Authors: John A. Buchan, Andrew Burkhart, Phillip R. Ross, Peter J. Stern
Categories: Research, Upper extremity infections, Acute phase reactants, C-reactive protein, Erythrocyte sedimentation rate, White blood cell count, Soft tissue infections, Incision and drainage, Septic arthritis, Diagnostic biomarkers
Source: International Journal of Emergency Medicine
Authors: John A. Buchan, Andrew Burkhart, Phillip R. Ross, Peter J. Stern
Upper extremity soft tissue infections are commonly encountered in emergency department settings and often require timely interventions. Acute phase reactants, such as C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and white blood cell count (WBC), are commonly used in diagnostic workups. However, the utility of these biomarkers in the management of upper extremity infections remains uncertain. This study evaluates the diagnostic value of these acute phase reactants in a broad cohort of patients requiring intervention for upper extremity infections.
A retrospective review over five years identified 103 patients with upper extremity infections who required incision and drainage. The study found variable sensitivity and reliability of acute phase reactants, particularly for infections distal to the wrist. The average ESR was 42 mm/hr, CRP 64.1 mg/L, and WBC 10.5 × 10³/uL. Of patients with normal acute phase reactants, all had infections located at or distal to the wrist. Notably, patients with culture-negative infections did not exhibit significant differences in laboratory values compared to those with positive cultures.
Our study highlights the limitations of acute phase reactants as diagnostic tools for upper extremity infections, particularly in cases involving infections distal to the wrist. Despite the frequent elevation of these markers in infections requiring surgical intervention, they should not be relied upon as sole determinants of management. A thorough clinical examination and history remain critical in guiding treatment decisions. This study underscores the need for caution when using these biomarkers to rule in or rule out infection in the emergency department.
Infections of the upper extremity are a common problem encountered in the emergency department. Evaluating patients with the acute onset of swelling and pain requires a broad differential including trauma, infection, and inflammatory or crystalline arthropathy. However, an accurate and timely diagnosis is crucial as soft tissue infections in the upper extremity frequently necessitate intervention. In search of objective information the work up will often include imaging and obtaining laboratory values such as white blood cell count (WBC), C-reactive protein (CRP), or Erythrocyte sedimentation rate (ESR).
For infections such as septic arthritis of the knee, hip and shoulder it has been reported that ESR and CRP have a sensitivity of > 90%, and can be useful in tracking patient’s response to treatment [1, 2]. Obtaining these laboratory tests has therefore become a routine portion of the workup of many Orthopaedic conditions upon presentation to the emergency department [3]. However in infections of the hand and upper extremity there is less agreement on the usefulness of acute phase reactants. In small joints, such as interphalangeal joints of the hand, CRP has been shown to be significantly less likely to be elevated than in large joint septic arthritis [4]. Even among patients with a septic wrist there is conflicting evidence about the usefulness of acute phase reactants. Some have shown similar ESR, CRP and WBC among patients with the final diagnosis of a septic wrist, gout or cellulitis while others have found only CRP to be useful in differentiating wrist inflammation from septic arthritis [5, 6]. In another study of patients with hand infections requiring surgical drainage, excluding paronychia, CRP was normal in 79% of these patients. ESR was normal in 33% and only slightly elevated in 14% [7].
More recent studies have presented evidence that laboratory values do in fact have reasonable sensitivity in the diagnosis of infections of the upper extremity. Gauger et al. showed that in patients with operatively treated upper extremity soft tissue infections CRP was elevated in 90% of patients – whereas ESR and WBC was less likely to be elevated [8]. This was further corroborated by Blumenthal et al. who showed the sensitivity of at least one acute phase reactant being elevated was 81% in patients with culture positive upper extremity abscesses [9]. However these recent studies only included patients treated in the operating room or those with culture positive infections. Many patients with soft tissue infections are treated outside of the operating room in the emergency department. Additionally, there are a subset of patients that despite obtaining cultures, no organisms are isolated. So called “cultures negative infections” have been a known entity in total joint arthroplasty and can be attributed to a variety of causes such as laboratory technique, organism factors or patient factors such as preoperative antibiotics [10, 11]. This study expands on previous work by evaluating the utility of acute phase reactants in a broader group of patients which more closely reflects patients presenting to the emergency department with the need for intervention for an upper extremity infection.
Approval from the institutional review board was obtained and a retrospective review was conducted of patients presenting to the emergency department of a regional health system which includes a level one trauma center. Patients were identified through a search of the electronic medical record using Current Procedural Terminology (CPT) billing and International Classification of Diseases 10th revision (ICD-10) codes pertaining to soft tissue infections of the upper extremity over a 5-year period from 2019 to 2024.
