Dynamics Of C-Reactive Protein In The Early Postoperative Period As A Predictor Of Infectious Complications And A Tool For Optimizing Antibiotic Therapy
Ochakovskaya, I. N.; Onopriev, V. V.; Dovlatbekyan, N. M.; Zhuravleva, K. S.; Zamulin, G. Y.; Durleshter, V. M.
Show abstract
Objective. To evaluate the diagnostic and prognostic significance of C reactive protein (CRP) level dynamics within the first five days after surgery for the early detection of surgical site infections (SSI) and to identify independent risk factors, taking into account regional specifics of surgical management (types of surgeries, duration of procedures), as well as the local hospital microbial landscape. Materials and Methods. A single-center retrospective cohort analysis of data from 127 patients who underwent surgical procedures between 2022 and 2024 was conducted. CRP levels on postoperative days 1, 3, and 5 were assessed, and delta values were calculated. Descriptive statistics, ROC analysis, and multivariate logistic regression were used to identify predictors of SSI. Results. Patients with SSI lacked the physiological decrease in CRP levels by day 5. The most informative indicator was the CRP level on day 3: a threshold of >106 mg/L was associated with a high risk of SSI (AUC=0.76; sensitivity 85%, specificity 63%). Independent predictors of SSI included surgery duration (OR=1.015 per 1 min; p<0.001) and the increase in CRP between days 3 and 5 (delta CRP3-5: OR=1.027; p=0.023). A combined model (clinical parameters + CRP) demonstrated the highest predictive ability (AUC=0.79). Conclusion. Monitoring CRP dynamics, particularly on days 3 and 5, is a highly informative and accessible method for the early diagnosis of SSI. A CRP threshold of >100 mg/L on day 3 and its subsequent increase should serve as a trigger for in-depth diagnostic investigation and rationalization of antimicrobial therapy. Keywords: C reactive protein, postoperative complications, surgical site infection, antibiotic therapy, predictive factors, diagnosis
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Prospective Cohort Study of Surgical Site Infections Following Single Dose Antibiotic Prophylaxis in Caesarean Section at a Tertiary Care Teaching Hospital in Medchal, India 96%
- Association between Intraoperative End-Tidal Carbon Dioxide and Postoperative Organ Dysfunction in Major Abdominal Surgery: A Retrospective Cohort Study 95%
- An effect of the COVID-19 pandemic: significantly more complicated appendicitis due to delayed presentation of patients! 94%
Similar papers in this journal
- Evaluating the Association and Predictability of Complex Medication Regimen Scores with Clinical Outcomes Among the Critically Ill 92%
- Effect of Virtually Led Value-Based Preoperative Assessment on Safety, Efficiency, and Patient and Professional Satisfaction 92%
- Screening for Right Ventricular Dysfunction in the Emergency Department Using a Smartphone ECG Analysis Application: An External Validation Study with Acute Pulmonary Embolism Patients 91%
Similar papers in this journal
- COMPARISON OF sPLA2-IIA PERFORMANCE WITH HIGH-SENSITIVE CRP, NEUTROPHIL PERCENTAGE, PCT AND LACTATE TO IDENTIFY BACTERIAL INFECTION: A PROSPECTIVE STUDY 95%
- Standard blood laboratory results in SARS-CoV-19 positive patients: do they show a typical pattern? 93%
- Transciptomic Analysis of the Effect of Remote Ischaemic Conditioning in an Animal Model of Necrotising Enterocolitis 92%
Similar papers in this journal
- Expression of nitric oxide synthase and nitric oxide levels in peripheral blood cells and oxidized low-density lipoprotein levels in saliva as early markers of severe dengue 90%
- Effects of statins on lipid profile of kidney transplant recipients: a meta-analysis of randomized controlled trials 89%
- Clinical Properties And Diagnostic Methods Of COVID-19 Infection In Pregnancies: Meta-Analysis 89%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.