Table 1. Most of the patients were ages between 61 and 70 with a mean age of 69.9 years. The most common comorbidities seen in our patient were hypertension (four patients) and diabetes mellitus (two patients). All these cases had more than one intra-articular injection at the outpatient facility. The most common intra-articular agents that were injected were cortisone and Synvisc.

Table 1

Demographic and Clinical Characteristics of the Patients (n = 10)

Demographics

Average age (years), 69.9

Male, 8 (80%)

Female, 2 (20%)

Caucasian, 9 (90%)

Comorbidities

Hypertension, 4 (40%)

Diabetes mellitus, 2 (20%)

Chronic kidney disease, 1 (10%)

Number of intra-articular injections received prior to admission

0 -1, 4 (40%)

2 - 3, 4 (40%)

4 - 5, 2 (20%)

Agent injected intra-articularly

Cortisone, 3 (30%)

Synvisc, 3 (30%)

Hyalagan, 1 (10%)

Euflexxa, 1 (10%)

Unknown, 2 (10%)

Incubation period (time in days from last intra-articular injection to initial symptoms)

0 - 5, 3 (30%)

6 - 11, 2 (20%)

12 - 17, 1 (10%)

More than 17, 1 (10%)

Unknown, 3 (30%)

Clinical symptoms on presentation

Joint pain, 10 (100%)

Swelling, 10 (100%)

Decrease range of motion, 3 (30%)

Fever (> 100.5 F), 2 (20%)

Erythema, 1 (10%)

The mean incubation period for these patients was 11.9 days. All patients in our case series present with joint pain and swelling, with other reported symptoms including fever (two patients) and decrease range of motion (three patients). Mean ESR and CRP levels were 52.6 mm/h and 10.3 mg/dL respectively. Two patients had positive calcium pyrophosphate deposition (CPPD) crystals and one had positive urate crystals as it appears in Table 2. The most common organism isolated in cultures was Streptococcus mitis (three patients) shown in Table 2, with other organisms all being oral flora. Five patients did not have an organism identified on cultures. Of these five patients, one patient started on empirical antibiotics (ampicillin-sulbactam) prior to joint tap, and one patient had negative cultures on the first admission but with a positive culture on the second admission. In one patient, antibiotics were stopped due to low suspicion of septic arthritis.

Table 2

Characteristics of Synovial Fluid (n = 10)

Patient number	ESR (mm in first h)	CRP reference value (0 - 0.74 mg/dL)	Appearance of synovial fluid	Fluid WBC count, reference value (0 - 1,000 cells/UL)	Crystals	Result of synovial fluid culture
1	111	26	Turbid, yellow	39,850	Calcium pyrophosphate dihydrate	Streptococcus mitis/oralis
2	60	2.99	Cloudy, yellow	97,250	Calcium pyrophosphate dihydrate	Streptococcus mitis/oralis/Actinomyces odontolyticus
3	44	20	Turbid, red	42,451	-	Streptococcus mitis/oralis/ Streptococcus gordonii
4	71	17.85	Turbid, yellow	11,933	-	Cx was negative on the first admission, Gemella morbilliform on the second admission
5	10	1.95	Cloudy, yellow	18,336	-	No growth
6	51	5.9	Turbid, yellow	-	Urate	Abiotrophia defectiva
7	9	-	Cloudy, yellow	19,000	-	No growth
8	36	12.22	Turbid, orange	50,000	-	No growth
9	76	14.6	Cloudy, yellow	98,901	-	No growth
10	58	1.9	Turbid, red	15,380	-	Streptoccus gordonii

ESR: erythrocyte sedimentation rate; CRP: C-reactive protein; WBC: white blood cell.

The average length of stay for these ten patients was 8 days. All patients underwent surgical intervention for septic arthritis. Three of ten patients needed more than one incision and drainage of the involved joint, with one patient needing three incision and drainages. Eight of ten patients were treated with antibiotics for 21 days. Five patients were treated with ceftriaxone, two with vancomycin and one with oral amoxicillin-clavulinic therapy. Seven of ten patients received a PICC line for outpatient antibiotic therapy. Six patients required placement at rehabilitation facility.

One patient was re-admitted for bacteremia with Gemella morbilliform. This patient’s first admission had a negative culture and antibiotics were stopped upon discharge. On second admission he was admitted with septic shock. He was started on vancomycin and ceftriaxone. Because Gemella morbilliform is a potential cause for endocarditis, patient had a transesophageal echocardiography (TEE) which was non-conclusive in ruling out endocarditis. Patient was continued on intravenous antibiotics (vancomycin) for 4 weeks.

Discussion

In our case review, we describe 10 of the 41 cases of septic arthritis caused by intra-articular injection reported by Ross et al at an outpatient center in NJ [11]. Typically intra-articular joint injections have side effects reported in 2-10% of cases, with most common being injection site inflammation [8]. The incidence of septic arthritis caused by intra-articular injection incidence is about 13.3% of all septic arthritis [12].

Unlike most septic arthritis, which is primarily caused by Staph aureus and few gram negative organisms, all organisms identified at this facility were part of normal oral flora. In a case series review by Xu et al, the most common organism identified in septic arthritis after intra-articular injection was coagulase-negative staph, which is a common organism seen in skin flora [8]. Xu et al described tissue coring with epidermis and dermis into the joint as the likely pathogenesis behind this [8]. Another two case series of septic arthritis caused by intra-articular injections report that one patient had a staph aureus infection, and the second one had Neisseria mucosa [9]. In the case series review conducted by Ross et al, which included our ten patients, septic arthritis was deemed to be due to poor sterile techniques utilized while performing intra-articular injections [11]. According to Department of Health (DOH) report [11], the main reason outlined for this outbreak was poor compliance with standard infection control precautions during intra-articular injections including unhygienic preparation of the injection, lack of hand hygiene during administration, use of multidose vial of local anesthetic for multiple patients, intra-articular administration by general physician without special training, and high volume of patients [13].

Four out of 10 patients did not have growth on synovial fluid cultures, and one of these patients had received empirical antibiotics prior to joint tap. Inability to identify organisms in septic arthritis is fairly common, with literature suggesting its occurrence as nearly 58.8% [14-15]. This false negative result can be a challenge for physicians as it can affect appropriate therapy, leading to further clinical complications. Pseudoseptic arthritis is rare complication of intra-articular injections which should also be considered in appropriate clinical settings [16]. Pseudoseptic arthritis is an inflammatory arthritis which presents about 72 h after intra-articular injection. Symptoms are similar to septic arthritis; however joint cultures are sterile but inflammatory [16]. Mechanisms of injury are not fully understood. The diagnosis of pseudoseptic arthritis is usually made by exclusion of other causes.

An interesting observation of our ID physicians reviewing this outbreak is that interstate healthcare facilities often do not fall under jurisdiction of county or state DOH. Health departments can make recommendations but cannot compel private facilities to close. In this case, the outpatient facility reopened soon after their state inspection was completed.

Conclusions

Intra-articular injections to treat osteoarthritis typically are associated with minor side effects and a good alternative of pain control in opioid epidemic. But due to improper technique, poor procedure preparation including early setup of injections and inadequate training of personnel, this simple procedure can be associated with serious complications reflected in this case series. Internists especially primary care physicians should refer their patients needing these injections to appropriate providers such as orthopedic surgeons or rheumatologists.