Patients were included in this review if they required an incision and drainage for a soft tissue infection at or below the elbow. This included procedures performed at bedside in the emergency department and those performed in the operating room. Patients were considered to have an infection if positive cultures were obtained or there was documented purulence during drainage. Simple paronychia and felons were excluded as nearly all could not be included on account of not having complete laboratory information. If documentation was ambiguous or not specific about seeing purulence, then the patient was excluded. Patients were also excluded if they had a previous history of operative intervention or orthopaedic fixation. Patients were excluded if they were < 18 years old or ESR, CRP, and WBC were not obtained within 24 h of presentation. Information collected included demographic information as well as history of intravenous drug use, smoking status, and diabetes. Immunocompromised patients that were included in this study were defined as patients on immunosuppressive therapy for organ transplants, inflammatory arthritis patients taking disease modifying antirheumatic agents or chronic steroids, chemotherapy, and acquired immunodeficiency syndrome(AIDS). Normal laboratory values of ESR < 15 mm/hr, CRP < 10.0 mg/L, WBC 3800–10,800/uL were based on our institution’s standards. There was one patient identified in this review with a necrotizing soft tissue infection and was excluded from subsequent analysis. Sensitivity was calculated for each marker based upon true positives being patients meeting the inclusion criteria for this study. A two tailed students t-test was used to compare laboratory values between groups of patients with significance set at a p-value of < 0.05.
Of the 782 patients identified by CPT/ICD-10 codes over a 5-year period, there were 103 that had a soft tissue infection of the upper extremity after patients with prior operative intervention, simple paronychia or felons, and incomplete laboratory data were excluded. Incomplete laboratory data was the most common reason for exclusion. The diagnoses included were soft tissue abscess, flexor tenosynovitis and septic arthritis. The most common co-morbidities were smoking (59%) and a history of IV drug use (31%) (Table 1). There was no significant difference in the average ESR, CRP and WBC among patients who were immunocompromised and those that were not.
Table 1Demographics and Co-MorbiditiesTotal Patients103Gender (Male: Female)61:45Age (STD)46.4 (14.92) Comorbidities Diabetes18 (17%) Smoking61 (59%) IVDU32 (31%) Immunocompromised12(12%)
The most common locations for a soft tissue abscess was the finger and hand (Table 2).
Table 2Location of soft tissue abscessInfection TypeFingerHandCarpusForearmSoft Tissue Abscess (n = 66)322932
Among patients with the final diagnosis of septic arthritis the wrist was the most common location (n = 7), followed by the metacarpophalangeal joint (n = 2), and the elbow and proximal phalangeal joint each with one occurrence. The average ESR was 42 mm/hr, CRP 64.1 mg/L, and WBC was 10.5 × 10^3/uL among all diagnoses, specific laboratory values for each diagnosis are listed in Table 2. WBC had the p0-Eight patients had completely normal acute phase reactants and an additional 14 patients only had one elevated marker. All patients with normal acute phase reactants had infections of the hand or fingers in this cohort. Among specific diagnoses average ESR was the highest for septic arthritis (66.1 mm/hr) while CRP was the highest for flexor tenosynovitis (101.3 mg/L) (Table 3).
Table 3Average lab values for each diagnosisInfection Type n Average ESR (mm/hr)Average CRP(mg/L)Average WBC(x10^3^/ul)Number with Normal LabsSoft Tissue Abscess6639.148.410.45Joint/Septic Arthritis1166.281.410.11Tenosynovitis2641.7101.311.62
Methicillin-resistant Staphylococcus aureus (MRSA) was the most common organism isolated. It was the most common isolate from the abscess group and the second most common in the septic arthritis group (Table 4).
Table 4Most common organism isolated by diagnosis n Soft tissue abscess Methicillin resistant Staphylococcus aureus17 Methicillin sensitive Staphylococcus aureus8 Streptococcus pyogenes5Septic arthritis Methicillin sensitive Staphylococcus aureus3 Culture Negative3 Methicillin resistant Staphylococcus aureus2Tenosynovitis Streptococcus pyogenes5 Methicillin sensitive Staphylococcus aureus5 Methicillin resistant Staphylococcus aureus5
Patients with MRSA isolated from cultures had a significantly higher average CRP (89.5 mg/L) compared to CRP from all other culture positive infections (54.2 mg/L) (p < 0.05). There was no significant difference noted between WBC and ESR for MRSA infections compared to other infections. Of all the patients who had cultures taken, 16 patients had negative cultures and no cultures were sent on 20 patients with the majority of those patient’s having an I&D at bedside in the emergency department. Average acute phase reactants were not significantly different between patients with culture negative infections vs. those with culture positive infections.
