Low albumin level is more strongly associated with adverse outcomes and Staphylococcus aureus infection than hemoglobin A1C or smoking tobacco

Abstract Postsurgical deep musculoskeletal infections are a major clinical problem in Orthopaedic Surgery. A serum‐based nomogram, which can objectively risk‐stratify patients, and aid surgeons in delineating infection risk associated with orthopedic surgical interventions, would be immensely helpful. Here, we constructed a multi‐parametric nomogram based on serum anti‐Staphylococcus aureus antibody responses, patient characteristics including demographics and standard clinical tests. This nomogram was formally tested in a prospective cohort study comparing 303 hospitalized patients with culture‐confirmed S. aureus infection compared with a cohort of 223 healthy screened preoperative patients. Serum anti‐S. aureus antibody responses, standard of care clinical tests, and patient demographic data were utilized to perform multivariate logistic regression analysis to quantify the presence of infection and adverse outcome using odds ratios (OR) and to assess predictive ability via area under the ROC curve (AUC). At enrollment, high anti‐S. aureus IgG titers were predictive of infection. Remarkably, low serum albumin was found to be significantly associated with infection (OR = 479.963, 95% CI 61.59 ‐ 3740.33, p < 0.0001) and this finding was surprisingly higher than BMI or HbA1c‐associations. Combining all risk factors in the nomogram yielded a diagnostic AUC of 0.949 for predicting S. aureus infection. Our results indicate that a serum‐based multi‐parametric nomogram can be useful in diagnosing S. aureus infections, and importantly, malnourishment is significantly associated with these infections.


| INTRODUCTION
Deep musculoskeletal infections after orthopedic operations represent a major clinical problem and are gradually increasing each year with increasing surgical volume. 1 The most commonly isolated pathogen in orthopedic infection is Staphylococcus aureus. 2 Overall, Methicillinresistant Staphylococcus aureus infects 120,000 patients with bloodstream infections and contributes to 20,000 deaths annually. 3 With an increasing number of joint arthroplasties performed in the United States, there has been a commensurate increase in peri-prosthetic joint infection (PJI) which places a large economic burden on the American health care system, costing $566 million in 2009. 4 S. aureus is the leading cause of PJI. 5 With rising antibiotic resistance, a need exists for nonantibiotic, immune-based approaches to treat drug-resistant infections. 6 Host factors play a key role in infection, as we have been unable to reduce infection rates for elective procedures below 1%-2%. 7 Many patient-specific factors seem to influence susceptibility to infection, including comorbid conditions 8 such as obesity and diabetes, all of which can affect immune responses to S. aureus infection. 9 According to the 2018 International Consensus Meeting on Musculoskeletal Infection, high BMI received unanimous agreement as a risk factor for surgical site infection and PJI. 10 Albumin is a commonly used marker for nutritional status. Serum albumin level >3.5 g/dl is frequently used as an optimization parameter in the clinic before major elective surgery. 10 We have previously identified multiple immunodominant S. aureus antigens, which provoke robust antibody responses in patients with musculoskeletal infections. Antibody titers against these S. aureus antigens can be measured by immunoassay of a patient's sera. 11 To date, the only way to confirm the diagnosis of deep musculoskeletal infection with S. aureus is to obtain a fluid/tissue sample followed by microbial culture, which can be a lengthy process

| Humoral response and patient factors associated with infection
The sera for all 526 patients were tested for antibody responses against these aforementioned S. aureus antigens in our custom Luminex immunoassay. Interestingly, we observed that high levels of serum IgG titers were associated with the presence of infection

| Humoral response and patient factors associated with adverse outcome
High levels of all antigens were also predictive of adverse outcome (

| Additive risk factors associated with increased infection and adverse outcomes
With an increasing number of risk factors, the odds ratio of infection increased ( Table 5). The same is true of the adverse outcomes (Table 6).

| DISCUSSION
The primary aim of this study was to analyze and compare the im- There are several strengths of our study. It is prospective with large sample size and uses both clinical demographic data with quantifiable humoral immune response measurements to analyze S. aureus infection risk. Our data were collected before the COVID-19 pandemic. This is a tremendous strength, as we do not know how COVID infection may alter the immune system, which could act as a confounding variable. However, our study has limitations. The infection group is a heterogeneous population of hospitalized patients, not just orthopedic-related infections. This heterogeneous population may also be considered a strength. Additionally, there was some missing data -not all laboratory data (hemoglobin A1c, albumin) or outcome data were available for every patient.

| CONCLUSIONS
Of the host risk factors evaluated, low albumin was most strongly associated with S. aureus infection and adverse outcome-more so than high BMI or hemoglobin A1c. By combining all host risk factors, we were able to achieve an AUC of ∼0.95 for predicting S. aureus infection.

ACKNOWLEDGMENTS
This study was supported by AO Trauma North America Resident