RISK FACTORS FOR INFECTION IN KNEE ARTHROPLASTY: A CASE-CONTROL STUDY

Abstract

Introduction: Deep infection continues to be the first most important early complication in knee arthroplasty. It is usual to apply standard prophylaxis to all patients, but it is not usual to use special measures in those of them who present a higher risk. Moreover, sometimes these patients are even not identified.

Purpose: To analyse statistically significant risk factors for deep infection in patients with a knee arthroplasty.

Patients and Methods:

Design: Case-control study.

Observational and retrospective comparison of incidence or prevalence of all risk factors described in the literature. These factors have been classified according to the period of risk in: epidemiologic; pre, intra and postoperative; and distant infections.

Case series: 32 consecutive patients with a deeply infected knee arthroplasty operated in the same Department of a University General Hospital.

Control series: 100 randomly selected patients, operated in the same hospital and period of time, with no deep infection in their knee arthroplasty along follow-up.

Pearson was used for comparison of qualitative variables and ANOVA for quantitative ones.

Results: The following risk factors were significantly more frequent (p< 0.05) in the patients with an infected knee arthroplasty:

• Preoperative conditions: previous surgery in the same knee (25% vs 9%), chronic therapy with glucocorticoids (19% vs 4%), immunosuppressive treatments (16% vs 3%), and non-rheumatoid inflammatory arthritis (13% vs 0%). Patients in this case-control did not present a significant difference in the prevalence of rheumatoid arthritis, diabetes, obesity (BMI> 30), chronic liver diseases, or alcohol addiction.

• Intraoperative facts: a prolonged surgical time (149 min vs 108 min) and intraoperative fractures. Differences were not found in the amount of bleeding or need for transfusion.

• Postoperative events: secretion of the wound longer than 10 days (48% vs 0%), wound haematoma (36% vs 6%), new surgery in the knee (30% vs 0%), and deep venous thrombosis in lower limbs (10% vs 1%).

• Distant infections (risk for haematogenous seeding): deep cutaneous (27% vs 3%), generalized sepsis (7% vs 0%), upper and lower urinary tract (27% vs 5%), pneumonias and bronchopneumonias (27% vs 5%), and diverse abdominal focus (17% vs 1%). On the contrary, significant differences were not found in the prevalence of severe oral or dental infections.

• Epidemiologic characteristics: significant differences were not found in gender or in the prevalence of any aetiology.

Conclusion: To identify significative risk factors for deep infection in knee arthroplasty is important:

1. to control and minimize these risk factors when present

2. when this is not possible not possible, to implement additional prophylactic measures.