While ESR, CRP and WBC are routinely ordered on patients presenting with a concern for an upper extremity infection these results should be interpreted with caution when making clinical decisions for patients. Our results found that WBC was not reliably elevated in patients with an upper extremity infection necessitating surgical management which is consistent with prior literature [8, 9]. However our results differed slightly in that ESR was more commonly elevated than CRP in our results (81% vs. 78%) as compared to previous reports showing the opposite. However, practically we feel this distinction is of little clinical value.
In contrast to previous studies, we choose to include patients with documented purulence in the absence of positive cultures as well as patients receiving treatment outside of the operating room which is common at our institution for treatment of uncomplicated upper extremity infections. This approach allows us to include a broader range of patients, more similar to the undifferentiated patients presenting to the emergency department. One such patient in our review presented 2 days after a puncture wound to the middle finger while gardening. The patient had all four Kanavel signs on presentation and upon operative debridement in the operating room was found to have purulence within the flexor tendon sheath [12]. The patients’ cultures however never became positive, potentially because they had received antibiotics at the outside hospital many hours before transfer and subsequent surgical management. Notably the patient’s inflammatory markers were only mildly elevated, ESR 31 mm/hr, CRP 10.3 mg/L, and WBC 13.3 × 10^3^/uL, and the decision to proceed with intervention was largely based on the physical exam.
Patients presenting without elevated acute phase reactants reinforce the need for clinical examination and history. Even in patients with fulminant infections of the upper extremity, acute phase reactants can be normal. Another otherwise healthy patient who was incarcerated presented for a week or more of worsening finger swelling and was found to have a large index finger abscess complicated by underlying osteomyelitis and cultures positive for S. aureus. The infection was so advanced that amputation was offered, yet the patient’s ESR, CRP and WBC were not elevated. Other studies have also not found ESR and WBC to be typically elevated in osteomyelitis of the hand suggesting that not all hand infections trigger a systemic response detectable by typical methods [13]. In patients with culture positive upper extremity soft tissue abscesses, ESR had the best negative predictive value of 0.22 suggesting that these inflammatory markers do not play a role in ruling out an upper extremity abscess [9]. This may be explained by the intrinsic limits of ESR and CRP as serum biomarkers. ESR is a nonspecific marker of inflammation that can rise over 24–48 h. Anything that affects fibrinogen or red blood cells can impact its value; for example, it may be decreased in patients with sickle cell disease or severe liver disease and conversely elevated in patients with chronic renal insufficiency [14]. CRP can rise more quickly than ESR and has a half-life of approximately 19 h but similar to ESR is affected by noninfectious inflammatory states like cardiac ischemia or smoking [15]. Surprisingly we also report that acute phase reactants in this cohort were not significantly different from those who were not immunocompromised.
Every patient in our study without elevated acute phase reactants had infections that were distal to the wrist. The level of systemic inflammation required to meaningfully affect ESR, WBC and CRP may not be achieved with infections of the hand despite these patients having infections requiring intervention. This could explain the difference in the findings between Houshian et al. [7] which found the majority of their patients did not have an elevation in CRP compared to more recent studies which reported good sensitivity of CRP in upper extremity infections [8, 9]. The Houshian et al. [7] cohort only included patients with infections of the wrist and distal.
It is crucial to understand the limitations of acute phase reactants as a screening tool specifically in upper extremity infections. While in total joint arthroplasty patients ESR and CRP may be an effective screening tool as to which patients require further workup, this must also be weighed against the morbidity of missing an infected total joint arthroplasty [16, 17]. Similarly, in infections of the upper extremity such as flexor tenosynovitis, a delayed diagnosis can result in inhibited blood flow to tendons, progression of the infection, and tendon necrosis [18, 19]. Therefore our bias would be to error on the side of clinical exam and history to arrive at a diagnosis rather than rely on a screening test such a ESR or CRP to rule out infections in the undifferentiated patient.
The limitations of this study include its retrospective design as interpreting sensitivity from a retrospective cohort should be approached with caution as the composition of the group reviewed can bias the results [20, 21]. The time since the onset of symptoms could impact the elevation of acute phase reactants, however gxiven the limitations of documentation in this retrospective review we could not consistently obtain this information. In the context of this study, there are likely patients with uncomplicated infections not having a complete set of acute phase reactants which have been excluded. Additionally, the patients collected were only those undergoing some sort of surgical intervention and as such there was no true negative group with which to calculate specificity, positive or negative predictive value.
We provide a different prospective on the usefulness of acute phase reactants in evaluating patients in the emergency department with concern for upper extremity infections. Despite being elevated in the majority of cases, providers must be cautioned that it is not uncommon for acute phase reactants to remain normal in the setting of infections that require intervention.